1 ARRESTIN TRANSLOCATION IN ROD PHOTORECEPTORS: THE SIGNALING CASCADE AND THE MECHANISMS By WILDA ORISME A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008
2 2008 Wilda Orisme
3 To my Mother and Father, Lucienne and Gesner Orisme
4 ACKNOWLEDGMENTS I would like to thank Dr. W. Clay Smith fo r being a wonderful mentor and advisor. I appreciate the long discussions a nd invaluable advice for m y progres sion as a scientist. For the many times experiments did not work for me, it was always great to see a reassuring smile that there are better days ahead. I would like to thank Dr. J. H ugh McDowell for being an awesome scientist and teaching me to be more critical. I would like to thank Charles Shelemar (may he rest in peace) for the endless laughs reminding me life is better enjoyed with bliss. I would like to thank my committee members Dr. Sue Semple-R owland, Dr. Barbara Battelle, and Dr. Stephen Sugrue for providing direction a nd support for my thesis researc h. I would like to thank my lab mates Donald Dugger, Dr. Susan Bolch, Dr. Jian Li, Dr. Paul Hargrave, Dr. Robert Cohen, and Dr. Anatol Adrendt. It wa s great to be part of one of the be st research groups. I have learned beyond my field through our discussi ons. I would like to thank my sister and brothers, Carole Theophin, Smith Orisme, Charlemagne Orisme, Eb en Orisme, and Lequin Orisme. I am blessed and grateful to be your sister. You have taught me what excellence is and how to continue to achieve it. Among the many things that I have learned from you, you have taught me success is not a measure of degree but a measure of influence. I would like to thank my mother and father for the many lessons you taught me, one of which is to maximize from the minimum of any situation. I dedicate th is milestone in my life to my mother and father, Lucienne and Gesner Orisme.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES.........................................................................................................................8 ABSTRACT.....................................................................................................................................9 CHAPTER 1 INTRODUCTION AND BACKGROUND........................................................................... 11 Introduction................................................................................................................... ..........11 Phototransduction............................................................................................................ 11 Inactivating the Photoresponse........................................................................................ 11 Arrestin Translocation.....................................................................................................12 Phosphoinositides in th e Photoreceptor Cell................................................................... 14 G-Proteins in P hotoreceptor Cell.....................................................................................15 2 THE SIGNALING CASCADE.............................................................................................. 19 Materials and Methods...........................................................................................................19 Immunocytochemistry..................................................................................................... 19 Treatment of Tadpoles and Mouse Organ Cultures with Agonists and Antagonists...... 21 Confocal Microscopy and Quantification .......................................................................23 Results.....................................................................................................................................24 Arrestin Translocates in Respons e to Activators of PKC and PLC................................ 24 Inhibitors of PKC and PLC Modify Arrestin Localization............................................. 25 A G-protein Mediates Arrest in Translocation to OS....................................................... 26 3 THE MECHANISM OF ARRESTIN TRANSLOCATION .................................................. 40 Materials and Methods...........................................................................................................40 Construction of Arrestin cDNAs....................................................................................40 Generation of Transgenic Animals .................................................................................41 Expression in Tissue Culture........................................................................................... 42 Heterologous Expression and Purification of Arrestin.................................................... 42 Rhodopsin Binding Assay. ..............................................................................................43 Microtubule Binding Assay.............................................................................................43 Potassium Cyanide Treatment of Tadpole Eyes.............................................................. 44 Results.....................................................................................................................................45 Arrestin Binding to Microtubules Is Indepe ndent of Charge Order in the C-term inus... 46 Arrestin Translocation in ATP-depleted Photoreceptors................................................ 47
6 4 DISCUSSION.........................................................................................................................57 G-proteins in Rod Photoreceptors..........................................................................................57 Role of PKC and PLC.............................................................................................................59 Arrestin Translocation is Energy Dependent..........................................................................60 Proposed Pathway...................................................................................................................62 5 SUMMARY AND PERSPECTIVE.......................................................................................67 LIST OF REFERENCES...............................................................................................................70 BIOGRAPHICAL SKETCH.........................................................................................................77
7 LIST OF TABLES Table page 3-1 Synthetic oligonucleotide primers used to construct arrestin with a scrambled Cterminus....................................................................................................................... .......41
8 LIST OF FIGURES Figure page 1-1 Rhodopsin inactivation pathway........................................................................................ 17 1-2 Arrestin localization in rod photoreceptors........................................................................18 2-1 Localization of phospholipase C and PKC in rod photoreceptors...................................... 29 2-2 Arrestin translocation to the OS is induced by PKC and PLC agonists. ........................... 30 2-3 Arrestin translocation is slowed in eyes treated wi th PKC and PLC inhibitors. ............... 32 2-4 Localization of G-proteins in rod photoreceptors. .............................................................34 2-5 Arrestin translocation is sensitive to pertussis toxi n ..........................................................35 2-6 D2-dopamine receptor localizati on in rod photoreceptors................................................. 36 2-7 Arrestin localization in response to a dopamine receptor agonist..................................... 37 2-8 Arrestin localization in response to a D2-dopamine receptor antagonist ..........................38 3-1 Binding of arrestins to rhodops in in rod disc membranes................................................. 49 3-2 Translocation of xAr and xA r-scr in transgenic tadpoles.................................................. 50 3-3 Native a rrestin and arrestin with a scra mbled c-terminus bind fluorescently labeled mi crotubules ......................................................................................................................51 3-4 Arrestin and tubulin lo calization in tissue culture ............................................................52 3-5 Arrestin localization in response to ATP depletion in transgenic tadpoles .......................53 3-6 Arrestin localization in respons e to ATP depletion in m ouse eyes ..................................55 4-1 Proposed phosphoinositide gatekeeper pathway for arrestin translocation....................65 4-2 Model for arrestin translocation......................................................................................... 66
9 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ARRESTIN TRANSLOCATION IN ROD PHOTORECEPTORS: THE SIGNALING CASCADE AND THE MECHANISMS By Wilda Orisme December 2008 Chair: W. Clay Smith Major: Medical SciencesNeuroscience Light adaptation of rod photoreceptors induces an accumulation of arrestin from the rod inner segments (IS) to the rod outer segments (OS) where arrestin binds to light-activated, phosphorylated rhodopsin to quench the photorespons e. The signaling molecules necessary to initiate arrestin translocation to the OS and th e mechanisms for arrestin translocation have not been fully elucidated. Here, for the first time, we demonstrate that components of the G-protein linked phosphoinositide pathway play a role in stimul ating arrestin translocation. First, we show that arrestin transloc ation can be stimulated by activators of phospholipas e C (PLC) and protein kinase C (PKC) in the absence of light. Surprisingly, arrestin translocates to the OS within 10-15 min after exposure to activators of PLC and PKC. Conversely, arrestin translocation to the OS is significantly slowed by inhi bitors of PLC and PKC, reducing th e amount of arrestin translocating to the OS. Second, we show that arrestin transl ocation to the OS is significantly reduced in photoreceptors that have been treated with pertussis toxin, which is a G-protein poison. Collectively, our results suggest that arrestin transl ocation is initiated by a G-protein coupled cascade through PLC and PKC signaling. In the second part of this study, we inve stigated the mechanism by which arrestin translocates in response to light Previous studies have suggest ed that arrestin translocation
10 between the OS and IS is a passive process, resu lting from arrestins binding affinity for lightactivated, phosphorylated rhodopsin in the outer segments and affinity for microtubules in the inner segments. The central tenet of this model is that arrestins translocation is an energy independent process. We have previously shown that arrestin does not fu lly distribute throughout the OS in response to microtubule poisons nor do es arrestin return back to the IS under these conditions, demonstrating there is a role for cyto skeleletal elements in arrestin translocation between the OS and IS. In this study, we addr ess the discrepancies between these findings regarding the mechanism for arrestin translocation. We find that trea tment with potassium cyanide inhibits arrestin transl ocation to the OS in response to light, suggesting that arrestin translocation does require ATP. We also found that we could e ngineer an arrestin with a scrambled C-terminal 30 amino acids that retain ed its binding for both activated rhodopsin and microtubules, but yet was unable to translocate in response to light The results obtained from both investigating the signaling cascade and the m echanisms of arrestin translocation indicate that arrestin transloca tion between the IS and OS is more complicated than previously proposed, and likely involves both diffusion and motor-assisted processes.
11 CHAPTER 1 INTRODUCTION AND BACKGROUND Introduction Phototransduction The phototransduction cascade is the process by which light is converted into a change in photoreceptor cell membrane potential. This cascad e is initiated by the li ght-sensitive G-protein coupled receptor rhodopsin. The inactive form of the chromophore in rhodopsin isomerizes from 11cis retinal to alltrans retinal in response to light. Th is isomerization and accompanying conformational changes of rhodopsin promotes activation of the Gprotein transducin. The alpha subunit of GDP-bound transducin then binds to light-activated rhodopsin and exchanges GTP for GDP. As a result of transducin activation, phosphodiesterase cleaves cGMP, decreases the intracellular concentration of cG MP, and results in the closure of the cyclic nucleotide-gated Na+/K+/Ca2+ channels. The closure of these channels cau ses a change in the ionic balance of the photoreceptor, resulting in hyperpol arization of the cell and a neural response (Burns and Baylor 2001). Inactivating the Photoresponse The phototransduction cascade is inactivated at several key points. Guanylate cyclase synthesizes cGMP to reopen the Na+/K+/Ca2+ channels. RGS9 facilitates GTP hydrolysis of activated transducin restori ng inactive GDP-bound transducin. Rhodopsin kinase, a G-protein coupled receptor kinase, phosphoryl ates light-activated rhodopsin. Arrestin then binds to lightactivated, phosphorylated rhodopsin and prevents transducin from binding rhodopsin, thereby quenching the photoresponse (Fig. 1-1). Arrestin is a 48 kDa protein whose localization in the photor eceptor cell is dependent upon the lighting conditions. In dark-adapted eyes, a rrestin localizes primarily in the rod inner
12 segments. In response to light, ar restin translocates to the r od outer segments (OS) passing through the connecting cilium (Fig.1-2). Since arrestin plays an important role in the inactivation of the photoresponse by binding to light-activated, phosphorylated rhodopsin, it would be intuitive for arrestin to be lo calized in the rod outer segments. The fact that arrestin is differentially partitioned and moves in response to light suggest an additional function for arrestin, perhaps in light adaptation (Strissel et al. 2006) or in energy conservation (McGinnis et al. 1992). Arrestin Translocation This current study investigates arrestin transl ocation to the outer segments, examining both the m echanism by which arrestin moves and th e signaling pathway that initiates arrestin translocation. Currently, there are two proposed mechanisms for arrestin transl ocation: (1) passive distribution of arrestin between the OS and IS through binding partner interactions (Peet et al. 2004, Nair et al. 2 005, Strissel et al. 2006), and (2) the active use of cytoskeletal elements for arrestin translocation (McG innis et al. 2002, Peterson et al 2005, Reidel et al. 2008). The model proposing passive distributi on of arrestin suggest s arrestin transloca tion to the OS is dependent upon arrestins binding affinity for light-activated, phosphorylat ed rhodopsin and that relocalization to the IS during da rk-adaptation is dependent upon arrestins binding affinity for microtubules (Nair et al. 2004, Na ir et al. 2005, Hanson et al. 2007, Nair et al. 2007). In support of this hypothesis, Nair et al. (2005) showed that arrestin can translocat e to the OS in mouse retinas depleted of ATP by potassium cyanide tr eatment. This evidence suggests arrestin can translocate to the OS in an energy independent manner in their mammalian system. Other studies have investigated potential contributions to arre stin translocation by cytoskeletal elements in a more active process (Williams et al. 1988, McGinnis et al. 2002, Peterson et al. 2005, Reidel et al. 2008). In light-adapted rod photoreceptors treated with thiabendazole to disrupt
