UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Expression and characterization of retinal disease linked ABCR mutants Safarpour, Azien 2001

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2001-0510.pdf [ 11.46MB ]
Metadata
JSON: 831-1.0090157.json
JSON-LD: 831-1.0090157-ld.json
RDF/XML (Pretty): 831-1.0090157-rdf.xml
RDF/JSON: 831-1.0090157-rdf.json
Turtle: 831-1.0090157-turtle.txt
N-Triples: 831-1.0090157-rdf-ntriples.txt
Original Record: 831-1.0090157-source.json
Full Text
831-1.0090157-fulltext.txt
Citation
831-1.0090157.ris

Full Text

EXPRESSION AND CHARACTERIZATION OF RETINAL DISEASE LINKED ABCR MUTANTS by A Z I E N S A F A R P O U R B.Sc. (Hons.), The University of British Columbia, 1999 A T H E S I S S U B M I T T E D IN P A R T I A L F U F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F S C I E N C E in T H E F A C U L T Y O F G R A D U A T E S T U D I E S (Department of Biochemistry and Molecular Biology) We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A September 2001 © Az ien Safarpour, 2001 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of OcWYvnST'j T£>vc)^T] The University of British Columbia Vancouver, Canada Date QcToUgr 2-^ i ^ DE-6 (2788) ABSTRACT The retinal A B C transporter, A B C R , is a 220 kDa glycoprotein localized in the outer segments of the photoreceptor cells. A s a member of the A B C superfamily, A B C R is organized into two tandem halves each containing a transmembrane domain and a nucleotide binding domain. Studies have suggested that A B C R may either act as an all-trans-retinal extruder or a retinylidene-phosphatidylethanolamine flippase. Eighty-nine mutations in ABCR gene have been identified and associated with a number of retinal degenerative diseases including Stargardt's disease. In this study the effects of Stargardt's disease causing mutations (D846H, T1526M, R2038W, R2077W and R2106C) on the structure and function of A B C R were investigated. A s a first step toward this goal, the five mutant constructs were expressed in the mammalian C O S cell expression system. When the expression levels of the proteins were compared, it was found that the R2038W and R2077W mutants were expressed at 57-59% of the wild type while R2106C and T1526M expressed at 71% and 81% of the wild type protein, respectively. The expression level of the D846H mutant varied between 31-68% of the wild type, depending on the transfection method used. The recombinant proteins were purified from C H A P S solubilized C O S cell by immunoaffinity chromatography. Immunofluorescence microscopy of the transfected C O S cells revealed that the expressed wild type A B C R was localized in the vesicles, whereas almost 100% of both D846H and T1526M, 77% of R2038W, 80% of R2077W and 34% of R2106C showed an endoplasmic reticulum/Golgi labeling pattern. Calnexin interaction studies showed that all of the mutants examined co-purified with calnexin to a greater extent than the wild type A B C R . The ATPase activity of purified proteins reconstituted in liposomes rich in phosphatidylethanolamine was measured. In the presence of 50 u M all-frvms-retinal, the basal activity of the wild type A B C R was increased by 1.6 fold; however, no stimulation was seen in T1526M and R2038W variants. The ATPase activities of both D846H and R2077W were impaired, whereas the R2106C mutant showed a retinal-stimulated activity ii similar to the wild type. The az ido-ATP labeling study showed that the wild type, T 1 5 2 6 M and R2106C bound A T P , whereas the D846H and R2077W did not. iii T A B L E O F C O N T E N T S Abstract i i Table of Contents iv List of Tables vi i List of Figures vi i i List of Abbreviations x Acknowledgements xi i i Dedication xiv 1. I N T R O D U C T I O N 1 1.1 The Human Eye 1 1.2 The Retina 1 1.3 The Photoreceptor Cells 4 1.4 The Rod Outer Segment and the Disk 6 1.5 Phototransduction 7 1.5.1 The Dark Current 7 1.5.2 Photoexcitation and Recovery 10 1.5.3 The Visual (Retinoid) Cycle 14 1.6 Retinal Pigment Epithelium 14 1.7 Macular Degeneration and Stargardt's Disease 20 1.8 The Superfamily of A B C Proteins 22 1.8.1 The Structure of A B C Transporters 22 1.8.2 A B C A Subfamily and A B C R 25 1.9 Possible Functions of A B C R 29 1.10 Thesis Investigation 30 2. M A T E R I A L A N D M E T H O D S 34 2.1 Materials and Buffers 34 2.2 Molecular Biology Techniques 35 2.3 Generation of 3F4 Coupled Sepharose Beads 37 2.4 Generation of the Constructs 37 iv 2.5 Expression of the Constructs in C O S Cells '41 2.5.1 Maintenance of C O S Cells 41 2.5.2 Transfection of C O S Cells Using Calcium Chloride 41 2.5.3 Transfection of C O S Cells Using SuperFect Reagent 42 2.5.4 Harvesting of the C O S Cells 42 2.5.5 Purification of A B C R from Solubilized C O S Cells 43 2.5.6 Immunofluorescence Labeling of C O S Cells 43 2.6 C O S Cell Membrane Preparation 44 2.7 Bovine Rod Outer Segment Preparation 45 2.8 Immunoaffinity Purification of A B C R from R O S 46 2.9 Extraction of Phospholipids from R O S 46 2.10 Determination of Lipid Phosphorus Content 47 2.11 Reconstitution of A B C R Proteins into Lipid Vesicles 48 2.12 A T P Hydrolysis of Reconstituted Proteins 49 2.12.1 ATPase Assay 49 2.12.2 Activity Calculations 49 2.13 Az ido-ATP Labeling of C O S Membranes 50 2.13.1 Membrane Preparation 5 0 2.13.2 Az ido-ATP Assay 51 2.14 Protein Electrophoresis and Western Blotting 52 2.15 Protein Determinations 53 3. R E S U L T S 54 3.1 Selection of the A B C R Mutants 54 3.2 Generation of Constructs 56 3.3 Transfection and Expression of W i l d Type and Mutant A B C R 56 3.4 Purification of Wi ld Type and Mutant A B C R from Solubilized C O S Cells 58 3.5 Immunofluorescence Microscopy 62 3.6 Calnexin Association 65 3.7 Reconstitution of A B C R s into Liposomes 65 3.8 The ATPase Assay 67 3.9 A z i d o - A T P Labeling 74 v 4. D I S C U S S I O N 77 4.1 Expression of W i l d Type and Mutant A B C R s in C O S Cells 77 4.2 Purification and Reconstitution of A B C R Variants 78 4.3 The ATPase Assay 79 4.4 Analysis of the Expressed Wi ld Type and Mutant A B C R s 81 4.4.1 Wi ld Type A B C R 81 4.4.2 D846H 83 4.4.3 T 1 5 2 6 M 84 4.4.4 R2038W 85 4.4.5 R2077W 86 4.4.6 R2106C 87 4.5 Summary 88 5. R E F E R E N C E S 90 vi L I S T O F T A B L E S Table 1 Major rod outer segment plasma membrane and disk membrane proteins 8 Table 2 Synthetic oligonucleotides designed and used for site directed mutagenesis 39 Table 3 Expression levels of mutant A B C R s relative to wild type A B C R in C O S cells 60 vii L I S T O F F I G U R E S Fig 1 Anatomy of the human eye 2 Fig 2 Organization of the retina 3 Fig 3 The rod and cone photoreceptor cells 5 Fig 4 Dark current in the rod photoreceptor cells 9 Fig 5 The process of phototransduction in the rod outer segment 11 Fig 6 Schematic diagram depicting the phagocytosis of the rod outer segment by the retinal pigment epithelium 15 Fig 7 The visual cycle and the role of retinal pigment epithelium 18 Fig 8 Normal and Stargardt's disease affected macula 21 Fig 9 Prototype domain arrangements in A B C proteins 23 Fig 10 Putative topological models of A B C R 27 Fig 11 Possible functions of A B C R 31 Fig 12 Strategy used to generate D846H construct 40 Fig 13 Putative model of A B C R showing the five Stargardt's disease mutations 55 Fig 14 Generation of the selected constructs 57 Fig 15 Expression of the wild type and the five mutant A B C R constructs in C O S cells 59 Fig 16 Purification of wild type and mutant A B C R s from solubilized C O S cells 61 Fig 17 Immunofluorescence labeling of C O S cells transfected with wild type and mutant A B C R constructs 63 Fig 18 Co-purification of calnexin with mutant A B C R s 66 Fig 19 A B C R proteins reconstituted in liposomes 68 Fig 20 The effect of all-frans-retinal on the A T P hydrolysis of reconstituted R O S A B C R 69 Fig 21 Basal and retinal stimulated A T P hydrolysis of R O S A B C R reconstituted in two different lipid mixtures 70 Fig 22 The effect of a\\-trans-ret'ma\ concentration on the ATPase activity of reconstituted R O S A B C R 71 viii Fig 23 Basal and retinal stimulated ATPase activities of R O S A B C R and wild type A B C R from transfected C O S cells Fig 24 A T P hydrolysis of reconstituted wild type and mutant A B C R s purified from transfected C O S cells F ig 25 A z i d o - A T P labeling and yield of wild type and mutant A B C R s ix L I S T O F A B B R E V I A T I O N S A B C A T P binding cassette A B C R retinal A B C transporter A D P adenosine 5'-diphosphate A L D P transporter protein responsible for adrenoleukodystrophy A T P adenosine 5'-triphosphate B C A bicinchroninic acid B E S N,N-bis(2-hydroxyethyl)-2-aminoethanesulfonic acid B H T butylated hydroxytoluene bp base pairs B S A bovine serum albumin c A M P adenosine 3',5'-cyclic monophosphate c D N A D N A reverse-transcribed from an m R N A template (copy D N A ) C F T R cystic fibrosis transmembrane regulator c G M P guanosine 3' ,5'-cyclic monophosphate C H A P S 3 - [(cholamidopropyl)-dimethylammonio] -1 -propanesulfonate C R B P cellular retinal-binding protein CY3 fluorescent cyanine dye D M E M Dulbecco's modified Eagle medium d N T P deoxyribonucleoside triphosphate D O P E 1,2-dioleoylphosphatidylethanolamine D T T dithothrieitol E C L enhanced chemiluminescence E D T A ethylaminediamine tetraacetic acid F C S fetal calf serum G A P GTPase-activating protein G C A P guanylate cyclase-activating protein G D P guanosine 5'-diphosphate G T P