MIAMI UNIVERSITY the Graduate School Certificate for Approving The

Home , Laminar organization, Regeneration in humans

MIAMI UNIVERSITY The Graduate School

Certificate for Approving the Dissertation

We hereby approve the Dissertation Of

Mayur Madhavan

Candidate for the Degree:

Doctor of Philosophy

Director Dr. Katia Del Rio-Tsonis

Reader Dr. Phyllis Callahan

Reader Dr. Paul F. James

Dr. Paul A Harding

Graduate School Representative Dr. Joseph Carlin Abstract

Mechanisms of Transdifferentiation and Regeneration by Mayur Madhavan

Hundreds of millions of people across the world suffer from severely impaired vision and as the population of the world grows older, this problem will become more severe. One avenue of therapy for these patients may be to induce healthy cells of the eye to regenerate cells that may be lost or damaged due to disease. The research in this dissertation focuses on regeneration via a process called transdifferentiation, wherein differentiated cells undergo dedifferentiation, proliferate and then redifferentiate to replace lost or damaged cells. Very few animals can regenerate ocular tissues and two such animals have been used in my research. The adult newt has been used to study the process of lens regeneration whereas the embryonic chick has been used to study retina regeneration.

Our studies using the newt have helped us identify a novel non-immunological role for complement components C3 and C5. Our studies show that these molecules are expressed during a variety of regenerative tissues and our results suggest that these molecules might convey positional information during regeneration. We have also studied the role of the transcription factor Pax-6 during lens regeneration and show that it is required for the proliferation of cells and differentiation of lens fibers.

We have established the chick as a good model system to study regeneration. We have also shown that the chick is not only a good model to study transdifferentiation but also to study the activation of progenitor cells. The work described in this dissertation also dissects the Fibroblast Growth Factor signaling cascade that is involved during transdifferentiation and also examines the roles of transcription factors such as Pax-6 and Microphthalmia during retina regeneration. MECHANISMS OF TRANSDIFFERENTIATION AND REGENERATION

A DISSERTATION

Submitted to the Faculty

of Miami University

in partial fulfillment of

the requirements for the degree of

Doctor of Philosophy

Department of Zoology

Mayur C. Madhavan

Miami University

Oxford, Ohio

2005

Dissertation Director: Katia Del Rio-Tsonis

Table of Contents

Chapter 1 General Introduction

1.1 Aim of this Research 1 1.2 Development of the Eye 2 1.3 Regeneration in the Adult Newt 4 1.4 Overview of Other Regenerative Processes in the Newt 7 1.5 Retina Regeneration in the Embryonic Chick 13 1.6 Organization of this Dissertation 16 1.7 References 17

Chapter 2 Involvement of Complement Molecules in Regeneration

2.1 Introduction 21 2.2 Methodology 23 2.3 Results 25 2.4 Discussion 33 2.5 References 35

Chapter 3 The Role of Pax-6 in Lens Regeneration

3.1 Introduction 38 3.2 Methodology 40 3.3 Results and Discussion 44 3.5 References 58

ii Chapter 4 Regeneration and Transdifferentiation in the Embryonic Chick

4.1 Introduction 61 4.2 Methodology 64 4.3 Results 67 4.4 Discussion 84 4.5 References 86

Chapter 5 Conclusions

5.1 Overview 89 5.2 Phases of Transdifferentiation 89 5.3 Questions of Immediate Interest 92 5.4 Future Directions 93 5.5 References 95

iii List of Figures

Chapter 1 General Introduction

Figure 1 Development of the eye 2 Figure 2 Anatomy of the eye 3 Figure 3 Lens regeneration in the adult newt 8 Figure 4 Retina regeneration in the adult newt 10 Figure 5 Limb regeneration in the newt 11 Figure 6 Tail regeneration in the newt 12

Chapter 2 Involvement of Complement Molecules in Regeneration

Figure 1 Protein expression of C3 and C5 during lens regeneration 26 Figure 2 Protein expression of C3 and C5 during retina regeneration 27 Figure 3 Expression of newt C3 and C5 protein during limb regeneration 29 Figure 4 Expression of C3 in the regenerating tail 30 Figure 5 Expression of C3 in the spinal cord and vertebra during tail 31 regeneration Figure 6 Confrontation Assay 32

Chapter 3 The Role of Pax-6 in Lens Regeneration

Figure 1 Pax-6 is associated with proliferating cells 46 Figure 2 Pax-6 expression does not correlate with cells undergoing 47 apoptosis Figure 3 Valproic Acid or Pax-6 morpholino treatment decreases levels of 49 Pax-6 Figure 4 Downregulation or knock-downn of Pax-6 inhibits lens 51 regeneration

iv Figure 5 Valproic acid treatment or morpholino injection reduces 53 proliferation in vivo and in vitro. Figure 6 Valproic acid treatment or Pax-6 morpholino treatment affects 55 the initiation of differentiation Figure 7 Effect of Valproic acid or Pax-6 morpholino treatments on late 57 stage events

Chapter 4 Regeneration and Transdifferentiation in the Embryonic Chick

Figure 1 FGF-2 induces retina regeneration in two distinct ways 69 Figure 2 Cells in the CB/CMZ are a source of regenerating retina 71 Figure 3 Regeneration gives rise to all cell layers of the retina 75 Figure 4 Distribution of FGFRs in the RPE 77 Figure 5 FGF-2 signals through ERK 78 Figure 6 Overexpression of constitutively active MEKDD 80 Figure 7 Overexpression of Pax-6 82 Figure 8 Overexpression of Mitf in the presence of FGF-2 83

Chapter 5 Conclusions

Figure 1 95

v List of Tables

Chapter 4 Regeneration and Transdifferentiation in the Embryonic Chick

Table 1 List of Retina Markers 72

Chapter 1

General Introduction

1.1 Aims of this research

The World Health Organization and the International Agency for the Prevention of Blindness estimate that around the world there are 45 million people who are blind and approximately a 131 million who have severely impaired vision. Even though a large majority of these cases are in developing and underdeveloped countries, blindness still effects large numbers of people in developed countries as well. According to the 2000 census, approximately 1 million people in the United States are blind and an additional 2.5 million people suffer from severe visual impairment (Friedman, 2002, [Vision Problems in the U.S. Prevalence of Adult Vision Impairment and Age-Related Eye Disease in America, Prevent Blindness America]). The major causes of blindness are age related retinal degeneration and cataracts of the lens.

One avenue of therapy for these diseases could be the induction of regeneration to replace damaged or lost tissues in the eye. Humans cannot naturally regenerate their retina or lens, and therefore, it is necessary to study animal models that have this incredible ability in order to determine the mechanisms involved, which can then be used to induce regeneration in humans. We use two model systems, the embryonic chick and the adult newt. The chick is a valuable animal model to study the mechanisms of retina regeneration whereas the newt is the best model to study lens regeneration. Regeneration in both these systems takes place through a process called transdifferentiation. Transdifferentiation is the process by which a differentiated cell undergoes dedifferentiation (loss of differentiated cell characteristics), proliferates and then differentiates into new cell types. During retina regeneration in the chick, cells of the

1 Retinal Pigment Epithelium (RPE) transdifferentiate to replace the lost retina whereas in the newt, pigmented cells of the dorsal iris transdifferentiate to regenerate a new lens. The focus of this research is to understand basic mechanisms underlying regeneration. In order to do this, we concentrate on mechanisms that may initiate regeneration as well as mechanisms underlying developmental processes that may be recapitulated during regeneration.

1.2 Development of the Eye

Figure 1. Development of the eye. A) Formation of the neural tube (nt) by the invagination of the outer layer of the embryo. B) The neuroepithelium (ne) that surrounds the neural tube forms optic vesicles on either side. C) The optic vesicle invaginates giving rise to the optic cup (oc). The interaction between the optic cup and the surface ectoderm (se) causes lens induction. D) The outer layer of the optic cup becomes the RPE while the inner layer becomes retina. The lens pinches of the surface ectoderm and becomes part of the developing eye.

During early vertebrate development, cells of the ectoderm invaginate to give rise to the neural tube (Figure 1A). The cells of the neural ectoderm (the cells that surround the neural tube) then grow outward on either side, producing the optic vesicles (Figure 1B). The optic vesicles grow towards the surface ectoderm while invaginating upon itself to form a double layered optic cup. Contact and molecular signaling between the optic vesicle and the surface ectoderm induces the formation of the lens from the surface ectoderm (Figure 1C). In addition, signals from the surface ectoderm (including fibroblast growth factor-2 = FGF-2) induce the inner layer of the optic cup to become neural retina (Pittack et al, 1997) and several transcription factors including Pax-6 and Chx-10 are essential for neural specification (Belecky-Adams et al., 1997). On the other

2 hand, the outer layer of the optic cup is under the influence of factors from the underlying mesenchyme such as activin which in turn could activate transcription factors like Microphthalmia (Mitf) that are able to specify the RPE fate (Fuhrmann et al, 2000). Pax- 6 is also essential for a cascade of events that causes the lens to grow and differentiate (for review see Lang, 2004). At the end of the developmental process, the eye that is generated has multiple cell types which are illustrated in Figure 2.

Figure 2. Anatomy of the Eye. The eye contains the following layers: 1) Sclera: the outermost layer of the eye 2) Cornea: the transparent part of the anterior portion of the eye, it is continuous with the sclera 3) Iris: the pigmented (colored) part of the eye 4) Lens: focuses light onto the retina and consists of two layers. An outer epithelial layer called the lens epithelium and a crystallin layer made of lens fibers. 5) Ciliary Epithelium and CMZ (Ciliary Marginal Zone): the area between the iris and retina. Is generally only two cell layers thick and houses stem /progenitor cells 6) Retina: consists of at least 7 different cell types that are arranged in 3 nuclear layers and 2 plexiform layers. The innermost layer is the Nerve Fiber Layer (NFL), which consists of axons of ganglion cells which eventually become the optic nerve. The Ganglion Cell Layer (GCL) is the innermost nuclear layer and consists of only ganglion cells. The next layer is the Inner Plexiform Layer (IPL) and consists of synapses between ganglion cells and process from cells of the Inner Nuclear Layer (INL). The INL consists of 3 cells types: amacrine cells, bipolar cells and horizontal cells. The Outer Nuclear Layer (ONL) consists of rod and cone photoreceptors and these photoreceptors synapse with cells of the INL in the Outer Plexiform Layer (OPL). The entire retina is traversed by support glial cells called Müller glia.

3 7) RPE: The RPE consists of a single layer of pigmented cells that act as support cells for the retina 8) Choroid: the layer between the RPE and Sclera.

In this study, we examine signaling molecules like Fibroblast Growth Factor (FGF) that induce transdifferentiation and complement components that may play a role in the early phases of regeneration and also study the roles of developmentally important transcription factors such as Pax-6 and Mitf in the processes of retina and lens regeneration.

1.3 Regeneration in the adult newt

The adult newt is one of the few animals that can regenerate almost every part of its body. Even though we shall discuss multiple regenerative processes in this dissertation, the main focus of our newt work is lens regeneration. Its ability to regenerate the lens was first described, independently by Collucci and Wolff in the 1890s. In the ensuing 110 years, research using the newt has established a number of interesting phenomena with respect to its ability to regenerate its lens, but the molecular mechanisms that trigger and sustain this process are still unknown. (reviewed in Del Rio- Tsonis and Tsonis., 2003; Tsonis et al., 2004c)

One of the interesting facts of lens regeneration is that even though the environment around both the dorsal and ventral iris is almost uniform, the lens always regenerates from the dorsal iris and not from the ventral iris. This would suggest that the dorsal iris has some intrinsic property that allows the cells to regenerate and that the ventral iris does not have the machinery required for regeneration and/or has an inhibitory mechanism in place. This is supported by the fact that ventral irises cultured in vitro for prolonged periods of time can transdifferentiate into lentoid structures (Eguchi et al., 1974). Therefore, exploring molecules that are preferentially expressed in the dorsal iris or in the ventral iris during regeneration might shed some light on the mechanisms underlying this process.

4 There are three sets of molecules that are preferentially expressed in the dorsal iris. The first set of molecules are signaling molecules such as Fibroblast Growth Factor (FGF) and their Receptors (FGFR), and also morphogens like Sonic Hedgehog (Del Rio Tsonis et al., 1998a; Thitoff et al.., 2003; Tsonis. et al., 2004b). Members of the FGF family play roles not only in the induction of lens development but also in the differentiation of lens fibers and lens epithelium. Members of the FGF family are expressed during lens regeneration and inhibiting FGFR-1 during lens regeneration has been shown to inhibit the differentiation of the lens vesicle (Del Rio Tsonis et al., 1998a). Sonic Hedgehog on the other hand has been shown to play a role in proliferation and differentiation during lens regeneration (Tsonis et al., 2004b). A second set of molecules differentially regulated in the dorsal iris are cell cycle regulators like Cyclin Dependant Kinases (CDK) and Retinoblastoma (Tsonis et al., 2004a). These molecules are thought to play a role in the re-entry of cells into the cell cycle once the signal to undergo transdifferentiation has been transduced. Inhibiting the function of CDKs result in the inhibition of lens regeneration and therefore provides further evidence that these molecules play an important role in lens regeneration (Tsonis et al., 2004a). A third set of proteins implicated in lens regeneration are transcription factors like Prox-1 and Pax-6 (Del Rio-Tsonis et al., 1995; Del Rio-Tsonis et al., 1999). The exact role for these transcription factors in lens regeneration is still unknown. My research has focused on dissecting the roles of two sets of molecules: Signaling molecules of the Complement family, namely C3 and C5, and the transcription factor Pax-6.

Complement components are a family of molecules that are part of an ancient form of innate immunity. The nine members of this family (C1-C9) are generally found in blood and work together to destroy invading foreign cells. Upon activation, the complement system can bring about a variety of responses including the formation of the Membrane Attack Complex, triggering the release of histamine by C3a (the cleavage product of complement C3); C3a and C5a (the cleavage product of C5) have anaphylotoxic activity (the ability to attract macrophages). Though complement components have been traditionally known to function in the immune system, a number of recent studies have shown that some members have non-immunological functions. A

5 number of complement components act as differentiation signals through domains that interact with extracellular matrix proteins such as collagen, fibronectin and integrins (Hautanen and Keski-Oja, 1983; Leivo and Engvall, 1986; Kiss et al., 1989). Some members of the complement family have also been implicated in promoting proliferation. C5a can be mitogenic to human neuroblastoma cells whereas C3 can promote proliferation of B-cells in vitro (Servis and Lambris, 1989; O’Barr et al., 2001). In another set of experiments, C5a has been implicated in the induction of apoptosis in a different line of neuroblastoma cells (Farkas et al., 1998). In addition to these reported non-immunological functions for the complement family, C5 deficient mice were shown to have impaired liver regeneration suggesting complement components may also be involved in regeneration (Mastellos et al., 2001). In addition to these data, work from our laboratory demonstrated that C3 is expressed by the cells of the regenerating axolotl and newt limb (Del Rio-Tsonis et al., 1998b; Kimura et al. 2003). This onset of C3 expression during regeneration suggests that complements may play a role in regeneration. We also hypothesize that complement components may have a general role during regeneration and may be involved in regeneration of various organs. Therefore we will examine their role in retina, limb and tail regeneration to ascertain if they are involved in these processes as well.

The second molecule that was the focus of my work is the transcription factor Pax-6. Pax-6 is expressed in the dorsal iris during lens regeneration (Del Rio-Tsonis et al., 1995) and it may play a role during regeneration because during development it is crucial for lens induction. Further Pax-6 mutant mice (small eye (sey) mutants) fail to form a normal lens placode (Hogan et al., 1986). In addition Pax-6 is known to regulate other molecules crucial to lens formation, such as the Sox, Fox and Mab family of genes (Chow and Lang, 2001). Pax-6 has also been implicated to interact with Retioblasatoma (Rb) protein and Pax-6 is able to regulate the expression of various crystallins and therefore plays a role in the differentiation of lens fibers (reviewed by Cvekl and Piatigorsky, 1996). In the adult newt, Pax-6 is absent in the iris prior to lentectomy, but Pax-6 transcripts appear in both the dorsal and ventral iris within the first five days after the lentectomy, during the onset of regeneration and then is localized specifically to the

6 dorsal iris as the process of regeneration continues (Mizuno et al., 1999). This expression pattern suggests that Pax-6 might play a role in the process of regeneration.

In an effort to uncover the roles of complement components and Pax6 during regeneration I have addressed the following aims.

Aim 1: Determine a potential role for complement components in regeneration. In order to study the role of complement components in regeneration, we performed immunohistochemistry to verify the presence of C3 and C5 during lens regeneration. Functional studies were also performed by inhibiting these molecules using an in vitro assay.

Aim 2: Study the cellular mechanisms that are regulated by Pax-6. Expression studies of Pax-6 protein were performed during various stages of regeneration. These studies were supplemented by inhibiting Pax-6 using Valproic acid and morpholinos. Assays were also performed to assess if Pax-6 is correlated with proliferation, apoptosis or differentiation.

1.4 Overview of Regeneration in the Newt Lens Regeneration Lens regeneration in the adult newt occurs via the transdifferentiation of the Pigmented Epithelial Cells (PEC) of the dorsal iris. The lens vesicle never regenerates from the PEC of the ventral iris. Once the lens has been surgically removed (the procedure is referred to as a lentectomy), the PEC of the dorsal iris proliferate, dedifferentiate and give rise to a lens vesicle by day 10 post lentectomy (Figure 3C). The cells of the lens vesicle start to differentiate, and, by day 15 post-lentectomy, the lens vesicle differentiates into lens epithelium cells and lens fiber cells (Figure 3D). The lens vesicle continues to grow and regeneration is considered complete by day 25-30. After this point the regenerated lens grows to reach the size of the original lens.

Figure 3: Lens regeneration in the adult newt. (A) Illustration show a cross section of an intact eye showing a lens (l) and the cornea (c). (B) The dorsal iris (di) and ventral iris (vi) are left intact on day 0 post-lentectomy. (C) By day 10 post lentectomy, the dorsal iris undergoes transdifferentiation to give rise to a lens vesicle (lv). (D) By day 15 the lens vesicle starts to differentiate into a lens epithelium (le) and lens fibers (lf). (E) The lens vesicle continues to differentiate and the lens fiber and lens epithelium becomes more evident by day 20. (F) Lens continues to grow and is considered complete between day 25- 30.

8 Retina Regeneration

The retina regenerates due to the transdifferentiation of the Retina Pigmented Epithelium (RPE). The RPE is the pigmented epithelium located behind the retina and serves primarily to nourish the retina and to reflect light onto the photoreceptors of the retina. To facilitate the surgical removal of the retina (retinectomy), the lens also has to be removed and therefore at the end of the retinectomy only the RPE is left behind (Figure 4B). Approximately 10 days after the surgical removal of the retina (retinectomy) the RPE begins to proliferate and depigment (Figure 4C). Within the next 10 days the dedifferentiated cells form a neuroepithelium (Figure 4D). By day 30 post- retinectomy, the neuroepithelium becomes more organized and the cells begin to differentiate into the various cells of the retina (Figure 4E). Around 40 days post- retinectomy, the retina is laminar and differentiated and shortly thereafter the retina is completely restored (Figure 4F).

Figure 4: Retina regeneration in the adult newt. (A) Illustration shows a cross section of an intact eye showing a lens (l) and the retina (c). (B) The RPE is left intact on day 0 post-retinectomy and the retina and lens have been removed. (C) By day 10 post lentectomy, the RPE dedifferentiates and proliferates. (D) By day 20 a neuroepithelium (ne) is evident. (E) Around day 30 the neuroepithelium starts to become laminar and differentiate. (F)The retina continues to differentiate until it is fully regenerated around day 40.

10 Limb Regeneration

Figure 5: Limb regeneration in the newt. Limb regeneration in the newt occurs within 5 weeks after amputation. First the epithelium around the wound closes to form the wound epithelium (w.e.) and the cells underneath the wound epithelium dedifferentiate within the first week to form the blastema (bl). Cells of the blastema proliferate (week 2) and by around 3 weeks a palette shaped structure forms. Within the next week the palette starts to differentiate to give rise to digits and by week 5 post-amputation all the cell types in the limb have been regenerated.