13 microtubules, arrestin translocated only to th e base of the OS, as opposed to being fully distributed throughout the OS (Peterson et al. 20 05, Reidel et al. 2008). Further, disruption of microfilaments also caused a delay in the return of arrestin to the IS during dark adaptation in mouse (Peterson et al. 2005, Reidel et al. 2008). Collectively, thes e two studies both showed an effect on arrestin translocation when cytoskeletal elements are disrupted, suggesting an important contribution is made by the cytoskel eton for arrestin translocation. In photoreceptors, the principal cytoskeletal elements are localized along the axoneme and in the inner segments. Microtubules are the prim ary component of the axoneme, projecting from the inner segments basal bodies as nine ciliary do ublets and then as single t microtubules into the distal outer segments. Microfilaments are al so found along the axoneme, and in the calycal processes which surround the outer segments (Chait in et al. 1986, Williams et al. 1988, Lin et al. 2004). As expected, there are motor proteins associ ated with these cytoskel etal elements, namely myosin, kinesin, and dynein. These motor proteins have been str ongly implicated in transporting proteins synthesized in the IS to the OS, such as opsin, and are important in disc and membrane morphogenesis. There is a small body of evidence suggesting participation of these motors in arrestin translocation. For example, in conditional knock-outs of KIF3A, the motor domain of kinesin III, arrestin translocation to the OS was prevented, primarily localizing in the IS (Marszalak et al. 2000). However, it is not possible to conclude th at arrestin uses kinesin II for transport from this study since there was also an accumulation of phosphorylated opsin in the IS in the KIF3A knockout mice. As a result the arres tin may have localized to the IS due to its affinity for phosphorylated opsin. A role for myosin motors in arrestin translocation has also been demonstrated in invertebrates. Lee and Montel l (2004) demonstrated that Drosophila arrestin co-precipitated
14 with Myosin III (NINAC) in the presence of phosphoinositides. Although Lee and Montell found no direct protein-protein in teraction between myosin and a rrestin, their findings suggested that arrestin may use a myosin/microfilament motor process for tran sportation. In support of this idea, flies lacking NINAC showed significant defect s in the light-driven tran slocation of arrestin. Although Myosin IIIA and arres tin co-localize in vertebrate photoreceptors, arrestin translocation in vertebrate organisms appears to use a mechanism different than that of Drosophila since vertebrate visual arrestins lack affinity for the phosphoinositides that are required for its association with My osin IIIA (Smith et al. 2006) Nonetheless, the presence of Myosin IIIA in the distal inner segments of Xenopus laevis photoreceptors (Lin et al. 2004) and its co-localization with arrestin in this region are intriguing. Phosphoinositides in the Photoreceptor Cell In addition to our incomplete understa nding of the mechanism by which arrestin translocates, there is also a gap in our knowledge of how light triggers arrestin translocation. In 2003, Mendez et al. showed transducin signaling is not necessary for arres tin translocation using a transducin knock-out m ouse. They further demonstrated that arrestin coul d also translocate in rhodopsin kinase knock-out animals, adding additiona l evidence that transl ocation of arrestin was not dependent on arrestin binding to light-a ctivated phosphorylated rhodopsin (Mendez et al. 2003). However, in animals that lacked the ability to form 11cis retinal, such as the RPE65 knockout mouse, arrestin translocation was abse nt, demonstrating a requirement for rhodopsin, and presumably rhodopsin activation, to initia te translocation (Mendez et al. 2003). More recently, Strissel et al. (2006) have shown that arre stin translocation does not occur at very low levels of light, but rather is initiated at a threshold where approximately 3% of the total molecules of rhodopsin are bleached. At this thre shold, a 30-fold excess of arrestin moves into the outer segments compared to the number of activated rhodopsin molecules. This excess of
15 arrestin moving at a threshold level is a hallmark of a signaling cascade activating the translocation of arrestin. Si nce rhodopsin is required for arrestin transl ocation, but not transducin, we reasoned that th e signaling for arrestin tran slocation likely occurs through rhodopsin activation of an altern ative G-protein coupled cascade. Several components of such cascades have been identified in the photoreceptor, many of which do not have a known function. G-Proteins in Photoreceptor Cell To identify the possible molecules associated with arrestin translocation, we sought molecules that are expressed in the rod photoreceptor, but have not specifically been linked to any pathway. The phosphoinositide pathway is one logical choice since many studies have identified components of the phos phoinositide pathway in photorecep tor cells (e.g., Kaplan et al. 1987 and Newton et al. 1993). The phosphoi nositide pathway consists of a Gi or Gq guanine nucleotide-binding protein which activate s phospholipase C (PLC). PLC cleaves phosphatidylinositol 4, 5 bis-phosphates (PIP2) to produce diacylglycerol (DAG) and inositol trisphosphate (IP3). IP3 activates calcium channels for calcium release in the cell. Free calcium and DAG can activate conventional PKCs to phosphorylate various target proteins in the cell (Newton et al. 1995). Heterotrimeric G-proteins, including G11 and Gi, are activated through a GDP to GTP exchange, activating its effect or protein PLC to cleave PIP2. There are several isozymes of PLC that have been identified in the photorecepto r: beta1-4, gamma1-2 and theta1-2. Both PLC 4 and PLC 1 have been identified in the OS with a possible role in the visual resp onse (Ghalayini et al. 1992, Ghalayini et al. 1998, Jiang et al. 1998). PLC 4 is suggested to modulate both the visual response in mice and cell to cell signali ng (Jiang et al. 1998). The role of PLC 1 in the
16 photoreceptor has been shown to differ depending on lighting conditions, with greater activity in light-adapted OS (69%) compared to dark-adapted OS (46%) (Ghalayini et al. 1992, Ghalayini et al. 1998). Interestingly, PLC 1 activity was found to be dependent on arrestin concentration and calcium added to the OS preparations (Ghalayini et al. 1992). Like the PLCs, the rod photoreceptor cell also expresses a wide range of protein kinase C isozymes. The protein kinase C family consists of the conventional protein kinase Cs, the novel protein kinase Cs, and the atypi cal protein kinase Cs. The c onventional PKCs require both Ca2+ and DAG for activation, whereas novel PKCs requi res only DAG. Atypical PKCs do not require Ca2+ or DAG. PKC isozymes that have been identif ied in the OS are alpha, beta, gamma, theta, and epsilon (Kelleher et al 1985, Newton et al. 1993, Greene et al. 1995, Kosaka et al. 1998, Udovichenko et al. 1996, Udovichenko et al. 1997, Williams et al. 1997). Among the PKC isozymes, PKC accounts for at least 40% of Ca2+-dependent PKC activity in the OS (Williams et al. 1997). These studies highlight the abundance of PLC and PKC isoforms in rod photoreceptors. Yet in most cases, but their role s are not fully known. Given the evidence provided in some of the studies indicating th at lighting conditions modulate PLC a nd PKC activity, it is reasonable to hypothesize that PLC and PKC may be coupled to rhodopsin activa tion, and that this coupling could be part of the signal that triggers arrestin transl ocation. In the first part of this study, we test this hypothesis, investigating the potential for phosphoinositide signaling as the initiator for arrestin translocation.
17 R h R* R* P P P ATP(n)ADP(n) RK R* P P P ARR ARR all-trans retinol P P P P P P opsin 11-cis retinal opsin ARR T* T RDH OH PP2A o Figure 1-1. Rhodopsin inactivation pathway. In re sponse to light, the ch romophore 11-cis retinal isomerizes to all-trans retinal therefore activating rhodopsin (R). Light-activated rhodopsin (R*) activates heterotrimeric G-prot ein, transducin (T-alpha) to initiate the phototransduction cascade. To quench th is response, rhodopsin kinase (RK) phosphorylates light-activated r hodopsin creating a high affinity complex for arrestin binding. Arrestin (ARR) binds to light-a ctivated phosphorylated rhodopsin (R*P), creating a stearic hindrance for a transducin -rhodopsin complex. All-trans retinal is released from rhodopsin and reduced to all-trans retinol by retinol dehydrogenase (RDH). The release of chromophore is coupled with the release of arrestin. After the release of all-trans retinol, which generates phosphorylated opsin, phosphatase 2A (PP2A) cleaves the phosphates from phos phorylated opsin. Opsin can then be regenerated to rhodopsin with th e addition of 11-cis retinal.
18 IS OS Dark Light Figure 1-2. Arrestin localization in rod photoreceptors. Arrestin is one of the few proteins in the rod photoreceptors that undergo a change in localization in response to light. Xenopus laevis tadpoles were dark adapted for at least 12 hrs or light adapted for 60 min, and cryosections immunostained for arrestin (xAr 1-6, red); nuclei were stained with Sytox green. In dark-adapted rod photorecept ors, arrestin is predominately localized in the inner segments and along the axoneme (arrows). In response to light, arrestin is predominantly localized to the rod outer segments.
19 CHAPTER 2 THE SIGNALING CASCADE In the first part of this study, we investigated the proteins that may pl ay a role in signaling arrestin translocation. We used imm unocytochemistry for detecti ng various proteins in the rod photoreceptors to determine th eir localization. We also used pharmacological agents for agonizing and antagonizing components of the phosphoinositide pathway. Materials and Methods Immunocytochemistry Wild-type Xenopus laevis tadpoles (stages 50-54, either lab reared or obtained from Xenopus Express) were dark adapted overnight in th e dark or under dim red lights. For darkand light-adaptation studies, the tadpoles were either left in the dark or were light adapted for 60 min under laboratory lighting (approximately 850 lux). After dark adaptation or light adaptation, the tadpoles were fixed in 3.7% formaldehyde and 73 % methanol in deionized water overnight at 4oC. The tadpoles were rehydrated through serial dilutions of me thanol incubating for 30 min on ice in 60% methanol/40% phosphate buffered saline (PBS), 40% methanol/60% PBS, 20% methanol/80% PBS, and 100% PBS. After rehydr ation, the tadpoles were cryoprotected in 30% sucrose in PBS overnight at 4oC. The eyes were dissected from the tadpoles and embedded in Optimal Cutting Temperature media (OCT) and s ectioned at 12 m. For immunocytochemistry, the sections were rinsed with PBS for 30 min to remove residual OCT and processed through the following series of treatments to optimize antibo dy penetration. The sections were incubated in freshly prepared 0.1% NaBH4 for 30 min at ro om temperature, followed by 1% Trton-X100 in PBS for 30 min. The sections were then denatu red with 6 M guanidinium hydrochloric acid in 50 mM Na2PO4, pH 7.0 for 20 min. The sections were rinsed with several changes of deionized water, and then blocked with 1% reduced gamm a globulin fetal bovine se rum or reduced gamma
20 globulin horse serum/ 0.2% Triton-X100 in PBS for 2 hr. All antibodies used in this study were diluted in 1% reduced gamma globulin horse or fetal bovine serum with 0.2%Triton-X100 in PBS. The following antibodies a nd dilutions were used: antiXenopus arrestin [xAr1-6: mouse monoclonal (Peterson et al. 2003) 1:50], anti-arrestin [SCT-128 mouse monoclonal; gift from Paul Hargrave (1:50)], anti-Gi-1 G-protein (R4) [mouse monocl onal (1:100 Santa Cruz, #sc13522)], anti-Gi-2 G-protein (L5) [mouse monoclonal (1:100, Santa Cruz, #sc-13534)], anti-Gi-3 G-protein (C-10) [rabbit polycl onal (1:100 Santa Cruz, #sc-262) ], anti-Gi-o G-protein (A2) [mouse monoclonal (1:100 Sant a Cruz, #sc-13532)], anti-G 11 G-protein (D-17) [rabbit polyclonal (1:100 Santa Cruz, #sc-394)], anti-transducin alpha G-protein (K-20) rabbit polyclonal (1:100 Santa Cruz, #sc-389)], anti-phospholipase C (PLC)1 (1249) [rabbit polyclonal (1:100 Santa Cruz, #sc-81)], anti-PLC 4 (C-18) [rabbit polyclonal (1:100 Santa Cruz, #sc-404)], antiprotein kinase C(A9) [mouse monoclonal (1:100 Santa Cruz, #sc-17804), and D2DR (H-50) rabbit polyclonal (1:100 Santa Cruz #sc-9113)]. The sections were incubated with the primary antibody for 18-24 hr at room temperature in a hydrated chamber with slow rotation. After incubation, the sections were washed with thr ee changes of PBS (30 min each). The secondary antibody was added to the sections and incubated for 18-24 hr. In this study, we used both Texas Red-X goat anti-mouse (Invitrogen, #T6390) a nd Texas Red-X goat an ti-rabbit (Invitrogen, #T6391) at 1:200 dilution. (These are shown in the images using the red channel.). We also used Alexa Fluor 647 goat anti-rabb it (Invitrogen, A21245) seconda ry antibody, which also fluoresces red, but at a longe r wavelength. To make the di stinction between the two fluorophores, we show Alexa Fluor 647 fluorescence using the blue channel. Two nuclear stains were used in this study and were incubate d with the secondary antibody: 0.2 M SYTOX Green (Invitrogen, shown in green) and 4', 6diamidino-2-phenylindole (DAPI) (Invitrogen,