guanosine 5'-triphosphate H E P E S N-2-hydroxyethylpiperazine-N '-2-ethanesulphonic acid x HH1 highly hydrophobic sequence Ig immunoglobulin IRBP interphotoreceptor retinal-binding protein kb kilobase K m Michaelis constant L B Luria-Bertani broth L R A T lecithin-retinol acyltransferase M D R multi-drug resistance M H C major histocompatibility complex M R P multidrug resistance associated protein N B D nucleotide binding domain P A G E polyacrylamide gel electrophoresis PBS phosphate buffered saline P B S - T phosphate buffered saline with 0.05% Tween 20 P C phosphatidylcholine P C R polymerase chain reaction P D E phosphodiesterase P E phosphatidylethanolamine PI phosphatidylinositol PS phosphatidylserine R* metarhodopsin II rds retinal degeneration slow R G S regulator of G protein signaling R O S rod outer segment R P E retinal pigment epithelium S D P E 1 -stearoyl-2-docosahexaenoylphosphatidyl ethanolamine SDS sodium dodecyl sulfate S U R sulfonylurea receptor T apy transduction complex T A P transporter associated with peptide presentation T B S Tris-buffered saline T M D transmembrane domain X I Tris [hydroxymethyl] aminoethane volume per volume weight per volume xii A C K N O W L E D G E M E N T S I would like to thank Dr. Robert Molday for giving me the opportunity to work on this interesting project and encouraging me to further my knowledge. I would like to acknowledge Jinhi A h n for being there for me since the beginning and teaching me the techniques that I know today. Special thanks to Dr. Ross MacGil l ivray and Dr. Michael Murphy for being on my committee. I cherish the experience that I gained as a graduate student and owe it to every member of the Molday lab, past and present. I would like to thank Laurie Molday for creating a most pleasant environment in the lab and for helping me with my technical questions, Chris Loewen for encouraging me and helping me during my debut as a new member of the lab, Stephanie Bungert for being a good friend full of energy, and for her encouragement, Jason Wong for his statistical hockey discussions, Dan Hornan for cycling in the "storm the wall" event, with an injured leg and a bad flu, Theresa H i i for understanding and sharing my frustration over the Lions-gate project, Ansgar Poetsch and Winco W u , for their help and Andrew Ho for not giving me a hard time during my teaching assistantship. I would also like to thank my parents and brother for always being by my side and encouraging me, especially during the most difficult times. I could have not achieved this without their support and understanding. Je tiens aussi a remercier Murray pour avoir ete toujours la pour moi et pour m'avoir supporte, meme dans les moments les plus difficiles. xiii To: my family xiv INTRODUCTION 1.1 T H E HUMAN E Y E The human eye (Fig 1) is an elongated sphere of about 1 inch in diameter that transduces light in the range of 400 nm to 700 nm into neural discharges, allowing visual perception. The eye consists of three layers or coats and two chambers. The white fibrous outer layer known as the sclera forms part of the supporting wall of the eyeball. Near the front of the eye, the sclera becomes the cornea, the transparent external surface. The middle thin, dark brown layer is the choroid which contains many blood vessels. Toward the front, the choroid, first, thickens and forms the ciliary body which controls the shape of the lens and, finally becomes a thin, muscular diaphragm called the iris. This colored circular muscle controls the amount of light entering the eye by altering the size of the pupil. The inner layer of the eye is the retina. It consists of a thin layer of neural tissue lining the back of the eye ball. The lens, attached to the ciliary body by ligaments, divides the eye cavity into two chambers. The large cavity behind the lens is the vitreous humor and is filled with gelatinous material. The chamber between the cornea and the lens is called aqueous humor and is filled with an alkaline watery solution. 1.2 T H E RETINA The retina is approximately 0.5 mm thick (http://retina.umh.es/Webvision) and lines the back of the eye. A l l vertebrate retinas are composed of three layers of cells and two layers of synapses (Fig 2). The outer layer, closest to the choroid, contains the photoreceptor cells (rods and cones), the inner layer contains the bipolar, horizontal and amacrine cells and the innermost layer contains ganglion cells whose fibers become the optic nerve. The first area of synapse occurs between the photoreceptor cells and both the vertically running bipolar cells and the horizontally oriented horizontal cells. The second synapse connects the bipolar cells to the ganglion cells. 1 Fig 1. Anatomy of the human eye. The eye consists of three layers: the sclera, choroid and the retina. Toward the front of the eye, the outermost sclera becomes the transparent cornea and the middle layer choroid becomes the ciliary body and iris. The iris is responsible for limiting the amount of light entering the eye, and the ciliary body determines the shape of the lens. The lens focuses the entering light onto the retina, the innermost layer of the eye. The photoreceptors of the retina are responsible for converting light into electrical signals which are relayed to the brain via the optic nerve. Figure from http://retina.umh.es/Webvision. 2 Fig 2. Organization of the retina. The retina is composed of three layers of cells and two layers of synapses. The closest layer to the choroid contains the photoreceptor cells, the rods and the cones. The photoreceptor cells communicate by way of electrical synapse with each other and with the second layer of cells. The horizontal cells, bipolar cells and the amacrine cells make up the second cell layer of the retina. The innermost layer consists of ganglion cells whose fibers become the optic nerve. Figure obtained from http://retina.umh.es/Webvision. 3 The photoreceptor cells of the retina detect light and transmit electrical signals to the bipolar cells. The bipolar cells then pass the nerve impulses, in form of action potentials, to the ganglion cells whose fibers exit the eye as the optic nerve and enter the central nervous system. The location where the optic nerve exits the eye is the blind spot and is devoid of photoreceptor cells. The human retina also contains a very special region called the fovea, where vision is most acute. The fovea is characterized as an oval-shaped, blood vessel-free, yellowish area with a depression that contains only cone photoreceptor cells. The fovea is at the center of an area known as the macula. 1.3 T H E PHOTORECEPTOR C E L L S There are two types of photoreceptor cells in the retina: the rods and the cones, named according to the shape of their light sensitive compartments. The rods are extremely light sensitive and mediate vision at low light intensities, allowing the detection of even the slightest motion at night. The cones on the other hand, are considerably less sensitive to illumination and respond only when the light is brighter. They are also responsible for colored and detailed vision. In the human retina the rod photoreceptor cells outnumber the cones everywhere but in the fovea. Both cone and the rod cells can be divided into four regions: the outer segment, the inner segment, the cell body and the synaptic terminus (Fig 3). The outer segment contains an array of flattened membranous disks. In the rods, the disks are physically separated from the plasma membrane, whereas in the cones, the disks are continuous with the plasma membrane (Cohen, 1968). The inner segment is connected to the outer segment by a thin cilium. The inner segment contains the cellular organelles such as Golgi apparatus, mitochondria and endoplasmic reticulum (ER). The cell body contains the nucleus and the synaptic terminus has synaptic vesicles containing the neurotransmitter glutamate. The release of neurotransmitter is high in the dark and is reduced in a graded fashion in the light. 4 Plasma membrane Disks Outer segment Connecting cilium Nucleus Rod Inner segment Cell body Synaptic Terminus Cone Fig 3. The rod and cone photoreceptor cells. There are two types of photoreceptor cells in the retina: the rods and the cones. Each photoreceptor cell consists of four regions: the outer segment, the inner segment, cell body and the synaptic terminus. The outer segment of the rods contains an array of light sensitive membranous disks that is surrounded by a plasma membrane. The outer segment plasma membrane of the cones is continuous with the disk membrane. The outer segment of both rods and cones is connected to the inner segment by a thin connecting cilium. The cell body contains the nucleus and the synaptic terminus releases glutamate. Figure adapted from Dose (1995). 