Following an amputation of a limb, the epithelium around the wound closes in on itself to form the Wound Epithelium (WE) (Figure 5). Within a week after the amputation, muscles, bones, connective tissue and other cells adjacent to the wound epithelium begin to undergo dedifferentiation and form a mass of undifferentiated cells called the blastema. The blastema continues to grow and by week 3 begins to flatten out

11 into palette. The cells within the blastema begin to differentiate and form the various layers of tissue that make up the limb. By week 4 post-amputation digits of the limb are visible and by the 5th week post-amputation all the structures have been replaced. The limb will continue to grow until it reaches its original size.

Tail and Spinal cord Regeneration

Figure 6. Tail Regeneration in the newt. The column on the left depicts the external appearance of the intact and regenerating tail while the column on the right depicts cross sections of the tail that correspond to the stages in the left hand column. During the first week after amputation, a wound epithelium is formed around the site of amputation. The cells beneath the wound epithelium including neurons and glia of the spinal cord, cartilage of the vertebrae and muscles dedifferentiate to form the blastema. Over the course of 5 weeks the blastema grows and differentiates to replace all the different cell types of the tail.

Tail regeneration represents a unique regeneration system since not only muscle, cartilage, bone and connective tissue can be replaced via transdifferentiation, but spinal cord can also regenerate via activation of ependymal cells (a reservoir of stem-like cells that line the spinal cord). It has been reported that ependymal cells of the spinal cord can also transdifferentiate into muscle (Echeverii and Tanaka, 2002). However, externally, tail regeneration resembles the process of limb regeneration. A wound epithelium is formed immediately after the amputation and the cells underneath the wound epithelium dedifferentiate to form a blastema (Figure 6). The cells from the muscles, vertebrae, and

12 spinal cord contribute to the formation of the blastema. Over the first 3 weeks the blastema grows and by 5 weeks post-amputation the process of regeneration is complete.

1.5 Retina Regeneration in the Embryonic Chick

The ability to regenerate the retina after injury is limited to a handful of organisms. In addition to urodeles, birds and fish have been reported to be able to regenerate their retinas (Del Rio Tsonis and Tsonis, 2003) and recently a limited amount of regeneration has been observed in rodents (Ooto et al., 2004). Among these animals the embryonic chick presents itself as a unique model system. The embryonic chick has been extensively used to study many developmental processes and therefore a number of techniques and tools exist that can be used to determine molecular pathways involved in retina regeneration.

Retina regeneration in the embryonic chick was first reported in 1965 by Coulombre and Coulombre (1965). This ability is restricted to a window of time between Embryonic day 3.5 (E3.5) and E4. Regeneration takes place via transdifferentiation of the Retinal Pigment Epithelium (RPE) in the posterior region of the eye. The pigmented cells of the RPE undergo dedifferentiation and loose their pigment. These cells proliferate and give rise to a neuroepithelium that will later differentiate to form all the different layers of the retina. However, the layers of the newly formed retina had reversed polarity and the ganglion cells, which are normally closest to the lens were now placed farthest away from the lens. Furthermore, the transdifferentiated RPE does not replace itself and therefore there is no RPE to nourish this new retina which eventually degenerates after E15. Coulombre and Columbre (1965) also observed a second layer of retina that regenerated from the anterior margin of the eye. The origin of these cells that gave rise to this regenerated retina remained unknown. Also, in these studies, the authors could induce regeneration only when a piece of retina was left in the eye (Coulombre and Coulombre, 1965). At that time the identity of the molecules (produced by the retina) that induced regeneration was unknown.

13 In the late 1980s and early 1990s, Park and Hollenberg performed experiments that identified the factors that could induce retina regeneration in the embryonic chick. In their experiments, Park and Hollenberg used an inert polymer called Elvax to deliver doses of various growth factors into the retinectomized eye. The only factors that could induce regeneration were Fibroblast Growth Factor (FGF)-1 and FGF-2. They also tried using other growth factors such as Insulin, Insulin like Growth Factor and Transforming Growth Factor β, but none of these were able to induce regeneration. FGF-2 was the most efficient at inducing transdifferentiation of the RPE. FGF-1 used at a very high concentration would not only induce transdifferentiation but also induced the regeneration of the retina from the anterior margin of the eye (Park and Hollenberg, 1989 and 1991).

There have been no studies since Park and Hollenberg’s experiments in 1991 that explore retina regeneration in the embryonic chick. As a result the molecular mechanisms that underlie this FGF induced retina regeneration remains unknown. Furthermore, all these studies were based purely on histological analysis of regenerating retina, and it is not clear if the regenerated retina has all the cell types of a normal retina. My research addresses some of these issues. In addition, I evaluated the effects of transcription factors, Pax-6 and Microphthalmia (Mitf) on the processes of regeneration and transdifferentiation.

Pax-6 is a transcription factor that contains paired as well as homeobox domains. It has been implicated in the normal development of various organs including the brain, pancreas and the eye (For review Ashery-Padan and Gruss, 2001). Its role in the eye is especially important and it is considered one of the “master regulators” of eye development since ectopic expression of Pax-6 homologs in Drosophila and Xenopus yield ectopic eyes (Chow et al., 1999). Though such drastic phenotypes are not seen in mammalian systems, hypomorphic Pax-6 mutations in mice results in a small eye phenotype and in humans can cause aniridia (Ton et al., 1991). The role of Pax-6 has therefore been extensively studied and functions of Pax-6 in eye development have been elucidated. During early development, Pax-6 is expressed throughout the optic cup, both

14 in the neuroepithelium (which will give rise to the retina) as well as the presumptive RPE (Grindley et al., 1995). As development progresses the expression of Pax-6 is limited to the developing retina and lens (Belecky-Adams et al., 1997). Its role in the development of the lens is discussed later. In the retina, Pax-6 is required to maintain the multipotent state of retinal progenitors in the neuroepithelium. Pax-6 is also responsible for the induction of various basic Helix loop Helix transcription factors that result in the differentiation of various retinal cell types. The only retinal cell type that is independent of Pax-6 during this phase is the amacrine cell (Marquardt et al, 2001). Given the functions of Pax-6 during eye development, it is reasonable to assume that this molecule is a good candidate to be upregulated during retina regeneration in the embryonic chick. We hypothesized that this upregulation maybe a direct or indirect consequence of the FGF signal that induces regeneration. This is supported by experiments performed in chick neural tubes which show that Pax-6 can be regulated by members of the FGF family (Bertrand et al., 2000)

In addition to upregulating pro-retinal genes like Pax-6, FGF-2 might also be necessary to downregulate genes that specify the RPE fate. One such candidate gene is Microphthalmia (Mitf). Mitf is expressed in the developing RPE and is required for the proper specification and maintenance of its phenotype (Mochii et al., 1998). It has also been shown in vitro that the inhibition of Mitf in cultured RPE, results in the transdifferentiation of the RPE into retina (Mochii et al., 1998, Iwakiri et al., 2005). Moreover, Pax-6 and Mitf can physically interact with each other and cause the downregulation of each other and their targets (Planque et al. 2001). These data suggest that during transdifferentiation of RPE into retina, Mitf must be downregulated to allow for the RPE to dedifferentiate and then Pax-6 upregulation allows the dedifferentiated cells to differentiate into retinal cells.

In order to address these relevant issues the following aims were proposed.

Aim 1: To establish the embryonic chick as a model to study retina regeneration.

15 A number of experiments have been done to characterize the spatial and temporal pattern of retina regeneration. These experiments included the use of histology as well as immunohistochemical markers to trace the emergence of various cell types during regeneration.

Aim 2: To understand the FGF-2 signal that induces regeneration. After establishing the embryonic chick eye as a model to study retina regeneration, we explored the nature of the FGF-2 signal by immunohistochemically identifying FGF Receptors (FGFR) that may be involved in transducing the signal to transdifferentiate. We also assessed if this signal is processed through a Mitogen Activated Protein (MAP) Kinase pathway. These experiments helped to establish the molecular nature of the FGF signal.

Aim 3: To study the effect of Pax-6 and Microphthalmia on retina regeneration. Since Pax-6 and Mitf are crucial to the normal development of the retina and RPE we hypothesize that their expression will have to be modulated to facilitate the transdifferentiation of the RPE. Both of these factors have been analyzed during the retina regeneration process.

1.6 Organization of this dissertation

The following chapters in this dissertation contained detailed explanations of the experiments that were performed in order to address the aims of this research. Chapter 2 will discuss the role of complement components in the regeneration process, Chapter 3 will address the role of Pax-6 in lens regeneration and Chapter 4 will focus on retina regeneration in the embryonic chick. Chapter 5 will be a general discussion of the results from all the experiments in the preceding chapters and their implications for the field of regeneration. Chapter 5 will also discuss future directions for this research.

16 1.7 References

Ashery-Padan, R. and Gruss, P. (2001). Pax6 lights-up the way for eye development. Curr Opin Cell Biol. 13(6):706-14.

Belecky-Adams, T., Tomarev, S., Li, H. S., Ploder, L., McInnes, R. R., Sundin. O. and Adler, R. (1997). Pax-6, Prox 1, and Chx-10 homeobox gene expression correlates with phenotypic fate of retinal precursor cells. Invest. Ophthalmol. Vis. Sci. 38, 1293-1303.

Bertrand, N., Medevielle, F.,and Pituello, F (2000). FGF signaling controls the timing of Pax6 activation in the neural tube. Development 127(22):4837-43.

Chow, R.L., Altmann, C.R., Lang, R.A. and Hemmati-Brivanlou, A.(1999). Pax6 induces ectopic eyes in a vertebrate. Development. 126(19):4213-2

Chow, R.L. and Lang, R.A. ( 2001). Early eye development in vertebrates. Annu. Re. Cell Dev. Biol. 17:255-96

Coulombre, J.L. and Coulombre, A.J. (1965). Regeneration of neural retina from the pigmented epithelium in the chick embryo. Dev Biol. 12(1):79-92.

Cvekl, A. and Piatigorsky, J. (1996). Lens development and crystalline gene expression: many roles for Pax-6. Bioassays 18:621-30

Del Rio-Tsonis, K., Washabaugh, C.H. and Tsonis, P.A. (1995). Expression of Pax-6 during urodele eye development and lens regeneration. Proc. Natl. Acad. Sci. USA 23;92(11):5092-6.

Del Rio-Tsonis, K., Jung, J.C., Chiu, I.M. and Tsonis, P.A. (1997). Conservation of fibroblast growth factor function in lens regeneration. Proc. Natl. Acad. Sci. USA 94(25):13701-6.

Del Rio-Tsonis, K., Trombley, M.T., McMahon, G., Tsonis, P.A. (1998a). Regulation of lens regeneration by Fibroblast Grow Factor Receptor 1. Developmental Dynamics 213:140-146

Del Rio-Tsonis, K., Tsonis, P.A., Zarkadis, I.K., Tsagas, A.G., Lambris, J.D.(1998b). Expression of the third component of complement, C3, in regenerating limb blastema cells of urodeles. J Immunol. 161(12):6819-24.

Del Rio-Tsonis, K., Tomarev, S.I. and Tsonis, P.A. (1999). Regulation of Prox 1 during lens regeneration. Invest. Ophthalmol. Vis. Sci. 40(9):2039-45.

Del Rio-Tsonis, K. and Tsonis, P.A. (2003). Eye regeneration at the molecular age. Developmental dynamics 226:211-224

17 Echeverri, K. and Tanaka, E.M. (2002). Ectoderm to mesoderm lineage switching during axolotl tail regeneration. Science. 298(5600): 1993-6.

Eguchi, G., Abe, I. and Watanabe, K., (1974). Differentiation of lens-like structures from Newt iris epithelial cells in vitro. Proc. Natl. Acad. Sci. U.S.A.:71(12):5052-6.

Farkas, I., L. Baranyi, M. Takahashi, A. Fukuda, Z. Liposits, T. Yamamoto, and H. Okada. (1998). A neuronal C5a receptor and an associated apoptotic signal transduction pathway. J.Physiol. 507:679

Friedman, D.S. (2002) Vision Problems in the U.S. :Prevalence of Adult Vision Impairment and Age-Related Eye Disease in America, Prevent Blindness America.

Fuhrmann, S., Levine, E. M. and Reh T. A. (2000). Extraocular mesenchyme patterns the optic vesicle during early eye development in the embryonic chick. Development 127, 4599-4609.

Grindley, J.C., Davidson, D.R and Hill, R.E. (1995). The role of Pax-6 in eye and nasal development. Development 121: 1433-1442.

Hautanen A and Keski-Oja J (1983) Interaction of fibronectin with complement component C3. Scand J Immunol.17(3):225-30

Hogan BL, Horsburgh G, Cohen J, Hetherington CM, Fisher G, Lyon MF. (1986). Small eyes (Sey): a homozygous lethal mutation on chromosome 2 which affects the differentiation of both lens and nasal placodes in the mouse. J Embryol Exp Morphol. 97:95-110

Iwakiri, R., Kobayashi, K., Okinami, S. and Kobayashi, H. (2005). Suppression of Mitf by small interfering RNA induces dedifferentiation of chick embryonic retinal pigment epithelium. Exp, Eye Res. 81(1):15-21.

Kimura, Y., M. Madhavan, M.K. Call, W. Santiago, P.A. Tsonis, J.D. Lambris, and K. Del Rio-Tsonis . 2003. Expression of complement 3 and complement 5 in newt limb and lens regeneration. J. Immunol. :170(5):2331-9.

Kiss I, Deak F, Holloway RG Jr, Delius H, Mebust KA, Frimberger E, Argraves WS, Tsonis PA, Winterbottom N, Goetinck PF..(1989) Structure of the gene for cartilage matrix protein, a modular protein of the extracellular matrix. Exon/intron organization, unusual splice sites, and relation to alpha chains of beta 2 integrins, von Willebrand factor, complement factors B and C2, and epidermal growth factor. J Biol Chem 264(14):8126-34.

Lang, R.A.(2004). Pathways regulating lens induction in the mouse. Int. J. Dev. Biol.: 48(8-9):783-91

18 Leivo, I. and Engvall, E (1986) C3d fragment of complement interacts with laminin and binds to basement membranes of glomerulus and trophoblast. J Cell Biol. 103(3):1091- 100.

Marquardt T, Ashery-Padan R, Andrejewski N, Scardigli R, Guillemot F, Gruss P. (2001). Pax6 is required for the multipotent state of retinal progenitor cells. Cell 105(1):43-55.

Mastellos, D., Papadimitriou, J. C., Franchini, S., Tsonis, P. A. and. Lambris, J. D (2001). A novel role of complement: mice deficient in the fifth component of complement (C5) exhibit impaired liver regeneration. J.Immunol. 166:2479.

Mizuno N, Mochii M, Yamamoto TS, Takahashi TC, Eguchi G, Okada TS(.1999). Pax-6 and Prox-1 expression during lens regeneratin from Cynps iris and Xenopus cornea: evidence for a genetic program common to embryonic lens development. Differentiation 65:141-149

Mochii M, Mazaki Y, Mizuno N, Hayashi H, Eguchi G. (1998). Role of Mitf in Differentiation and Transdifferentiation of Chicken Pigmented Epithelial Cell. Developmental Biology 193:47-62.

O'Barr, S. A., J. Caguioa, D. Gruol, G. Perkins, J. A. Ember, T. Hugli , and N. R. Cooper. 2001. Neuronal expression of a functional receptor for the C5a complement activation fragment. J.Immunol. 166:4154.

Ooto, S., Akagi, T., Kageyama, R., Akita, J., Mandai, M., Honda, Y. and Takahashi, M.(2004) Potential for neural regeneration after neurotoxic injury in the adult mammalian retina. Proc. Natl. Acad. Sci. U S A. 14;101(37):13654-9.

Park, C.M. and Hollenberg, M.J. (1989) Basic fibroblast growth factor induces retinal regeneration in vivo. Dev Biol. :134(1):201-5.

Park, C.M. and Hollenberg, M.J. (1991). Induction of retinal regeneration in vivo by growth factors. Dev Biol. :148(1):322-33.

Planque N, Leconte L, Coquelle FM, Martin P, Saule S (2001). Specific Pax- 6/microphthalmia transcription factor interactions involve their DNA-binding domains and inhibit transcriptional properties of both proteins. J Biol Chem. 3;276(31):29330-7

Pittack, C., Grundwald, G,B. and Reh, T.A. (1997). Fibroblast growth factors are necessary for neural retina but not pigmented epithelium differentiation in chick embryos. Development.:124(4):805-16.

Servis, C. and Lambris, J.D. (1989). C3 synthetic peptides support growth of human CR2-positive lymphoblastoid B cells. J Immunol. 142(7):2207-12

19 Thitoff, A.R., Call, M.K., Del Rio-Tsonis, K. and Tsonis, P.A..(2003). Unique expression patterns of the retinoblastoma (Rb) gene in intact and lens regeneration-undergoing newt eyes. Anat Rec. 271A(1):185-8.

Ton, C.C., Hirvonen, H., Miwa, H., Weil, M.M., Monaghan, P., Jordan, T., van Heyningen, V., Hastie, N.D., Meijers-Heijboer, H. and Drechsler, M. (1991) Positional cloning and characteization of a paired box and homeobox containing gene form the aniridia region. Cell 67:1059-1074

Tsonis, P.A., Madhavan, M., Call, M.K., Gainer, S., Rice, A.and Del Rio-Tsonis, K. (2004a) Effects of a CDK inhibitor on lens regeneration. Wound Repair Regen. 2004 Jan- Feb;12(1):24-9.

Tsonis, P.A., Vergara, M.N., Spence, J.R., Madhavan, M., Kramer, E.L., Call, M.K., Santiago, W.G., Vallance, J.E., Robins, D.J. and Del Rio-Tsonis, K. ( 2004b). A novel role of the hedgehog pathway in lens regeneration. Developmental Biology 15;267(2):450-61.

Tsonis, P.A., Madhavan, M., Tancous, E.E., Del Rio-Tsonis, K. (2004c). A newt's eye view of lens regeneration. Int. J. Dev. Biol. 48(8-9):975-80.

Chapter 2

Involvement of Complement Molecules in Regeneration This chapter has been formatted for the Journal of Immunology and part of this work was published in Kimura et al. (2003) Journal of Immunology :170(5):2331-9

2.1 Introduction

Complement Components and Regeneration The complement system is comprised of several serum proteins, membrane-bound receptors, and regulatory proteins that constitute a phylogenetically ancient mechanism of innate immunity (Sunyer et al., 1998; Lambris et. al., 1999). The functions of the complement system in host defense and inflammation are mediated mainly through the sequential activation and proteolytic cleavage of serum proteins. Complement activation occurs through three distinct pathways (classical, alternative, and lectin), all of which converge in the activation of C3, the third component of complement. C3 can interact with a wide spectrum of factors and therefore is a e to mediate a wide variety of functions (Lambris et al., 1998; Lambris, 1998; Mastellos and Lambris, 2002). In addition to interacting with complement proteins, C3 fragments interact with several proteins that are involved in differentiation, such as fibronectin and integrins (Hautanen, and Keski-Oja. 1983; Leivo and Engvall. 1986). Other complement components share homologies with domains of extracellular matrix proteins, such as collagen binding domains, which might indicate that complement components could be involved in interactions with the extracellular matrix (Kiss et al., 1989). This suggests that complement components may be able to perform non-immunological functions. For example, C3 is expressed in myoblasts (Legoedec et al., 1995; Legoedec et al., 1997) and is also associated with proliferation and growth of B cells in vitro (Servis and Lambris, 1989). C5, the fifth component of complement, has also been found to have novel noninflammatory functions in various tissues. Studies in a human neuroblastoma cell line have suggested that C5a (a fragment of C5) participates in apoptotic signal

21 transduction pathways through its binding to the neuronal C5aR (Farkas, 1998a and Farka, 1998b). It has also been shown that the terminal complex system C5b-9 (membrane attack complex), in sublytic doses, can induce DNA synthesis and cell proliferation in cultured mouse fibroblasts (Halperin et al., 1993), human aortic smooth muscle cells (Niculescu et al., 1999), oligodendrocytes (Rus et al., 1997), and glomerular epithelial cells in the absence of other growth factors (Shankland et al., 1999). Also, sublytic concentrations can activate monocytes and induce cytokine release through activation of NF-κB signaling pathways, which are critical for the cell cycle transition into DNA synthesis (Kilgore et al., 1997).