21 shown in blue). After the sec ondary antibody incubation, the sect ions were washed with PBS with three changes at 30 min in tervals. They were then covere d with Mowiol (Peterson et al. 2003) to reduce autofluorescence and seal ed with a cover slip (No. 0). Treatment of Tadpoles and Mouse Organ Cu ltures with Agonists and Antagonists Arrestin-GFP tadpoles (4-6 weeks) were dark adapted overnight in 0.1x TPR (Peterson et al. 2003). Arrestin-GFP tadpoles were produced by m a ting a male expressing arrestin fused with GFP at the C-terminus driven by the rod opsin promoter (Peterson et al. 2003). The following day, dark-adapted arrestin-GFP tadpoles were tr eated with 1 M phorbol 12, 13-diacetate (PDA) (Sigma, #P9143) for 10 min in the dark. The PDA was prepared at 30 mM in 33% dimethylsulfoxide, and diluted in tadpole Ringers such that the final DMSO concentration was 3.3%. PDA is a diacylglycerol analog that has be en shown to activate protein kinase C (Martin et al. 1999). The tadpoles were then fixed in 3.7% formaldehyde in 73% methanol immediately after treatment. After the tadpoles were in fixative overnight, they were re hydrated through serial rinses of PBS and cryprotected in 30% sucros e in PBS overnight as described above. Mouse organ cultures were treated with a similar phorbol ester, phorbol 12-myristate 13-acetate (PMA). The mouse organ cultures were prepared by trea ting C57/BL6J mice with 1.2 mg/ml proteinase K for 15 min at 37oC. The reaction of the proteinase K was quenched by adding Dulbeccos Modified Eagles Medium (D MEM) containing 10% fetal calf serum for 5 min at 37oC. The eyes were thoroughly rinsed w ith serum-free media before carefully removing excess muscle tissue, the cornea, and the sclera and the hayloi d vessels (Reidel et al. 2006) around the eye fully exposing the retina. The retina was then cultu red in DMEM supplemented with F12 and 10% fetal calf serum, l-glutamine, penicillin, and streptomycin at 37C with 5% CO2. The cultures were dark adapted for 12 hr and then trea ted with 100 nM PMA for 10 min. The eyes were
22 immediately fixed with 4% paraformaledhyde in Soerensens phosphate buffer and prepared for immunocychemistry. Dark-adapted arrestin-GFP tadpoles and retina l organ cultures were both treated with a phospholipase C activator, 2,4,6-trimethyl-N(m-3 -trifluoromethylphenyl) benzenesulfonamide (m-3M3FBS, Calbiochem, #525185). Arrestin-GFP tadpole eyes were placed in Niu-Twitty buffer, (310 M Na2HPO4, 150 M KH2PO4, 58 mM NaCl, 670 M KCl, 340 M Ca(NO3)2, 830 M MgSO4, 2.4 mM NaHCO3, 340 M CaCl2 ,pH7.3) and then treated with 1 M m3M3FBS (3.3% DMSO final concentration) for 15 min in the dark. The eyes were transferred to a new tube and fixed immediately after incubation. The eyes were rehydrated and cryoprotected as described above. The retina l organ cultures were treated with m-3M3FBS for 20 min and fixed in 4% paraformaledhyde in Soerensens ph osphate buffer. After fixation, the eyes were sectioned and prepared for immunohistochemistry. Arrestin-GFP tadpoles were also treated with a dopamine receptor agonist. Following overnight dark adaptation in 0.1x tadpole ringers overnight, the tadpoles were treated with 10 M (-)-quinpirole hydrochloride (Sigma, #Q102), a D2-dompamine receptor agonist, for 10 and 30 min. The control tadpoles and quinpirole-treated tadpoles were fixed immediately after each time point. Chelerythrine and U73122 (1-(6-(17 -3-methoxyestra-1, 3,5(10)-trien-17-yl) amino) hexyl)-1H-pyrrole-2, 5-dione) are inhibitors of PKC and PLC, respectively. Dark-adapted arrestin-GFP eyes and mouse retin al organ cultures were treated with 10 M chelerythrine chloride (Calbiochem, #220285) and 10 M U73122 (C albiochem, #662035) for 4 hr in the dark. For arrestin-GFP eyes, the eyes were incubated with the inhibitor in Niu-Twitty buffer. Mouse organ cultures were incubated in DMEM-F12 suppl emented with 10% fetal calf serum. A set of
23 untreated and treated eyes remained in the dark a nd a set of untreated and treated eyes were light adapted for 60 min. To investigate if arrestin is utilizing a G-protein regulated pathway, we used pertussis toxin and sulpiride, which are G-protein and D2-dopamine receptor antagonists, respectively. ArrestinGFP eyes were treated with 15 g/ml pertu ssis toxin (Sigma, #P7208) and 10 M (s) -(-) sulpiride (Sigma, #S7771) for 4 hr in the dark and then light adapted for 60 min. A set of eyes remained in the dark in Niu-Tw itty buffer, treated and untreated, while another set of eyes were light adapted for 60 min. To determine if sulpirid e penetrated into the photoreceptors during this incubation, dark-adapted wild-type ey es were also treated with 1 M 3H-sulpiride (-) [methoxy-3H] (Perkin Elmer, #NET775250uC) for 4 hr. A set of control eyes and su lpiride treated eyes remained in the dark and a set were light adap ted for 60 min. The eyes were fixed, rehydrated, and sectioned as described earlier. To detect the tritiated sulpiride, sections were slowly dipped in diluted Kodak emulsion in total darkness. The slides were placed ver tically to drain excess emulsion for 2 hr at room temperature. Afterwards the slides were placed in slide box wrapped in foil to prevent any light exposure. In this study, the eyes were exposed to the emulsion at 4oC for 3 months. After 3 months, the slides were wa rmed to room temperature and then placed in Kodak D19 developer (Kodak, #1464593) for 2.5 min. Im mediately afterwards the slides were rinsed in water for 30 sec and then fixed for 3 min in Kodak fixer (Kodak, #1971746). The slides were rinsed under gentle running water for 15 min and then dried in a dust free area. The slides were then visualized with an Olympus BH-2 Epifluorescent microscope. This procedure was adapted from Pardue (1985). Confocal Microscopy and Quantification The slides that were prepared for this study were visualized by confocal microscopy using three d ifferent microscopes (1024ES, BioRad; Zeiss LSM Multiphoton Laser Scanning Confocal
24 Microscopy; Olympus Spinning Disc Confocal Mi croscope) through filters optimized for FITC, Texas Red and Cy 5 fluorescence. The slides were collected through a series of z-sections at 0.5 m intervals and projected to a 2D image. The images were then visualized through the image software Photoshop 6.0. To quantify the fluoresce nce of GFP in the images, the images were gray scaled and inverted. The i nverted images were then impo rted to Scion Image version 4.0.2. The fluorescence in the OS and IS were measured by multiplying the encircled area and the density of that same area. Th e relative fluorescence in the rod outer segment and the rod inner segment of each photoreceptor were then graphed using Microsoft Office Excel 2003. In all cases, the calculated fluorescence in the OS was measured from a minimum of least two images collected from each eye from at least three animals (n 12). Average values were statistically compared using Students two-taile d t-test for unpaired values. In some experiments, arrestin fluorescence was quantified along the length of a phot oreceptor. For this measurement, a line was drawn through the center of the photorecepto r cell and the fluorescence intensity along the line measured. This measurement was perfor med on 10-15 photoreceptors per image collected from each eye of at least three animals. (n 6). Results Arrestin Translocates in Response to Activators of PKC a nd PLC Since phospholipase C (PLC) and protein kinase C (PKC) biochemical activity have been previously demonstrated in rod outer segments (OS) (e.g. Kaplan et al 1987 and Newton et al. 1993), we hypothesized that perh aps these enzymes may also function to signal arrestin translocation. If so, then PLC and PKC should be localized to rods and in a location that could report information to initiate arrestin translocation. Antibodies against PLC and PKC are localized in the ellipso id region of the IS (Fig.2-1), but are not expressed uniformly throughout the IS like arrestin (Fig. 1-2). This localizat ion at the interface between the inner segment and
25 outer segment is a reasonable location for regul ating movement of arrestin between the two compartments of the rod. Since the components of th e phosphoinositide pathway ar e expressed in the rod photoreceptors, we wanted to determine if act ivating PLC and PKC could initiate arrestin translocation to the outer se gments. Arrestin-GFP tadpoles were treated with phorbol 12, 13diacetate (PDA), an activator of PKC, under dim red lights. Arrestin translocated to the OS in response to the activator of PKC within 10 min of exposure (Fig.2-2B). Arrestin returned back to the IS after 30 min or longer expos ure (data not shown). To our know ledge, this is the first time arrestin translocation has been stimulated in th e absence of light. Similarly, we were able to stimulate arrestin translocation using an act ivator of PLC, m-3M3FBS, applied for 15 min (Fig.2-2C). Again, arrestin did not remain in the OS during longer exposure to m-3M3FBS (data not shown). To ensure this phenomenon was not specific to Xenopus laevis photoreceptors, darkadapted mouse retinal organ cultures were treated with a similar activator of PKC, phorbol 12myristate 13-acetate (PMA). Like Xenopus, a significant amount of arre stin translocated to the OS in response to PMA treatment in the absence of light (Fig.2-2E). Arrestin also translocated to the OS in response to m-3M3FBS in mouse retinas within 20 min. (Fig.22F). The confluence of these results from both Xenopus and mouse photoreceptors suggests PKC and PLC may play a role in signaling arrestin translocation to the OS. Inhibitors of PKC and PLC Modify Arrestin Localization Arrestin translocation to the OS in response to PKC a nd PLC activation strongly supports our hypothesis that arres tin translocation is associated with the phosphoinositide pathway. If this is an essential pathway to initiate arrestin tr anslocation, then inhibiti on of PKC and PLC should affect arrestin translocation during light ad aptation. Arr-GFP eyes and mouse retinal organ
26 cultures were treated with chelerythrine or U73122, which are PKC and PLC inhibitors, respectively. In the presence of these inhibitors, the light-induced translocation of arrestin was significantly slowed (Fig.2-3). In Xenopus the principal effect is that arrestin localized to the base of the OS (Fig.2-3B and C) as opposed to being fully distributed throughout the OS. This effect was most pronounced for the PLC inhibitor (Fig.2-3C and G). In mouse eyes, arrestin was localized to both the IS and OS in mouse retinas treated with th e inhibitors, unlike the untreated eye where most of the arrestin was in the OS (F ig.2-3D-F). Thus, arrestin translocation in both Xenopus and mouse photoreceptors was significantly slowed by these inhi bitors of PKC and PLC, although the effect on Xenopus was relatively minor compared to mouse. Perhaps the difference may be attributed to different degrees of drug penetration or to the differences in cell biology of the two species. A G-protein Mediates Arresti n Translocation to OS Since the above data indi cate th at the signaling cascade for arrestin translocation to the OS appears to use both PKC and PLC, we wanted to further investigate if this pathway utilizes a Gprotein to couple to PLC and PKC. The logical first consideration might be transducin, a Gprotein that is abundant in the r od photoreceptors (Fig.2-4F). Howe ver, translocation of arrestin appears to be independent of th is G-protein, since mice with a knockout of the transducin alpha subunit have relatively normal patterns of arrestin translocation (Mendez et al. 2003). Another possibility is a Gi-like G-protein, for which activity has been demonstrated in rod photoreceptors (Zhang et al. 2003, Koulen et al. 2005, Gotow et al. 2007). To assess the possible involvement of Gi-like proteins in arrestin tr anslocation, we immunostained fo r several G-proteins in rod photoreceptors and found them to lo calize differentially in the IS. G i-1 and G i-2 are primarily localized in the IS (Fig.2-4A and B) whereas G i-3 is mostly localized in the ellipsoidal region of the IS and along the axoneme of the OS (Fig.2-4C). Like G i-1 and G i-2, G o and G 11 are
27 distributed throughout the IS and also along the proximal region of the axoneme (Fig.2-4D and E). Thus, there are several G-proteins in the rod photoreceptors that could potentially activate PLC. To distinguish between some of these Gproteins as candidates for signaling arrestin translocation, we treated eyes from transgenic Arr-GFP tadpoles with pertussis toxin. Pertussis toxin targets inhibi tory G-proteins: Go, Gi, Gt, and Ggust (Wong et al. 1999). In transgenic tadpole eyes pretreated with pertussis toxin (15 g/ml), the localization of arrestin in response to light adaptation was significantly differe nt than in untreated eyes (Fi g.2-5). In these eyes, pertussis toxin treatment significantly reduced arrestin translocation to the OS in light-adapted photoreceptors. These results provide evidence that arrestin translocation to the OS may utilize a G-protein, and that this G-protein is pertussi s-toxin sensitive. To further investigate if arrestin translo cation to the OS is Gprotein regulated, we attempted to activate the G-protein by co-opt ing a G-protein-coupled receptor other than rhodopsin that is expressed in the retina. The dopamine receptor has been extensively studied and has been shown to play a role in light adaptation (Dearry et al. 1990, Shulman et al. 1996, Witkovsky et al. 1988, Muresan and Besharse 1993). Because its expression has been detected in photoreceptor cells, along with am acrine cells and Mueller cells (Muresan and Besharse 1993), the D2 dopamine receptor is a good target for act ivating arrestin translocation. Our own immunolocalization studies with a D2 receptor-specific antibody shows localization of this dopamine receptor in the IS in Xenopus similar to previously publis hed studies (Fig.2-6). To activate this receptor, we trea ted dark-adapted Arr-GFP tadpoles with 10 M quinpirole, a D2dopamine receptor agonist, and then assessed arres tin localization in the absence of light (Fig.27). In these treated animals, arrestin-GFP translocated rapidly to the OS in response to the
28 dopamine agonist and its localizatio n in the OS was sustained for at least 30 min. Our ability to stimulate a non-rhodopsin G-protein-coupled receptor to initiate arrestin translocation fits our hypothesis, but it also raises the possibility that dopamine receptors could be involved in arrestin translocation. To determine if arrestin actu ally uses the dopamine receptor to initiate translocation to the OS in response to light, we treated Arr-GFP tadpole ey es with sulpiride, a D2-dopamine antagonist, and then assayed for any effect on light-d riven arrestin translocation (Fig.2-8). Arrestin translocati on in light-adapted photoreceptors was indistinguishable between sulpiride-treated and untreated eyes Because of these negative results, we assessed penetration of the sulpiride reagent using tritiated sulpir ide (Fig.2-8C and D). The clear presence of the 3Hlabeled sulpiride throughout the retina demonstrat es that arrestin transl ocation does not rely on the D2-dopamine receptor to initiate translocation, but rather that we were able to co-opt the D2dopamine receptor with quinpirole treatment to initiate arrest in translocation.
29 -PLC -PKC B. C. A.IS OS -PLC Figure 2-1. Localization of phospholipase C and PKC in rod photoreceptors. Phospholipase C and PKC are localized in the ellipsoid region of Xenopus photoreceptors. Cryosections were immunostained for (A) Phospholipase C gamma (1:100) (B) Phospholipase C beta (1:100) and (C) Protein Kinase C (1:100). The nuclei were stained with nuclear stain sytox green. The retinal sect ions were visualized by confocal microscopy.