5 1.4 T H E ROD OUTER SEGMENT AND T H E DISK Phototransduction in the rod cells occurs specifically in the outer segment, which consists of hundreds of membranous disks. Disks are formed at the connecting cil ium by successive evagination of the plasma membrane such that each newly made disks become stacked one on top of another (Steinberg et ah, 1980). A s new disks are generated and move apically, they grow in diameter and detach from the plasma membrane into separate disk structures. To maintain the outer segment at a steady length, old disks are pushed up toward the tip of the outer segment, where they are eventually released from the outer segment and engulfed by the retinal pigment epithelium (reviewed in Bok, 1985). The lipids and proteins that make up the disk membrane are synthesized in the inner segment and transported to the base of the outer segment, where they are incorporated into the basal folding of the outer segment (Young and Droz, 1968; Young, 1968). Although the disk membrane arises from the folding of the plasma membrane, the protein and phospholipid compositions of the two membranes are somewhat different. Boesze-Battaglia and Albert (1992) have shown that phosphatidylethanolamine (PE) accounts for 11% of the total phospholipids in the plasma membrane, whereas it accounts for 42% in the disk membranes. The amount of phosphatidylserine (PS) in the plasma membrane (24%) is twice that in the disk membrane (14%). Phosphatidylcholine (PC) in the plasma membrane represents 65% of the total phospholipids but 45% of total disk phospholipids. Phosphatidylinositol (PI) represents a very small percentage of the total lipids in both the plasma membrane (<1%) and the disk membrane (1-2%). There is also a remarkable difference in the fatty acyl components of the membrane phsopholipids between the disks and the surrounding plasma membranes. When compared to the disk membrane, the plasma membrane is high in 18:1 and 18:2 fatty acids but low in 22:6. Furthermore, as disks move toward the apical end of the outer segment, the ratio of cholesterol to phospholipids decreases (Boesze-Battaglia et al, 1990); however, the phospholipid headgroup and fatty acyl composition of the disk membranes do not change. Although the protein compositions of the two membranes differ (Molday and Molday, 1987), rhodopsin is the main integral membrane protein found in both membranes. Some 6 of the proteins that are predominantly present in the ROS plasma membrane are: the cGMP-gated channel (Cook et al., 1987), the Na + /Ca 2 + -K + exchanger (Reid et al, 1990), and the GLUT-1 glucose transporter (Hsu and Molday, 1991). To understand the protein distribution in the disk membrane, it should be recognized that each disk encloses a compartment called intradiskal space (or lumen) and that, a single disk is composed of two regions: a flat lamellar region and a hairpin loop-like rim region (Falk and Fatt, 1969). Immunocytochemical labeling studies have localized three integral membrane proteins to the rim region of the disks: peripherin/rds (Molday et al, 1987), rom-1 (Moritz and Molday, 1996) and the 220 kDa glycoprotein (ABCR) (Illing et al, 1997). Similar studies have confined rhodopsin to the lamellar region of the disks (Hicks and Molday, 1986). The disk membrane also contains the guanylate cyclase (Liu et al., 1994) involved in the recovery from phototransduction and all-/r<ms-retinol dehydrogenase (Ishiguro et al, 1991). Table 1 summarizes the major proteins found in the two membranes and their possible functions. 1.5 PHOTOTRANSDUCTION 1.5.1 The dark current In the dark, the resting membrane potential of the rod photoreceptor cells is ~ - 40 mV whereas that of the neurons is — 70 mV. This slight polarization of the rod membrane is due to the dark current (reviewed by Yau, 1994) (Fig 4). Dark current is characterized by the inward flow of Na+, Ca 2 + and M g 2 + into the outer segment through the cyclic nucleotide-gated channel (cGMP-gated channel) on the plasma membrane and the outward flow of K + . In the dark, the cGMP-gated channel is held open by a relatively high concentration of cGMP, which is maintained through a balance between synthesis of cGMP by guanylate cyclase and hydrolysis of cGMP to 5'-GMP by phosphodiesterase. A Na + /Ca 2 + -K + exchanger localized in the plasma membrane of rod outer segment, also contributes to the dark current as it transports four Na + ions inside the outer segment and exports one Ca 2 + and one K + (Fig 4). To counter the inward flow of Na + and maintain its concentration at steady state, a Na + /K + ATPase located on the plasma membrane of the 7 Table 1 Major rod outer segment plasma membrane and disk membrane proteins L Protein Molecular weight , D a Function Plasma membrane proteins Rhodopsin 36 000 (38 000) Phototransduction cGMP-channel a subunit 63 000 (79 600) Phototransduction cGMP-channel P subunit 240 000 (155 000) N a + / C a 2 + - K + exchanger 230 000 (130 000) Cation, exchange Glucose transporter 50 000 Glucose transport Disk membrane proteins Rhodopsin 36 000 (38 000) Phototransduction Guanylate cyclase 112 000 Phototransduction Peripherin/rds 35 000 (39 000) Outer segment integrity Rom-1 37 000 Outer segment integrity A B C R 220 000 (257 000) Retinoid transport Retinol dehydrogenase 33 500 Retinal reduction a Table reproduced from Molday, 1998. b Molecular weight values were determined by S D S - P A G E ; values in parentheses were calculated from sequence. 8 Na7Ca2+-K+ exchanger 4 N a • Fig 4. Dark current in the rod photoreceptor cells. Dark current consists of the entrance of cations mainly through cGMP-gated channel and their exits through different channels or pumps. N a + enters the rod outer segment through the cGMP-gated channels and the N a + / C a 2 + - K + exchanger and is pumped out by a N a + / K + ATPase in the inner segment. C a 2 + also enter the outer segment through the cGMP-gated channel but is extruded through the N a + / C a 2 + - K + exchanger in the outer segment. In the inner segment C a 2 + is pumped out through a C a 2 + ATPase . K + is extruded by the N a + / C a 2 + - K + exchanger in the outer segment and a voltage-gated channel in the inner segment, but enters via the N a + / K + ATPase. Although M g 2 + is transported through the cGMP-gated channel, its efflux mechanism needs to be determined. A s shown, in the dark there is a flow of cations that enters the rod photoreceptor cell and results in the polarization of the cell membrane. Figure adapted from Warren, 1999. 9 inner segment extrudes Na and imports K . Another component of the dark current is the outflow of the K + ions through the N a + / C a 2 + - K + exchanger and the voltage-gated K + channel located on the inner segment plasma membrane. K + is transported back inside the rod cell via N a + / K + ATPase. Finally, the C a 2 + that enters the outer segment through the cGMP-gated channel, is extruded through N a + / C a 2 + - K + exchanger in the outer segment (Cervetto et al, 1989; Kim et al., 1998). The rod inner segment plasma membrane does j 2~1~ ~F 2"i" 2"i" not contain the Na /Ca - K exchanger, instead Ca is exported by a Ca ATPase (Krizaj and Copenhagen, 1998; Morgans et al., 1998). In the absence of light, the dark current partially polarizes the rod cell membrane, resulting in the release of the neurotransmitter glutamate from the synaptic terminal. Glutamate exerts an inhibitory effect on the bipolar cells, preventing their polarization. In the presence of light, the influx of ions through the cGMP-gated channel is reduced or stopped, while their efflux continues. The result, is a hyperpolarization of the rod cell membrane followed by a decrease in the release of inhibitory neurotransmitter. 1.5.2 Photoexcitation and recovery Visual detection in the rods begins when a photon of light is absorbed by the light sensitive pigment: rhodopsin (reviewed in Palczewski and Saari, 1997). Rhodopsin, a member of the G protein-coupled receptor family, consists of a protein moiety called opsin and a chromophore derived from vitamin A called 11-czs-retinal. 11-c/s-retinal is bound to opsin via a Schiff base linkage and together they form rhodopsin (Wang et al, 1980). When a photon of light is absorbed by rhodopsin, 11-cz's-retinal is isomerized to a\\-trans-retinal. This isomerization which takes place in less than 20 ps (Hayward et al., 1981), causes a change in the conformation of rhodopsin (Farahbakhsh et al, 1993). The active form of rhodopsin, called metarhodopsin II or R*, triggers the transduction cascade (Fig 5) by exposing its transducin binding site. Transducin (T ap y) is a multisubunit peripheral membrane protein consisting of three subunits: a or T a (39 kDa), (3 or Tp (36kDa) and y or T y (8kDa) (Hamm and Gilchrist, 1996). The a subunit (T a ) contains both a binding site for GTP/GDP and a catalytic site for the hydrolysis of bound GTP. The P and y subunits form the TpT which remain membrane bound. T a associates with TpY when GDP is bound, but may separate and be released into the space between disks when GTP is attached. 10 Fig 5. The process of phototransduction in the rod outer segment. A photon of light isomerizes 1 l-cis-retinal attached to rhodopsin to all-rrarcs-retinal to produce metarhodopsin II, the activated form of rhodopsin. Metarhodopsin II activates transducin by catalyzing the exchange of G D P for G T P on the a subunit of transducin ( T a ) . The activated transducin ( T a - G T P ) dissociates from its py subunits and in turn activates cGMP-phosphodiesterase (PDE) by removing P D E ' s inhibitory y subunits. Activated P D E catalyzes the hydrolysis of c G M P to 5 ' - G M P . The decrease in the c G M P concentration causes the cGMP-gated channels to close, preventing N a + and C a 2 + to enter the cell and causing the cell to hyperpolarize. The C a 2 + concentration decreases as a result of constant extrusion of C a 2 + by the N a + / C a 2 + - K + exchanger. Figure reproduced from M a h (1999). 