Work from our laboratory has previously shown that C3 is expressed in the blastema during limb regeneration in the axolotl (Del Rio-Tsonis et al., 1998). The blastema is a cell population that gives rise to the various cell types that reconstitute the limb, through processes that involve dedifferentiation, transdifferentiation into different phenotypes, and extensive tissue remodeling. C3 was not expressed in the intact or developing limb, indicating its specificity for the regeneration process. Furthermore, this study also found expression of C3 in dedifferentiated newt muscle cells in vitro. These findings were the first to indicate that complement might have a novel, possibly nonimmunologic, role in regenerative processes. This potential role for complement in tissue regeneration was supported by the recent observation that C5-deficient mice exhibit defective liver regeneration after acute toxic injury (Mastellos et al., 2001). In mice, the liver can regenerate by proliferation of the remaining hepatic cells. Mice lacking C5 are not able to repair their liver properly unless they were provided with C5a (Mastellos et al., 2001).

To expand our understanding of the role of complement components in regenerative processes, we decided to use the newt Notophthalmus viridescens because of its extensive repertoire of regenerative capabilities. In addition to its ability to regenerate eyes tissues, the newt can also regenerate its limbs and tail. The association of complement components with regeneration of very diverse tissues (for example liver and limb) suggests that they may not be involved in programs that rebuild these tissues but

22 perhaps play a role early in the induction of regeneration. We therefore hypothesized that if complement components did indeed play a role in the initial stages of induction of regeneration, then complement components should be expressed in various types of regenerative processes. We studied the expression of complement components C3 and C5 during regenerative processes in different systems. Our findings show that at least one of the complement components, C3 or C5, is associated with cells undergoing transdifferentiation during lens, retina, spinal cord, tail and limb. We also used an in vitro system to show that C3 and C5 may play a role in cell signaling during regeneration.

2.2 Methodology

Animals and Surgeries: Newts (N. viridescens) were purchased from Mike Tolley Newt Farm (Nashville, TN). Before any surgical procedures or animal sacrifice, all animals were anesthetized with 3- aminobenzoic acid hydrochloride (MS222; Sigma-Aldrich, St. Louis, MO) and were handled in compliance with the regulations of the Institutional Animal Care and Use Committee. Four to six animals/group were used for all experiments. Lentectomy: A slit was made along the length of the cornea using a scalpel. The lens was then removed by applying gentle pressure around the cornea using a pair of forceps. The animals were allowed to recover and collected at various time points. Limb amputation: All amputations for immunohistochemical analysis were distal amputations. The newt forelimbs were amputated along the radial-ulnar plane above the wrist using a scalpel (distal amputation). Bone and any other protruding tissues were trimmed back using a pair of surgical scissors. Animals were allowed to recover and tissues were collected 1 and 3 weeks later. For confrontation assays, proximal amputations were made just above the elbow of the forelimb. Tail Amputation: The lower one third of the tail was amputated using a scalpel and any protruding tissue was trimmed away using scissors.

23 Retinectomy: A lentectomy was first performed and the inside of the eye was then rinsed with Ringer’s solution using a Pasteur pipette. The retina was then removed using the same pipette.

Tissue Processing and Immunohistochemistry Animals were anesthetized and then sacrificed. Tissues used for lens, limb and tail regeneration were collected and immediately immersed in OCT (Fisher Scientific) and quick frozen in an ethanol: dry ice bath. Tissues used for retina regeneration were first fixed in 4% formaldehyde before being embedded in OCT. Frozen samples were sectioned and processed for immunohistochemistry as follows: tissues were blocked using 10% goat serum in 1X PBS for 1 hour followed by incubation with primary antibody solution for 1 hour at room temperature. Neurofilament -200 (Sigma) C3 and C5 antibodies (gifts from Dr. Lambris, University of Pennsylvania) were diluted 1:100 in 1X PBST containing 10% goat serum. The slides were then washed three times in PBST and were incubated in secondary antibody for 2 hours at 370C. Goat anti-Rabbit IgG conjugated to either Alexa Fluor 546 (Invitrogen) or rhodamine (Santa Cruz Biotechnology, Santa Cruz, CA) was used to detect the rabbit polyclonal antibodies against C3 and C5. Goat anti-mouse Alexa Fluor 488 (Invitrogen) diluted 1:100 was used to detected Neurofilament. Secondary antibodies were diluted 1:100 in PBST containing 10% goat serum. Slides were washed extensively in PBS and coverslipped. Vectashield (Vectorlabs, Burlingame, CA) was used as mounting medium in order to preserve the fluorescence of the samples.

Confrontation Assay The forelimbs of animals were amputated in order to collect blastemas for the confrontation assay. One amputation was made at the distal position, immediately above the wrist whereas the other forelimb was amputated more proximally just above the elbow. The animals were allowed to recover and blastemas were collected between 18 and 20 days post amputation. The animals were anesthetized and the blastemas along with the wound epithelia were removed and placed in Calcium/Magnesium Free Hank’s Solution. The wound epithelium was dissected away and the cells of the blastemas were

24 labeled by placing them in cell permeable dyes (dissolved in Serum Free Dulbecco’s Modified Eagle Medium (DMEM)) for 20 minutes. 10µM Cell Tracker dye antigen ((5) and (6) carboxy-2’,7’ dichlorofluorescien diacetate succinimdyl ester form Invitrogen) was used to label the proximal blastema. PKH26 labeling dye (Sigma) was used to mark the distal blastemas. The blastemas were then juxtaposed against each other at the basal 0 surface and cultured in confrontation media for 4 days at 25 C and 5% CO2. The confrontation medium contained DMEM with 10% Fetal Calf serum, antibiotics and 0.1% Agarose 1000. The medium was made fresh and antibodies against C3 or C5 were diluted 1:100 into the media before the agarose gelled.

2.3 Results

Expression of Complement Components during Lens Regeneration C3 was not seen in the regenerating lens vesicle at any stage of lens regeneration (Figure 1, D–I). However, it was present in the stroma and pigmented epithelium of the iris and in the cornea. In contrast, C5 was present in the regenerating vesicle at 10 (Figure 1, M–O), and 20 days (Figure 1, P–R) post-lentectomy. C5 was also present in the cornea at all stages of regeneration (not shown). C3 and C5 were absent from the intact lens (Figure 1, A–C for C3 and J–L for C5); however, they were present in the corneal tissue (not shown). No reactivity was detected in the retina at any stage with either of the Abs (not shown). In situ hybridization experiments confirmed that C3 and C5 were being transcribed in the regenerating lens vesicle and cornea (Kimura et al., 2003).

25 FIGURE 1. Protein expression of C3 and C5 during lens regeneration. The first column contains differential interference contrast (DIC) images (magnification, X20 in A–L and X40 in M–U) of the sections in columns two and three. Column two contains fluorescence images, while the pictures in column three are overlays of the DIC and fluorescence images in columns one and two. The intact lens shows no expression of C3 protein (A–C). C3 (red) is expressed in the stroma and pigmented cells of dorsal iris at days 10 (D–F) and 20 (G–I) post-lentectomy. Sections have been counterstained with DAPI (blue) to visualize nuclei. C5 protein expression (red) is shown in J–R. There is no C5 expression in the lens of an intact eye (J– K). C5 expression is seen in the dorsal iris and lens vesicle at days 10 (M–O) and 20 (P–R) post-lentectomy. There was no fluorescent signal when preimmune serum was used instead of either anti-C3 or anti- C5 Ab (S-U). c, Cornea; di, dorsal iris; l, lens; lv, lens vesicle; and vi, ventral iris. Arrowheads indicate areas of C3 or C5 protein expression.

26 Figure 2: Protein expression of C3 and C5 during retina regeneration. The left column (A, C, E, G, I and K) contain DIC images while the right column (B, D, F, H, J and L) contains the corresponding fluorescence images of C3 immunoreactivity (red) and DAPI (blue). C3 is expressed in the transdifferentiating retina as early as day 10 post retinectomy (A and B) and continues to be expressed at day 20 (C and D). As the regenerated retina becomes more differentiated the expression of C3 protein is decreased and by day 41 post-retinectomy there is little C3 protein in the retina (E and F). Similar expression of C5 was also observed. C5 is expressed strongly in transdifferentiating tissues between day 10 (G and H) and day 20 (I and J) post- retinectomy. The expression of C5 is diminished by day 41 (K and L). Arrowheads indicate areas of C3 or C5 protein expression.

Expression of Complement Components during Retina Regeneration

Both C3 and C5 were observed in regenerating retina. As early as day 10 post- retinectomy both C3 and C5 were present in depigmented RPE cell that are undergoing transdifferentiation (Figure 2A-B for C3 and G-H for C5). This trend continues and at day 20 both C3 and C5 continue to be expressed in transdifferentiating tissues (Figure 2 C-D for C3 and I-J for C5). As the regenerated retina becomes more differentiated and around day 41 post-retinectomy, the expression of C3 (Figure 2E-F) and C5 (Figure 2K- L) decreases.

Expression of Complement Components during Limb Regeneration To analyze the expression patterns of C3 and C5 in limb regeneration, we examined sections taken from blastemas 1 or 3 weeks after amputation. These stages represent an early and a late blastema. Immunohistochemical staining of regenerating limbs using the newt-specific anti-C3 and anti-C5 antibodies indicated that C3 and C5 had non-overlapping expression patterns. Even though C3 was present in the WE, most of the protein was found in the blastema, as was evident in the early regeneration stage, the 1-week blastema (Figure 3C and D). This pattern continued in later stages of limb regeneration (Figure 3E and F). In contrast, C5 was present exclusively in the WE, with no expression in the blastema at any stage of regeneration (Figure 3I–L). Neither protein was found in the intact limb (Figure 7A and B for C3, and G and H for C5). In addition, in situ hybridization for C3 and C5 demonstrated that these molecules are being made locally, which suggests that systemic activation of complement is not a prerequisite for its involvement in limb regeneration (Kimura et al., 2003).

28 FIGURE 3. Expression of newt C3 and C5 proteins during limb regeneration. Expression of C3 (A–F) and C5 (G–N) proteins during limb regeneration. Column one contains DIC images (magnification, x20); column two shows the corresponding fluorescence images. Intact muscle in the limb (A and B) shows no C3 expression. C3 (red) is predominantly expressed in the blastema as early as 1 wk (C and D) post-amputation and is maintained even at 3 wk post-amputation (E and F). C5 is not expressed in intact muscle (G and H) but is expressed in the WE of the regenerating limb at 1 wk (I and J) and 3 wk (K and L) post-amputation. Using preimmune serum instead of anti-C3 or anti-C5 Ab yields no fluorescent signal (M and N). Sections have been counterstained with DAPI (blue). bl, Blastema; m, muscle; and we, wound epithelium. Arrowheads indicate areas of C3 or C5 protein expression

29 Expression of Complement Components during Tail Regeneration C5 expression was not detected in intact or regenerating tails. However, C3 was expressed in the regenerating blastema as well as the wound epithelium as early as 1 week post-amputation (Figure 4C-E). C3 continued to be expressed through week 2 (Figure 4F-H) but the expression in the blastema diminished by week 3 (Figure 4I-K) and at this time the expression was limited only to the wound epithelium. C3 was not expressed in intact tails. When the cross sections of the same stages of tail regeneration were studied, C3 expression was seen in the spinal cord and vertebrae by week 1 post- amputation (Figure 5A and B). This expression continued through weeks 2 (Figure 5 C and D) and 3 (Figure 5 E and F) post-amputation.

Figure 4: Expression of C3 in the regenerating tail. The first column is composed of DIC images (A, C, F and I) of the section shown in the middle column (B, D, G and J). The third column (E, H and K) is a close up view of section in the middle column. C3 (red) is not expressed in intact tails (A and B), but is expressed at week 1 (C- E), and week 2 (F-H) post-amputation in the wound epithelium (we) as well as the blastema (bl). By week 3 (I-K) C3 expression is restricted to the wound epithelium. Magnification is X20

30 Figure 5: Expression of C3 in the spinal cord and vertebra during tail regeneration. C3 is expressed in tissues of the spinal cord (arrow) and vertebra as early as week 1 post amputation (A and B respectively). Expression in the spinal cord and vertebrae continue through weeks 2 (C and D) and 3 (E and F) post- amputation. Neurofilament-200 (green) was used to detect neurons in the spinal cord and all sections were counter-stained with DAPI. Magnification is X40

31 Confrontation Assays The expression of C3 and C5 in multiple regenerating systems suggests that they may have a non-immunological role in these processes. In order to ascertain if C3 or C5 had unique roles during regeneration, we used an in vitro assay that has been widely used to study signaling and cell-cell interactions during limb regeneration (da Silva et al., 2002).

Proximal and distal blastemas were juxtaposed and cultured for 5 days. At the end of 5 days the blastemas were observed. If these interactions are left alone, the proximal blastema engulfs the distal blastema (Figure 6B). In order to assess if C3 or C5 play a role in these interactions and signaling events, we disrupted their activity using antibodies against each protein. When C3 or C5 antibodies were added to the culture medium, the proximal blastema was unable to engulf the distal blastema (Figure 6C and D respectively). This result suggests that C3 and/or C5 may be involved in the signaling pathways that regulate cell-cell interactions in these blastemas or may have a role in specifying position along the proximo-distal axis. Both of these possibilities clearly indicate a novel and new non-immunological function for complement molecules in the process of regeneration.

Figure 6. Confrontation Assay. (A) Proximal blastemas (green) were juxtaposed with distal blastemas (red). (B) The cells were cultured for 5 days and by the end of this period the proximal blastema engulfs the distal blastema. (C) Addition of C3 antibody to the culture medium blocks the engulfment of the distal by the proximal blastema. (D) Similarly, treatment with C5 antibody prevented engulfment. Magnification is X4

2.4 Discussion

Previously published reports from our laboratory showed that C3, a central component of the complement system, is expressed in the regenerating blastema in axolotls. This finding suggested that complement may exert a novel, non-inmmunologic role in complex processes such as limb regeneration in urodeles. Also work from our laboratory had previously demonstrated by in situ hybridization that a newt blastema cell line expresses C3 mRNA, suggesting that these cells are actually synthesizing C3 locally (Del Rio Tsonis et al., 1998). Extending this intriguing observation to another urodele species that possesses the capacity to regenerate several of its body parts, we investigated the involvement of complement in multiple regenerating systems.

During lens regeneration in the newt, C3 and C5 protein expression patterns vary from their RNA expression. C3 or C5 was not expressed in the intact lens but C3 protein was found in the stroma and PECs of the dorsal iris, while C5 was mostly found in the regenerating lens vesicle. It is interesting to note that, even though both C3 and C5 mRNA were expressed in the regenerating lens vesicle, we detected only the C5 protein. Theses results suggest that there is a regulatory mechanism in place that controls C3 translation, or a rapid degradation or processing of the protein product. However, another possibility is that, in addition to the locally produced C3 and/or C5, circulating complement molecules move to these tissues during the process of regeneration Increased levels of C3 and C5 is not limited to lens regeneration and we have observed increases in complement components during retina regeneration as well.

The results of the present study indicate that C3 protein is present throughout the process of limb regeneration in the newt N. viridescens and is found mainly in the blastema, with some staining of the WE as well, matching the pattern that we have reported previously for the axolotl, but with a more distinct pattern in the blastema. In addition, we observed that C5 is strongly expressed in the WE at all stages of regeneration. Similarly, we found that C3 was the complement component expressed in

33 the blastema of regenerating tails and was expressed by cells in the spinal cord and vertebra. It is possible that these complement molecules play a role in the proliferation and establishment of the blastema (limb and tail), neuroepithelium (retina) and lens vesicle (lens). It is known that complement components play a role in mitogenic events, as has been reported for C3 during the proliferation of a macrophage-monocyte progenitor lineage (Mastellos et al., 2001) and during the proliferation and growth of B cells in vitro (Legoedec et al., 1995). Several studies have suggested that the terminal complexes (C5b-9) can exert a mitogenic effect on various cell types, including mouse fibroblasts (Halperin et al., 1993), human aortic smooth muscle cells (Niculescu et al., 1999), and glomerular epithelial cells (Shankland et al., 1999), when administered in sublytic doses (Kilgore et al., 1997). Moreover, the anaphylatoxic peptide C5a has been shown to be mitogenic for undifferentiated human neuroblastoma cells (O’Barr et al., 2001). It is also possible that these molecules play a role in early dedifferentiation processes, or even in later processes such as differentiation. Reca et al. (2003) have found expression of several complement receptors and complement components in normal human early stem/progenitor cells as well as in lineage-differentiated hematopoietic cells. This finding is especially interesting because cells that surround the vertebral ventricle express C3 during regeneration and in mammals, ependymal cells that line ventricles of the central nervous system are thought to behave like stem/progenitor cells.

Although our studies point to a possible role of complement components in regenerative processes, their actual role in a pure developmental sense, such as pattern formation, remains unclear. Our results from the confrontation assays are a preliminary step to understanding the role of these complement molecules. Abrogated engulfment of the distal blastema by the proximal blastema in the presence of C3 or C5 antibodies suggests that blocking these complement components disrupts some signaling process that is required to instruct each blastema if it is a proximal or distal identity and thus specify its position along a proximo-distal axis. In this respect, it is interesting to note that a molecule with homology to CD59, a protein that binds and inhibits membrane attack complex formation, has recently been labeled as an indicator of blastemal cell

34 identity during limb regeneration (da Silva et al., 2002), i.e., its expression depended on the position of the cells along the proximo-distal axis of the limb. Such discoveries, in conjunction with our data, further support a novel role for complement components as molecules that have important signaling function during regeneration.

Finally, it is interesting to speculate that complement components might be involved in tissue regeneration by binding to cell surface receptors and triggering signaling pathways that modulate cell-cell interactions and/or adhesion. All these could lead to proliferation, differentiation, or positional identity and therefore future experiments will focus on elucidating the exact molecular role of complements on the regenerative process.

2.5 References Da Silva, M., S.P.B. Gates and J.P. Brockes. 2002. The newt orthologue of CD59 is implicated in proximodistal identity during amphibian limb regeneration. Dev Cell 3(4):547-55.

Del Rio-Tsonis, K., P. A. Tsonis, I. K. Zarkadis, A. G. Tsangas, and J. D. Lambris. 1998. Expression of the third component of complement, C3, in regenerating limb blastema cells of urodeles. J.Immunol. 161:6819.

Farkas, I., L. Baranyi, M. Takahashi, A. Fukuda, Z. Liposits, T. Yamamoto, and H. Okada. 1998. A neuronal C5a receptor and an associated apoptotic signal transduction pathway. J.Physiol. 507:679.

Farkas, I., L. Baranyi, Z. S. Liposits, T. Yamamoto, and H. Okada. 1998. Complement C5a anaphylatoxin fragment causes apoptosis in TGW neuroblastoma cells. Neuroscience 86:903.

Halperin, J. A., A. Taratuska, and A. Nicholsonweller. 1993. Terminal Complement Complex C5b-9 stimulates mitogenesis in 3T3 Cells. J.Clin.Invest. 91:1974.

Hautanen, A. and J. Keski-Oja. 1983. Interaction of fibronectin with complement component C3. Scand J Immunol. 17:225.

Kilgore, K. S., E. Schmid, T. P. Shanley, C. M. Flory, V. Maheswari, N. L. Tramontini, H. Cohen, P. A. Ward, H. P. Friedl, and J. S. Warren. 1997. Sublytic concentrations of the membrane attack complex of complement induce endothelial interleukin-8 and monocyte chemoattractant protein-1 through nuclear factor-kappa B activation. Am.J.Pathol. 150: 2019.

Kiss, I., F. Deak, R. G. Holloway, H. Delius, K. A. Mebust, E. Frimberger, W. S. Argraves, P. A. Tsonis, N. Winterbottom, and P. F. Goetnick. 1989. Structure of the gene for cartilage matrix protein, a modular protein of the extracellular matrix. J. Biol. Chem. 264:8126.

Lambris, J. D. 1988. The multifunctional role of C3, the third component of complement. Immunology Today 9: 387.