30 DA Untreated 1 M PDA 1 M m-3M3FBSA. B. C. DA Untreated 100 nMPMA 100 nMm-3M3FBS D. IS OS IS OS IS OS E. F. Figure 2-2. Arrestin tran slocation to the OS is induced by PKC and PLC agonist s. Dark-adapted arrestin-GFP transgen ic tadpoles were treated with 1 M PDA (phorbol 12, 13diacetate) for 10 min in the dark (B) or explanted eyes from Arr-GFP transgenic tadpoles were treated with 1M m-3M3FBS for 15 min in the dark (C). After PDA and m-3M3FBS exposure, the eyes were fi xed, cryosectioned, and GFP visualized by confocal microscopy. Mouse eye organ cu ltures were treated with 100 nM PMA (phorbol 12-myristate 13-ace tate) (E) or 100 nM m-3M3F BS (F) in organ culture media under dim red lights. The eyes were fixed, cryosectioned, and immunostained for arrestin (red), and nuclei stained with DAPI (blue). Th e sections were visualized by fluorescent microscopy. Control eyes were treated with only the DMSO carrier (A, D). The average fluorescence in the IS a nd OS was quantified from a minimum a two images from each eye of three different tadpoles (n 12), and the fraction of fluorescence in the outer segments plotted (G). The differences between untreated tadpoles eyes and agonist-treated eyes are statistically si gnificant (p<0.01).
31 Fraction of Total Fluorescence in OS 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Untreated 10 min1 uM PDA 15 min 1uM PLC Figure 2-2. Continued DA G.
32 Untreated 10 M Chelerythrine10 M U-73122 A. B. C. LA Untreated 10 M Chelerythrine10 MU-73122 LA IS OS IS OS IS OS D. E. F. Figure 2-3. Arrestin transl ocation is slowed in eyes treated with PKC and PLC inhibitors. Darkadapted Arrestin-GFP transgenic tadpoles (A-C) and mouse organ cultures (D-F) were treated with 10 M Chelerythrine (B E) or 10 M U73122 (C, F) for 4 hrs and then light adapted for 1 hr. The eyes were fixed as described in the Methods and prepared for confocal microscopy. Arre stin-GFP was visualized by confocal microscopy using the endogenous fluorescence of the GFP. Mouse sections were immunostained for arrestin (red channel) and the nuclei stained with DAPI (blue channel). For the Xenopus photoreceptors, the fraction of fluorescence along the length of the photoreceptor was measured (G).
33 OS ISPosition along the Photoreceptor -20 0 20 40 60 80 100 120 140 160 180 200 02040 6080 100120140 Untreated 10uM Chelerythrine 10uM U73122 1 0 Total Fluorescence in OS Figure 2-3. Continued G.
34 -Gi-3 -Gi-1 -Gi-2 -Gi-o -T -G 11A. B. C. D. E. F.RIS ROS RIS ROS Figure 2-4. Localization of G-proteins in rod photoreceptors. Dark-ada pted wild-type tadpole eyes were immunostained for various G-protei ns to determine thei r localization in the rod photoreceptors. Cryosect ions were immunoprobed for (A) G-protein Gi-1 (B) Gprotein Gi-2, ( C ) G-protein Gi-3, and ( D ) G-protein Gi-o (red channel), or ( E ) G-protein G11-alpha, and ( F ) G-protein GT-alpha (blue channel). The nuclei were stained with Sytox green. The sections were visu alized using confocal microscopy.
35 UntreatedDA LAPertussisToxinA. B. C. D.IS OS Figure 2-5. Arrestin tran slocation is sensitiv e to pertussis toxin Dark-adapted Arrestin-GFP transgenic tadpole eyes were treated with 15 g/ml of pertussis toxin for 4 hrs in NiuTwitty buffer in the dark. Both treated eyes ( B, D ) and untreated eyes (A, C ) were retained in the dark ( A, B ), or were light adapted for 60 min ( C, D ). All eyes were fixed, cryosectioned, and GFP visualized by confocal fluorescence microscopy.
36 DA LAIS OS Figure 2-6. D2-dopamine receptor localization in rod photoreceptors. Dark-adapted (DA) and light-adapted (LA) wild-type tadpole eyes were fixed and prepared for immunocytochemistry. Cryosections were immunostained with D2-Dopamine Receptor antibody (red channel), and nuclei we re stained with Sytox-green (green channel).
37 Untreated 10 min Quinpirole 30 min Quinpirole OSISB. C. A. Fraction of Total Fluorescence in OS 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1 Untreated 10 min 10uM Quinpirole 30 min 10uM QuinpiroleDA Figure 2-7. Arrestin localization in response to a dopamine receptor agonist. Dark-adapted arrestin-GFP tadpoles were trea ted with 10 M quinpirole, a D2-dopamine receptor agonist, for 10-30 min under dim red lights (B, C). The tadpoles were fixed and cryosections were visualized by fluorescen t and confocal microscopy. The control eyes received no quinpirole (A). The fraction of fluorescence in the outer segments of the photoreceptors was measured (D), and av eraged across at leas t 9 images from at least 3 animals (n=9). The difference between untreated tadpol es and quinpirole treated eyes are statistica lly significant. (p<0.05). D.
38 Untreated3H-SulpirideSulpirideArr-GFPC. D. A. B.IS OS Figure 2-8. Arrestin localization in response to a D2-dopamine receptor antagonist Darkadapted arrestin-GFP transgenic tadpole eyes were treated with 10 M sulpiride, a D2-dopamine receptor antagonist, for 4 hrs and then light adapted for 1 hr ( B ). The eyes were fixed after treatment and GFP distribution was visualized by confocal microscopy. To assess penetration of sulpir ide into the eyes, w ild-type tadpole eyes were incubated with 1M 3H-sulpiride for 4 hrs and light adapted for 1 hr ( D ). The eyes were fixed and cryosectioned, and dipped in photoemulsion to detect the radiolabel (see Methods). Developed slides were visu alized through an epifluorescent microscope. Control eyes were untreated (A, C ).The relative fluorescence Arr-GFP in the outer segments of photoreceptors fr om light-adapted eyes was quantified (E ); statistical comparison shows there is no si gnificant difference between treated and untreated samples (p>0.05).
39 LA 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1 Untreated 10uM SulpirideFraction of Total Fluorescence in OS Figure 2-8. Continued
40 CHAPTER 3 THE MECHANISM OF ARRESTIN TRANSLOCATION The signaling pathway we are proposing for arre stin translocation to the OS requires ATP. However, Nair et al. (2005) pr oposed arrestin translocation betw een the OS and IS is an ATP independent process. According to their hypothesis, arrestin re mains in the IS based on its affinity for microtubules (Nair et al. 2005) and it translocates to the OS based on its high affinity for light-activated, phosphorylated rhodopsin (N air et al. 2005) Because of our findings indicating the involvement of a G-protein signaling cascade that included PKC activation, we reinvestigated whether arrestin translocation is ATP independent in our animal model. To determine if arrestin translocation uses an ATP dependent pathway, we treated Arr-GFP eyes with potassium cyanide (KCN) to deplete the ATP. A second tene t of the Nair et al. (2005) proposed model is that arrestin translocation is dependent upon its affinity for light-activated, phosphorylated rhodopsin in the OS and its affinity for microtubules in the IS. Because the Cterminus of arrestin is essentia l for translocation, but not for binding to rhodopsin (Peterson et al. 2005), we tested the Nair et al. (2005) hypothesis by creating an arrestin mutant that has a scrambled C-terminus (Arr-scr) to generate an arrestin that was still able to bind to both microtubules and light-activated, phosphorylat ed rhodopsin and assess ed its effect on translocation. Materials and Methods Construction of Arrestin cDNAs. Xenopus visual ar restin with the C-terminal 33 amino acids scrambled was prepared using overlapping synthetic oligonu cleotides (Table 1) as follows Sense Primer 1 (200 pmol) was annealed to anti-sense Primer 2 (200 pmol) and filled using Pfu polymerase (Stratagene). The resultant 85 bp product was cleaned (Qiagen Nucleotide Removal kit), denatured by heating (94oC), annealed to anti-sense Primer 3 containing an Nhe I site, and subsequently filled using
41 Pfu polymerase to complete the cDNA for the scrambled C-terminus. This DNA was then mixed with arrestin cDNA coding for the remainder of the protein (generated by PCR using Primers 4 and 5), and overlapping PCR performed with Pfu polymerase to anneal and fill the complete arrestin cDNA with a scrambled C-termi nus. This product was then cut with XhoI and Nhe I and used to replace the wild-type arrestin cDNA in the arrestin-green fl uorescent protein (GFP) transgenic vector previously de scribed (Peterson et al. 2003). For heterologous expression of Xenopus arrestin with the native C-terminus and with a scrambled C-terminus in yeast, the arrestin was amplified with Primer 6 paired with Primer 7 (for wild-type arrestin) or paired with Primer 8 (for the scrambled C-terminus), incorporating flanking Eco RI sites along with a 5 His(6) tag. This cDNA was then cloned into the Eco RI site of pPIC-Za (Invitrogen). The proper constr uction of all cDNAs was confirmed by DNA sequencing. Table 3-1. Synthetic oligonucleot ide primers used to construct arrestin with a scrambled Cterminus. Primer 1 5-GGAGCCAAAGAAAGTGAAGC CGAAATGGAATTTGAAGGT GACGAACGTCC-3 Primer 2 5-CCTCCTTTTCTGCAAATTCTTGCTTATCGAGTTCTGGACGTTCGTCACCT-3 Primer 3 5-GCGCTAGCCTCGTCCAGCTCCTTCTTCTCCACTTGATCCTCCTTTTCTGCAA-3 Primer 4 5-GCCTCGAGATGAGTGGTGAAAAGAAATCCAGA-3 Primer 5 5-TTCCATTTCGGCTTCACTTTCTTTGGCTCC-3 Primer 6 5GCCTCGAGGAATTCACCATGCATCATCATCATCATCACAGTGG-3 Primer 7 5-GCGAATTCTTATTTCTCATCATCCTCCTCTTC-3 Primer 8 5-GAGAATTCTCACTCGTCCAGCTCCTTCTTCTCCAC-3 Generation of Transgenic Animals Transgenic Xenopus tadpoles were prepared using the method of Kroll and Am aya (Kroll and Amaya 1996) as previously described (P eterson et al. 2003). Resultant tadpoles were
42 screened for green fluorescence at one week, and subsequently reared an additional 2-3 weeks prior to experimentation. For microscopic stud ies, tadpoles were dark adapted overnight and then exposed to laboratory lights (approximate ly 850 lux) for 45 min. Tadpoles were fixed and processed for confocal microsc opy as previously described (Pet erson et al. 2003), imaging the endogenous fluorescence from the GFP fusion proteins. Expression in Tissue Culture. For expression in HEK293 cells, arrestin-GFP and scrambled arres tin-GFP were excised with XhoI and Not I, and cloned into the XhoI/Not I sites of pDsRed2-N1 (Clontech) to place the fusion protein under the control of the cytome galovirus promoter. Plasmid DNA for xAr-GFP and xArScr-GFP was introduced into HEK cells growing on slide chambers using Effectine transfection reagent (Q iagen). Two days post-transfection, the cells were fixed with 4% paraformaldehyde in PBS for 20 min at RT. The s lides were washed with 50 mM glycine in PBS for 5 min and then incubated for 5 min in 1% Triton-X100 in PBS for 5 min. After washing with 50 mM glycine in PBS, the slides were incubated in a 1:500 dilution of antitubulin (Serotec, #MCA77G) for 3-4 hrs. The cells were then wash ed with 50 mM glycine in PBS, and secondary goat anti-rat antibody conj ugated to Texas Red-X (Invitrogen, T6392) added for 30 min. After the slides were washed and sealed, they were visualized by confocal microscopy, imaging tubulin (red channel) and the e ndogenous fluorescence of the GFP (gr een channel). Lebercilin (a kind gift from Dr. R. Roepman) has been shown to have a significant binding affinity for tubulin (den Hollander et al. 2007) and was used as a positive control for microtubule co-localization. Heterologous Expression and Purification of Arrestin. The cDNA for native arrestin and arrestin wi th a scrambled C-term inal arrestin were electroporated into Pichia pastoris for expression as previously described (McDowell et al. 1999). Arrestins were purified by sequential a ffinity chromatography, first pressing the yeast
43 pellet in 50 mM sodium phosphate (pH 8.0) w ith 300 mM NaCl, 10 mM imidazole, and 1 mM benzamidine and then applying the soluble extract to a GraviTrap His column (GE Life Sciences). The arrestin was then eluted from the washed column with 50 mM sodium phosphate (pH 8.0), 300 mM NaCl, 500 mM imidazole, dilute d with two volumes of 10 mM HEPES (pH 7.0), 15 mM NaCl, and then applied to a HiTrap heparin column (GE). After washing the column with 10 mM HEPES containing 250 mM NaCl the protein samples were eluted with a salt step of 750 mM NaCl, and subsequently dialyzed against 10 mM HEPES with 100 mM NaCl (pH 7.0) prior to storage at -80oC. Rhodopsin Binding Assay. Arrestin binding to rhodopsin was measur ed using a pull-down a ssay with mem brane fragments from rod outer segments as descri bed (Sommer et al. 2007). Membranes containing unphosphorylated and phosphorylated rhodopsin were prepared from rod outer segments isolated from bull frogs ( Rana catesbeiana) retinas, using the methods described for bovine retinas (McDowell, 1993). Microtubule Binding Assay. Binding to microtubules was assayed using taxo l-stabilized m icrotubules prepared from rhodamine-labeled tubulin (Cytos keleton, Inc). Microt ubules were polymerized and stabilized with taxol according the manufacturers recommendation, and centrifuged (100,000 x g, 30 min) through a 50% glycerol cushion to separate the microtubules from unpolymerized tubulin. Histagged arrestin and scrambled a rrestin (20 g/ea) were immob ilized on a 200 L slurry of NiNTA agarose (Invitrogen) in 10 mM HEPES (pH 7.0) with 100 mM NaCl. A control slurry was prepared using 20 g of bovine serum albumin (Fr action V, Sigma). After rinsing the resin to remove unbound arrestin, polymerized rhodamine microtubules (5 g) were added to each slurry, and incubated for 30 min at room temperature. Subsequently, the sl urry was applied to a
44 microcolumn, retaining the sample flow-through, and then the column was washed with HEPES buffer (1 mL) to remove unbound microtubules. The columns were then eluted with 10 mM HEPES (pH 7.0) with 100 mM NaCl and 500 mM imidazole. Samples of the flow through and eluate were then submitted to fluorescence sp ectroscopy (PTI 810/814 fluorimeter) to measure the amount of rhodamine present in the sample, using 540 nm excitation (0.37 nm bandpass) and collecting emission spectra 560-660 nm (0.5 sec integration per 1 nm step, 0.25 nm bandpass). Potassium Cyanide Treatment of Tadpole Eyes. Arrestin-GFP (Arr-GFP) tadpoles were dark adap ted for at least 12 hrs and then treated with 0.05% benzocain e until euthan ized, at which point the eyes were removed with a scalpel. Their eyes were then placed in MOPS-Twitty buffer (50 mM MOPS, 58 mM NaCl, 670 M KCl, 340 M Ca(NO3)2, 830 M MgSO4, 2.4 mM NaHCO3, 340 M CaCl2), pH7.5 with or without 5 mM KCN under dim red lights. The eyes were incubated in 5 mM KCN for 1 hr in the dark and then light adapted for 1 hr under laboratory lighting. In eyes where the cyanide was removed, after treatment with 5 mM KCN for 1 hr in the dark, the eyes were washed three times with MOPS-Twitty buffer, and then incubate d in either MOPS-Tw itty or MOPS-Twitty containing 5 mM ATP for 2 hr in th e dark. Eyes either remained in the dark or were light adapted for 1 hr under laboratory lighting. Both dark-adapted eyes and light-adapted eyes were then fixed in methanolic formaldehyde and processed for conf ocal microscopy as described above. A set of unfixed eyes were also collected for measurem ent of ATP levels using the Cell TiterGloMax (Promega, #G7570) system. For this measurement, eyes were processed in the same way as above for microscopy except that instead of fixing in formaldehyde the eyes were sonicated in 100 L MOPS-Twitty buffer, allowed to sit for 30 mi n (25oC) to allow the luciferase activity to stabilize and then luminescence measured, follo wing the manufacturers protocol to determine ATP levels.