11 Upon binding to the activated rhodopsin (R*), transducin exchanges its G D P for G T P on its T a , producing the active form T a - G T P which dissociates from Tp Y and R*. The T a - G T P diffuses and binds to a photoreceptor specific phosphodiesterase (PDE) , while the freed R* catalyzes another round of G T P / G D P exchange on a second transducin molecule. Thus, a single R* may activate hundreds of molecules of transducin. Phosphodiesterase (PDE) is a peripheral membrane protein consisting of four subunits (aPy2), two catalytic and two inhibitory (Farber, 1995). The a and p catalytic subunits are inhibited in the dark by the y subunits which bind very tightly (Deterre et al, 1988). Studies seem to suggest that two T a - G T P interact and activate the aPy2 holoenzyme, by removing its inhibitory constraint (Fung and Griswold-Prenner, 1989; Fung et al, 1990). The activated P D E complex hydrolyzes c G M P to 5 ' - G M P and effectively reduces the concentration of c G M P in R O S . The decrease in the intracellular c G M P concentration causes the cGMP-gated channels on the plasma membrane, that were maintained open in the dark, to close (Molday, 1998; Zagotta and Siegelbaum, 1996). The closure of the channel prevents the flow of N a + , C a 2 + a n d M g 2 + into the R O S , causing the rod cell membrane to become hyperpolarized. Since the N a + / C a 2 + - K + exchanger continues to extrude C a 2 + , the C a 2 + concentration decreases from ~ 500-700 n M in dark to -30-50 n M in light (Sampath et al, 1998). Hyperpolarization of the rod membrane decreases the release of inhibitory neurotransmitter glutamate from the synaptic terminus, this in turn polarizes the bipolar cells and results in the generation of action potentials that is transmitted to the brain. After photoexcitation, the photoreceptor cells return to their dark steady state by turning off the transduction pathway, regenerating c G M P and reopening the cGMP-gated channel. R* is inactivated by two sequential processes: the phosphorylation of its C-terminus by rhodopsin kinase (Udovichenko et al, 1997) followed by the binding of arrestin. Arrestin is a 48 kDa cytosolic protein that prevents phosphorylated rhodopsin from interacting with transducin (Krupnick et al, 1997; Zhang et al, 1997). T a - G T P is inactivated through its intrinsic GTPase activity, which hydrolyzes bound "GTP to G D P . In vivo, this hydrolysis is further accelerated by a membrane-bound factor that-acts as a GTPase-activating protein (GAP) , the most important being R G S 9 (He.er al, 1998). P D E activity is inhibited as T a -12 G D P dissociates and re-associates with Tp y . Alternative mechanisms of P D E inactivation have also been suggested including phosphorylation of P D E y by a specific kinase (Tsuboi et al, 1994 a,b). The next step in the recovery from light is the regeneration of the c G M P to its original level in order to re-open the cGMP-gated channel. c G M P is re-synthesized by guanylate cyclase present in the outer segment. A s the level of c G M P increases, the cGMP-gated channel re-opens and N a + , C a 2 + and M g 2 + enter, bringing the membrane potential back to ~ - 40 m A . It has been shown that the changes in C a 2 + concentration may trigger several negative feedback pathways (Gray-Keller and Detwiler, 1994). First guanylate cyclase which synthesizes c G M P is inhibited by high C a 2 + concentration (Lolley and Racz, 1982). Following illumination however, low C a concentration leads to the activation of guanylate cyclase. This process seems to be mediated through a specific family of C a 2 + -binding proteins termed guanylate cyclase activating proteins or G C A P s (Koch and Stryer, 1988; Palczewski et al, 1994; Dizhoor et ah, 1995). C a 2 + concentrations also regulate the affinity of the cGMP-gated channel for c G M P through calmodulin, another Ca 2 + -b ind ing protein (Hsu and Molday, 1993; Chen et al, 1994). For example a fall in the C a 2 + concentration increases the affinity of the channel for c G M P . Similarly C a 2 + , through recoverin, affects the phosphorylation of activated rhodopsin (Chen et al, 1995; Kawamura, 1993). At low C a 2 + levels, recoverin, a C a 2 + binding protein, seems to interfere with the inhibition of rhodopsin kinase, which may lead to shorten the lifetime of R* and eventually increase the intracellular C a concentration. After the inactivation of R* by phosphorylation and arrestin binding, in order for a new photon to be absorbed, rhodopsin must be regenerated (Amer and Akhtar, 1973; Bridges, 1976). This is a multistep process that requires dephosphorylation of the protein moiety (opsin) and the conversion of the chromophore all-frans-retinal to 11-czs-retinal. The transduction mechanism of vertebrate cone was not discussed here; however it very much resembles that of the rods. The visual pigments in cones have similar architecture as rhodopsin and also use 11-c/s-retinal as their chromophore (Nathans et al, 1986). 13 1.5.3 The visual (retinoid) cycle Following the phototransduction, 11-cw-retinal must be regenerated and must bind to opsin in order to re-form rhodopsin. In the vertebrate retina, pigment regeneration requires the cycling of retinoids between the photoreceptor cell and the surrounding retinal pigment epithelium (RPE), in a process termed the visual or retinoid cycle (Crouch et al, 1996). In the cones, the cone pigment regeneration occurs in less than 10 min (Rushton and Henry, 1968), whereas in rods, it lasts 20-30 min (Ripps and Weale, 1969). Following the photoexcitation of rhodopsin and the isomerization of 11-cz's-retinal to all-/r<ms-retinal, all-^nms-retinal is released from opsin and reduced by all-zT-tf/w-retinol dehydrogenase (Ishiguro et al, 1991) to its alcohol counterpart: all-zrans-retinol. It is suggested that interphotoreceptor retinoid-binding protein (IRBP) is the carrier protein that transports retinoids between the photoreceptor and the R P E cells (Lai et al, 1982; L i o u et al, 1982; Flannery et al, 1990). Within the R P E , \\-trans-xe\mo\ is converted to 11-c/s-retinal (section 1.6). 1 l-cw-retinal is then released from R P E cells and returned to the photoreceptor cells via IRBP. Once inside the photoreceptor cell, 11-c/s-retinal combines with opsin through a Schiff base linkage and re-forms rhodopsin. 1.6 RETINAL PIGMENT EPITHELIUM The outer segment of both the rods and the cones are closely associated with a monolayer of cuboidal cells called retinal pigment epithelium (RPE). R P E is located between the capillary bed of the choroid and the photoreceptor cells. The cells of R P E contain pseudopodial attachments that envelop the outer segments. Two of the important functions of R P E include: i) the phagocytosis of rod and cone outer segment fragments that are shed and ii) the uptake, processing and release of retinoids involved in the visual cycle (reviewed in Bok, 1993). The phagocytosis function of R P E in the renewal process of outer segments was discovered by autoradiography. Young and Bok (1969) showed that when radioactive frog outer segment disks reached the apex of the outer segment, they became detached (a process called disk shedding) and were rapidly phagocytized by R P E (Fig 6). Although 14 Fig 6. Schematic diagram depicting the phagocytosis of the rod outer segment by the retinal pigment epithelium. A monolayer of retinal pigment epithelium (RPE) is adjacent to the outer segment of the photoreceptor cells (only rods are shown). Although the exact mechanism is still unknown, the phagocytosis process begins as peusodopodial attachments of R P E cells envelop and internalize the old discarded disks. Inside the R P E , the discarded disks form phagosomes. To break down and digest the contents of the phagosomes, lysosomes fuse with them, forming larger structures called phagolysosomes. 15 <Tmnnnnnnnnnnnnt?nnnnnmiiiiiiii/inl.1...Ti E J5 <D <u |E£ co Q) D 3 Q _ O or: Q_ <D a: CO 16 the mechanism underlying disk shedding and internalization by RPE is still obscure, some investigators suggest that RPE is an active partner in separating parts of outer segments (Spitznas and Hogan, 1970). Others, however, believe that 10-30 disks surrounded by the plasma membrane are shed first and then ingested by RPE (Young, 1971). Once internalized in the RPE as phagosomes, they fuse with lysosomes to form phagolysosomes. Then larger lysosomes (previously formed by fusion of smaller ones) interact with phagolysosomes (Bosch et al, 1993). The uptake, processing and release of retinoids, involved in the visual cycle, are other important functions of RPE (Fig 7). As explained earlier, all-rrans-retinol is transported to RPE via IRBP. Once inside, all-Zrara-retinol binds to cellular retinal-binding protein or C R B P (Bok et al, 1984) which belongs to a family of fatty acid binding proteins. In a first reaction, all-/rtms--retinol is esterified to all-frcmy-retinyl ester by the enzymatic activity of lecithin-retinol acyltransferase (LRAT) (Saari and Bredberg, 1989). The resulting retinyl ester serves as a stable storage form of retinoids in the RPE cells. This ester is hydrolyzed and isomerized to 11-c/s-retinol by an enzyme called isomerohydrolase (Fig 7, reaction 4) (Deigner et al, 1989; Rando, 1991; Winston and Rando, 1998). Alternatively, all-Zrcws-retinyl ester first may be hydrolyzed to all-Zra/w-retinol (the form that entered RPE) by a retinyl hydrolase (Fig 7, reverse of reaction 3) and then isomerized to 11-c/s-retinol (Fig 7, reaction 5) (Stecher et al, 1999). 