Lambris, J. D., A. Sahu, and R. Wetsel. 1998. Chemistry and Biology of C3, C4 and C5. In The Human Complement System in Health and Diseases. 1st ed. J.E. Volanakis and M. Frank, Marcel Dekker Inc., NY, p. 83.

Lambris, J. D., K. B. M. Reid, and J. E Volanakis. 1999. The evolution, structure, biology, and pathophysiology of the complement system. Immunology Today 20: 207.

Legoedec, J., P. Gasque, J-F. Jeanne, and M. Fontaine. 1995. Expression of complement alternative pathway by human myoblasts in vitro. Biosynthesis of C3, factor B, factor H and factor I. Eur.J.Immunol. 25: 3460.

Legoedec, J., P. Gasque, J.-F. Jeanne, M. Scotte, and M. Fontaine. 1997. Complement classical pathway expression by human sceletal myoblasts in vitro. Mol.Immunol. 34:735.

Leivo, I. and E. Engvall. 1986. C3d fragment of complement interacts with laminin and binds to basement membranes of glomerulus and trophoblast. J Cell Biol. 103:1091.

Mastellos, D., and J. D. Lambris. 2002. Complement: more than a “guard” against foreign pathogens. Trends in Immunology 23 : 485-491.

Mastellos, D., J. C. Papadimitriou, S. Franchini, P. A. Tsonis, and J. D. Lambris. 2001. A novel role of complement: mice deficient in the fifth component of complement (C5) exhibit impaired liver regeneration. J.Immunol. 166:2479.

Niculescu, F., T. Badea, and H. Rus. 1999. Sublytic C5b-9 induces proliferation of human aortic smooth muscle cells: role of mitogen activated protein kinase and phosphatidylinositol 3-kinase. Atherosclerosis 142:47.

Paczkowski, N. J., A. M. Finch, J. B. Whitmore, A. J. Short, A. K. Wong, P.N. Monk, S.A. Cain, D. P. Fairlie, S. M. Taylor. 1999. Pharmacological characterizarion of antagonists of the C5a receptor. Br.J.Pharmacol. 128:1461.

Reca R, M. Majka, D. Mastellos, D. Zhao, A. Janowska-wieczorek, J. D. Lambris. 2001. New unexpected stimulatory effects of complement proteins on human hematopoiesis: evidence for crosstalk between G-protein coupled C3a complement receptor (C3aR) and alpha-chemokine SDF-1 receptor (CXCR4). Blood, 98:64a

Rus, H., F. Niculescu, T. Badea, M. L. Shin. 1997. Terminal complement complexes induce cell cycle entry in oligodendrocytes through mitogen activated protein kinase pathway. Immunopharmacology 38:177.

Servis, C. and J. D. Lambris. 1989. C3 synthetic peptides support growth of human CR2- positive lymphoblastoid B cells. J.Immunol. 142:2207.

Shankland, S. J., J. W. Pippin, and W. G. Couser. 1999. Complement (C5b-9) induces glomerular epithelial cell DNA synthesis but not proliferation in vitro. Kidney Int. 56:538.

Sunyer, J. O., I. K. Zarkadis, and J. D. Lambris. 1998. Complement diversity: a mechanism for generating immune diversity. Immunological Today 19:510 Tsonis P. A. 1999. Regeneration of the vertebrate lens and other eye structures. In Embryonic Encyclopedia of Life Sciences, Nature Publishing Group, London. www.els.net.

37 Chapter 3

The Role of Pax-6 in Lens Regeneration This chapter has been formatted for Developmental Biology and is currently under review.

3.1 Introduction The newt is one of the few adult vertebrates that can regenerate the lens after damage or removal. The event of regeneration is characterized by the process of transdifferentiation, whereby terminally differentiated pigment epithelial cells of the dorsal iris dedifferentiate, proliferate and then differentiate into lens cells (Del Rio-Tsonis and Tsonis, 2003; Tsonis and Del Rio-Tsonis, 2004; Tsonis et al., 2004a). The process of transdifferentiation is rapid with proliferation ensuing as early as 4 days after lentectomy (Eguchi and Shingai, 1971; Reyer, 1977) and regeneration being completed within 25-30 days post-lentectomy. Notably, only cells of the dorsal iris can regenerate the lens, and under normal conditions the ventral iris pigmented epithelial cells never contribute to the regenerative process. This implies that there is some inherent difference between the dorsal and the ventral iris. Therefore many studies have focused on molecules that show differential expression in the dorsal or the ventral iris during regeneration.

Despite the fact that the end product is the same, induction of newt lens regeneration differs considerably from induction of lens development where interactions between the surface ectoderm and the optic cup trigger differentiation of the lens vesicle (Fisher and Graigner, 2004; Lang, 2004). However, the expression of genes involved in embryonic lens induction and in adult newt iris alludes to the fact that similar players might be involved in lens regeneration as well. Interestingly, certain events, such as proliferation and expression of key developmental genes do take place in the ventral iris as well and this adds to the mystery related to regulatory events. Indeed, lens differentiation-controlling genes such as FGFs, Pax-6, Sox-2, MafB and Prox-1 have been found in both dorsal and ventral irises of intact as well as regenerating newt eyes (Del Rio-Tsonis et al., 1995, 1997, 1999; Mizuno et al., 1999; Hayashi et al., 2004).

The present studies were undertaken in order to answer questions pertaining to the role of Pax-6 in lens regeneration. Pax-6 is a known master regulator for eye development and mediator of ectopic lens formation during development (Halder et al., 1995; Altmann et al., 1997; Cvekl and Piatigorsky, 1996; Gehring and Ikeo, 1999; Chi and Epstein, 2002) and therefore its contribution to lens regeneration must be quite critical. Previous studies by us and others have established that the Pax-6 gene is expressed during the process of lens regeneration but a direct association with events of regeneration has never been established (Del Rio-Tsonis et al., 1995: Mizuno et al., 1999; Hayashi et al., 2004). In the present studies we have utilized an antibody to Pax-6 protein that has enabled us to co-localize the protein with certain cellular events. At the same time we have used two different treatments to knockdown expression of Pax-6 and further study the events of lens regeneration associated with its expression. We used valproic acid (VPA), a teratogen that has been previously shown to down-regulate members of the Pax family during Xenopus and chick development (Pennati et al., 2001 and Whitsel et al., 2002) and morpholino technology.

Our results show that Pax-6 is a regulator of cell proliferation and lens fiber differentiation. Proliferating cells in both dorsal and ventral iris are positive for Pax-6 and down-regulation of Pax-6 reduces proliferation and lens regeneration from the dorsal iris, but does not abolish the inductive process of dedifferentiation that eventually leads to regeneration from the dorsal iris. In addition, Pax-6 regulates the initiation of crystallin expression and thus is involved in the induction of lens fiber differentiation as well. These results indicate that while Pax-6 might not be involved in the induction (or the dedifferentiation) process of lens regeneration, it is nevertheless associated with the events of proliferation and differentiation.

39 3.2 Methodology

Animals Adult newts (Notophthalmus viridescens) were obtained from Mike Tolley Newt Farm (Nashville, TN). 0.1% 3-aminobenzoic acid (Sigma, St. Louis, MO) solution was used as an anesthetic for surgical procedures and euthanasia. Lentectomies were performed by first making a slit across the cornea using a scalpel and then applying gentle pressure to the eye to remove the lens. Animals were allowed to recover and regenerate for various periods of time and were then sacrificed to collect tissues.

BrdU Analysis Using a glass micropipette, animals were injected with 1µl of a 10mM BrdU solution (Roche, St. Louis, MI) 24 hours prior to being sacrificed. Eyes were fixed in 4% formaldehyde solution for 10 hours, washed extensively in PBS and then placed in 30% sucrose overnight. They were then embedded in OCT medium (Andwin Scientific, Warner Center, CA). Ten micron sections were used for immunohistochemical detection of BrdU. To study the distribution of BrdU and Pax-6 positive cells in the iris, sections were collected from the entire length of the iris from 4-6 eyes. The sections were washed in PBS, incubated in 2N HCl for 30 minutes and washed again in PBS. The sections were then blocked for 30 minutes using a blocking solution made of 10% Goat serum in PBS. Rat Anti-BrdU (Sigma, St. Louis, MO) and Mouse anti-Pax-6 antibodies (Developmental Studies Hybridoma Bank, Iowa City, IA) were diluted 1:100 and 1:10 respectively in blocking solution and used to detect BrdU and Pax-6. Anti-Rat FITC (Jackson ImmunoResearch Laboratories, Inc, West Grove, PA), Anti-Rat Alexa Fluor 488, Anti- mouse Alexa Fluor 405 (Molecular Probes, Eugene, OR) and anti-mouse biotin (Jackson ImmunoResearch Laboratories, Inc, West Grove, PA) and streptavidin conjugated Alexa Fluor 350 were used to detect the primary antibodies. The slides were coverslipped using Vectashield (Vector labs, Burlingame, CA) and analyzed using an Olympus BX-51 epiflourescence microscope or an Olympus Laser Confocal Microscope.

40 For analysis of proliferation in VPA-treated and untreated animals, tissues, including intestines, were processed as described above. For all analysis of BrdU incorporation in VPA-treated and control animals, 8 random sections from at least 4 separate animals were used and the number of BrdU positive cells and the total number of cells were counted. A student’s t-test was performed to assess statistical significance.

TUNEL Assay and Immunohistochemistry Tissues were fixed in 4% formaldehyde solution, cryoprotected in 30% sucrose and then embedded in OCT. Ten micron sections were used for TUNEL and immunohistochemical analysis. TUNEL was performed using the In situ cell death detection kit TMR-red (Roche, Indianapolis, IN). Sections used for the TUNEL assay were permeabilized using an ice cold solution of 1% Sodium Citrate and 0.1% Triton X- 100 for 15 minutes. For the rest of the assay we followed the manufacturer’s recommendations or as described previously (Tsonis et al., 2004b).

Sections for immunohistochemistry were first washed in PBS, followed by a 5 minute wash in 1% Saponin in PBS and then three more washes in PBS. The sections were then blocked using 10% goat serum in PBS and 0.1% Triton X-100 and incubated in primary antibody solution overnight at 40C. Monoclonal antibodies against Pax-6 were obtained from the Developmental Studies Hybridoma Bank, Iowa City, IA. α-crystallin and β-crystallin were a gift from Dr. Eguchi (Sawada et al., 1993). Anti-Mouse Alexa Fluor 488 or Anti-Mouse Alexa Fluor 546 (Molecular Probes, Eugene, OR) were used as secondary antibodies. Anti-mouse Alexa Fluor 405 (Molecular Probes, Eugene, OR) or Anti-mouse biotin (Jackson ImmunoResearch Laboratories, Inc, West Grove, PA) and streptavidin conjugated Alexa Fluor 350 (Molecular Probes, Eugene, OR) were also used to detect the anti-Pax6 antibody.

41 Valproic Acid treatment Animals treated with VPA received a daily oral dose of 180µg of VPA (6µl of a 30mg/ml VPA solution). The VPA solution was made in water and therefore controls animals were given 6µl of water instead of the VPA solution. The animals used in these studies did not show a large variation in size and were between 1.5 to 2 grams of weight. For QPCR, Pigmented Epithelial Cells (PEC) of the iris were treated in vitro for 5 days with 10mM VPA.

Morpholinos A Special Delivery Morpholino designed to inhibit the translation of Pax-6 was purchased from GeneTools, LLC (Philomath, OR, USA). The oligonucleotide sequence for the Pax-6 targeted morpholino was: TGTCTCCCTTATGTAGTCCCTCATG. A Special Delivery Standard Control morpholino was also purchased from GeneTools, LLC. Both morpholino oligonucleotides were labeled with lissamine.

For in vitro analysis, 14-day old newt PEC from the iris (cultured as described in Grogg et al., 2005) were transfected with either the Pax-6 specific morpholino or the standard control morpholino using the special delivery method as per manufacturer’s instructions. Forty-eight hours after transfection, 150 µl of 10mM BrdU was added to each plate and the transfected cells were harvested 24 hours later. The cells were fixed in 4% paraformaldehyde and 3% sucrose for 30 minutes at room temperature followed by three washes in PBS. The cells were then washed with 2N HCl for 20 minutes followed by three washes in PBS. Primary antibodies for Pax-6 and BrdU were diluted 1:10 and 1:100 respectively in PBS containing 0.3% Triton X-100 and 10% normal goat serum (Sigma-Aldrich, St. Louis, MO, USA). The cells were incubated with the primary antibodies overnight at 40C and then washed with PBS. The cells were incubated for 2 hours at room temperature with secondary antibodies diluted 1:100 in PBS with 0.3% Triton X-100 and 10% normal goat serum followed by 3 washes in PBS. Coverslips were then added to the cells using Vectashield mounting medium (Vector Labs, Burlingame, CA, USA). Twenty-four fields of view from 6 independent experiments were analyzed

42 for each treatment group and a student’s t-test was performed to determine statistical significance.

For in vivo analysis, 10 µl of the Pax-6 morpholino or the standard control morpholino was mixed with 5.6 µl of endoporter (GeneTools, LLC, Philomath, OR, USA) and 1 µl of this solution was injected into the vitreous chamber of a lentectomized adult newt. In all experiments, one eye was injected with Pax-6 morpholino while the other eye from the same animal was treated with the control morpholino. For proliferation studies, the morpholino/endoporter injections were made 10 days or 20 days post-lentectomy and 1 µl of a 10mM BrdU solution was injected into the eye 12 days post-lentectomy or 24 days post-lentectomy respectively. These animals were collected 24 hours after BrdU was administered and processed as previously described above. For histological data, the morpholino/endoporter solution was injected twice on days 4 and 10 post-lentectomy. These animals were collected 15 days-post-lentectomy and processed as previously described.

Histology and Regeneration Index Eyes that were used for histological examination were fixed in Bouin’s solution and embedded in paraffin. Ten micron sections that spanned the entire length of the iris were H&E stained and observed using an Olympus BX-51 microscope. A regeneration index value of 0 or 1 was assigned to each regenerating eye. Lens vesicles that showed a lens vesicle appropriate in size and lens fiber differentiation for that stage of regeneration were given a regeneration index of 1. If the regenerating lens vesicle was either too small or showed insufficient differentiation, the eye was given a regeneration index of 0. The averages of the individual eyes from each experimental and control group were used to represent the regeneration index of that group. The regeneration index of the VPA- treated eyes was the average from 25 eyes while the regeneration index from the control group was the average from 22 eyes. For the morpholino experiments, the regeneration index included the average from 15 eyes for both the control and Pax-6 morpholino treatments.

43 Real-Time PCR RNA was isolated from VPA-treated and untreated PEC using TRI REAGENT® (Molecular Research Center, INC.) according to manufacturer’s instructions. Seventy five micrograms of RNA was used to synthesize cDNA using iScript™ cDNA Synthesis Kit (BioRad). All Real-Time PCRs were performed using the iCycler™ (BioRad). For each Real-Time PCR reaction run in triplicate, 2 microliters of cDNA, 800 nM primers, and iQ™ SYBR® Green Supermix (BioRad) were used. Primers were designed from the previously published sequence for Pax-6, forward 5’- CTGGGCAGGTATTACGAG; reverse 5’- GTCTCTGATTTCCCAGGC. For reference we used the newt rpL27, forward 5’-TACAACCACTTGATGCCA Reverse 5’- CAGTCTTGTATCGTTCCTCA. The data were analyzed using the Pfaffl method (Pfaffl, 2001).

3.3 Results and Discussion

Pax-6 Expressing cells of the iris proliferate during lens regeneration Considering the expression of Pax-6 mRNA in intact irises and during the process of lens regeneration (Del Rio-Tsonis et al., 1995; Mizuno et al., 1999; Hayashi et al., 2004; Grogg et al., 2005) we designed a series of experiments to examine possible correlations between Pax-6 expression and cellular processes such as proliferation and apoptosis. In order to study a potential correlation between Pax-6 expression and proliferation during lens regeneration, eyes were injected with 1µl of 10mM BrdU 24 hours prior to their enucleation and then analyzed for patterns of cell proliferation and Pax-6 expression. Starting at day 5 post-lentectomy proliferating cells can be detected in both the dorsal and the ventral iris (Figure 1A). All these BrdU-positive cells are also positive for Pax-6 protein (Figure 1B). This trend continues throughout the process of lens regeneration and can be seen at day 7 (Figure 1D, E), 10 (Figure 1G, H), 15 (Figure 1J, K), 20 (Figure 1M, N) and 25 (Figure 1P, Q) post-lentectomy. Proliferating and Pax- 6 expressing cells were also seen in the ventral iris up to 15 days post-lentectomy (Figure 1J, K). Both proliferation and Pax-6 expression is rarely detected in the ventral iris after day 15.

Not every region in the dorsal half of the iris-ring has the same potential for eliciting lens transdifferentiation. In fact, the central portion of the dorsal iris is the one with the most potential for regeneration (Mikami, 1941). Prompted by such observations, we then determined the spatial distribution of the dividing (and Pax-6-positive) cells by evaluating serial sections through the entire dorsal and ventral iris. We divided the sections into 3 regions, one that encompassed the middle third of the iris and two that flanked this central region. Each region was approximately 60µm long on average. During the early stages of regeneration, when the dorsal iris is proliferating, dedifferentiating and giving rise to a new lens vesicle, Pax-6 expression and cell proliferation are mainly limited to the central portion of the iris and only a small fraction of the proliferation occurs in the regions that flank the central portion of the iris. A similar pattern of Pax-6 expression and proliferation is also seen in the ventral iris (Figure 1C, F, I and L). This pattern of Pax-6 expression and cell proliferation tends to suggest that subsets of cells located in the central portion of both irises have the ability to respond to an injury and proliferate. This correlates perfectly with the ability of this part of the iris to elicit lens regeneration (Mikami, 1941). Pax-6 expression as well as cell proliferation in the ventral iris are dramatically reduced after day 15 post-lentectomy. However, a very small number of cells continue to proliferate and are randomly dispersed through the iris on days 20 and 25 post-lentectomy (Figure1O and R respectively). On the other hand, a large number of cells in the dorsal iris continue to express Pax-6 and proliferate between days 20 and 25 post-lentectomy (Figure 1M-R). By day 25 post-lentectomy, as the regenerating lens grows in size, the wave of Pax-6 expression and cell proliferation spreads more laterally and cells that are proliferating and expressing Pax-6 are distributed more evenly through the naso-temporal plane of the regenerating lens (Figure1P-R). This is expected because of the lens growth and differentiation.

In addition to cell proliferation, we also studied the correlation between Pax-6 expression and apoptosis. As seen in Figure 2, on day 7 and day 10 post-lentectomy no apoptosis was detectable (Figure 2A-H). Between days 15 to 25 post-lentectomy, limited

45 apoptotic events were observed in the regenerating lens, however no apoptosis was noticed in Pax-6 expressing cells (Figure 2I-P).

Figure 1. Pax-6 is associated with proliferating cells in the dorsal and ventral iris during lens regeneration. Pax-6 expressing cells are shown in blue while proliferating cells that have incorporated BrdU are shown in green. (A) Cells in both the dorsal (di) and ventral iris (vi) begin to proliferate within 5 days after lentectomy. (B) All proliferating cells of the dorsal and ventral iris also express Pax-6 (arrow). (C) Proliferation and Pax-6 expression in the dorsal and ventral iris is mostly limited to the central part of the iris and is reduced in the peripheral nasal and temporal regions. These trends continue through the earlier phases of the regenerative process and can be seen on days 7 (D-F), 10 (G-I), 15 (J- L) post-lentectomy. Proliferation in the ventral iris reduces after day 15 and is limited to the dorsal iris and lens vesicle (lv) by day

46 20 post-lentectomy (M-O). (P-R) By day 25 post-lentectomy the wave of proliferation (P) and Pax-6 expression (Q) moves more laterally along the lens epithelium (le) of the regenerating lens and is spread more evenly through all three regions of the iris (R).