45 ATP Depletion of C57BL/6 Mouse Eyes. Wild-type C57BL/6 mice were dark adapted for at least 4 hr and then placed in isoflurane for euthanizing. The eyes were promptly removed after sacrificing them by cervical dislocation. All animal procedures followed the rules and regulations set forth by the Institutional Animal Care and Use Committee at the University of Fl orida. The eyes were either placed in high glucose DMEM supplemented with 10% fetal ca lf serum or glucose free DMEM supplemented with 2 mM deoxyglucose. They were maintained at 37oC with 5% CO2 for 60-90 min and then light adapted for 60 min. Both darkand light-ada pted eyes were fixed in 4% paraformaldehyde and 0.25% gluteraldehyde in 1x PBS for 2 hr at ro om temperature. Afterwards, the eyes were transferred to 4% paraformaledhyde in PBS and fixed overnight at 4oC. After fixation, the eyes were cryoprotected in 30% sucrose in PBS for at least 4 hr and embedded in OCT. Sections of the eyes were immunostained for arrestin (SCT-128, 1:50) and visualized by confocal microscopy. Results Previous investigations have shown that the Cterminus of arrestin plays a significant role in arrestin translocation. For example, arrest ins that lack the C-terminal approximately 30-35 amino acids can be found partially localized to the rod outer segm ents in dark-adapted tissues (Smith et al. 1994, Peterson et al. 2003, Nair et al. 2005). Further, these same truncated arrestins show a significant reduction in their movement to the outer segments during light adaptation, primarily localizing to the base of the outer segments in both mous e (Nair et al. 2005) and Xenopus photoreceptors (Peterson et al 2005). Noting that the C-terminus of the visual arrestins are significantly acidic (mouse rod arrestin has twelve aspartate or glutamate residues, and Xenopus rod arrestin has seventeen as partate or glutamate residues in the C-terminal 35 amino acids), we investigated whether the net charge of the C-terminus was significant in promoting
46 translocation of arrestin since the positioning of the aspartate and glutamate residues are not strongly conserved in arrestin from various species, even within mammals. Arrestin Binding to Microtubules Is Independe nt of Charge Order in the C-terminus For this study, a Xenopus arrestin cDNA was constructed in which the C-term inal 33 amino acids were randomized and then added back to the arrestin cDNA lacking the codons for these 33 residues (see Methods). Following e xpression and purification of the native and scrambled arrestins, a pull-down assay was performed using rhodopsin in disc membranes isolated from bullfrogs to assess binding of th e arrestin to rhodopsin. This assay showed that both arrestin (xAr) and arrestin with a scramble C-terminus (xAr-Scr) retained good binding for light-activated, phosphorylated r hodopsin (p-Rho*) compared to rhodopsin (Rho) (Fig.3-1). Because this arrestin with a scrambled C-te rminus retained its se lectivity for pRho*, we were curious as to how this ar restin would translocate in rod photoreceptors, predicting that its translocation would be indistinguishable from na tive arrestin. Consequen tly, transgenic tadpoles were generated that expressed a fusion of G FP with xAr-Scr (see Met hods). Figure 3-2 shows that the xAr-Scr-GFP localizes prim arily to the OS in dark-adapted tadpoles (Fig.3-2C), with no evidence of change in localization in response to 60 minutes of li ght adaptation (Fig.3-2D). In contrast, in tadpoles expressing a fusion of G FP to native arrestin (x Ar-GFP), the arrestin localized primarily to the rod inner segments in dark-adapted eyes (Fig.3-2A), with more than 80% translocating to the outer segments in response to light adaptation (Fig.3-2B). This surprising result led us to investigate whether scrambling the C-terminus of arrestin had affected the binding of arre stin to microtubules. For this assessment, we measured the association of xAr and xAr-scr to rhodamine-labeled microtubules (Fig.3-3). This assay showed that both xAr and xAr-scr bound indistinguishable amounts of microtubules. In contrast, the bovine serum albumin used as a control sample re tained only background le vels of microtubules.
47 Considering this apparent similarity in micr otubule binding for xAr and xAr-scr, we also wanted to investigate the in vivo binding of these arrestins to microtubules. For this analysis, HEK-293 cells were transiently transfected with xAr-GFP and xAr-Scr-GFP driven by a cytomegalovirus promoter and co-localization with microtubules assessed by staining of tubulin with Texas red (Fig.3-4). Both xAr-GFP a nd xAr-Scr-GFP showed a relatively diffuse distribution of GFP fluorescence (g reen channel) compared to th e filamentous pattern shown by tubulin staining (red channel). N onetheless, there was significant co-localization of the two fluorescent signals shown by the yellow signal (c ombined red and green channels), indicating some association with microtubules. In contra st, lebercilin, a microtubule binding protein with higher affinity for microtubules than arrestin (den Hollander et al. 1996 ), showed essentially complete overlap between the lebercilin-G FP fluorescence and the microtubule fluorescence (Fig.3-4G-I). These results provide additional evidence that scrambling the C-terminus does not appear to alter the association of arrestin with microtubules since the microtubule binding pattern is indistinguishable between xAr and xAr-Scr. Arrestin Translocation in ATP-depleted Photoreceptors The above observations showing disruption of light-dependent translocation in an arrestin that retains binding for both r hodopsin and microtubules argues that the m achinery driving translocation might be more complex than the two-partner binding hypothe sis (Nair et al. 2005). To help provide clarity, we recapitulated their ATP depletion study using explanted Xenopus eyes expressing arrestin-GFP. Analysis of the AT P levels in eyes treated for 60 min with 5 mM KCN, revealed that ATP levels were depleted by three orders of magnitude compared to the untreated controls (data not shown). In these da rk-adapted ATP-depleted eyes, arrestin partitions to the inner segments (Fig.3-5C). However, upon light exposure, arrestin does not move to the
48 outer segments in cyanide-treated eyes (Fig.3-5D), in contrast to the untreated controls where the arrestin localizes to the outer segments (Fig.3-5B). One possible explanation for this observation is that the photoreceptor cells have been damaged in some non-specific manner by the cy anide treatment that would make them permanently refractory to light stimulation. To test this idea, the eyes were washed with several changes of buffer without cyanide, and then suppl emented with 5 mM ATP. In these animals, light adaptation led to transloca tion of arrestin to the outer se gments to an extent that was statistically indistinguishable from un treated controls (Fig.3-5F and G). Our ATP-depletion results indicat e that arrestin translocati on does require ATP, which is in striking contrast to Nair et al. (2005) where they found arrestin translocation was not affected by ATP depletion. In order to determine whether th is was due to a species difference, we decided to reinvestigate the ATP depende nce of arrestin translocation in mouse photoreceptors. Mouse eyes were placed either in high glucose media supplemented with 10% fetal calf serum or glucose-free media supplemented with 2 mM deoxyglucose for 60-90 min. There was no difference in arrestin localization between the da rk-adapted photoreceptors of high glucose or glucose-depleted eyes (Fig.3-6C and D). However, arrestin translocation to the OS in response to light adaptation was significantly reduced in ey es that were incubate d in glucose-free media compared to the controls in which a significan t amount of arrestin was localized to the OS (Fig.3-6E). These results suggest that in both frog and mouse rod photoreceptors, an ATPdependent process plays an important role in the proper translocation of a rrestin in response to light.
49 Figure 3-1. Binding of arrestins to rhodopsin in rod disc membranes. Arrestin with a scrambled C-terminus (xAr-scr), EAEMEFEG DERPELDKQEFAEKEDQVEKKELDE, and native arrestin (xAr), EQEDDMVF EEFARDPLKGELQAEEKEEEEDDEK, were pulled down in a centrifugation assay usi ng bullfrog rhodopsin (Rho) in rod disc membranes that was kept in the dark (R), light activated (R*), phosphorylated (pR), and both phosphorylated and light activated (pR*). The bar graph shows the binding of arrestin relative to the binding of wild-type arrestin to pR*; bars show the average of three replicates with SEM. Inset shows a representative gel used for the calculations. xAr Rho
50 Figure 3-2. Translocation of xAr a nd xAr-scr in transgenic tadpoles. Native arrestin-GFP transgenic tadpoles (xAr, A, B ) and transgenic tadpoles expressing arre stin with a scrambled c-terminus fused to GFP (xAr-scr, C, D ) were dark adapted overnight (DA) or light adapted for 45 min (LA). GFP fluorescence in cryosections of fixed tissue was imaged by confocal microscopy. The bar graph shows the fraction of fluorescence in the OS for the dark samples ( black bars ) and the light samples ( gray bars ); averages were obtained from 6 images collected from three separate animals ( SEM).
51 Figure 3-3. Native arrestin and a rrestin with a scrambled C-term inus bind fluorescently labeled microtubules Native Xenopus arrestin with an N-terminal His(6) tag (xAr, 2 M) and arrestin with a scrambled C-terminus xA r-scr, 2 M) immobilized on nickel-agarose were mixed with microtubules formed from polymerized rhodamine-labeled tubulin (see Methods ). Rhodamine fluorescence was measured for flow-through (dark gray bars ) and eluted fractions ( light gray bars). Bovine serum albumin (BSA, 2 M) was included as a control for non-specific bindi ng. Bars represent the mean for three replicates SEM. Inset shows a western blot perfo rmed on the flow through and eluted samples, probing w ith an antitubulin antibody.
52 Tubulin Merged GFP Figure 3-4. Arrestin and tubulin localization in tissue culture HEK-293 cells were transiently transfected with native arrestin-GFP (xAR, A-C ), arrestin with a scrambled Cterminus-GFP (xAR-scr, D-F ), or lebercilin-GFP ( G-I .). Fixed cells were immunostained for tubulin. GFP fluorescence ( green channel ), and tubulin fluorescence ( red channel ) were imaged by confocal mi croscopy. Merged images of the two channels for each sample are shown on the right ( C, F, I ).
53 A. B. C. D. E. F.IS OS IS OS IS OS Figure 3-5. Arrestin localization in respons e to ATP depletion in transgenic tadpoles Darkadapted Arrestin-GFP transgenic tadpole eyes were treated with 5 mM KCN in MOPS-Twitty for 1hr ( white bars )and kept in the dark for 1 hr ( C ) or were light adapted for 1 hr ( D ). To remove KCN, the eyes we re washed and then supplemented with 5 mM ATP for 2 hrs ( light gray bars) (E ) and were then light adapted for 1 hr ( F ). Untreated control eyes ( dark gray bars) were dark and light adapted ( A, B ). GFP in cryosections of fixed tissue was visuali zed using confocal microscopy. The relative amount of fluorescence in the OS was quantif ied; bars represent the mean from at least eight eyes SEM ( G ). There is a statistically significant difference between light-adapted KCN-treated eyes and light-a dapted control eyes (p<0.05), but not a statistically significant difference between control eyes and eyes in which the KCN was washed out and supplemented with exogenous ATP (p>0.05).
54 Figure 3-5. Continued
55 +glucose -glucoseIS OS ONL OPL IS OS ONL OPL IS OS ONL OPL IS OS ONL OPL DA LA Figure 3-6. Arrestin localization in re sponse to ATP depletion in mouse eyes Wild-type mice were dark-adapted for at least 4 hrs. Th eir eyes were then removed and placed in either high glucose DMEM supplemented with 10% fetal bovine serum ( A, B ) or placed in glucose-free DMEM supplemented with 2 mM deoxyglucose ( C,D ). Lightadapted (LA) eyes were incubated at 37oC at 5% CO2 for 60-90 min and then exposed to light for 1 hr. Cryosections of fixed ey es were immunostained for arrestin (SCT128) and visualized by confocal microscopy. The relative amount of fluorescence in the OS was quantified; bars represent the mean from at least 2 images collected from at least two separate eyes SEM (E ).