11-c/s-retinol, from either of the two pathways, is oxidized to 11-cz's-retinal by the action of retinol dehydrogenase (Fig 7, reaction 6). 11 -c/s-retinal is then delivered to the photoreceptor cells via IRBP. Retinoids progress through the visual cycle in the RPE by binding to different retinoid binding proteins (Bok, 1993). As a consequence of aging, lipofuscin material (nondegradable end products of phagocytosis) may accumulate in the aging RPE and may compromise the function of RPE. In some cases, the excessive accumulation of this material, rich in polyunsaturated fatty acid and vitamin A , may result in age-related macular degeneration, the leading cause of retinal blindness among seniors (Young, 1987). Lipofuscin may also accumulate in the RPE of patients with retinal diseases such as Stargardt's disease. 17 Fig 7. The visual cycle and the role of retinal pigment epithelium. A photon of light isomerizes 11-cz's-retinal of rhodopsin to all-zra/«-retinal producing metarhodopsin II (reactionl). All-frvms-retinal separates from opsin and is reduced to all-zr<ms'-retinol by all-/ro«5-retinol dehydrogenase {reaction 2). All-fra/w-retinol is transported to the RPE where it is esterified to all-/>aw,s-retinyl ester by lecithin-retinol acyltransferase (LRAT) (reaction 3). All-Zrans-retinyl ester is a substrate for an enzyme called isomerohydrolase, which couples the hydrolysis of the ester to the isomerization to 11-c/s-retinol (reaction 4). Alternatively, the ester can be hydrolyzed back to all-fra/w-retinol (reverse of reaction 3) and then isomerized to 11-c/s-retinol (reaction 5). 11-c/s-retinol is oxidized to 11-c/s-retinal by the action of 11-cz's-retinol dehydrogenase (reaction 6) and then transported back to the rod outer segment, where it binds to opsin to regenerate rhodopsin. The retinoid binding proteins (CRBP and IRBP) are not shown in the figure. Figure based on McBee et al 2000. 18 19 1.7 M A C U L A R DEGENERATION AND STARGARDT'S DISEASE Over the last decade, a diverse collection of retinal-specific genes have been identified that, when mutated, result in retinal degeneration (the RetNet website summarizes the current genetic knowledge of the various retinal diseases). This should not be of great surprise when one considers the large number of proteins that are involved in the visual cascade and visual cycle alone. For example mutations in rhodopsin and catalytic a and (3 subunits of P D E result in autosomal recessive retinitis pigmentosa (Rosenfeld et al, 1992; McLaughl in et al, 1993; Huang et al, 1995). Similarly, mutations in guanylate cyclase, peripherin/rds and A B C R lead to Leber's congenital amaurosis, autosomal dominant retinitis pigmentosa and Stargardt's disease, respectively (Perrault et al, 1996; Farrar et al., 1991; Wells et al. ,1993; Allikmets et al. ,1997a). Stargardt's disease, which was first described by Kar l Stargardt in 1909, is a recessive form of macular degeneration with a juvenile to young-adult onset. This disease is characterized by the loss of central vision, progressive atrophy of the foveal R P E and the appearance of yellow-orange flecks around the macula (Fig 8). These flecks seem to arise from the accumulation of lysosomal material similar to lipofuscin within the aging R P E cells. The abnormal and premature accumulation of lipofuscin may be the result of any one or more of the following defects: i) increased turnover of photoreceptor cells, ii) abnormal phagocytosis of R O S , iii) abnormal degradative activity in the R P E and iv) defective visual cycle. In 1997, a study by Allikmets et al. (1997a) identified ABCR as the gene responsible for Stargardt's disease. Immediately, this gene was cloned and the resulting protein, called A B C R , was identified as an A B C transporter. To date mutations in the A B C R gene have been linked to a number of retinal diseases such as: Stargardt's disease, autosomal recessive cone-rod dystrophy, retinitis pigmentosa, fundus flavimaculatus and the controversial age-related macular degeneration (Cremers et al, 1998; Martinezmir et al, 1998; Allikmets et al, 1997a,b). 20 A ~ ~ — ^ Fig 8. Normal and Stargardt's disease affected macula. (A) The macular region of a normal individual. (B) The macula of a Stragardt's disease patient. One of the clinical features of Stargardt's disease is the appearance of yellowish/orange deposition of lipofuscin material around the macula, as shown. Diagram from Yannuzzi et al, 1995. 21 1.8 T H E SUPERFAMILY OF ABC PROTEINS The superfamily of ATP-binding cassette (ABC) proteins which includes P-glycoprotein, multidrug resistance associated protein (MRP) and cystic fibrosis transmembrane conductance regulator (CFTR), is one of the largest protein families with more than 1000 members (Higgins, 1992). These proteins, which can be found in all living cells from bacteria and yeast to human, have a variety of cellular functions. Although most A B C proteins act as active transporters, others may act as ion channels, channel regulators, proteases, receptors and sensing proteins (Higgins, 1995). As transporters, A B C proteins accommodate the transport of a large variety of substrates including: ions, heavy metals, amino acids, peptides, steroids, glucocorticoids, bile acids, anticancer drugs and antibiotics (Higgins, 1992; Dean and Allikmets, 1995; Kuchler and Thorner, 1992). Several mammalian A B C proteins are also of great importance as mutations in their genes result in genetic diseases. Cystic fibrosis for example is an autosomal recessive disorder that has been linked to mutations in CFTR. This A B C transporter acts as a c A M P activated chloride channel in a variety of epithelial tissues. Similarly, mutations in A L D P , SUR1 and A B C R each result in the following genetic diseases: Adrenoleukodystrophy, persistent hyperinsulinemic hypoglycemia of infancy and Stargardt macular dystrophy, respectively (Klein et al, 1999). 1.8.1 The structure of ABC transporters Members of A B C superfamily are defined by the presence of an A B C unit, also known as nucleotide-binding domain (NBD). N B D is a 200-250 amino acid residue segment characterized by two short motifs, Walker A and Walker B , that are involved in A T P binding (Fig 9 A). A third motif, called A B C signature or motif C, with the consensus sequence L S G G Q is located between Walker A and B (Schneider and Hunke, 1998). Together these three conserved motifs are hallmarks for A B C proteins. As transporters, A B C proteins contain a six membrane spanning segment, denoted T M D for transmembrane domain. The TMDs and NBDs may be arranged in different orders (Fig 9 B). Half size transporters have one N B D and one T M D arranged in either (TMD-NBD) or (NBD-TMD) configuration. Some examples of half size A B C transporters are TAP1 and 22 Fig 9. Prototype domain arrangements in ABC proteins. (A) Linear representation of the nucleotide binding domain (NBD) . Shown are the Walker A and B motifs and the signature sequence (motif C) of A B C transporters. The amino acids are designated by one letter code, 'h' stands for hydrophobic amino acids and 'x' can be varied. Figure reproduced from Schneider and Hunke, 1998. (B) The different organizations of the nucleotide binding domains (NBD) and the transmembrane domains ( T M D ) in A B C proteins. The following configurations are shown: N B D , N B D - N B D , N B D - T M D , T M D -N B D , ( T M D - N B D ) 2 and T M D 0 ( T M D - N B D ) 2 . Figure adapted from Kle in et al, 1999. 23 Walker A GxxGxGKT/S Walker B hhhhD NH 2 I C O O H Motif C L S G G Q Q Q R R K B NBD N B D ^ J TMD 7 \ f~ ^ ^^  ~^f-"^^^^^^ _^ & ^^^^^^ '"1 7 TMDO n n n r TMD1 TMD2 24 T A P 2 that are involved in the antigen presentation by the human major histocompatibility complex ( M H C ) (Kelly et al, 1992). In order to actively transport degraded peptide, T A P 1 and T A P 2 need to come together and form a heterodimer. A full size A B C protein is one that has two N B D s and two T M D s . The domain arrangements in these proteins are either duplicated forward ( T M D - N B D ) 2 or duplicated reverse ( N B D - T M D ) 2 . P-glycoprotein, C F T R and A B C R are three examples of full size transporters. Some A B C proteins such as S U R , and some of M R P s contain the configuration of a full size transporter; however, they also possess one extra five helical T M D at their N-terminus denoted as T M D o (Cole and Deeley, 1998). A B C superfamily also includes members that do not contain any T M D s , consisting only of either one or two N B D s . Based on sequence similarity, 48 human A B C transporters have been identified and classified into seven subfamilies to date (http://www.humanabc.org). The M R P / C F T R and A B C A subfamilies are the largest, each with twelve members. The M D R / T A P subfamily is next with eleven members, followed by the White (five members), the A L D (four members), the G C N 2 0 (three members) and the O A B P subfamily (one member). Many of these subfamilies have a structural counterpart in yeast. For example, the M D R , M R P and A L D subfamilies in humans seem to be present in yeast. Thus, the baker's yeast Saccharomyces cervisiae, the first eukaryotic genome to be sequenced, has proven to be an invaluable model for studying A B C proteins and some of their mutational effects. It should be mentioned however, that this yeast is not a good model for the present study as it lacks the A B C A subfamily, the focus of this study. 