Figure 2. Pax-6 expression does not correlate with cells undergoing apoptosis. (A, E, I and M) DIC images of sections from different stages of lens regeneration. (B, F, J and N) show TUNEL fluorescence. (C, G, K and O) show Pax-6 immunofluorescence and (D, H, L, and P) show overlays of TUNEL and Pax- 6. (A-D) On day 7 post-lentectomy there are no apoptotic cells (red) in the dorsal iris (di) or the ventral iris (vi). The red staining seen in these sections is due to stromal tissue that autoflouresces. (C) Pax-6 (blue) is

47 expressed in the both irises at this stage. Similarly no apoptosis is noticed on day 10 post-lentectomy (E-H). Apoptotic cells (arrow) are detected in the dorsal iris and lens vesicle (lv) between days 15 (I-L) and 25 (M-P) post-lentectomy, but apoptotic cells do not express Pax-6 (L and P respectively).

Down-regulation of Pax-6 Our observations suggested that Pax-6 is clearly associated with proliferation of the iris pigment epithelial cells during lens regeneration. In order to extend these observations and evaluate their importance for the process of lens regeneration, we undertook experiments where Pax-6 expression was down-regulated. We used Valproic Acid (VPA), a teratogen that has been previously shown to down-regulate members of the Pax family during Xenopus and chick development (Pennati et al., 2001 and Whitsel et al., 2002), as well as morpholinos against Pax-6.

Down-regulation of Pax-6 via VPA We evaluated the efficiency of VPA to down-regulate Pax-6 using two different methods (see Methods). First, PEC from the iris were treated with VPA and downregulation of Pax-6 was evaluated via QRT-PCR. As can be seen in Figure 3A expression of Pax-6 was significantly affected by VPA (p<0.005) in PEC. Second, during lens regeneration, we treated animals for 15 days post-lentectomy and then examined for the presence of Pax-6 protein by immunohistochemistry. Likewise, in VPA-treated animals, the Pax-6 protein was barely detectable in normally Pax-6-expressing cells of the dorsal and ventral iris and of the retina (Figure 3F, G, L, M).

Down-regulation of Pax-6 via morpholinos To assess the efficiency of morpholinos to down-regulate Pax-6, we checked for the presence of Pax-6 specific morpholinos (which were tagged with lissamine, see methods) and for the absence of Pax-6 in those same cells via immunohistochemistry in PEC from the iris (Figure 3B-E). In addition, Pax-6 morpholino treated lentectomized eyes, showed a significant reduction of Pax-6 protein in both regenerating lens vesicles (Figure 3H-K) and in the retina (Figure 3N-Q). Not all cells that have Pax-6 morpholino

48 showed a reduction in Pax-6 levels, however, there was a significant reduction in the number of cells expressing Pax-6. This expression of Pax-6, in a small number of cells that have Pax-6 morpholino, could be explained by the fact we might be detecting residual Pax-6 and that the morpholino might not have been in the cell long enough to completely shut down translation of Pax-6.

Figure 3. Valproic Acid (VPA) or Pax-6 morpholino treatments decrease levels of Pax-6. (A) Real time PCR analysis of Pax-6 expression in cultured PEC. The data is presented as relative change of Pax-6 mRNA levels normalized to untreated cells. The asterisk indicates statistical significance (p<0.005). (B, C, D and E) PEC cultures treated with morpholinos (red). (B) Cells treated with a control morpholino show robust Pax-6 immunofluorescence (blue) whereas, cells treated with a morpholino against Pax-6 shows extremely reduced Pax-6 immunofluorescence (C). (D, E) Pax-6 immunofluoresence from (B) and (C) respectively. (F and G) Sections of 15 day regenerating lens showing Pax-6 immunoflourescence in the dorsal iris and lens vesicle from control (F) and VPA-treated eyes (G) respectively. Note that expression is diminished in the VPA-treated eye. (H and I) Pax-6 immunofluorescence in the dorsal iris and lens vesicle

49 of day 13 regenerating lens treated with morpholinos (red). Pax-6 is reduced in animals treated with Pax-6 morpholino (I) compared to those treated with the control morpholino (H). (J and K) Pax-6 immunofluorescence from (H) and (I) respectively. Pax-6 immunofluorescence is also diminished in the retina (L-Q) of animals treated with VPA (M) or with Pax-6 morpholino (O and Q) when compared to their respective controls (L, N and P). Panels D, E, J, K, P and Q are of the Pax-6 channel extracted from B, C, H, I, N and O respectively. Morphological effects of Pax-6 down-regulation To study the effects of Pax-6 down-regulation during regeneration, we treated newts continuously with VPA starting immediately after lentectomy or with morpholinos 4 and 10 days post-lentectomy, and examined the morphology of the regenerating lens 15 days later. Usually by day 15, untreated controls already have an established lens vesicle with active lens fiber differentiation (Figure 4A). Treatment with VPA caused a marked reduction in the size of regenerating lens vesicles when compared to control animals. On day 15 post-lentectomy, in a majority of VPA-treated eyes, dedifferentiation of the dorsal iris tip was obvious but differentiation of a lens vesicle was rather retarded when compared to the control group (Figure 4B). This retardation of regeneration was reflected in the regeneration index (as described in the methods section) of VPA-treated animals when compared to controls. The VPA-treated group had a regeneration index of 0.0, indicating that regeneration was affected in all animals used in this experiment (n=25) (Figure 4C).

We obtained similar results when Pax-6 was down-regulated via morpholinos. In this case, also, very small vesicles regenerated (Figure 4E) when compared to control morpholino treated eyes (Figure 4D). The regeneration index of the Pax-6 morpholino treated eyes was 0. 27 compared to control-morpholino of 0.80 (n=15) (Figure 4F).

Figure 4. Downregulation or knock-down of Pax-6 inhibits lens regeneration. (A) A control animal on day 15 post-lentectomy shows a well developed lens vesicle. (B) Treatment of animals with Valproic acid (VPA) for 15 days after lentectomy reduced the size of the regenerating lens vesicles. An example of a severely reduced lens vesicle is shown in B. (C) Treatment with VPA reduced the regeneration index of animals as assessed on day 15. (D) Fifteen day lens vesicle from an animal treated with control morpholino (C-Mo) shows a well developed lens vesicle. (E) However, animals treatedwith Pax-6 morpholino (Pax6- Mo) have a lens vesicle that is diminished in size. (F) Injection of eyes with Pax-6 morpholino reduced the regeneration index as assessed on day 15.

Pax-6 controls proliferation of cells in the regenerating iris A reduction in the size of the lens vesicle as a result of VPA or Pax-6 morpholino treatment suggests that Pax-6 down-regulation causes a reduction in proliferation or an increase in cell death. To test these two hypotheses, we treated lentectomized animals with VPA for a total of 13 days, or in the case of morpholinos, we treated the eyes at day 10 (a time when proliferation is at a high peak in normal regenerating lenses) and

51 collected 3 days later. Twenty-four hours prior to collecting tissues, BrdU was injected into the eyes of these animals. BrdU analysis revealed that proliferation was reduced when Pax-6 was down-regulated. On day 13, only 25% of the cells in the lens vesicles of VPA-treated animals were proliferating as compared to controls in which 53% of all cells in the lens vesicle were proliferating (Figure 5A-C). This 50% reduction in proliferation is statistically significant (p<0.001) and suggests that down-regulation of Pax-6 might be responsible for reduced proliferation. It could also be possible that the effect of VPA on lens regeneration was due to a general down-regulation of cell division and not necessarily through a Pax-6 mediated down-regulation. To exclude this possibility, we also examined proliferation in epithelial cells of the small intestine, where Pax-6 was not detected (not shown). This is in accordance to Larsson et al. (1998) where they report via a LacZ line that epithelial cells in the lower GI tract do not express Pax-6. The small intestine of VPA-treated animals did not show a reduction in proliferation suggesting that the reduction of proliferation in the regenerating lens vesicle was specifically mediated via down-regulation of Pax-6 and not due to a suppression of cell proliferation in general. (Figure 5D-F).

In Pax-6 morpholino treated eyes, a statistically significant reduction of proliferating cells was also apparent, with only 28% of the lens vesicle cells scoring for BrdU reactivity compared to 50% of the cells in control morpholino-treated eyes (Figure 5 G-I). In addition, morpholino treated PEC where also analyzed for cell proliferation activity, showing that upon addition of Pax-6 specific morpholinos, proliferation (21%) was reduced approximately two-thirds compared to control-treated morpholino cultures (60%) (Figure 5J-L).

Animals treated with VPA or Pax-6 morpholino did not show any increase in apoptosis and the levels of apoptosis on day 13 were comparable between the two groups (not shown). These sets of data strongly argue that Pax-6 is indeed associated with proliferation during lens regeneration and is not involved in survival or apoptosis.

Figure 5. Valproic acid (VPA) treatment or morpholino injection reduces proliferation in vivo and in vitro. Pax-6 expressing cells are shown in blue while cells that are proliferating and incorporating BrdU are shown in green. (A) Lens vesicles of control animals show a large number of cells that are proliferating and incorporate BrdU at day 13 post-lentectomy. (B) Lens vesicles from animals that were treated with VPA for 13 days after lentectomy have few cells that proliferate and incorporate BrdU. (C) VPA- treatment severely retards cell proliferation and the difference from control is statistically significant (p < 0.001). (D and E) Cells from the small intestine show a large number of proliferating cells that incorporate BrdU in both control (D) and VPA-treated animals (E). (F) The difference in the levels of proliferation in the small intestine of VPA-treated and control animals is not statistically significant. (G) Animals treated with control morpholino (C-Mo) (red) also show a large number of proliferating cells on day 13 post- lentectomy, and the percentage of cells proliferating are comparable to those in A. (H) Treatment with a Pax-6 morpholino (Pax6-Mo) (red) however, decreases the number of proliferating cells seen on day 13 post-lentectomy. (I) Morpholino knock-down of Pax-6 results in a statistically significant reduction in proliferation (p<0.01) when compared to animals treated with a control morpholino. In vitro, cultured PEC

53 transfected with the control morpholino (J) show more proliferation than those transfected with the Pax-6 morpholino (K). (L) The number of proliferating cells transfected with the Pax-6 morpholino is significantly less than the number of proliferating cells transfected with the control morpholino (p<0.01). Asterisk in E, I and J denote statistical significance. All error bars are S.E.M.

Down-regulation of Pax-6 delays differentiation in the regenerating lens During lens development, Pax-6 is required for the production of various crystallins and differentiation of lens fibers (Gopal-Srivastava et al., 1998). We, therefore, wanted to determine whether Pax-6 plays a similar role during regeneration as well. Newts were treated with VPA for 13 days post-lentectomy, at which time lens fiber differentiation begins, and then α and β crystallin expression was assayed using immunohistochemical methods. Control lens vesicles showed expression of both crystallins whereas lens vesicles from animals that were treated with VPA for 13 days post-lentectomy did not express either α or β crystallin (compare Figure 6A and F to 6B and G respectively). In addition, Pax-6 morpholino treated eyes (lentectomized eyes treated with morpholinos from day 10-13) also lacked crystallin expression when compared to control morpholino treated vesicles which expressed both α or β crystallin (compare Figure 6D and I to 6E and J respectively).

It could be argued that the lens vesicle in VPA-treated animals on day 13 post- lentectomy might be too small and had not reached a size that would be needed before crystallin expression begins. In order to rule out this possibility, we performed a second set of experiments where animals were treated with VPA everyday for 15 days post- lentectomy and then assayed for crystallin expression. Consistent with our previous results, animals that had lower levels of Pax-6 did not show any α or β crystallin expression (Figure 6C and H). However, at this stage, the lens vesicles from VPA- treated animals were similar in size to a control day 13 lens vesicle that expresses both α or β crystallin (compare Figure 6A and F to 6C and H). These results clearly show, that in addition to playing a role in proliferation, Pax-6 is also required for the expression of crystallins in the regenerating lens and that the effects on differentiation are not due to a the reduced size of lens vesicles. These experiments also suggest that these two roles may

54 be independent of each other, since animals treated with VPA for 15 days were able to proliferate sufficiently to generate a lens vesicle comparable to a control day 13 lens vesicle but were still not able to initiate crystallin expression.

Figure 6. Valproic Acid (VPA) treatment or Pax-6 morpholino treatments affects the initiation of differentiation. (A) Lens vesicles from control animals on day 13 post-lentectomy express α-crystallin (green). (B) However the lens vesicle from a VPA-treated animal shows no α-crystallin expression at the same stage. (C) Lens vesicles from animals that were treated with VPA for 15 days have lens vesicles similar in size to those seen in day 13 controls (compare with A), but they do not express α-crystallin. Morpholino (red) mediated knock down of Pax-6 also yielded the same result. (D) Lens vesicle from an animal treated with control morpholinos (C-Mo), between days 10 and 13, expresses α-crystallin whereas (E) lens vesicles of a similar size from animal treated with Pax-6 morpholino (pax6-Mo) do not show any α-crystallin expression. (F- H) A similar trend is seen with the expression of β-crystallin (green). Control animals express β-crystallin at day 13 (F) post-lentectomy but lens vesicles from VPA-treated animals at day 13 (G) and day 15 (H) do not express any β-crystallin. Similarly animals treated with control morpholino express β-crystallin 13 days post-lentectomy (I) but animals treated with Pax-6 morpholino do not express β-crytsallin (J). All error bars are S.E.M.

However, it is not clear if defects in the initiation of differentiation might influence proliferation during the later stages of regeneration. In order to address this question, we lentectomized animals and allowed them to recover for 15 days after lentectomy. By this time, differentiation would have been initiated in the lens vesicle. Treatment with VPA (180µg/day/animal) was initiated at day 15 in order to study the effects of Pax-6 down-regulation after differentiation had been initiated. VPA treatment was continued until day 25 at which time we assayed for BrdU incorporation and cell proliferation. For morpholino experiments, lentectomized eyes were treated from day 20-

55 25 and eyes assayed for cell proliferation. BrdU was injected in both treatments 24 hours before collection. Animals that were treated with VPA showed fewer proliferating cells in the lens vesicle than that of control animals (compare Figure 7A to B). Animals that had reduced levels of Pax-6 showed approximately 25% reduction in cell proliferation (Figure 7C). The regenerating lens of VPA-treated animals still expressed α and β crystallins (not shown). Similarly, Pax-6 morpholino treated eyes showed a 23% reduction in cell proliferation in regenerating lenses when compared to the control morpholino treated eyes (69% compared to 92% in controls, Figure 7D, E and F). However, these regenerating lenses still expressed α and β crystallins (Figure 7G-J). These results suggest that once lens fiber differentiation has begun and crystallins are produced, Pax-6 is not required for the maintenance, stability and continued production of these proteins.

Figure 7. Valporic Acid (VPA) or Pax-6 morpholino treatments decrease proliferation of lens epithelium cells. When VPA was administered later between days 15 and 25, proliferation in the lens epithelium decreased (B) compared to the lens epithelium of control animals (A). (C) The percentage of cells proliferating in the lens epithelium of VPA-treated animals is 25% lower than those in control and this difference is statistically significant (p<0.01, asterisk). Animals treated with morpholinos (red) between days 20 and 25 post-lentectomy were also assessed for proliferation defects. (D) Animals treated with control morpholinos (C-Mo) did not show any reduction in proliferation and are comparable to control animals in A. (E) Animals treated with Pax-6 morpholino (Pax6-Mo) however show reduced proliferation. (F) There are approximately 35% more cells proliferating in the lens epithelium of control morpholino treated animals than Pax-6 morpholino treated animals. This difference is statistically significant (p<0.01). Pax-6 is however, not required for the maintenance of crystallin expression after the onset of differentiation. α-crystallin (G and H) and β-crystallin (I and J) are present in day 25 lens vesicles of morpholino treated eyes. There seems to no qualitative difference in crystallin expression in lens vesicles from control morpholino treated (G and I) and Pax-6 morpholino treated (H and J) eyes. All error bars are S.E.M.

57 Our experiments have shown that the primary role of Pax-6 during regeneration is the regulation of proliferation in both the dorsal and ventral iris. In addition to this, Pax-6 regulates differentiation and crystallin expression in the regenerating lens vesicle. Since dedifferentiation was not inhibited, our data suggest that Pax-6 is not involved in the induction of lens regeneration from the dorsal iris. These observations are in agreement with in vitro/ in vivo studies where it was found that over-expression of Pax-6 in the ventral iris PECs was not sufficient to coax this regeneration-incompetent tissue to transdifferentiate to lens (Grogg et al., 2005). Our results therefore negate the long held view that over-expression of a master-regulator gene such as Pax-6 will be sufficient to induce regeneration in regeneration incompetent tissues types and suggest that there are other molecular mechanisms that induce regeneration.

3.4 References

Altmann, C.R., Chow, R.L., Lang R.A.and Hemmati-Brivanlou A. (1997). Lens induction by Pax-6 in Xenopus laevis. Dev. Biol. 185: 119-23.

Chi, N. and Epstein, J.A. (2002). Getting your Pax straight: Pax proteins in development and disease. Trends Genet. 18(1):41-7

Cvekl, A. and Piatigorsky, J. (1996). Lens development and crystallin gene expression: many roles for Pax-6. Bioessays. 18(8):621-30.

Del Rio-Tsonis, K., Washabaugh, C.H. and Tsonis, P.A. (1995). Expression of Pax-6 during urodele eye development and lens regeneration. Proc. Natl. Acad. Sci. USA 23;92(11):5092-6.

Del Rio-Tsonis, K., Jung, J.C., Chiu, I.M. and Tsonis, P.A. (1997). Conservation of fibroblast growth factor function in lens regeneration. Proc. Natl. Acad. Sci. USA 94(25):13701-6.

Del Rio-Tsonis, K., Tomarev, S.I. and Tsonis, P.A. (1999). Regulation of Prox 1 during lens regeneration. Invest. Ophthalmol. Vis. Sci. 40(9):2039-45.

Del Rio-Tsonis, K. and P.A.Tsonis (2003). Eye regeneration at the molecular age. Dev. Dyn. 226(2):211-24.

Eguchi, G. and Shingai, R. (1971). Cellular analysis on localization of lens forming potency in the newt iris epithelium. Develop. Growth. Differ. 13: 337-49.

Fisher, M. and Grainger, R.M. (2004). Lens induction and determination. In: Development of the ocular lens (Lovicu, F.J and Robinson, M.L, editors) Cambridge University Press, Cambridge, pp 27-47.

Gehring, W.J. and Ikeo, K. (1999). Pax 6: mastering eye morphogenesis and eye evolution. Trends Genet. 15(9):371-7.

Gopal-Srivastava, R., Cvekl, A. and Piatigorsky, J. (1998). Involvement of the retinoic acid/retinoid receptors in regulation of murine alphaB-crystallin/small heat shock protein gene expression in the lens. J. Biol. Chem. 273: 17954-61.

Grogg, M.W., Call, M.K., Vergara M.N, Okamoto, M., Del Rio-Tsonis, K. and Tsonis, P.A. (2005). BMP inhibition-driven regulation of six-3 underlies induction of newt lens regeneration. Nature (in press).

Halder G., Callaerts P., Gehring W.J. (1995). Induction of ectopic eyes by targeted expression of the eyeless gene in Drosophila. Science 267(5205):1788-92.

Hayashi, T., Mizuno, N., Ueda, Y., Okamoto, M. and Kondoh, H. (2004). FGF2 triggers iris-derived lens regeneration in newt eye. Mech. Dev. 121(6):519-26.

Lang, R.A. (2004). Pathways regulating lens induction in the mouse. Int. J. Dev. Biol. 48: 783-91.

Mikami, Y. (1941). Experimental analysis of the Wolffian lens regeneration. Jap. J. Zool. 9: 269-302.

Mizuno, N., Mochii, M., Yamamoto, T.S., Takahashi, T.C., Eguchi, G., Okada, T.S. (1999). Pax-6 and Prox 1 expression during lens regeneration from Cynops iris and Xenopus cornea: evidence for a genetic program common to embryonic lens development. Differentiation. 65(3):141-9.

Pennati, R., Groppelli, S., de Bernardi, F. and Sotgia, C. (2001). Action of valproic acid on Xenopus laevis development: teratogenic effects on eyes. Teratog. Carcinog. Mutagen. 2001;21(2):121-33

Phaffl, M.W. (2001). A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 29: 2002-7.