56 Relative of Fluorescence in OSDA 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 12 +glucose -glucoseLA Figure 3-6. Continued
57 CHAPTER 4 DISCUSSION In this study, we investigated which molecules may play a role in the signaling cascade to initiate light-dependent arrestin translocation an d the mechanism by which arrestin translocates. To briefly summarize, our results show that arre stin translocation appear s to utilize a G-protein regulated pathway initiated by rhodopsin activation. Further, our results show that at least a portion of the light-driven translocati on is an energy dependent process. G-proteins in Rod Photoreceptors Our studies show that the initial step in signa ling arrestin translocation appears to utilize a G-protein regulated pathway. The ev idence supporting this conclusion is several fold. First, we demonstrated that arrestin translocation is sensitive to pertussis toxin, showing a highly significant reduction in light-drive n arrestin translocation after treatment with pertussis toxin (Fig.2-5). Pertussis toxin is known to aff ect only certain classes of G-proteins (Gi, Go, and Gt), suggesting that the G-protein that might be part of the transloc ation cascade could be a member of one of these groups. Obviously, Gt is abundantly present in photoreceptors, but the dependence of arrestin transloc ation on this G-protein, transducin, can be excluded based on the experimental observation that arrestin tr anslocates normally in mice in which T expression is knocked out (Mendez et al. 2003). Although pertu ssis toxin is relatively selective for ADPribosylation of select G-prot eins, it is known to bind to other carbohydrate moieties (Wong and Rosoff 1999, Alfano et al. 2001, Garc ia et al. 2001). Consequently, our dem onstration of pertussis sensitivity for arrestin translocation is not definitive, but it does lend support for our hypothetical model. Additional support for the G-prot ein mediated cascade is prov ided by our studies showing the localization of several candidate G-protei ns in rod photoreceptors. Our immunohistochemical
58 investigations show that several types of G-proteins particularly in the Gi class, are found in photoreceptors (Fig.2-4). For all of the Gi proteins, the localization is strongest in the inner segments and along the axoneme. Inte restingly, the localization of Gi3 is stronger in the ellipsoid region of the IS, similar to the distribution observed for PLC, PLC, and PKC (Fig.2-1). The localization of these G-proteins along the axoneme is intriguing since this localization would make the G-proteins accessible to rhodopsin in the OS. Although G11 is also expressed in the IS and axoneme, we do not consider it likely to be involved in th e signaling pathway for arrestin translocation since G 11 is not pertussis toxin sensitive. The presence of multiple G-proteins in this region of the rod photoreceptor highlights the complexity of the processes that are occurring at this location in this highly sp ecialized cell. Our results demons trating the presence of multiple G-proteins in rods confirm previous observations of Gi-like activity in the retina (Zhang et al. 2003, Koulen et al. 2005, Gotow et al. 2007,). Because Gi G-proteins are able to couple to dopamine receptors (Fan et al. 2004, Lane et al. 2008), which are also expressed along the membra nes of the IS (Fig.2-6), we reasoned that another way to demonstrate that arrestin tr anslocation is initia ted through a G-protein transduction cascade would be to activate dopamine receptors to stimulate Gi activation. Using quinpirole, a D2-dopamine receptor agonist, arrestin transl ocation was initiated in the absence of light, moving to the OS rapidly and remaining in the OS with prolonged exposure to quinpirole (Fig.2-7). In contrast to the PLC and PKC agonists which caused only transient translocation of arrestin to the base of the outer segments, ar restins localization to the OS in response to quinpirole was sustained for at least 30 minutes and was fully distributed within the outer segments. Although we are not sure of the reason behind these differences, arrestins persistence in the outer segments may be attributed to the stability of the quinpirole compared to the PKC and PLC agonists, as well as to the position of the dopamine receptor as an initiator in the
59 cascade. Since the dopamine receptor is at the t op of the cascade, it may take longer for the cascade to desensitize, therefore re ducing the rate of arrestin retu rning to the IS. It should be emphasized that we are not proposing that normal light-driven arrestin tran slocation is initiated by dopamine receptor activation. Instead, our studies only demonstrate that we are able to co-opt the D2-receptors with a high dose of r eceptor agonist to initiate arre stin translocation. The lack of effect on arrestin transloc ation by sulpiride (Fig. 2-8), a dopamine receptor antagonist, supports this view that dopamine receptors do not normally initiate a rrestin translocation. However, we will also not exclude the possibility of a role for the dopamine receptor modulating the photoresponse, perhaps altering the sensitivity of transl ocation signaling. The co-localization of the dopamine receptor with the multiple Gi G-proteins (Fig. 2-6 and Fig. 2-4) in the inner segments and its capability to couple to Gi G-proteins (Weber and Schlicker 2001) suggest there is role for the receptor in th e rods (Fan et al. 2004, Lane et al. 2008). D2-like dopamine receptors can couple to the G protein G i, (Weber and Schlicker 2001) which subsequently increases cAMP phosphodiesterase ac tivity. The role of the dopamine receptor can be to restore the ionic balance back to the da rk-adapted state in rod photoreceptors (Akopian et al. 1996) by reducing Na+,K+-ATPase activity (Shulman et al. 1996). Role of PKC and PLC In support of our hypothesis that arrestin tran slocation is signaled through a G-protein regulated pathway, we also provide evidence im plicating PKC and PLC in initiating a rrestin translocation. Fig. 2-2 demonstrates that activators of both PKC and PLC can initiate arrestin translocation to the OS in the absence of light Apparently, the function of PKC and PLC in initiating this translocation is essential since antagonists of both PKC and PLC slow the lightactivated movement of a rrestin to the OS (Fig. 2-3). Although the results are consistent with a possible role for PKC in this path way, it is important to note that the activators us ed in this study
60 are not completely specific, and that they can also potentially activate chimaerins and diacylglycerol kinases (Kazanietz 2005, Yang and Kazanietz 2007). Chimaerins have a diacylglycerol binding site whic h can be activated by PLC and subsequently stimulate RhoGTPase activity (Kazanietz 2005, Yang and Kaza nietz 2007). The activity of Rho-GTPs are typically associated with modul ating the dynamics of actin reorganization in tumorogenesis and development. Although no role for actin in ar restin translocation was demonstrated in Xenopus (Peterson et al. 2003), a dependence on microfilaments for the transloc ation of arrestin to the IS during DA was demonstrated in mouse (Reidel et al. 2008). Our immunohistochemical anal ysis also supports our hypot hesis showing PKC and PLC co-localization in the IS (Fig. 2-1) and reported activity in the OS. Although, we did not detect a significant difference in PKC and PLC local ization between lightand darkadapted photoreceptors (data not shown), PLC activity has been shown to increase by 23% in light adapted photoreceptors (Ghalayini et al. 1992, Gh alayini et al. 1998) suggesting a role for PLC during light adaptation. It is intere sting that rhodopsin has been id entified as the major substrate in the photoreceptor cell for PKC, contributing to at least 50% of rhodopsin phosphorylation at lower levels of illumination (New ton et al. 1993, Williams et al. 1997). Whether this is relevant to arrestin translocation, or whether PKC activ ity couples to arrestin translocation through another substrate is not revealed by our studies. Arrestin Translocation is Energy Dependent The first part of this study shows that arres tin translocation involve s a G-protein regulated pathway that would require both GTP and ATP. Ho wever, arrestins transl ocation to the OS has been proposed as an energy independent mechanis m, requiring only protei n-protein interactions (Nair et al. 2005). Contrary to Na ir et al. (2005), we demonstrate in the second part of this study that arrestin translocation is an energy-dependent process that does not solely depend on protein-
61 protein interaction. The evidence is as follows. First, our results show arrestin does not translocate to the OS in ATP-depleted tadpole eyes or mouse eyes (Fig. 3-5 and Fig. 3-6) Arrestin is predominately localized in the IS in light-adapted photorecep tors under ATP-depleted conditions. Importantly, translocation of arre stin can be recovered by the addition ATP, demonstrating the viability of the ATP-depleted explanted eyes (Fig. 3-5). Our results are strikingly different from Nair et al. (2005), where they showed arres tin is still able to translocate in ATP-depleted eyes. We are unsure as to the source of the di screpancy, although the differences in results may be attributed to more effective penetration of KCN since we were able to deplete ATP levels by three orders of magnitude and whereas the previous study only achieved 100-fold reduction of ATP. Regardless of the source of th e difference, it is clear that in both frog and mouse photoreceptors that arrestin translocation is dependent on ATP. This ATP dependence further supports our hyp othesis that arrestin transloca tion uses a G-protein regulated pathway for the initial distribution to the OS. In the second part of this st udy, we investigated the other ma jor tenet of the Nair et al. (2005) hypothesis, namely that arrestin loca lization was solely dependent on binding to rhodopsin in the outer segment and binding to microtubules in the inner segments. For this investigation, we created an arre stin that had the C-terminal 33 amino acids scrambled, based on the previous demonstration from several laboratories that the absence of the C-terminal 30-35 amino acids affected the transloc ation of arrestin (Sm ith et al. 1994, Peterson et al. 2005, Nair et al. 2005). When heterologously expressed an d characterized for binding to rhodopsin and microtubules, this scrambled arrestin showed excellent selectivity for pRho* (Fig. 3-1) and retained its binding for microtubules (Fig. 3-3). This lack of eff ect on microtubule binding is not surprising since Nair et al. (2005) have shown th at microtubule binding is not a function of the C-terminus, but rather appears to involve much of the same surface on arrestin that is utilized for
62 binding to pRho*. When we introduced the scrambled arrestin and na tive arrestin into HEK cells, the result showed there was only a modest degree of co -localization of arrestin and tubulin, consistent with arrestins low affinity for micr otubules, unlike lebercilin which showed nearly perfect colocalization with tubu lin (Fig. 3-4). However, when we introduced this scrambled arrestin into Xenopus rod photoreceptors, our results show ed that the scrambled arrestin mislocalized to the OS in dark-adapted photore ceptors and remained in the OS in response to light (Fig. 3-2). These results contradict the Nair et al. (2005) model si nce if arrestins localization to the IS during da rk adaptation was based solely on its binding affinity for microtubules, then arrestin should have remained in the IS in the transg enic tadpoles expressing the scrambled arrestin. Furthermore, if arrestins translocation to the OS was based on its affinity for light-activated, phosphorylated rhodopsin, then th e portion of arrestin that was present in the IS should have translocated to the OS, yet ther e was no change in arre stin distribution during light adaptation. Thus, this scrambled arrestin offers incontrovertible evidence that the two partner binding hypothesis is insufficient to fully account for the light-dependent translocation of arrestin. Proposed Pathway Our ability to stimulate arres tin translo cation to the OS in the absence of light demonstrates the complexity associated with arre stin translocation. In lig ht of our findings, we are proposing that arrestin transl ocation to the OS requires two major steps. In our model, the initial distribution of arrestin in response to light is signa led by the phosphoinositide pathway with the final step in the phos phoinositide pathway serving as a g atekeeper to release arrestin to the OS (Fig. 4-1). It appears that arrestin can then rapidl y disperse throughout the basal onethird of the outer segment, probably by diffusi on. However, to rapidly and fully distribute throughout the length of the OS, a rrestin then requires cytoskeletal elements, since we have
63 previously shown that arrestin does not fu lly distribute throughout the OS when frog photoreceptors are treated with thiabendazole, a microtubule poison,. Under these conditions, arrestin strongly localizes to the base of the OS. A similar effect was also found in mouse retinas treated with thiabendazole (Reidel et al. 2008). In keeping with the Nair et al. (2005) model, it appears that binding to light-activ ated rhodopsin is important for re taining arrestin in the outer segment, since in our studies with the PLC a nd PKC agonists, arrestin moved to the outer segments but did not remain as occurs with light-activated translocation. This observation then leads to the obvious question about what is unique about the basal one-third of the OS compared to the distal two-thirds. Structurally, there is no obvious difference, except that there is a larger cytosolic space associated with the axoneme at the base of the OS (Besharse 1986, Besharse et al. 1990, In sanna et al. 2008), perhaps creating fewer barriers for diffusion of arrestin. It is intriguing that a recent i nvestigation of the kinesin motor proteins in rods revealed that a second kinesin, Kif17, is presen t along the singlet microtubules in the apical two-thirds of the OS, whereas Kinesi n II is present in the c onnecting cilium and along the doublet microtubules in the proximal one-thi rd of the axoneme (Insanna and Besharse 2008, Insanna et al. 2008). We also propose that arrestin translocation back to the IS during dark adaptation is also partially dependent on active transport (Fig. 4-2). This claim is based on previous demonstrations that arrestin tran slocation to the IS is perturbed when cytoskeletal elements are disturbed by thiabendazole treatme nt in frogs (Peterson et al. 2 005) and in mouse (Reidel et al. 2008). Our demonstration that the scrambled arrestin does not specifically localize in the inner segments, even though it retains binding to mi crotubules, argues that depolymerization of the microtubules by thiabendazole to remove the binding partner is an insufficient explanation for the mislocalization of arrestin.