1.8.2 ABCA subfamily and ABCR The A B C A subfamily includes members such as A B C A 1 , A B C A 2 , A B C A3 and A B C A 4 or A B C R . These proteins are full size transporters with the forward configuration ( T M D -NBD )2 . Further more, they contain a stretch of highly hydrophobic amino acids (HH1) linking the two tandem halves (Luciani et al, 1994). Members of this subfamily show 45-66% identity along their entire sequence, 55-61%) along their first N B D and 57-69%) identity along their second N B D (Broccardo et al, 1999). Among the A B C A subfamily, A B C R is the focus of this study. 25 A B C R , which was named the R i m protein, was first reported by Papermaster et al. in 1978 as a 290 kDa glycoprotein localized in the rim region of the frog outer segment disks. Two decades later, in 1997 the bovine homologue of the R i m protein was cloned and identified as a novel member of the A B C superfamily by Illing et al. (1997). A t the same time, Allikmets et al. (1997a) used a genetic approach and identified an A B C transporter gene (ABCR) as the causal gene for Stargardt's disease, and thus cloned the human ABCR gene. From 1997 until recently it was believed that A B C R was exclusively expressed in the rod photoreceptor cells; however, a study by Molday et al. (2000) using immunofluorescence and western blotting showed that A B C R is also expressed in foveal and peripherial cones. Although much needs to be learned regarding the structure of this protein, putative models have been proposed based on hydropathy profiles and sequence similarities to other members of the A B C superfamily. Fig 10 shows putative models of the A B C R proposed by Illing et al. (1997) and supported by Bungert et al. (2001) and that of Azarian and Travis (1997) and Sun et al. (2000). In both models A B C R consists of a single polypeptide chain arranged in two tandem halves, each containing a N B D proceeded by a T M D . In the model by Illing et al. and Bungert et al, (Fig 10, A ) the first (1) and the seventh (7) transmembrane segments are both separated from the following five membrane spanning segments by a large intradiskal loop. In many other A B C A members and in the second proposed model of A B C R (Fig 10, B), the highly hydrophobic segment (HH1) is thought to partially enter the membrane and form a hairpin loop. This HH1 segment in the model of Illing et al. and Bungert et al. is regarded as the seventh transmembrane segment. The result of HH1 spanning the membrane instead of looping in the membrane, is that the second intradiskal loop in Fig 10, A is located inside of the disk (in the lumen) whereas in the other model it is in the cytoplasmic side of the disk. Bungert et al. support their model by locating four /V-glycosylation sites on each of the putative hydrophilic intradiskal loops. In addition, their study also provided evidence that the two tandem halves of the A B C R interact through disulfide bonding. 26 Fig 10. Putative topological models of ABCR. (A) A representation of the model proposed by Illing et al. (1997) and supported by Bungert et al. (2001). This model is characterized by two large intradiskal loops that are located in the disk lumen. The highly hydrophobic segment (HH1, in white) between the first T M D and the second T M D is envisioned as the seventh transmembrane segment. (B) Topological model of A B C R proposed by Azarian and Travis (1997) and Sun et al. (2000). In this model the HH1 loops in and out of the disk membrane and causes a very short second intradiskal loop. The transmembrane segment seven from one model does not correspond to the same region in the second model. The predicted membrane spanning segments are numbered in both models. Figure adapted from Bungert et al, 2001. 27 A B 28 1.9 POSSIBLE FUNCTIONS OF ABCR Since its identification as an A B C transporter and its involvement in Stargardt's disease, it was speculated that A B C R may somehow be involved in the active transport of retinoids. The accumulation of lipofuscin materials seen in Stargardt's disease may then be the result of a nonfuctional protein. Although the exact function of A B C R is currently under investigation, different studies (below) have shed light on the putative function and the environmental requirements of this protein. In 1999, Sun et al. investigated the ATPase activities of purified and reconstituted A B C R in the presence of a number of compounds, including various geometric isomers of retinal. The basis of this experiment was that i f A B C R (in the presence of A T P ) encountered its substrate or its allosteric activator, its ATPase activity would be stimulated as it transports its substrate across the membrane. Similar studies conducted for C F T R and P-glycoprotein had been successful in identifying their substrates (Bear et al, 1992; Shapiro and Ling , 1994). As a result of the study by Sun et al (1997) retinoids, especially all-/rarcs-retinal, were identified as possible substrates of A B C R . Shortly after, Weng et al. (1999) reported the generation of an abcr knockout mice. These mice showed interesting features such as: delayed dark adaptation (delayed recovery from photoexcitation), light-dependent increase in all-fr-arcs-retinal, elevated levels of both phosphatidylethanolamine and protonated /V-retinylidene-PE (a complex of aW-trans-retinal and PE) in R O S , and accumulation of lipofuscin materials in R P E . These observations along with the clinical features of Stargardt's disease, especially the accumulation of lipofuscin in R P E , led the authors to suggest A B C R as a putative transporter of all-^nms-retinal or /V-retinylidene-PE. In a third study, A h n et al. (2000) measured the ATPase and GTPase activities of purified bovine A B C R reconstituted in different lipid environments and in the presence of different retinoids. A s a result of their study, they identified all-fra^s-retinal and /V-retinylidene-PE as possible substrates. Furthermore, they recognized that both P E and a reducing 29 environment are required to stimulate the ATPase activity of purified and reconstituted bovine A B C R . Following these studies two models for the function of A B C R have been suggested (Fig 11, A and B). Both A and B of Fig 11 show the release of all-trans-retinal following photoexcitation of rhodopsin. In A , A B C R acts as a transporter and actively extrudes retinal from the disk membrane, making it available to all-trans-retmol dehydrogenase. Panel B shows that some of the released all-trans-retmal may react with P E in the disk membrane to form JV-retinylidene-PE. The /V-retinylidene-PE that is in the outer leaflet of the disk membrane, is readily accessible to a\l-trans-ret'mo\ dehydogenase, whereas the N-retinylidene-PE in the inner leaflet is not. In this case, A B C R acts as a flippase and flips /V-retinylidene-PE from the inner segment to the outer segment. If the putative functions of A B C R is compromised due to a mutation in its gene, a\\-trans-retinal and/or 7V-retinylidene-PE may accumulate first in the R O S and then in the R P E cells, where they would form the precursors of lipofuscin. The death of R P E and photoreceptors would shortly follow, resulting in Stargardt's disease. 1.10 THESIS INVESTIGTION At the start of this project, Lewis et al. (1999) published a list of 89 variations in ABCR gene that were considered disease causing. The goal of this project was to study the effect of five of these A B C R variants on the structure and possibly the function of A B C R . The first step toward this goal was the selection of the mutations, followed by the generation of the desired A B C R mutant constructs via site-directed mutagenesis. These constructs were then used to transfect monkey kidney C O S cells. The different expression levels of the mutants relative to the wild type protein were revealed by western blotting. Furthermore, the localization patterns of the expressed proteins (mutants and wild type) to different compartments in the C O S cells were examined by immunofluorescence microscopy. Finally, the degree of interactions between the wild type or mutant A B C R s 30 Fig 11. Possible functions of ABCR. (A) Function of A B C R as an all-trans-retinal extruder. After being released from rhodopsin (not shown) all-zra/w-retinal is actively extruded from the disk membrane by A B C R and becomes exposed to all-frtws-retinol dehydrogenase. Once reduced to all- /ra /M-retinol, it enters the visual cycle. (B) Function of A B C R as a retinylidene-phosphatidylethanolamine flippase. After the release of all-trans-xttina\ from rhodopsin (not shown), some of the a\\-trans-retina\ complexes with P E in the disk membrane to form ,/V-retinylidene-PE. The ./V-retinylidene-PE that forms on the intradiskal/luminal side of the disk membrane is not accessible to all- /ra /M-retinol dehydrogenase. A B C R acting as a flippase, flips the VV-retinylidene-PE that forms on the inner leaflet of the disk to the outer leaflet where it can become accessible to all-trans-retinol dehydrogenase. The flipping action of A B C R is shown by the solid arrow. The broken arrows represent the paths taken by different retinoid compounds as they progress through the visual cycle. Figure is based on Weng et al, 1999. 