Reyer, R.W. (1977). The amphibian eye: Development and regeneration. In: The visual system in vertebrates (Crescitelli, F., editor). Springer-Verlag, Berlin, pp 309-390.

Sawada, K., Agata, K., Yoshiki, A., Eguchi, G. (1993). A set of anti-crystallin monoclonal antibodies for detecting lens specificities: beta-crystallin as a specific marker for detecting lentoidogenesis in cultures of chicken lens epithelial cells. Jpn. J.

59 Ophthalmol. 37(4):355-68.

Tsonis, P.A. and Del Rio-Tsonis, K. (2004). Lens and Retina Regeneration: Transdifferentiation, Stem Cells and Clinical applications. Exp. Eye Res. 78: 161-172.

Tsonis, P.A., Madhavan, M., Tancous, E.E., Del Rio-Tsonis, K. (2004a). A newt's eye view of lens regeneration. Int. J. Dev. Biol. 48(8-9):975-80.

Tsonis, P.A., Madhavan, M., Call, M.K., Gainer, S., Rice, A. and Del Rio-Tsonis K (2004b). Effects of a CDK inhibitor on lens regeneration. Wound Repair Regen. 2(1):24- 9.

Whitsel, A.I., Johnson, C.B., and Forehand, C.J. (2002). An in ovo chicken model to study the systemic and localized teratogenic effects of valproic acid. Teratology 66(4):153-63.

Chapter 4

Regeneration and Transdifferentiation in the Embryonic Chick This chapter has been formatted for Development and part of this work was published in Spence and Madhavan et al. (2004) Development:131(18):4607-21

4.1 Introduction

Coulombre and Coulombre (1965) were the first to show that in the embryonic chick retina could regenerate upon its complete removal but only if they placed a small piece of retina back into the eye. This ability was restricted to a small window of time between Hamburger and Hamilton stages 22 - 24.5 or between E3.5 - 4.5. They also showed that the retina regenerated from two different sources: via transdifferentiation of the Retina Pigmented Epithelium (RPE) and by an unknown population of cells in the anterior part of the eye. The identity of the molecules in the retina that induced regeneration remained unknown.

Almost 25 years later, it was discovered that members of the FGF family were able to induce regeneration in the embryonic chick. Park and Hollenberg (1991) used a slow release polymer, called Elvax, to deliver different growth factors including various FGFs, TGF-β, Insulin, IGF-1 and 2 and NGF-β into the eye post-retinectomy but only FGFs could regenerate a laminar retina by E11. They demonstrated that FGF-2 could cause transdifferentiation of the RPE in the posterior region of the eye while FGF-1 at high doses could induce cells in the anterior region of the eye to regenerate a retina (Park and Hollenberg, 1989 and 1991). Like Coulombre and Coulombre, Park and Hollenberg observed that retina regenerated from transdifferentiating RPE seemed to have reversed polarity where the photoreceptor layer was closest to the lens. They also observed that the RPE did not replenish itself and therefore the regenerated retina did not receive

61 nourishment and degenerated around E19. On the other hand, the retina that regenerated from the anterior margin of the eye had the correct polarity and had an intact RPE. The identity of the cells in the anterior of the eye that gave rise to a new retina remained unknown until the work described in this dissertation was published (Spence and Madhavan et al., 2004).

Work from other labs on post-hatch chickens have shown that retinal stem cells located in the Ciliary Body (CB) and retinal progenitor cells located in the Ciliary Marginal Zone (CMZ) can proliferate in response to growth factors and generate various retinal neurons (Fischer and Reh, 2000; 2003). We hypothesize that a similar embryonic population of cells in our system might be responsible for the regeneration of retina from the anterior margin of the eye. We performed experiments to trace the lineage of the regenerated neurons to stem/progenitor cells located in the anterior of the embryonic chick eye. Due to the lack of clear morphological or molecular markers to distinguish between the CB and the CMZ at this stage we refer to this region as the CB/CMZ.

Previous studies on retina regeneration in the embryonic chick mainly used histological observations to describe the process of regeneration (Coulombre and Coulombre, 1965 and Park and Hollenberg, 1989). Although these investigators observed that a laminar retina was formed, they did not determine if all the different cell types in each retina layer were regenerated. Our experiments demonstrate that FGF-2 can induce regeneration from both the RPE as well as from stem/progenitor cells in the CB/CMZ and that all main cell types of the retina are regenerated.

Although studies published prior to the work described in this dissertation focused on identifying molecules that can induce transdifferentiation of the RPE, very little is understood about the molecular mechanisms that limit this ability to a short window in development. RPE from older animals have been shown to transdifferentiate in vitro suggesting that the ability of the RPE to transdifferentiate might be present long after this window and therefore the inability of the RPE to transdifferentiate in vivo at later stages might be due to the insensitivity of cells to an FGF signal. Although FGF Receptor

62 (FGFR) expression patterns have been described in the eye, a clear understanding of the FGFRs expressed by the RPE between E3.5 and 4.5 does not exist (Ohuchi et al., 1994 and Tcheng et al., 1994). Although the signaling cascades used by FGF receptors in the RPE are not known, the expression of constitutively active Mitogen activated Erk Kinase (MEK) during development can convert RPE into retina (Galy et al., 2002). These results indicate that activation of Mitogen Activated Kinases cascade by FGF-2 could regulate the process of transdifferentiation. The downstream targets of these kinases still remain unknown. Two potentially downstream molecules that have been implicated in the processes of transdifferentiation are the transcription factors Pax-6 and Microphthalmia (Mitf).

Pax-6 is a paired box - homeobox containing transcription factor that plays a role in the normal development of various organs including brain, pancreas and eye (Ashery- Padan and Gruss, 2001). Its role in the eye is especially important and it is considered the “master regulator” of eye development since ectopic expression of Pax-6 homologs in Drosophila and Xenopus yield ectopic eyes (Halder et al., 1995, Chow et al., 1999). Though such drastic phenotypes are not seen in mammalian systems, hypomorphic Pax-6 mutations in mice results in a small eye phenotype and in humans can cause aniridia (Ton et al., 1991). During early development, Pax-6 is expressed throughout the optic cup, both in the neuroepithelium (which will give rise to the retina) as well as the presumptive RPE (Grindley et al., 1995). As development progresses the expression of Pax-6 is limited to the developing retina and lens (Belecky-Adams et al., 1997). In the retina, Pax-6 is required to maintain the multipotent state of retinal progenitors in the neuroepithelium and is responsible for the induction of various basic Helix loop Helix transcription factors that result in the differentiation of various retinal cell types (Marquardt et al, 2001). Recent studies have also shown that over-expression of Pax-6 in the RPE, during development, can cause transdifferentiation as late as E14 (Azuma et al., 2005). However, at the present time, a concrete connection between FGF-2 signaling and Pax-6 upregulation during regeneration has not been established.

63 In addition to upregulation of pro-retinal genes like Pax-6, transdifferentiation of RPE might also require the downregulation of genes that specify the RPE phenotype. One such candidate gene is Mitf. Mitf is expressed in the developing RPE and is required for the proper specification and maintenance of RPE (Mochii et al., 1998). Naturally occurring mutations in Mitf cause RPE to transdifferentiate into retina in the quail and mouse (Mochii et al., 1998b, Bumsted and Barnstable, 2002). It has also been shown in vitro, that down-regulation of Mitf in cultured RPE, results in transdifferentiation of the RPE into retina (Mochii et al., 1998a, Iwakiri et al., 2005). Moreover, Pax-6 and Mitf can control the regulation of the other target genes (Planque et al. 2001). These data suggest that reciprocal changes in Mitf and Pax-6 levels might be crucial for transdifferentiation of RPE into neural retina.

The purpose of this study was to determine the molecular mechanisms that allow the RPE to regenerate retina, via transdifferentiation, in response to FGF-2 on E4 chick embryos. We show that the temporal and spatial expression pattern of FGFR-2 correlates well with the time during which FGF-2 can induce transdifferntiation. Results indicate that FGF signals through a Mitogen Activated Protein Kinase cascade and that activation of MEK in vivo is sufficient to regenerate retina. Furthermore, we demonstrate that the activation of Pax-6 post-retinectomy causes transdifferentiation while the downregulation of Mitf is necessary for the transdifferentiation process to take place.

4.2 Methodology

Chick Embryos White Leghorn chicken eggs were purchased form Berne Hi-Way Hatcheries, Berne, IN and from the Ohio State University, Columbus, OH and incubated in a humidified rotating incubator at 38oC.

Surgical Procedures Retinectomies were performed on E4 embryos as previously described (Coulombre and Coulombre, 1965). Briefly, a window was made in the egg using a pair

64 of forceps and the membranes surrounding the embryo were removed. A fine tungsten needle was used to make an incision dorsal to the lens and then micro-dissection scissors were used to make a semi-circular cut around the lens that allowed access into the optic cup. The retina was removed using fine forceps and either FGF-2 or control heparin coated beads were placed inside the optic cup. RCAS-Mitf (generously provided by Dr. M. Mochii), RCAS-MEKDD (gift from Dr. R.Latarjet) and RCAS-Pax6 (gift from Dr. P. Bovolenta) were injected subretinally on E3, one day before the retinectomy. All viruses were cultured as described below and viruses were mixed with Fast Green dye and injected at a concentration of 108-109 Colony Forming Units/µl. The embryos were returned to the incubator and processed for histology or immunohistochemistry.at E5, E7, E9, E11 or E15.

Preparation of FGF-2 beads Heparin coated polyacrylamide beads (Sigma, St. Louis, MO) were washed 3 times in PBS. Twenty of these beads were placed in 2µl of 1mg/ml FGF-2 (R&D Systems, Inc., Minneapolis, MN) solution. Control beads were soaked in PBS.

Histology Tissues used for histological analysis were fixed in Bouin’s solution for at least 1 day and then dehydrated through an ethanol series and embedded in paraffin. Samples were section at a thickness of 10µm and stained using Hematoxylin and Eosin (H&E).

Immunohistochemistry Embryos were fixed in 4% Formaldehyde and embedded in OCT medium (Sakura Finetek, Torrance, CA) and then frozen in a dry ice: ethanol bath. The embryos were sectioned at a thickness of 10µm. A standard protocol was used for immunohistochemistry. Briefly, sections were washed in PBS and then blocked for one hour. Immunohistochemical procedures using antibodies against transcription factors included a five minute wash in 1% Saponin (Sigma, St. Louis, MO) followed by PBS washes prior to the blocking step. Primary antibodies were diluted in blocking solution and incubated at 4oC overnight. The sections were then washed in PBS and incubated in

65 secondary antibody at 4oC for 2 hours and cover slipped using Vectashield mounting medium (Vector labs, Burlingame, CA). The samples were first analyzed using an Olympus Laser Confocal Microscope and then the coverslip were removed and the sections were stained using hematoxylin and eosin.

Antibodies The following antibodies were obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by The University of Iowa, Department of Biological Sciences, Iowa City, IA 52242. Anti-Pax6 (1:100), Anti-Napa 73 (1:100), Anti-visinin (7G4, 1:100), Anti-vimentin (1:100 ) Anti- BrdU(G3G4., 1:100) and Anti-AMV3C2 (1:100). Anti-Brn3a, (1:100) was purchased from Covance Research Products, Inc (Denver, PA). Anti-Chx-10 (1:1000 for immunohistochemistry), Anti-Mitf (1:100 for immunohistochemistry) and Anti-MMP115 (1:100 for immunohistochemistry) antibodies were generous gifts from Dr .T. Jessell (Columbia University, New York, NY) and Dr. M. Mochii (Himeji Institute of Technology, Hyogo, Japan). Antibodies against FGFR-1 (flg): 1:100 and FGFR-2 (bek): 1:100 were purchased from Jackson Immunoresearch Laboratiories Inc. (West Grove, CA). Antibodies against FGFR-3 were purchased from Chemicon International (Temecula, CA). Secondary antibodies include Goat anti-mouse FITC, Goat anti-mouse Cy3, Goat anti-mouse Biotin (Jackson Immunoresearch Laboratories Inc, West Grove, PA), Alexa Fluor 488 and 546 conjugated Goat anti-mouse antibody or Goat anti-rabbit antibody and streptavidin conjugated Alexa Fluor 350 (Invitrogen)..

Retroviral production Retroviruses were produced by infecting DF1 chicken fibroblasts (ATCC, Manassas, VA) with retroviral DNA using Lipofectamine reagent (Gibco Invitrogen

Corp., Carlsbad, California). Culture media containing retroviruses were collected and then concentrated using Millipore ultra collection device (Millipore, Billerica, MA). The viruses were titered by diluting the viral stock and infecting cultured DF-1 cells and then performing immunohistochemistry using the AMV3C2 antibody against the viral gag protein to determine Colony Forming Units/µl (CFUs/µl).

DiI and BrdU labeling DiI cell labeling paste (Molecular probes, Eugene, OR) was used to track cells of the CB/CMZ during regeneration. A retinectomy was performed as described. This exposed the lens and surrounding CB/CMZ region. Using a glass micropipette, DiI cell labeling paste was carefully transferred onto the surface of cells in this area. Embryos were collected either a few minutes after the surgery, to ensure that only cells in the CB/CMZ were labeled and that there was no transfer of DiI onto the RPE, or on E5 or E7 to track the cells that regenerate from the CB/CMZ. Regenerating eyes were labeled with BrdU by micropippeting 1µl of 10mM BrdU solution into the optic cup.

4.3 RESULTS

Spatio-Temporal Characterization of Retina Regeneration

Park and Hollenberg (1991) used Elvax beads to deliver FGF-2 in their chick retina regeneration studies. We instead used heparin coated Polyacrylamide beads to deliver FGF because heparin is known to increase binding of FGF-2 to its receptors (for review see Mohammadi et al., 2005). As this method was different, we first examined the spatial and temporal pattern of retina regeneration. We performed Retinectomies on E4 chick embryos and placed an FGF-2 or a control bead in the optic cup (Figure 1A). The embryos were collected at 3, 5, 7 or 11 days post-retinectomy (E7, 9, 11 and 15 respectively) and used for histological analysis. Embryos that received the control heparin beads did not regenerate (Figure 1B). At all time points examined, regeneration via transdifferntiation of RPE was seen in the posterior region of the eye along with retina that regenerated from the CB/CMZ from the anterior region of the eye. This observation is different from previous reports using FGF-2 where only transdifferentiation was observed. We believe the difference is due to the use of heparin coated FGF-2 beads which allows for the activation of cells from the CB/CMZ. The regeneration induced by this method is robust and by E11 both of the regenerated retinas show laminar organization (Figure 1C). The spatial and temporal pattern in both types of

67 regeneration are very similar (compare Figure 1E, H, K, N to 1F, I, L, O). RPE that transdifferentiated showed depigmentation as early as E5 (one day after surgery with FGF-2 treatment, not shown) and by E7 forms a neuroepithelium (Figure 1E). By this time a neuroepithelium was also formed from the CB/CMZ (Figure 1F and Figure 2). These neuroepithelia continue to grow (Figure 1 H and I) and start to show laminar organization by E11 (Figure 1K and L). At E11, the regenerating retina has 3 distinct retinal layers, similar to an E9-E11 retina during normal development (compare Figure 1 G, J to K, L). The reverse orientation of transdifferentiated compared to normally developing retina or to CB/CMZ regenerated retina are apparent at E11 and E15 (Compare Figure 1 K and N to Figure 1 J and M and Figure 1 L and O). All the eyes in figure 1 J-O are oriented with the lens placed towards the top of the page. The ganglion cells in normally developing eyes (Figure 1J (E11) and 1M (E15)) and in CB/CMZ regenerates(Figure 1L (E11) and Figure 1O (E15)) are positioned towards the lens, but the ganglion cells in the retina derived from transdifferentiating RPE is located away from the lens and photoreceptors of this retina are closest to the lens (Figure 1K (E11) and 1N (E15)). Even though there are differences in the orientation of the regenerated retinas, they are similar to a normally developing retina at E11 an E15.

68 Figure 1. FGF- 2induces retina regeneration in two distinct ways. (A) At E4 the retina is removed surgically leaving behind RPE and CB/CMZ. An FGF-2 soaked heparin bead (*) is then placed in the eye cup. (B) A heparin bead (not visible in this section) alone does not cause regeneration after 7 days (E11). (C) By E11, FGF-2 induces regeneration from the CB/CMZ (cr) and by the transdifferentiation of the RPE (td). (D-O) Histology of normally developing as well as regenerating retina at E7 (D-F), E9 (G-I), E11 (J-L) and E15 (M-O). D,G,J,M show normal development at each stage. Three days after retinectomy (E7) a transdifferentiated neuroepithelium (E) as well as a neuroepithelium generated from cells in the CB/CMZ (F) are present. At E9, the transdifferentiated neuroepithelium (H) as well as the neuroepithelium arising from the CB/CMZ (I) thicken and grow. (K, L ) Seven days post- retinectomy, at E11, the various retinal layers become visible. By E15, regenerated retinas (N, O) are laminated and resemble an E11 developing retina (J). Scale bar: 100 µm (A, B); 500 µm (C); 100 µm (DO). GCL, ganglion cell layer; IPL, inner plexiform layer; INL, inner nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; ne, neuroepithelium; l, lens; CB/CMZ, ciliary body/ciliary marginal zone; RPE, retinal pigmented epithelium.

Stem/progenitor cells in the CB/CMZ

Though we hypothesized that cells in the CB/CMZ were responsible for retina regeneration from the anterior margin, we did not have any evidence for this in our system. To test this hypothesis, we needed to show that retina regeneration was initiated from cells of the CB/CMZ and not from retinal cells that were not removed during surgery. We specifically marked the CB/CMZ (Kubo et al., 2003) with Collagen Type IX (Figure 2 A and D) and showed that the only tissue left behind was CB/CMZ. We then used adjacent sections to show that the CB/CMZ was rich in Pax-6/Chx-10 positive stem/progenitor cells (Figure 2B) (Belecky-Adams et al., 1997) and that cells in the CB/CMZ were proliferating (BrdU incorporation as shown in Figure 2C). We then reasoned that if cells in the CB/CMZ were indeed giving rise to new retina, then we would be able to trace the movement of newly formed cells from the CB/CMZ into the regenerating retina. We labeled cells in the CB/CMZ using a lipophyllic dye called DiI (Figure 2D). The dye remains in the membrane of these cells and is passed on to the progeny of these cells. When we collected embryos 1 and 3 days post-retinectomy (E5 and E7 respectively), we observed that the regenerated retinas were labeled with DiI (Figure 2E and F). The DiI staining was strongest near the CB/CMZ and faded over distance. These results demonstrate that Pax-6/Chx-10 positive stem/progenitor cells in the CB/CMZ proliferate and to regenerate retina.

Fig. 2. Cells in the CB/CMZ are a source of regenerating retina. (A) Collagen type IX immunohistochemistry performed on an E4 operated eye 4 hours post-retinectomy shows the CB/CMZ region has not been removed. (B) An adjacent section shows that the CB/CMZ is rich in cells that express Pax-6 (blue) and Chx-10 (red). White arrowheads indicate progenitor cells that co-express Pax-6 and Chx10. (C) Another adjacent section shows that within 4 hours of retinectomy, most cells in the CB/CMZ are proliferating and labeled with BrdU, suggesting that progenitor cells (like those shown in B) are proliferating. (D) The CB/CMZ was labeled with DiI (red) immediately after retinectomy. The collagen type IX (green) immunofluorescence confirms that the labeled region is the CB/CMZ. (E) At E5, 1 day after the CB/CMZ was labeled, DiI labeled cells have migrated from the CB/CMZ and generated a neuroepithelium. (F) Three days after retinectomy this phenomenon is still apparent. The asterisk indicates the region where the CB/CMZ was originally labeled with DiI. c, cornea; CB/CMZ, ciliary body/ciliary marginal zone; l, lens; RPE, retina pigmented epithelium; pe, pigmented epithelium; cr, retina that regenerated from the CB/CMZ. The red arrowheads indicate the site at which the incision was made during the retinectomy. Scale bar in A is for A-C.

All retinal cell types are present in the regenerating retina.

To determine if all the different cell types were produced during regeneration we used multiple cellular markers. A list of these markers can be found in Table 1.