64 Our model that signaling of th e phosphoinositide pathway initiates arrestin translocation is supported by several lines of evidence. In our study, we show co-localiza tion of the G-proteins, PLC, and PKC, suggesting these molecules may be working together to stimulate arrestin. Furthermore, the sensitivity of arrestin localiz ation to pertussis toxin and quinpirole treatment underscores the G-protein mediation in signaling a rrestin translocation. Fi nally, our ability to initiate translocation with PKC and PLC agonists, or to slow translocation with PKC and PLC antagonists, offer final evidence for involvement of PLC and PKC in arrestin translocation. Our model is also supported by other studies, partic ularly the work of Strissel et al. (2006) who reported that a critical threshold of light is necessary to releas e a suprastoichiometric amount of arrestin into the OS, an amount that exceeds the available activated rhodopsin by thirty fold. In summary, our current studies and the findings in previous studies show that arrestin translocation is most likely a combination of both passive and active tran sport that requires an energy dependent process (Fig. 4-2). The comb ination of our ATP de pletion studies and engineered arrestin studies strong ly suggests that arres tin translocation between the OS and IS is more complicated then simple diffusion. Arrestin translocation does require ATP to move to the outer segments in response to light, and binding to microtubules is insufficient to explain the return of arrestin to the inner segment during dark adaptation.
65 PKC* PKC* ATP ADP P P R h R* G GDP G* GTP GTP GDP G GDP P o o11cis alltransrhodopsin PLC PLC PLC* PLC* PLC PLC PKC PKCArrestin Translocates Figure 4-1. Proposed phosphoinositide gatekeeper pathway for a rrestin translocation. We are proposing arrestin translocation to the outer segments may involve the phosphoinositide pathway. We hypothesize that li ght-activated rhodopsin can initiate the cascade by activating a Gi G-protein. The G-protein can then activate phospholipase C (PLC) to cleave phosphatidy linositol 4, 5-bisphosphate (PIP2) to produce diacylglycerol (DAG) and inositol trisphosphate (IP3). IP3 can then activate Ca2+ release in the cell to initiate protein kinase C phos phorylation (PKC). Activated PKC can then phosphorylate some undetermined target protein(s) to subsequently release arrestin to translocate to the OS.
66 DA LA LA DA =Arrestin =Motor proteinsOS ISA. B. Figure 4-2. Model for arrestin tran slocation. A) We propose arrestin translocation to the OS may require both passive and active mechanisms. In response to light, arrestin passively translocates to the base of the OS, onl y requiring signaling from the phosphoinositide molecules identified in this study to open the gate. A rrestin distribution throughout the length of the OS is then a combin ation of passive diffusion and microtubuleassociated motor proteins such as kinesin th at then actively trans port arrestin from the base of the OS to the distal OS. B). Arrest in translocation back to RIS appears to rely on a microtubule-associated moto r proteins, such as dynein.
67 CHAPTER 5 SUMMARY AND PERSPECTIVE The purpose of this study is to better understa nd the molecular requirem ents for arrestin translocation and the molecules associated with signaling arrestin translocation. Collectively, our results show arrestin translocation to the OS is an energy dependent process that is G-protein regulated. We investigated the notion that arrestin transloca tion is only based on its binding affinity for microtubules in the IS and rhodopsin in the OS and found this model to be an oversimplification of the translocation process. In the first part of this study, we show th at the signaling cascad e initiating arrestin translocation appears to use a G-protein to activate the phosp hoinositide pathway. Although it is clear that PLC and PKC activation can initiate arrestin transloca tion, several important features remain obscure. For example, it is unclear which specific isozymes of PKC and PLC are participating in the pathway or which G-protein is mediating PLC activation. Determining these specific components would permit targeted disrup tion of the transloca tion process to better understand the functioning of arrestin translocat ion in photoreceptors. Another important point to address is to determine how PLC and PKC activation couples to arrestin to signal arrestin mobilization to the outer segments. We can determine the role of these proteins through process of elimination. First, PKC is the major isozyme of PKC in the rod photorecep tors. To determine if arrestin is using PKC in its pathway, we can knock-out PKC in mouse and observe arrestin translocation to the OS. If arrestin does use PKC then we should expect a reduction or lack of arrestin translocation. The pitfall associated with this appr oach is the photoreceptor cells abil ity to compensate for the lost enzyme. The same approach can be used for the other PKC isozymes in the photoreceptor cell to determine if there are associated with arrestin translocation.
68 There are a couple isozymes of PLC that has been shown to play a role in visual response: PLC and PLC PLC has been shown to be involved in cell to cell signaling, whereas PLC has been shown to differential activity in dark and light conditions. To determine if arrestin translocation is u tilizing PLC, tadpole eyes can be treated with specific inhibitors against PLC gamma. If arrestin is using PLC, we should expect a change in arrestin translocation to the OS. We could also try specific inhib itors against other isoz ymes of PLC. Single experiments with only one inhibitor can skew the results since the cell has the ability to compensate for loss of function. It would be intere sting to observe arrestin translocation in the presence of all the PLC inhibitors. We have also shown arrestin translocation is a G-protein re gulated pathway and we have identified several G-proteins in the rod photor eceptors. We can specifically target these Gproteins with activators and inhibitors to determ ine their role in signali ng arrestin translocation. For example, melittin is an activator of Gi and G11 but inhibits protein kinase C activity. These experimental results will elucidate if Gi G-proteins and PKC play a ro le in signaling arrestin. Just as there were undetermined factors in th e first part of the study, the second part of this study also showed some discrepancies. Our results showed arrestin translocation does not entirely depend on binding to microtubules and rhodopsin. The mislocalization of modified arrestins that retain both bindi ng to activated rhodopsin and to microtubules is a clear indication that more than these two binding partners are needed for proper tr anslocation of arrestin. Our evidence further shows that arrestin translocation requires ATP to fully distribute to the OS in response to light, which is di fferent from Nair et al. (2005) previous investigations. It is clear that our model does not provide a complete description of the translocation process. For example, the mechanism by which arre stin is retained in the inner segment in the dark needs to be elucidated. Such a determination might well be approached by in vivo chemical
69 cross-linking coupled with tandem mass spectrome tric analyses. These results will not only reveal what arrestin binds in the dark, but may also suggest a possible role for arrestin in darkadapted photoreceptors. We should be mindful of the possible binding partners for arrestin when we are conducting these experiments, There has been speculation in possi ble binding partners for visual arrestin including calm odulin, JNK, and Mdm2 (Song et al. 2006, Wu et al. 2006, Hanson et al. 2007). The suggested role of arrestin binding to thes e proteins is to faci litate trafficking of these proteins throughout the cell. In this study, we also showed that arrestin translocation to the OS requires ATP. Where ATP factors into the transloca tion process is not fully defined, although our study suggests that ATP could potentially function both in post-tr anslational modifications for signaling through PKC and as an energy source for microtubule-assisted distribution of arrestin to the distal outer segments. Investigations that titrate the minimu m levels of ATP required for arrestin translocate may help separate these two points of ATP requirement, and help us unde rstand the expenditure of ATP in the translocation process. Ultimately, the goal of our research is to unde rstand the function of arrestin translocation in the photobiology of photoreceptors. These results presented here provid e significant advances towards delineating the mechanism for translocati on, and have identified key points that should permit us to generate a system in which arrestin retains its function of quenching rhodopsin but is deficient in translocation. Analysis of the photoresponse and photoreceptor cell biology in such a system in future studies should reveal the function of arrestin translocation.
70 LIST OF REFERENCES 1. Akopian A and Witkovsky P. D2 dopami ne receptor-m ediated inhibition of a hyperpolarization-activated cu rrent in rod photoreceptors J Neurophysiol. (1996) Sep;76(3):1828-35. 2. Alfano M, Schmidtmayerova H, Amella CA, Pushkarsky T, Bukrinsky M. The Boligomer of pertussis toxin deactivates CC chemokine recepto r 5 a nd blocks entry of Mtropic HIV-1 strains. Am J Physiol Cell Physiol. (2001) May;280(5):C1233-41. 3. Besharse JC. Photosensitive me mbrane tur nover: differentiated me mbrane domains and cell-cell interaction. (198 6) In: Adler R Farber D edito rs. The retina: a model for cell biological studies. Part I. New York: Academic Press. p 297-352. 4. Besharse JC, Horst CJ. The photoreceptor c onnecting cilium. A model for the transition zone (1990). In: Bloodgood RA editor. Cili ary and flagellar membranes. New York: Plenum Publishing Corp. p 389-417. 5. Burns M.E. and. Baylor D.A, Activation, deactivation, and adapta tion in vertebrate, photoreceptor cells. Annu. Rev. Neurosci. 24, (2001), pp.779-805. 6. Chaitin MH and Bok D. Immunof erritin localization of actin in retinal photoreceptors. Invest. Ophthalmol. Vis. Sci. 27:(1986) pp 1764-1767. 7. Dearry, A. and Burnside, B. (1988) Stimulation of distinct D2 dopaminergic and alpha2adrenergic receptors induces light-adaptive pigment dispersion in teleost retinal pigment epithelium. J. Neurochem. 51: 1516-1523. 8. Dearry, A. and Burnside, B. Light-induced dopami ne release from teleost retinas acts as a light-adaptive signal to the re tinal pigment epithelium. J. Neurochem. (1989) 53: 870878. 9. Dearry, A., Edelman, J. L., Miller, S. and Burnside, B.,Dopamine induces light-adaptive retinomotor movements in bullfrog cones vi a D2 receptors and in retinal pigment epithelium via D1 receptors. J. Neurochem. (1990) 54: 1367-1378. 10. den Hollander AI, Koenekoop RK, Mohamed MD, Arts HH, Boldt K, Towns KV, Sedmak T, Beer M, Nagel-Wolfrum K, McKibbin M, Dharmaraj S, Lopez I, Ivings L, Williams GA, Springell K, Woods CG, Jafri H, Rashid Y, Strom TM, van der Zwaag B, Gosens I, Kersten FF, van Wijk E, Veltman JA, Zonneveld MN, van Beersum SE, Maumenee IH, Wolfrum U, Cheetham ME, Ueffing M, Cremers FP, Inglehearn CF, Roepman R. Mutations in LCA5, encoding th e ciliary protein lebe rcilin, cause Leber congenital amaurosis. Can J Physiol Pharmacol. (1996) May;74(5):559-64. 11. Dizhoor AM, Nekrasova ER, Philippov PP. The binding of G proteins to immobilized delipidated rhodopsin. Biochem Biophys Res Commun. ( 1989) Jul 14;162(1):544-9. 12. Eckmiller MS Microtubules in a rod specific cy toskeleton associated with outer segment incisures. Visual Neurosci.17;(2000) pp 711-722.
71 13. Fan SF, Yazulla S. Inhibitory in teraction of cannabinoid CB1 recep tor and dopamine D2 receptor agonists on voltage-gated currents of goldfish cones. Vis Neurosci. (2004) JanFeb;21(1):69-77. 14. Garcia JG, Wang P, Liu F, Hershenson MB, Borbiev T, Verin AD. Pertussis toxin directly activates endothe lial cell p42/p44 MAP kinases vi a a novel signaling pathway. Am J Physiol Cell Physiol. 2001 May;280(5):C1233-41. 15. Ghalayini AJ, Anderson RE. Activation of bovine rod outer segment phospholipase C by arrestin. J Biol Chem 1992 Sep 5;267(25):17977-82. 16. Ghalayini AJ, Weber NR, Rundle DR, Koutz CA, Lambert D, Guo XX, Anderson RE. Phospholipase Cgamma1 in bovine rod outer segments: immunolocalization and lightdependent binding to m embranes. J Neurochem. 1998 Jan;70:171-8. 17. Gotow T, Nishi T. Involvement of a Go-type G-pr otein coupled to guanylate cyclase in the phototransduction cGMP cascade of molluscan sim ple photoreceptors. Brain Res. (2007) May 4;1144:42-51. Epub 2007 Jan 26. 18. Greene NM, Williams DS, Newton AC. Kinetics and localiz ation of the phosphorylation of rhodopsin by protein kinase C. J Biol Chem. 1995 Mar 24;270(12):6710-7. 19. Greene NM, Williams DS, Newton AC. Identification of protein kinase C phosphorylation sites on bovine rhodopsin. J Biol Chem. 1997 Apr 18;272(16):10341-4. 20. Hanson SM, Cleghorn WM, Francis DJ, Vishnivetskiy SA, Raman D, Song X, Nair KS, Slepak VZ, Klug CS, Gurevich VV. Arrestin mobilizes signaling proteins to the cytoskeleton and redirects their activity. Nat Genet. (2007) Jul;39(7):889-95. Epub 2007 Jun 3. 21. Hillman, D., Li, D. and Burnside, B. D4-like dopam ine receptors mediate inhibition of dark-induced elongation in tele ost retinal cone inner-outer segments (CIS-COS). J. Neurochem. (1995) 64: 1326-1335. 22. Insinna C, Besharse JC. Intraflagellar tr ansport and the sensory outer segment of vertebrate photoreceptors. De v Dyn. (2008) May 19;237(8):1982-1992. 23. Jiang H Lyubarsky A, Dodd R, Vardi N, Pugh E, Baylor D, Simon MI, Wu D. Phospholipase C beta 4 is involved in m odulating the visual response in mice. Proc Natl Acad Sci U S A. 1996 Dec 10;93(25):14598-601. 24. Kaplan MW, Iwata RT, and Sears RC. Lengths of I mmunolabeled ciliary microtubules in frog photoreceptor outer segments. Exp. Eye. Res. (1987) 44: pp623-632. 25. Kazanietz MG.Targeting protein kinase C and "non-kinase" phorbol ester receptors: emerging concepts and therapeutic implications. Biochim Biophys Acta. (2005) Dec 30;1754(1-2):296-304. Epub 2005 Sep 12.