31 All-zrans-retinol All-frans-retinol dehydrogenase Intradiskal space Disk Membrane Cytoplasmic space All-trans-retinal B 32 with calnexin (a chaperone protein in the E R ) was investigated by western blotting. These above mentioned studies provided some insight regarding the folding of mutant proteins in comparison to the wild type A B C R . The next part of the study was to investigate the effect of the selected mutations on the A T P hydrolysis activity of A B C R . The expressed A B C R mutants were purified from detergent solubilized C O S cells by immunoaffinity chromatography and reconstituted in R O S lipids or brain polar lipids and D O P E vesicles. The ATPase activity of the reconstituted proteins (wild type and mutants) were measured in the presence and absence of a\\-trans-ret'ma\, the possible substrate of A B C R . The nucleotide binding abilities of four of the mutants were also briefly investigated by 8-azido-ATP labeling. 33 MATERIAL AND METHODS 2.1 MATERIALS AND BUFFERS Fresh bovine retinas were obtained from Schenk Packing Co. Inc. The Bluescript plasmids containing the following mutations: R2038W, R2077W, R2106C were generous gifts of J. Nathans. Wild type A B C R c D N A was previously inserted into pRK5 plasmid by J. Ahn. The primers designed for both PCR and sequencing were synthesized by Gibco B R L . The Pfu Turbo D N A polymerase and the QuickChange Site-Directed Mutagenesis Ki t were obtained from Stratagene. The 2'-deoxynucleoside 5'-triphosphates (dNTPs) and the alkaline calf intestinal phosphatase were from Amersham Pharmacia Biotech Inc. and New England Biolabs, respectively. The pcDNA3 plasmid was from Invitrogen and the restriction enzymes from Pharmacia, New England Biolabs, M B I Fermentas and Gibco B R L . The Klenow fragment and D N A T4 ligase were from M B I Fermentas. The QIAquick Extraction Kit , QIAquick PCR purification kit, QIAprep Spin Miniprep Ki t and the Q I A G E N Plasmid Maxiprep Kit were all from Qiagen. AW-trans-xeXm&X, soybean phospholipid and 3-[(3-cholamidopropyl)dimethylammonio]-l-propanesulsonic acid (CHAPS) were purchased from Aldrich, while brain polar lipids, 1,2-dioleoylphosphatidylethanolamine (DOPE) and l-stearoyl-2-docosahexaenoylphosphatidyl ethanolamine (SDPE) were from Avanti Polar Lipids. 8-azido-[a- 3 2P]ATP and [a-3 2 P ] A T P were from N E N Life Science Products. The goat anti-mouse Ig, goat anti-rabbit Ig conjugated to horseradish peroxidase and E C L reagent were from Amersham while the goat anti-mouse antibody conjugated to CY3 was obtained from Jackson ImmunoResearch Laboratories. Bovine serum albumin (BSA) and B C A protein assay were obtained from Pierce. A l l tissue culture media and antibiotics ( D M E M , fetal calf serum (FCS), trypsin-EDTA, penicillin, streptomycin) were purchased from Gibco B R L and all other reagents or chemicals were from Fisher Scientific, Sigma, B D H Inc., Gibco B R L , Bio-Rad, or Difco Laboratories. 34 The following buffers, medium and solution were prepared and extensively used: Buffer A: 50 mM HEPES, pH 7.5, 0.1 M NaCI, 1 mg/ml sonicated soybean phospholipid, 10 mM CHAPS, 10% glycerol, 1 mM DTT, 3 mM MgCl 2 . Solubilization buffer or buffer B: 50 mM HEPES, pH 7.5, 0.1 M NaCI, 1 mg/ml sonicated soybean phospholipid, 18 mM CHAPS, 10% glycerol, 1 mM DTT, 3 mM MgCl 2 , lx Complete Protease Inhibitor. Buffer C: 25 mM HEPES, pH 7.5, 0.14 M NaCI, 10% glycerol. Reconstitution buffer or buffer D: 25 mM HEPES, pH 7.5, 0.14 M NaCI, 10 % glycerol, 1 mM DTT, ImM EDTA. Phosphate buffered saline (PBS): 140 mM NaCI, 3 mM KC1, 10 mM phosphate, pH 7.4. Tris buffered saline (TBS): 20 mM Tris, 150 mM NaCI, pH 8.0. Luria-Bertani medium (LB): 10 g tryptone, 5 g yeast extract, 10 g NaCI. All-rrans-retinal solution: All-rratts-retinal was dissolved in ethanol in the dark, and its concentration was determined spectrophotometrically using the molar extinction coefficient 42 880. 2.2 M O L E C U L A R BIOLOGY TECHNIQUES Guidelines and protocols used for the general molecular biology techniques were either from Sambrook et al. (1989) or from the manufacturer supplying the product in use. Most of the restriction enzyme digestions were carried in either 15 or 50 ul volumes. Typically, 0.5 or 1 ul of the desired restriction enzyme was mixed with 1-4 jag of plasmid DNA in water, lx restriction enzyme buffer, and if required lx bovine serum albumin (BSA), and incubated for 1-2 hours at 37 °C. When double digestions were needed and the two enzyme activities remained high in a given buffer, the two enzymes were added together, otherwise two sequential digestions were carried out. The reactions were stopped by either running the samples on an agarose gel or by inactivating the enzymes. Sometimes, the DNA fragments were dephosphorylated by incubating at 37°C for 30 min with 0.5 unit of calf intestinal alkaline phosphatase. For creating blunt ends, either a blunt cutter restriction enzyme such as EcoRVox a restriction enzyme that produced 5'-overhangs such as NotI or Xbal was used. In the latter case, the generated 5'overhangs were filled-in by the Klenow 35 enzyme in the presence of 1 u.1 of 20 m M dNTPs , to produce blunt ends. The D N A fragments or the P C R products were separated by 0.6-1% agarose gel electrophoresis and visualized in the presence of ethidium bromide that was included in the gel. Desired fragments were excised from preparative gels, while under U V light. The D N A was then purified from the agarose gel using QIAquick Gel Extraction Kit and eluted in 50 u.1 water. The D N A concentrations were either determined spectophotmetrically or by running on agarose gel along side Lamda/Hindl l l ladder (MBI Fermentas). During ligation reactions the vector and the insert were added in the molar ratio of 1:2 to lx T4 ligation buffer and 1 ul of bacteriophage (T4) D N A ligase. If required, water was added to bring the final volume to 10 ul. Often, two control reactions were also ligated: one containing the insert only and one carrying the vector only. The reactions were carried out overnight at 1 5 ° C . Competent E. coli (DH5a) cells were prepared using CaCl2 method outlined in Sambrook et al. (1989). To transform the competent cells, 45 ul of the thawed E. coli cells were mixed with 4-6 u.1 of the ligated D N A and incubated on ice for 30 min. The cells were heat shocked at 4 2 ° C for exactly 90 seconds and placed back on the ice for another 2 min. After the addition of 0.2 ml of L B , the cells were allowed to incubate for 1 hour at 3 7 ° C while shaking (300 rpm). Following the incubation, 250 ul of the cells was spread on agar L B plates containing 50 u.g/ml of ampicillin. The plates were left overnight at 3 7 ° C in an incubator. To screen the cells for the desired construct, single colonies from agar plates were inoculated in 2 ml of LB/ampic i l l in and were allowed to grow overnight at 3 7 ° C while shaking. Plasmid D N A was extracted from the overnight culture using QIAprep Spin Miniprep Kit . To obtain larger amount of D N A (1-2 jag) a maxiprep was done on 500 ml of E. coli culture using the manufacturer's instructions. To screen the constructs for the correct plasmid D N A , restriction enzyme digestions were carried (as explained above) and a restriction enzyme map was often constructed. Glycerol stock solutions of the E. coli cells containing the desired recombinant plasmid were prepared by mixing 0.85 ml of cultured cells with 0.15 ml of sterile glycerol. The glycerol stocks were stored at -80 ° C . 36 2.3 GENERATION OF 3F4 COUPLED SEPHAROSE BEADS The R i m 3F4 antibody against an epitope near the C-terminus of bovine A B C R was generated and described by Illing et al, 1997. Approximately 1 mg of antibody per 1 ml of beads was dialysed in three changes of 20 m M borate buffer p H 8.4, at 4 ° C over a three day period. The Sepharose 2B beads (Pharmacia) were first washed three times with three volumes of water and then activated by C N B r as outlined by Cuatrecasas (1970). The activated beads were washed four times with cold borate buffer (20 m M , p H 8.4) by centrifugation in a table top centrifuge for 3-5 min, and subsequently incubated with the 3F4 antibody at 1 mg/ml beads. After 2-4 hours of incubation at 4 ° C on a rotating wheel, the 3F4-beads were washed twice with Tris Buffered Saline (TBS: 20 m M Tris, 150 m M N a C l , p H 8.0) containing 50 m M glycine (in order to react and block any remaining activated sites on the beads). The 3F4-beads were stored in an equal volume of T B S and 0.01% N a N 3 a t 4 ° C . 2.4 GENERATION OF T H E CONSTRUCTS Wild type A B C R c D N A was previously subcloned in p R K 5 plasmid by J. A h n and named A B C R / p R K 5 . The R2038W, R2077W and R2106C missense mutations were produced and inserted in bluescript plasmids by J. Nathans. These three mutants were first inserted into A B C R / p R K 5 plasmids and then the entire A B C R D N A containing the mutations were subcloned into p C D N A 3 . Briefly, all three mutations were located within a cassette defined by the restriction enzyme sites Hindlll and Aflll. The ABCR/bluescript constructs containing the cassettes were cut with the restriction enzymes Hindlll and Aflll. The resulting 1.03 kb fragments, each bearing one of the three mutations, were used to replace the equivalent fragments in three wild type A B C R / p R K 5 constructs, also cut with Hindlll and Aflll. Once inserted, the wild type and the three mutant A B C R D N A s were cut out of p R K 5 plasmid with NotI and Hindlll and inserted into p C D N A 3 vector cut with NotI and Xbal. Trie Hindlll restriction site on the insert and the Xbal site on the vector were both blunt ended before ligation, resulting in the loss of the two restriction sites in the final constructs. 