Table 1: List of Retinal Markers Marker Molecular Characteristics Cell Types Marked Pax-6 Transcription Factor GCL, Amacrine cells, and Horizontal cells Chx-10 Transcription Factor Bipolar Cells Visinin Calcium binding Protein Photoreceptors Brn3A Transcription Factor Ganglion cells Napa 73 Neurofilament NFL Mitf Transcription Factor Retina Pigmented Epithelium MMP 115 Matrix Protein Retina Pigmented Epithelium Pax-6 + Transcription Factors Neural Progenitor Cells Chx-10

Eyes collected on E7, E11 and E15 were used for double labeling experiments using the following antibodies: Napa73 and Brn3a to detect the Nerve Fiber Layer (NFL) and the Gangion Cell Layer (GCL), respectively; Pax-6 to detect horizontal, amacrine and ganglion cells along with Chx-10 to detect bipolar cells; Visinin and Mitf to detect photoreceptors and RPE, respectively. In addition to these, Vimentin was used to identify Müller glia and co-expression of Pax-6 and Chx-10 identified progenitor cells.

During development, the first set of cells to be clearly defined are ganglion cells and ganglion cells along with their fibers can be clearly seen at E7 (Figure 3 A1). Also at this stage photoreceptors are beginning to differentiate but they have not formed an organized layer (compare Figure 3 A3). Also visible at this stage are a large number of progenitor cells (Pax-6/Chx-10 positive cells in Figure 3 A2) present in the area that will form the Inner and Outer Nuclear Layers of the retina. Even though most of the neurons

72 in the retina are still being specified at this stage, Müller Glia cells are present throughout the retina (Figure 3 A4). However, during regeneration at E7, few differentiated cell types are present in the retina and most of the cells in the regenerated neuroepithelia are progenitor cells (Figure 3 B2 and Figure 3 C2). The retina that originates from transdifferentiation (Figure 3B1-B4) and also the one that regenerates from the CB/CMZ (Figure 3 C1-C4) have few nerve fibers that are Napa-73 positive, suggesting that ganglion cells are beginning to differentiate but have not yet started to express Brn3a (Figure 3 B1 and C1 respectively). Most cells in the neuroepithelium remain as progenitor cells and have not started to differentiate (Figure 3B2 and 3C2). The neuroepithelium derived from transdifferentiating RPE shows some visinin positive cells and no Mitf expression suggesting that Mitf is downregulated during transdifferentiation (Figure 3B3). Retina regenerating from the CB/CMZ show strong Mitf expression (Figure 3C3) in the RPE, but shows no photoreceptor differentiation (Figure 3C3). Glia area present through out the length of both regenerated neuroepithelia and seem to have no defects in their organization as they run perpendicular along the length of the neuroepithelia (Figure 3 B4 and C4).

By E11 of normal development all the different cell types in the retina are differentiated and ganglion cells (Figure 3D1) amacrine cells, bipolar cells horizontal cells (Figure 3D2) and photoreceptors (Figure 3D3) are clearly defined by this stage. Also by this time the regenerating retina has “caught up” temporally with normal developing retinas and retinas arising from transdifferentiating RPE (Figure 3E1-E4) and from the CB/CMZ (Figure 3 F1-F4) express all the cell markers for ganglion cells (Brn3a and Pax-6 as seen in Figure 3 E1 and E2 respectively for transdifferentiating retina and Figure 3 F1 and F2 respectively for retina that regenerated from the CB/CMZ), amacrine cells , horizontal cells (Pax-6 positive cells in Figure 3E2 and F2), bipolar cells (Chx-10 cells in Figure 3E2 and F2) and photoreceptors (Figure 3E3 and F3). The transdifferentiated retina does not have any RPE and therefore there is no Mitf expression (Figure 3E3), but Mitf is expressed in the RPE that is present behind the retina that has regenerated from thee CB/CMZ (Figure 3F3). Müller glia patterning resembles normal

73 development (Fig 3D4, 3E4 and 3F4). At E15 all cell layers are present and organized in both regenerates (Fig 3G1-3I4).

Fig. 3. Regeneration gives rise to all cells layers of the retina. Rows: de, normal retinal development; td, retinas that arose from transdifferentiation; cr, retinas that regenerated from the CB/CMZ. Immunohistochemistry using antibodies against Brn3a (ganglion cells) and Napa73 (NFL) are shown in the first column (A1-I1). Immunohistochemistry using antibodies against Pax-6 (ganglion cells, amacrine cells and horizontal cells) and Chx-10 (bipolar cells) are shown in column 2 (A2-I2). Immunohistochemistry using antibodies against visinin (photoreceptors) and Mitf (RPE) are shown in column 3 (A3-I3). Sections in column 4 (A4-I4) were stained for vimetin to detect Müller glia. Rows A, B and C show eyes at E7, D, E and F at E11 and G, H and I at E15. (Row A) At E7, during eye development, ganglion cells are starting to differentiate and an NFL is visible (A1), Pax-6 is mostly expressed in the region where ganglion cells are present while Chx-10 is expressed through a large part of the retina (A2). Photoreceptors have started to differentiate and Mitf expression in the RPE is noticeable (A3). At this stage Müller glia are also present (A4). (Row B) At E7 transdifferentiating retina shows some Napa73 staining but no Brn3a (B1). Pax-6 is expressed throughout the retina but its expression is very prominent in the ganglion cells. There are low levels of Chx-10 at this

75 stage (B2). There is no detectable Mitf expression in the retina that arises from transdifferentiation and visinin is expressed in the presumptive photoreceptor region. (B3). Vimentin expression is limited to Müller glia that span the retina (B4). (Row C) Retina that regenerates from the CB/CMZ shows a similar pattern of cell markers as the retina that arises from transdifferentiation. There is no Brn3a expression but there is some Napa73 staining (C1), and Pax6 is expressed throughout the retina, while there is a very low level of Chx10 expression (C2). However, visinin is not present and Mitf is expressed in the intact RPE (C3). Müller glia are present at this stage (C4). (Row D) By E11, developing retina has a clearly defined NFL, GCL (D1), INL (D2) and ONL (D3). Mitf continues to be expressed in the RPE (D3) and Müller glia are present through out the retina (D4). (Row E) Transdifferentiation results in all differentiated cell types by E11. Ganglion cells (E1), amacrine cells, bipolar cells, horizontal cells (E2), photoreceptors (E3) and Müller glia (E4) are present. The orientation of this layer, however, is flipped (E1-E4). (Row F) Regeneration from the CB/CMZ also results in production of all differentiated cell types by E11. The NFL, GCL (F1), INL (F2), and ONL (F3) have all the cells types seen in a normal retina. Here, the RPE is still intact and expresses Mitf (F3). Müller glia span the retina (F4). The orientation of this layer is similar to that of developing retina. These expression patterns continue to be maintained at E15 in the developing eye (G1-G4), the transdifferentiated retina (H1-H4) and the retina that regenerated from CB/CMZ (I1-I4) Scale bar: 50 µm.

FGFR distribution during regeneration

The experiments described in previous sections have helped establish the embryonic chick as a good model to study retina regeneration. We have shown that all the different cell types of the retina can be regenerated within 7 days after retinectomy when the surgery is performed at E4 and only if FGF-2 is provided. We therefore turned our attention to understanding this developmental time point restriction that allows regeneration via transdifferentiation to occur only on E4 embryos. We hypothesized that one of the reasons that the RPE is not able to respond to FGF signals after E4 might be due to the downregulation of receptors after E4. We used immunohistochemistry to study the distribution of FGFR-1, 2 and 3 between E4 and E7. At E4, FGFR-1, the major receptor for FGF-2, does not co-localize with the RPE marker MMP115 (Figure 4A) while FGFR-2 does (Figure 4D). By E5 FGFR-1 begins to be expressed in the RPE (Figure 4B) but the level of FGFR-2 (Figure 4E) is reduced in the RPE at this stage. By E7 there is no FGFR-2 (Figure 4F) immunofluorescence in the RPE while FGFR-1 (Figure 4C) continues to be expressed in the RPE. We also studied the expression of

76 FGFR-3 (not shown) and were not able to detect it in the RPE using immunohistochemical techniques at E4 (it was present in brain, a positive control tissue for FGFR-3). These results show that FGFR-2 is the FGF receptor abundantly present in the RPE during regeneration competent stages and it may be responsible for transmitting the FGF-2 signal that induces transdifferentiation.

Figure 4. Distribution of FGFRs in the RPE. Panels A-C show the distribution of FGFR-1 (red) at E4, E5 and E7. (A) FGFR-1 is present in the retina at E4 but is not present in the RPE and does not co-localize with the RPE marker MMP115 (green). (B) At E5, FGFR-1 is highly expressed in the retina and the RPE, where it co-localizes with MMP115. (C) By E7, FGFR-1 expression is reduced and very little FGFR-1 is seen in the RPE. Panels D-F show the distribution of FGFR-2 (red) at E4, E5 and E7. (D) At E4 FGFR-2 is present in the RPE and co-localizes with MMP115. (E) By E5 the FGFR-2 signal is reduced and by E7 the FGFR-2 signal is completely gone from the RPE.

MAPKs are Involved in Transdifferentiation.

FGFR receptors generally signal through a Mitogen Activated Protein Kinase (MAPK) cascade. In order to verify if this was the signaling cascade involved in transdifferentiation of the RPE, we used immunohistochemistry to detect a phosphorylated and activated form of the MAPK, ERK (Extra-cellular signal Regulated Kinase). We studied ERK because it is the effector molecule in the MAPK cascade and is capable of influencing transcription. Retinectomies were performed and either a control heparin bead or an FGF-2 coated bead was placed in the eye. The embryos were collected after 4 hours and an anti-phospho-ERK antibody was used to detect activated ERK. In control animals, p-ERK could not be detected in the RPE (Figure 5A) but in animals treated with FGF-2, we were able to detect a strong p-ERK signal. In addition to this, work from our laboratory shows that inhibition of MEK, the MAPK just upstream of

77 ERK, reduces regeneration (Unpublished results). These data together suggest that the FGF signal that induces transdifferentiation is transmitted through a MAPK cascade.

Figure 5. FGF-2 signals through ERK. Immunohistochemical analysis shows that ERK is not phosphorylated in the RPE of control animals. However, the RPE of FGF-2 treated animals show high levels of p-ERK within 4 hours of FGF-2 treatment.

Overexpression of a constitutively active MEK

Our studies suggest that transdifferentiation during regeneration occurrs through the action of MAPKs. In addition to this, previous work had shown that overexpression of a constitutively active form the MAPK, MEK, could cause transdifferentiation of the RPE during normal development (Galy A., et al., 2002). We therefore wanted to ascertain if constitutive activation of a MAPK would be sufficient to induce regeneration via transdifferentiation in the absence of FGF-2.

In order to constitutively activate MEK, we used an RCAS virus that expressed a constitutively active form of MEK. This RCAS virus will be referred to as RCAS- MEKDD. Embryos were injected with RCAS- MEKDD on E3 so that the virus could infect the RPE and begin expressing MEK by E4 when the retinectomies were performed. None of these embryos received FGF-2 treatment. The embryos were collected on E7 or E11 and were analyzed using immunohistochemical and histological techniques. Embryos infected by RCAS-MEKDD showed areas of transdifferentiation (arrow in Figurse 6 A and 6D) that did not express the RPE marker MMP115 (Figures 6B and 6E). Analysis of adjacent sections showed that the transdifferentiating areas were infected by the virus (AMV3C2 immunoreactivity in Figure 6H corresponds to the unpigmented and transdifferentiating areas seen in Figure 6G) and that transdifferentiating areas express constitutively active MEK. Another section adjacent to the one seen in Figure 6A was

78 analyzed for transcription factors that may be regulated during transdifferentiation. As seen in Figure 6J and K, regions that are transdifferentiating, and most likely express active MEK, also express Pax-6 but the expression of Mitf is reduced and is seen only in intact RPE.

By E11, a large part of the RPE has transdifferentiated (arrow in Figure 6C) and the new retina has started to differentiate. At this stage a clearly defined ganglion cell layer that expresses Brn3a and Napa-73 (Figure 6F). The inner nuclear layer contains three distinct types of cells that express Pax-6 (amacrine and horizontal cells) and Chx-10 (bipolar cells) although this layer does not appear as organized as those seen in eyes treated with FGF-2 (compare Figure 6I to Figure 3E2). In addition, there is also a very clearly defined outer nuclear layer that expresses the photoreceptor marker visinin (Figure 6L).

These results clearly show that activation of MEK is sufficient to set off a cascade of events that causes regeneration through transdifferentiation. These results also imply that expression of MEKDD during a regeneration incompetent stage like E5 should be sufficient to induce regeneration. We therefore performed retinectomies on E5 embryos and infected them with RCAS-MEKDD and analyzed the embryos 3 days later at E8. As seen in figure 6M, expression of active MEK was sufficient to induce transdifferentiation in these regeneration incompetent embryos.

Figure 6. Overexpression of constitutively active MEKDD in the absence of FGF-2 induces transdifferentiation. Retinectomies were performed at E4 on RCAS-MEKDD infected eyes shown in panels A-L. Eyes collected at E7 (A, B, D, E, G, H, J and K) showed regions of transdifferentiation. (A) E7 eye stained with H&E shows unpigmented areas that are transdifferentiating (arrow). (B) The same section as in A shows that transdifferentiating areas do not express the RPE marker MMP115. (D) Close up of the transdifferentiating area shown in (A). (E) Area shown in (D) shows no MMP115 immunoreactivity in transdifferentiating areas. (G-H) H&E staining (G) of section adjacent to the ones shown in (D and E) shows areas that are unpigmented and undergoing transdifferentiation and these areas are immunoreactive for the viral coat protein AMV3C2 (green) in H. (J-K) Pigmented RPE still expresses Mitf whereas the transdifferentiating areas express Pax6. Eyes collected on E11 (C, F, I and L) show areas of the RPE that had transdifferentiated into retina. (C) E11 eye that was stained with H&E shows large areas where RPE has transdifferentiated into retina (arrow). (F) The regenerated retina expresses the ganglion cell marker Brn3a as well as Napa-73, a marker for ganglion cell fibers. (I) Distinct layers of Pax-6 and Chx-10 expressing cells can be also be seen at E11. (L) The regenerated retina expresses the photoreceptor marker visinin. (M) Retinectomies were performed on E5 and the eyes were then infected with MEKDD and the eyes were collected on E8 and stained with H&E, transdifferentiation can be observed in these eyes (arrow). l = lens

80 Upregulation of Pax-6 Induces Regeneration

Our immunohistochemical data (Figure 3B2 and Figure 6K) as well experiments done during development from other laboratorys show that Pax-6 upregulation is required to induce transdifferentiation (Azuma et al., 2005). To test whether Pax-6 can induce regeneration via transdifferentiation in the absence of FGF-2, we overexpressed Pax-6, using an RCAS virus, in embryos that underwent retinectomies. By E7 these embryos showed transdifferentiation with unpigmented areas (arrow in Figure 7A) that did not express MMP115 (Figure 7B). At E11 large areas of the RPE had transdifferentiated (Figure 7C) and the transdifferentiated areas were immunoreactive for AMVC32 (Figure 7G) suggesting that areas that overexpress Pax-6 had transdifferentiated (Figure 7D). These embryos were then used to access the level of differentiation in the regenerated retina using various retinal markers. We were unable to detect Brn3a or Napa-73 in these eyes (not shown) suggesting that the cells in the regenerated retina had not differentiated and they remained as progenitor cells as accessed by the co-localization of Pax-6 and Chx-10 in the cells (Figure 7E and H). We were, however, able to detect some early ganglion cells that expressed Pax-6 (Figure 7H) but not Brn3a and also photoreceptors that expressed visinin (Figure 7I).

Figure 7. Overexpression of Pax-6. Eyes were infected with RCAS-Pax-6 24 hours prior to retinectomy. (A) H&E staining of an E7 eye shows areas of the RPE that are transdifferentiating (arrow). (B) RPE that is transdifferentiating does not express MMP115. (C) H&E staining of an E11 eye shows a large area of transdifferentiation. (D-I) are sections from an E11 eye that have been analyzed using immunohistochemistry and also stained using H&E. Transdifferentiating RPE (D) has been infected with RCAS-Pax-6 and express AMV3C2 viral marker (G). The newly generated retina (E) has not differentiated completely and is composed mostly of progenitor cells that co-express Pax-6 and Chx-10 (H). However, the regenerating retina (F) expresses the photoreceptor marker visinin (I). l = lens.

Mitf Downregulation is Necessary for Transdifferentiation

We had initially hypothesized that RPE specific transcription factors such as Mitf would have to be downregulated during transdifferentiation and our immunohistochemical data seem to support this hypothesis (see Figure 3B3 and Figure 6K). Also data from other laboratories show that the downregulation of Mitf is sufficient

82 to induce transdifferentiation in vitro (Mochii et al., 1998a, Iwakiri et al., 2005). However, we do not know if Mitf downregulation is necessary to induce transdifferentiation in vivo. We therefore overexpressed Mitf in the presence of FGF-2 and determined its effects on regeneration.

Embryos were injected subretinally with RCAS-Mitf on E3 and retinectomies were performed at E4. Analysis of the embryos showed that by E7 there was very little transdifferentiation (arrows in Figure 8A and C). We also found that areas that transdifferentiated were not infected with RCAS-Mitf and the RPE that did not transdifferentiate was infected with RCAS-Mitf (Figure 8B and D). This result demonstrates that Mitf downregulation is necessary for transdifferentiation and if Mitf expression is maintained transdifferentiation cannot take place.

Figure 8. Overexpression of Mitf in the presence of FGF- 2. (A) DIC image of an E7 eye that was infected with RCAS- Mitf and that received FGF-2 post-retinectomy. There is a small region of transdifferentiation (arrow) in the posterior of the eye. (B) Immunohistochemical analysis of the section in (A) reveals that RPE that is undergoing transdifferentiation (arrow) has not been infected with RCAS- Mitf whereas RPE infected by the virus (areas that co-localize MMP115 and AMV3C2) do not transdifferentiate even in the presence of FGF-2. (C) A close up view of the transdifferentiating area in A. (D) is a close up view of the same area in B.

83 4.4 Discussion

The work in this dissertation characterizes, for the first time, the spatial and temporal nature of embryonic chick retina regeneration. We have demonstrated that the embryonic chick can regenerate its retina through two distinct modes: Transdifferentiation of the RPE and by the activation of stem/progenitor cells located in the CB/CMZ. We have also shown that FGF-2 can induce both these modes of regeneration and that all the cell types of the retina can be regenerated within 7 days. We have begun to understand molecular mechanism and signaling cascades underlying regeneration through transdifferentiation.

During transdifferentiation, the RPE, in response to FGF-2, loses its characteristic pigmented phenotype, dedifferentiates, proliferates and forms a neuroepithelium that eventually gives rise to all of the layers of the retina. However, this ability to transdifferentiate is lost after E4.5 (Coulombre and Coulombre, 1965) (data not shown) and therefore the regeneration competent stage (E3.5-4.5) represents a time in development when the RPE has begun, but not completed differentiation and as a result is very plastic. Understanding the cellular dynamics of the RPE at this time could help us reprogram completely differentiated RPE to be more plastic.

Early in development, FGF-1 and FGF-2 from the surface ectoderm act on the FGFRs in the optic cup to specify RPE and retina domains (Pittack et al., 1997; Hyer et al., 1998). Previous studies using in situ hybridization and whole eye PCRs have shown that FGFRs are present in the eye and the expression of the FGFRs are at the highest on E4 and their expression diminishes after that (Tcheng et al., 1994; Ohuchi et al., 1994). Our studies on the other hand, have shown that the only FGFR localized to the RPE at E4 and earlier (E3.5 data not shown) was FGFR-2, suggesting that FGFR-2 may be involved in the initial setting up of the RPE domain. Therefore E4 represents a time in development when the RPE domain has already been specified (by an FGF signal from the ectoderm) and has begun to differentiate into a mature RPE. Around this stage (between E2.5 and E4) RPE differentiation is directed by signals from the mesenchyme

84 that induce the expression of RPE specific genes like such as Mitf, Wnt2b and the matrix protein MMP115, and suppress retina specific genes. (Jasoni et al., 1999; Fuhrmann et al., 2000). It now appears that the downregulation of FGFR2 is a part of the RPE differentiation program and therefore as the RPE continues to differentiate FGFR2 expression is decreased and as a result the RPE becomes insensitive to FGFs. Our data therefore lays the foundation for explaining why the RPE can transdifferentiate in response to FGF-2 on E4 and not later. Though our data suggests FGFR-2 is a candidate for the transmission of FGF signals that induce transdifferentiation, functional studies will be required to prove this conclusively.