72 26. Kelleher DJ, Johnson GL. Purification of PKC fr om bovine rod outer segment. J. Cyclic Nucleotide protein Phosphor. Res. 1985 10:579-591. 27. Keranen LM, Dutil EM, Newton AC. Protein kinase C is regulated in vivo by three functionally distin ct phosphorylations. Curr Biol. 1995 Dec 1;5(12):1394-1403. 28. Kosaka J, Suzuki A, Morii E, Nomura S. Di ff erential localization a nd expression of alpha and beta isoenzymes of protein kinase C in the rat retina. J Ne urosci Res.;54(1998):655663. 29. Koulen P, Liu J, Nixon E, Madry C. Interaction between mGluR8 and calcium channels in photoreceptors is sensitive to pertu ssis toxin and occurs via G protein betagamma subunit signaling. Invest Ophthalmol Vis Sci. (2005) Jan;46(1):287-91. 30. Kroll, K.L. and Amaya, E., 1996. Transgenic Xenopus embryos from sperm nuclear transplantations reveal FG F signalling requirements duri ng gastrulation. Developm ent 122, pp. 3173. 31. Lane JR, Powney B, Wise A, Rees S, Milligan G. G protein coupling and ligand selectivity of the D2L a nd D3 dopamine receptors. J Pharmacol Exp Ther. (2008) Apr;325(1):319-30. Epub 2008 Jan. 32. Lee SJ, Montell C. Light-dependent transloc ation of visual arrestin regulated by the NINAC myosin III. Neuron. 43(2004): 95-103. 33. Lin-Jones J, Parker E, Wu M, Dos A and Burnside B. Myosin 3A transgene expression produces abnormal actin filam e nt bundles in transgenic Xenopus laevis rod photoreceptors J. Cell Sci. 117: (2004) pp 5825-5834. 34. Mangini NJ, Garner GL, Okajima TI, Donoso LA, Pepperberg DR. Effect of hydroxylamine on the subcellular distributi on of arrestin (S -antigen) in rod photoreceptors. Vis Neurosci. : 11(1994):561-568. 35. Mangini NJ, Peperberg DR. Immunolocaliza tion of 48K in Rod Photoreceptors: Light and ATP Increase in OS Labeling. Inves. Ophth.Vis. Sci. 29 (1988)1221-1234. 36. Marszalek JR, Liu X, Roberts EA, Chui D, Martha JD, Williams DS, Goldstein LS. Genetic evidence for selective transport of ops in and arrestin by kine sin-II in mammalian photoreceptors. Cell.102 (2000):pp 175-87. 37. Martin G, Ahmed SH, Blank T, Spiess J, Koob GF, Siggins GR.., Chronic morphine treatment alters NMDA receptor-m ediated synaptic transmission in the nucleus accumbens. J. Neurosci. (1999) 19, 9081-9089, 38. McCormack, C.A. and Burnside, B. A ro le for endogenous dopami ne in circadian regualtion of retinal cone movement. Exp. Eye Res. (1993) 55: 511-520. 39. McDowell JH, Nawrocki JP, Hargrave PA. Phosphorylati on sites in bovine rhodopsin. Biochemistry. 1993 May 11;32(18):4968-74.
73 40. McDowell, J.H., Smith, W.C., Miller, R.L., Popp, M.P., Arendt, A., Abdulaeva, G. and Hargrave, P.A., 1999. Sulfhydryl reactivity dem onstrates different conforma tional states for arrestin, arrestin activat ed by a synthetic phosphopeptid e, and constitutively active arrestin. Biochemistry 38, pp. 6119. 41. McGinnis JF, Matsumoto B, Whelan JP, Ca o W (2002) Cytoskeleton participation in subcellular trafficking of si gnal transduction proteins in rod photoreceptor cells. J Neurosci Res 67:290. 42. Mendez A, Lem J, Simon M, Chen J. Light-dep endent translocation of arrestin in the absence of rhodopsin phosphoryla tion and transducin signali ng. J Neurosci. 23 (2003):pp 3124-3129. 43. Muresan Z, Besharse JC. D2-like dopamine recepto rs in amphibian re tina : localization with fluorescent ligands. J Comp Neurol. 1993 May 8;331(2):149-60. 44. Nair KS, Hanson SM, Kennedy MJ, Hurley JB, Gurevich VV, Slepak VZ. Direct binding of visual arrestin to microtubules determ ines the differential subcellu lar localization of its splice variants in rod photoreceptor s. J Biol Chem. 279 (2004):pp 41240-41248. 45. Nair KS, Hanson SM, Mendez A, Gurevi ch EV, Kennedy MJ, Shestopalov VI, Vishnivetskiy SA, Chen J, Hurley JB, Gu revich VV, Slepak VZ. Light-dependent redistribution of arrestin in vertebrate rods is an energyindependent process governed by protein-protein interacti ons. Neuron. 46 (2005): pp 555-567. 46. Nair KS, Hanson SM, Kennedy MJ, Hurley JB Gurevich VV, and Slepak VZ. Direct Binding of Visual Arrestin to Microtubules Determines the Differential Subcellular Localization of Its Splice Vari ants in Rod Photoreceptors. J Mol Biol. (2007) Apr 27;368(2):375-87. Epub 2007 Feb 22. 47. Newton AC, Williams DS. Rhodopsin is the major in situ substrate of protein ki nase C in rod outer segments of photoreceptors. J Biol Chem. 1993 Aug 25;268(24):18181-6. 48. Newton AC. Protein Kinase C: Structure, F unction, and Regulation. J. Biol. Chem. 1995 Dec :4828495-28498. 49. Ohguro H, Van Hooser JP, Milam AH, Palczewski K. Rhodopsin phosphorylation and dephosphorylation in vivo. J Biol Chem. 270 (1995): pp 14259-14262. 50. Pardue, ML. In situ hybidiztio. In Nucleic Acid Hybridization: A Practical Approach (B.D. Hames and S.J Higgins, eds.) IR L Press, Oxford (1985) pp. 179-202. 51. Peet JA, Bragin A, Calvert PD, Nikonov SS, Ma ni S, Zhao X, Besharse JC, Pierce EA, Knox BE, Pugh EN Jr (2004) Quantification of the cytoplasmic spaces of living cells with EGFP reveals arrestin-EGFP to be in disequilibrium in dark adapted rod photoreceptors. J Cell Sci 117:3049.
74 52. Peng YW, Rhee SG, Yu WP, Ho YK, Schoen T, Chader GJ, Yau KW. Identification of components of a phosphoinositide signaling pa thway in retinal rod outer segments. Proc Natl Acad Sci U S A. 1997 Mar 4;94(5):1995-2000. 53. Peterson JJ, Tam BM, Moritz OL, Shelam er CL, Dugger DR, McDowell JH, Hargrave PA, Papermaster DS, Smith WC. Arrestin migrates in photoreceptors in response to light: a study of arrestin localization using an arre stin-GFP fusion protein in transgenic frogs. Exp Eye Res. 76 (2003): pp 553-563. 54. Peterson JJ, Orisme W, Fellows J, McDowe ll JH, Shelamer CL, Dugger DR, Smith WC. A role for cytoskeletal elements in the lig ht-driven translocation of proteins in rod photoreceptors. Invest Ophthalmol Vis Sci. 2005 Nov;46(11):3988-98. 55. Reidel B, Orisme W Goldmann T, Smith WC Wolfrum U. Photoreceptor vitality in organotypic cultures of mature vertebrate retinas validated by light-dependent molecular movements. Vis.Research 46 (2006) 4464-4471. 56. Reidel B, Goldmann T, Giessl A, Wolfrum U The translocation of signaling molecules in dark adapting mammalian rod photoreceptor cells is dependent on the cytoskeleton. Cell Motility and Cyto. (2008). 57. Roof D, Adamian M, Jacobs D, and Hayes A Cytoskeletal specializations at the rod photoreceptor distal tip. J. Comp.Neurol.305 (1991): pp4093-4100. 58. Sagi-Eisenberg R, Traub LM, Spiegel AM, Zick Y. Protein ki nase C-mediated phosphorylation of retinal rod outer segment m embrane p roteins. Cell Signal. 1989;1(5):519-31. 59. Shulman LM and Fox DA.Dopa mine inhibits mammalian photoreceptor Na+,K+-ATPase activity via a selectiv e effect on the alpha3 isozyme. Pr oc Natl Acad Sci (1996) 93(15): 8034. 60. Smith WC, Milam AH, Dugger D, Arendt A, Hargrave PA, Palczewski K. A splice variant of arrestin. Mole cular cloning and localization in bovine retina. J Biol Chem. (1994) Jun 3;269(22):15407-10. 61. Smith WC, Peterson JJ, Orisme W, Dincul escu A.Arrestin translocation in rod photoreceptors. Adv Exp Me d Biol. (2006);572:455-64. 62. Sommer ME Farrens DL, McDowell JH, Webe r LA, Smith WC. Dynamics of arrestinrhodopsin interactions: loop movement is invol ved in arrestin activation and receptor binding. J Biol Chem. (2007) Aug 31;282(35):25560-8. Epub 2007 Jul 2. 63. Song X, Raman D, Gurevich EV, Vishnivetskiy SA, Gurevich VV.Visual and both nonvisual arrestins in their "inactive" conf or mation bind JNK3 and Mdm2 and relocalize them from the nucleus to the cytoplasm. J Biol Chem. 2006 Jul 28;281(30):21491-9. Epub 2006 May 31.
75 64. Strissel KJ, Sokolov M, Trieu LH, Arshavsky VY. Arrestin translocation is induced at a critical threshold of visual signaling and is superstoichi ometric to bleached rhodopsin. J Neurosci. 2006 Jan 25;26(4):1146-53. 65. Udovichenko IP, Cunnick J, Gonzales K, Take moto DJ.Phosphorylation of bovine rod photoreceptor cyclic GM P phosphodiesterase J. Biochem 1993 Oct 1;295:49-55. 66. Udovichenko IP, Cunnick J, Gonzalez K, Takemoto DJ. Functional effect of phosphorylation of the photoreceptor phosphodies terase inhibitory subunit by protein kinase C. J Biol Chem. 1994 Apr 1;269(13):9850-6. 67. Udovichenko IP, Cunnick J, Gonzalez K, Yakhni n A, Takemoto DJ. Protein kinase C in rod outer segments: effects of phosphoryla tion of the phosphodiesterase inhibitory subunit. Biochem J. 317 ( 1996):291-295. 68. Udovichenko IP, Newton AC, Williams DS. Contribution of protein kinase C to the phosphorylation of rhodopsin in intact retinas. J Biol Chem. 1997 Mar 21;272(12):79529. 69. Weber B, Schlicker E. Modulation of dopamine re lease in the guinea-pig retina by G(i)but not by G(s)or G(q)-protein-coupled receptors. Fundam Clin Pharmacol. (2001) Dec;15(6):393-400. 70. Weyand I, Kuhn H. Subspecies of arrestin from bovine retina. Equal functional binding to photoexcited rhodopsin but various isoe lectric focusing phenotypes in individuals. Eur J Biochem. 1990 Oct 24;193(2):459-67. 71. W illiams DS, Linberg DK, Vaughan DK, Fari ss RN, and Fisher SK. Disruption of microfilament organization and deregulati on of disk membrane morphogenesis by cytochalasin D in rod and cone photoreceptors. J. CompNeurol.272 (1988): pp161-176. 72. Williams DS, Liu X, Schlamp CL, Ondek B, Jaken S, Newton AC. Characterization of protein kinase C in photoreceptor outer segments. J Neurochem;69 (1997):1693-1702. 73. Witkovsky P, Stone S, Besharse JC. Dopamine modifies the balance of rod and cone inputs to horizontal cells of the Xenopus retina. Brain Res. (1988) May 24;449(1-2):3326. 74. Wong WS, Rosoff PM. Pharmacology of pe rtussis toxin B-oligomer. J Exp Med. (1999) Sep 6;190(5):597-605. 75. Wood JG, Hart CE, Massei GF, Girard PR, a nd Kuo JF. Distribution of protein kinase C immunoreactivity in rat retin a. J. Histochem 30,63-68. 76. Wu N, Hanson SM, Francis DJ, Vishnivetskiy SA, Thibonnier M, Klug CS, Shoham M, Gurevich VV. Arrestin binding to calmodu lin: a direct interaction between two ubiquitous signaling proteins. J Mol Biol. 2006 Dec 15;364(5):955-63. Epub 2006 Oct 3.
76 77. Yang C, Kazanietz MG Chimaerins: GAPs that br idge diacylglycerol signalling and the small G-protein Rac. Biochem J. (2007) Apr 1;403(1):1-12. 78. Zhang J, Wu SM. Goalpha labels ON bipolar cells in the tiger salamander re tina. J Comp Neurol. (2003) Jun 23;461(2):276-89.
77 BIOGRAPHICAL SKETCH Wilda Orisme was born November 9, 1982 in Port-au-Prince, Haiti to Lucienne Orisme and Gesner Orisme. She has one sister and f our brothers: Carole Theophin, Smith Orisme, Charlemagne Orisme, Eben Orisme and Lequin Or isme. She arrived to United States in 1985 and resided in New York City, New York. She grew up in New York City graduating in 1999 from A. Philip Randolph Campus High School. She then pursued her bachelors degree at Florida Memorial University in Miami, Florida in 2003. There she obtained a Bachelors of Science in Chemistry. After college, Wilda attended the University of Florida to obtain her Doctor of Philosophy in Biomedical Sciences with a con centration in Neuroscience. She fulfilled her requirements for a Ph.D in August 2008. Wilda pl ans to continue research and obtain a postdoctoral position.