Our data suggest that two very important molecular events are required for the induction of transdifferentiation. One is the downregulation of Mitf and the other is the increased expression of Pax-6. In other systems such as the Quail, reduction of Mitf expression is sufficient to cause spontaneous transdifferentiation of the RPE into neural retina in vitro and in vivo (Mochii et al., 1998a; 1998b). Overexpression of Mitf causes a hyperpigmented phenotype in RPE cells, as well blocks the response of RPE to FGF-2 in culture (Mochii et al., 1998b). Our data are in agreement with findings in the Quail, that Mitf downregulation is necessary for the induction of transdifferentiation. However, we have not established a direct link between FGF signaling and the downregulation of Mitf and unpublished observations from our laboratory suggest that FGF signaling is not required for Mitf downregulation post-retinectomy. Future experiments will focus on understanding the mechanisms underlying Mitf regulation post-retinectomy. Though we have not uncovered the mechanisms of Mitf regulation, we have shown Mitf downregulation is essential for transdifferentiation.

In addition to the role of Mitf in transdifferentiation, we have shown that FGF-2 signals through MAPKs and overexpression of constitutively active MEK can induce regeneration even in regeneration incompetent stages. This result is in aggrement with the observations of Galy et al., (2002) who induced transdifferentiation by overexpressing active MEK during early development. In their experiments, they also showed that activation of MEK was sufficient to reduce levels of Mitf. However, in our

85 system, removal of the retina itself is sufficient to reduce Mitf levels (unpublished observations from our laboratory). In addition to a reduction in Mitf levels, we have observed an increase in Pax-6 levels in the presence of FGF-2 or active MEK. Further, we have uncovered the significance of this increased Pax-6 expression by demonstrating that overexpression of Pax-6 in the RPE, even in the absence of FGF-2, was sufficient to regenerate a neuroepithelium. Unfortunately, this neuroepithelium was developmentally challenged and did not differentiate into all the cell types of the retina. There are two possible explanations for this result: one, Pax-6 has to be downregulated for differentiation to occur and sustained Pax-6 expression by the RCAS virus maintains these cells as progenitors (as evidenced by the large number of Pax-6/Chx-10 positive cells in figure 8H); a second explanation could be a delay in the differentiation process, we cannot however, resolve this issue since animals treated with RCAS-Pax-6 virus rarely survive beyond E11.

In conclusion, this work demonstrates that the embryonic chick is a good model system to understand the two different modes of regeneration and to dissect out the major molecular players involved in the process of transdifferentiation. Our research suggests that at E4 the RPE can transdifferentiate because FGF-2 can activate FGFR-2, which in turn activates MAPKs and potentially activates Pax-6. This, in conjunction with the downregulation of Mitf, induces transdifferentiation.

4.5 References Azuma N, Tadokoro K, Asaka A, Yamada M, Yamaguchi Y, Handa H, Matsushima S, Watanabe T, Kida Y, Ogura T, Torii M, Shimamura K, Nakafuku M (2004). Transdifferentiation of the retinal pigment epithelia to the neural retina by transfer of the Pax6 transcriptional factor. Hum Mol Genet. 2005 Apr 15;14(8):1059-68

Coulombre, J. L. and Coulombre, A. J. (1965). Regeneration of neural retina from the pigmented epithelium in the chick embryo. Dev. Biol. 12, 79-92.

86 Fischer, A. J. and Reh, T. A. (2000). Identification of a proliferating marginal zone of retinal progenitors in postnatal chickens. Dev. Biol. 220, 197-210.

Fischer, A. J. and Reh, T. A. (2003). Growth factors induce neurogenesis in the ciliary body. Dev. Biol. 259, 225-240.

Fuhrmann, S., Levine, E. M. and Reh T. A. (2000). Extraocular mesenchyme patterns the optic vesicle during early eye development in the embryonic chick. Development 127, 4599-4609.

Galy, A., Neron, B., Planque, N., Saule, S. and Eychene, A. (2002). Activated MAPK/ERK Kinase (Mek-1) induces transdifferentiation of pigmented epithelium into neural retina. Dev. Biol. 248, 251-264.

Grindley, J.C., Davidson, D.R and Hill, R.E. (1995). The role of Pax-6 in eye and nasal development. Development 121: 1433-1442.

Halder G., Callaerts P., Gehring W.J. (1995). Induction of ectopic eyes by targeted expression of the eyeless gene in Drosophila. Science 267(5205):1788-92.

Haynes T. and del Rio-Tsonis, K. (2004). Retina Repair, Stem Cells and Beyond. Current Neurovascular Research 1, 231-239.

Hyer, J., Mima, T. and Mikawa, T. (1998). Fgf1 patterns the optic vesicle by directing the placement of the neural retina domain. Development 125, 869- 877.

Hitchcock, P. F., Ochocinska, M. J., Sieh, A. and Otteson D. C. (2004).Persistent and injury-induced neurogenesis in the vertebrate retina. Prog. Retin. Eye Res. 23, 183-194.

Jasoni, C., Hendrickson, A. and Roelink, H. (1999). Analysis of chicken Wnt-13 expression demonstrates coincidence with cell division in the developing eye and is consistent with a role in induction. Dev. Dyn. 215, 215-224.

Kubo F., Takeichi, M. and Nakagawa, S. (2003). Wnt2b controls retinal cell differentiation at the ciliary marginal zone. Development 130, 587-598.

Marquardt T, Ashery-Padan R, Andrejewski N, Scardigli R, Guillemot F, Gruss P. (2001). Pax6 is required for the multipotent state of retinal progenitor cells. Cell 105(1):43-55.

87 Martinez-Morales, J. R., Dolez, V., Rodrigo, I., Zaccarini, R., Leconte, L., Bovolenta, P. and Saule, S. (2003). OTX2 activates the molecular network underlying retina pigment epithelium differentiation. J. Biol. Chem. 278,21721-21731.

Mochii, M., Mazaki, Y., Mizuno, N., Hayashi H. and Eguchi, G. (1998a). Role of microphthalmia in differentiation and transdifferentiation of chicken pigmented epithelial cell. Dev. Biol. 193, 47-62.

Mochii, M., Ono, T., Matsubara, Y. and Eguchi, G. (1998b). Spontaneous transdifferentiation of quail pigmented epithelial cell is accompanied by a mutation in the Mitf gene. Dev. Biol. 196, 145-159.

Mohammadi, M., Olsen, S.K. and Ibrahimi, O.A. (2005). Structural basis for fibroblast growth factor receptor activation. Cytokine and Growth Factor Reviews 16:107-137 Ohuchi, H., Koyama, E., Myokai, F., Nohno, T., Shiraga, F., Matsuo, T., Matsuo, N., Taniguchi, S. and Noji, S. (1994). Expression patterns of two fibroblast growth factor receptor genes during early chick eye development. Exp. Eye Res. 58, 649-658.

Park, C. M. and Hollenberg, M. J. (1989). Basic fibroblast growth factor induces retinal regeneration in vivo. Dev. Biol. 134, 201-205.

Park, C. M. and Hollenberg, M. J. (1991). Induction of retinal regeneration in vivo by growth factors. Dev. Biol. 148, 322-333.

Pittack, C., Grunwald, G. B. and Reh, T. A. (1997). Fibroblast growth factors are necessary for neural retina but not pigmented epithelium differentiation in chick embryos. Development 124, 805-816.

Spence, J.R., Madhavan, M., Ewing, J.D., Jones, D.K., Lehman, B.M. and Del Rio- Tsonis, K.(2004) The hedgehog pathway is a modulator of retina regeneration. Development:131(18):4607-21.

Tcheng, M., Fuhrmann, G., Hartmann, M. P., Courtois, Y. and Jeanny, J. C. (1994). Spatial and temporal expression patterns of Fgf receptor Genes Type 1 and Type 2 in the developing chick retina. Exp. Eye Res. 58, 351-358.

Ton, C.C., Hirvonen, H, Miwa, H., Weil, M.M., Monaghan, P., Jordan, T., van Heyningen, V., Hastie, N.D., Meijers-Heijboerm, H., Drechsler, M., et al. (1991) Positional cloning and characteization of a paired box and homeobox containing gene form the aniridia region. Cell 67:1059-1074

88 Chapter 5

Discussion and Conclusion

5.1 Overview The work in this dissertation dissects some of the molecular and cellular aspects of the regeneration process. In order to put our results into the larger context of regeneration via transdifferentiation, I will divide this process into three phase: 1) the initial phase when injury is detected and information of that injury is transmitted to the surrounding tissue, 2) the second phase where signaling molecules induce transdifferentiation in differentiated cells and 3) the final phase in which developmental processes are recapitulated to rebuild lost tissues. In this chapter I will illustrate how we have been able to gain insights into each of these phases using both the newt and chick as model systems and finally discuss future directions for this research.

5.2 Phases of Transdifferentiation

In order to provide a conceptual framework to understand the many facets of regeneration, I have divided the process into three phases. These phases are not necessarily temporally distinct steps of regeneration but merely three major set of events that occur during regeneration.

Phase I: Changes at the Site of Injury

During this early phase, immediately following injury, the animal has to detect the fact that tissue has been lost (lens or lens and part of the dorsal iris), then it has to identify the location of the lost tissue (position along the proximo-distal axis of the limb) and provide this information to cells that will transdifferentiate to replace the lost tissue. Also during this phase, the environment around the wound has to be modified to allow

89 regeneration. For example, the extracellular matrix has to be changed to allow cells to rapidly divide and migrate (reviewed by Tsonis et al., 2004). In addition to this, the immune reaction at this site has to be modified to allow for the clearing of cellular debris while minimizing scarring. In the case of higher vertebrates the formation of scar tissue has been a major barrier to regeneration and animals like the newt that regenerate show reduced scarring (reviewed in Stocum, 1999). All of these processes may occur independent of each other but should occur in all tissues that can regenerate. These mechanisms might be more universal and not limited to the lens or retina or limb. Therefore studying multiple regenerating tissues in an animal like the newt will yield the most information on molecules involved in this phase of the regenerative process.

Our studies with complement components in regeneration suggests that C3 and C5 play a role in the process of regeneration and since they are involved in multiple regenerative systems including the lens, retina, limb and tail, we propose that it is involved in Phase I of regeneration. However the exact function of these molecules is unknown. Their function does not seem limited to an immunological role (such as attraction of macrophages to the site of injury) and our work suggests that they have other roles in the process of regeneration. Disruption of C3 and C5 in culture, for example using antibodies against them, perturbs engulfment in confrontation assays (Chapter 2, Figure 6), suggesting that C3 and C5 may play some role in signaling positional information. Though further experiments (discussed below) are required to pinpoint the exact role of these molecules in transmitting positional information, the work described in this dissertation clearly demonstrates that C3 and C5 have a non-immunological signaling role during regeneration.

Phase II: Induction of Transdifferentiation.

During this phase, signaling molecules start a cascade of events that cause differentiated cells near the wound to transdifferentiate. The first reaction to this transdifferentiation signal is to undergo dedifferentiation and to proliferate to give rise to a population of progenitor like cells. Our studies with the chick system provide insights

90 into the mechanisms involved in the initiation of dedifferentiation and the conversion of the RPE into a population of progenitor cells.

During retina regeneration, FGF-2 binds to a receptor in the RPE and starts a MAPK cascade. Our results indicate that FGFR2 is the receptor that binds FGF-2 in the RPE and instructs it to transdifferentiate (Chapter 4, Figure 4). However, further functional studies are required to verify if FGFR2 is indeed the receptor that is responsible for relaying the transdifferentiation signal. The FGF signal is transmitted through a MAPK cascade that activates ERK (Chapter 4, Figure 5) and activation of MEK was sufficient to induce transdifferentiation in the absence of FGF-2 even in regeneration incompetent stages (Chapter 4, Figure 6).

One of the molecules that was upregulated during transdifferentiation and a possible target of the FGF signal was Pax6. Overexpression of Pax6, in the absence of FGF-2, was sufficient to induce transdifferentiation (Chapter 4, Figure 7). This result was expected as overexpression of Pax6 in development has also been shown to cause transdifferentiation as late as E14 (Azuma et al., 2005). It is therefore safe to say that the FGF signal upregulates the expression of Pax6, which leads to conversion of RPE into a progenitor population. Additionally, it was necessary to downregulate RPE specific transcription factors such as Mitf, since overexpression of Mitf even in the presence of FGF-2, prevented transdifferentiation of the RPE (Chapter 4, Figure 8). Taken together, our data suggest that the balance between Pax6 and Mitf steers the cell toward an RPE or retinal progenitor fate.

Phase III: Recapitulation of Development

Once a progenitor cell population has been derived from dedifferentiated cells, they can recapitulate developmental processes that can regenerate the lost tissue. Studying phase III will not only help us understand regeneration but will also shed light on developmental processes and the interaction of various molecules during development.

Experiments by Grogg et al. (2005) have already shown that expression of the “master gene” Pax-6 was not sufficient to induce lens regeneration, even though it is expressed in the iris during lens regeneration (Del Rio Tsonis et al., 1995) and perhaps performs a secondary function. We therefore performed experiments that disrupted Pax- 6 expression during lens regeneration and discovered that it controlled proliferation and lens fiber differentiation. These are functions associated with Pax-6 during lens development (Walther and Gruss, 1991; Grindley et al., 1995; Richardson et al., 1995; Duncan et al., 1998; Zhang et al., 2001; Van Raamsdonk and Tilgham, 2000; for review see Cvekl et al., 2004) and are clearly replicated during the regeneration process. These experiments are important because they demonstrate that animals like the newt may use novel methods of activating the regeneration program but they use evolutionarily conserved developmental pathways to regrow lost tissue.

5.3 Questions of Immediate Interest

Establishing the role of Complement Components in Regeneration

The evidence thus far suggests that complement components may have some sort of signaling role during regeneration. Results of the confrontation assay suggest that complement components may be relaying positional information, but it remains unclear if C3 and C5 expression label the blastema as proximal or distal. In order to address this we need to treat regenerating blastemas with recombinant C3 and/or C5 and then use the confrontation assay to determine if their proximal/distal identities have been changed. For example, if expression of C3 marks a blastema as proximal and the lack of C3 marks it as distal, then overexpressing C3 in a distal blastema will make it behave like a proximal blastema. As a result, this distal blastema treated with C3 will engulf another untreated distal blastema. Results from experiments such as this will provide a better understanding of the signaling role of complement components in regeneration.

92 In addition to these experiments, we can attempt to knock down C3 and C5 during lens regeneration to study their specific roles. During the course of this research, we have established the use of morpholino knock-down technology in the newt and this technology can now be applied to studying the roles of various molecules during lens regeneration. Morpholino oligos can be used to knock down levels of either C3 or C5 and the effect of this knock down could uncover the functions of C3 and C5 during lens regeneration.

FGFR-2 and Transdifferentiation of the RPE.

Though we have shown that FGFR-2 is present in the RPE at E4, we do not know which FGFR-2 isoform may be responsible for this restriction. RT-PCR analysis using primers specific for FGFR-2 isoforms will have to be used to determine which isoforms are present in the RPE and responsible for inducing transdifferentiation.

Additionally, we will need functional data that shows that activation of FGFR-2 by FGF-2 induces regeneration. We will have to misexpress or overexpress the correct FGFR-2 isoform in the RPE at E5 or later and then assay its ability to induce regeneration under the influence of FGF-2. However, we are not currently aware of an overexpression system that is readily available for FGFR-2 and therefore we will have to either make an RCAS-FGFR-2 construct or electroporate an expression vector containing FGFR-2 into the RPE and consequently assay if the RPE can respond to FGF-2.

5.4 Future Directions

Identifying Phase I molecules As we have seen from the work in this dissertation, molecules expressed in all multiple regenerative processes (lens, retina, limb and tail) may hold the keys to understanding the amazing regenerative capability of the newt. Therefore an attempt must be made to identify such molecules. Comparison of gene expression from various early stage regenerating tissues could be used to help identify genes expressed in multiple

93 regenerative processes. Housekeeping genes that might be identified during this screen can be eliminated by using the respective uninjured tissue as controls. At this time however, there are no commercially available newt microarrays and therefore we might try probing a microarray from another amphibian such as Xenopus. Xenopus laevis would be a good model for this comparision because pre-metamorphic stages can also regenerate retina, lens, tail and limb. If this strategy does not yield satisfactory results we might have to make our own microarray using newt cDNA libraries.

Global Changes in the RPE during Retina Regeneration in the Chick Even though we have implicated Pax-6 and Mitf as major players in the process of transdifferentiation, we do not know the identities of other molecules that are modulated during transdifferentiation. In order to get a picture of the global changes occurring in the RPE during transdifferentiation, a microarray analysis of tissues from different stages of regeneration should be performed. Laser Capture Microdissection could be used to dissect tissue in various stages of regeneration. As shown in figure 1, regions of intact RPE, regions that are depigmenting and regions that are forming a neuroepithelium could be dissected out and then used for microarray analysis. Comparison of these tissues would give us a comprehensive view of the changes that lead to conversion of a pigmented RPE into a depigmented layer and eventually into a neuroepithelium.

Figure 1: Scheme for Laser Capture Dissection (LCM) and Microarray Analysis. LCM can be used to dissect out intact RPE(I), Depigmenting RPE (II), Depigmented epithelium(III) and neuroepithelium (IV) and RNA from these tissues can used for microarray analysis.

5.4 References

Azuma N, Tadokoro K, Asaka A, Yamada M, Yamaguchi Y, Handa H, Matsushima S, Watanabe T, Kida Y, Ogura T, Torii M, Shimamura K, Nakafuku M (2004). Transdifferentiation of the retinal pigment epithelia to the neural retina by transfer of the Pax6 transcriptional factor. Hum. Mol, Genet. 2005 Apr 15;14(8):1059-68

Cvekl,A., Yang, Y., Chauhan, B.K., Cveklova, K. (2004) . Regulation of gene expression by Pax6 in ocular cells: a case of tissue-preferred expression of crystallins in lens. Int J Dev Biol. 2004;48(8-9):829-44.

Del Rio-Tsonis, K., Washabaugh, C.H. and Tsonis, P.A. (1995). Expression of Pax-6 during urodele eye development and lens regeneration. Proc. Natl. Acad. Sci. USA 23;92(11):5092-6.

Duncan, M.K., Haynes, J.I., Cvekl, A. and Piatigorsky, J. (1998). Dual roles for Pax-6: a transcriptional repressor of lens fiber cell-specific β-crystallin genes. Mol. Cell. Biol. 18: 5579-5586.

Grindley, J.C., Davidson, D.R and Hill, R.E. (1995). The role of Pax-6 in eye and nasal development. Development 121: 1433-1442.

Richardson, J., Cvekl, A. and Wistow, G. (1995). Pax-6 is essential for lens specific expression of ζ-crystallin. Proc. Natl. Acad. Sci. USA 92: 4676-4680.

Stocum ,D.L.(1999) Limb regeneration: re-entering the cell cycle. Curr Biol. 1999 Sep 9;9(17):R644-6.

Tsonis PA, Madhavan M, Tancous EE, Del Rio-Tsonis K (2004). A newt's eye view of lens regeneration. Int J Dev Biol. 48(8-9):975-80.

Van Raamsdonk, C.D. and Tilghman, S.M. (2000). Dosage requirement and allelic expression of PAX6 during lens placode formation. Development 127: 5439-5448.\

Walther, C. and Gruss, P. (1991). Pax-6, a murine paired box gene, is expressed in the developing CNS. Development 113: 1435-1449.

Zhang, W., Cveklova, K., Oppermann, B., Kantorow, M. and Cvekl, A. (2001). Quantitation of PAX6 and PAX6(5A) transcript levels on adult human lens, cornea and monkey retina. Mol. Vis. 7: 1-5