37 T1526M mutant was generated by using the QuickChange Site-Directed Mutagenesis Ki t (Stratagene) and by following the manufacturer's guidelines. Two complementary primers 38 bases in length and each bearing the required single nucleotide change were designed (Table 2). The forward and reversed primers were called A Z 3 + and A Z 3 - , respectively. Three different Polymerase Chain Reactions (PCRs) were carried out using 10, 20 and 50 ng of template wild type A B C R / p C D N A 3 . Each reaction was carried in 50 ul volume and contained: 250 u M dNTPs , 125 ng of each of the primers, 5 pi of lOx Pfu buffer and 2.5 units of Pfu Turbo. A negative control reaction was also carried out in which one of the primers was missing. The P C R reactions were carried out overnight in a P C R machine programmed at the following settings: 1- 9 5 ° C for 5 min (to denature the D N A double helix). 2- 9 5 ° C for 30 seconds (to denature the helix at the beginning of each cycle), 5 7 ° C for 1 min (to allow annealing of the primers) and 7 2 ° C for 30 min (to allow extension of the new strand by the polymerase). The cycle was repeated 15 times followed by a final 10 minute extension at 7 2 ° C . While the samples were stored at 4 ° C , usually 10 u.1 was removed and analyzed on a 1% agarose gel for the presence of the amplified plasmids. To differentiate between the old D N A template and the newly synthesized plasmid containing a possible nucleotide change, the remaining P C R products were incubated at 3 7 ° C with 1 ul of Dpnl restriction enzyme (10 units). This enzyme digests the methylated template D N A s and thus allows the direct use of the resulting P C R products in the transformation of competent E. coli cells. To generate D846H, the wild type A B C R / p C D N A 3 was used and two primers were designed complementary to the template D N A (Table 2): A Z 1 + (forward primer with a predicted melting temperature of 6 2 ° C ) and A Z 2 A - (reverse primer having a predicted melting temperature of 6 7 ° C ) . The A Z 2 A - primer contained the nucleotide change G instead of C , corresponding to the amino acid change D to H , and the restriction enzyme site Kpnl. A Z 1 + was complementary to a sequence ~ 2.1 kb upstream of the mutation, just 5' to an EcoRV site (Fig 12). The P C R reaction contained: 10 ng of template, 250 ng of both A Z 1 + and A Z 2 A - , 250 u M dNTPs , l x Pfu buffer and 2.5 units of Pfu Turbo. After a first heat-denaturation at 95 °C for 1 min, the following P C R cycle was repeated 25 times: heat-denaturation at 9 5 ° C for 1 min, annealing of primers for 1 min at 5 7 ° C , extension by the D N A polymerase for 2.5 min at 7 2 ° C . After the 25 rounds of P C R , the amplified 38 i J % fi o t t t o c cr I CO s o o < < o o < u o H H < < o < o o u u o < u H u < o 9 Co + o < o o < Si < 6 u a < o < < U a < < u < U U H O O t—1 u u o < < 0 a a a HI < a o o u < < H H a < + 01 cn N N < < U U H H H < < H O H H U H o o <n u H <l U u H O H O o H H O H < o < o o u I ro N < fi c3 cn CS o fi l-fi c o en <U o, di o fi <u fi cr 0) C/3 >H 1) CCJ 0) 0> o fi c I-a fi c o CCJ +-» fi " C f 00 Q uo fi CD l-fi H 39 rat m M | » m M 5 ' — — 1 GH 3' 3' 1 C - l 5' 5 - > EcoRV Kpnl AZ1 + (546) (2649) 1- P C R 2- Cut with EcoRV and Kpnl EcoRV 2103 bp c H I Q — j Kpn/ Subclone into modified A B C R / p C D N A 3 cut with EcoRV and Kpnl Figure 12. Strategy used to generate D846H construct. The reverse primer A Z 2 A -(dotted arrow) carries the nucleotide change and contains a Kpnl restriction enzyme site. The forward primer A Z 1 + is complementary to the 3' strand of the A B C R c D N A , upstream of EcoRV site. After 25 rounds of polymerase chain reactions, the amplified fragment was cut with EcoRV and Kpnl and subcloned into the modified A B C R / p C D N A 3 construct cut with the same enzymes. The numbers in ( ) correspond to the location of restriction enzyme sites in A B C R c D N A . 40 fragment was cut with EcoRVand Kpnl and purified from the P C R solution and restriction enzymes, by using the QIAquick P C R Purification Kit . This new fragment now defined by two restriction sites, was separated from template D N A in 1 % agarose gel. The resulting fragment was then inserted in the modified A B C R / p C D N A 3 plasmid (below) previously cut with EcoRV and Kpnl. In order to insert the fragment bearing the D846H mutation back into an equivalent fragment in the wild type A B C R / p C D N A 3 , some modifications needed to be done on the A B C R / p C D N A 3 . That is, the EcoRV and Kpnl sites in the multiple cloning site of this construct needed to be removed. This was done by cutting A B C R / p C D N A 3 both with Hindlll and NotI, blunting the NotI end and re-ligating the blunted ends. 2.5 EXPRESSION OF T H E CONSTRUCTS IN COS C E L L S 2.5.1 Maintenance of COS cells A l l media used for tissue culture were filter sterilized and special precaution was made to keep the tissue culture area sterile. COS-1 or C O S - 7 cells were grown in 5% CO2 and 3 7 ° C , in Dulbecco's Modified Eagle Medium ( D M E M ) which was supplemented with 10% fetal bovine serum, 50 units/ml penicillin and 50 u.g/ml streptomycin (from this point on, this D M E M will be referred to by as "complete"). To maintain the cell line, the cells were passaged almost twice a week as follows. First the old medium was removed and the cells were washed in 2 ml of t ryps in -EDTA. Then, to detach the cells from the 10 cm dish, 1ml of t ryps in -EDTA was added and the cells were incubated for 5 minutes at 3 7 ° C . Sometimes the C O S cells remained attached to the dish, and a gentle tapping of the dish was required to dislodge them. To stop trypsinization, 9 ml of new complete D M E M was added. The cells were re-plated at a dilution of 1:20 in complete D M E M and grown at 5% CO2, 37 °C until further passaging. 2.5.2 Transfection of COS cells using calcium chloride The wild type and the mutant human A B C R D N A s subcloned into the NotI blunted Xbal sites of the p C D N A 3 (Invitrogen) or the modified p C D N A 3 , were used to transfect mammalian C O S cells. One of the transfection methods employed was the calcium phosphate method of Chen and Okayama (1987). The cells were plated at a concentration 41 of 6x10 5 cells per 10 cm dish the night before the transfection. The next day, 30 jj,g of the A B C R / p C D N A 3 or mutant A B C R / p C D N A 3 construct was diluted with sterile water to a final volume of 372 ul. Then, 123 ul of I M C a C l 2 was added to the diluted D N A and mixed. While gently vortexing the D N A - C a C l 2 mixture, 495 ul of 2x B B S (50 m M N , N -bis-(2-hydroxyethyl)-2-aminoethanesulfonic acid (BES) , 280 m M N a C l , 1.4 m M Na2HP04, p H 6.95) was added dropwisely. The mixture was incubated for 20 min at room temperature and then added to the C O S cells while gently swirling the plate. The cells were incubated at 3 7 ° C and 5% C 0 2 overnight. The following morning the cells were washed once with 10 ml of filtered phosphate buffered saline (PBS) containing 5 m M E D T A , and incubated for another 24 hours at 3 7 ° C , 5% C 0 2 in 10 ml of fresh D M E M , before being harvested. 2.5.3 Transfection of COS cells using the SuperFect reagent The recombinant plasmid D N A (30 p.g) was diluted in sterile water to a final volume of 50 i l l . To further dilute the D N A , 500 ul of D M E M (without serum and antibiotics) was added. Then 60 ul of SuperFect transfection reagent (Qiagen) was added to the D N A / D M E M dropwisely and the mixture was allowed to sit for 5-10 min at room temperature. Meanwhile, the adherent C O S cells at 70-90% confluency per one 10 cm dish were washed with 10 ml of sterile P B S . After the incubation, 3 ml of complete D M E M was added to the DNA/SuperFect mixture and the resulting D N A / S u p e r F e c t / D M E M was immediately added to the washed C O S cells. The cells were allowed to incubate for 2 hours at 3 7 ° C and 5% C 0 2 , before being washed with P B S . Complete D M E M (10 ml) was then added and the cells were incubated for 48 hours in the 3 7 ° C incubator in the presence of 5% C 0 2 . 2.5.4 Harvesting the COS cells Forty-eight hours post-transfection, each 10 cm C O S cell dish was first washed in PBS and then scraped in P B S . If more than one dish was used per D N A sample, the cells were pooled together after being scraped. The cells were centrifuged at 2800 g for 10 min at 4 ° C and the pellet resuspended in 50-100 ul of P B S . The resuspended cells were then 42 added dropewisely to 0.5 ml of solubilization buffer (50 m M N a H E P E S , p H 7.5, 100 m M NaCI, 1 mg/ml sonicated soybean phospholipid, 18 m M C H A P S , 10% glycerol, 1 m M D T T and 3 m M M g C l 2 and lx complete protein inhibitor cocktail). Solubilization was carried for 30-40 min on ice with frequent, yet gentle vortexing. The solubilized cells were finally centrifuged at 40 000 rpm in T L A 100.4 rotor (Optima T L Ultracentrifuge, Beckman) for 10 min at 4 ° C . Approximately 50 ul of the supernatant representing total cell lysate was set aside for further analysis. 2.5.5 Purification of ABCR from solubilized COS cells The supernatant obtained after centrifuging the solubilized cells (above) was mixed with 100-150 |uil of a 50% suspension of R i m 3F4-Sepharose 2B beads, in 0.4 u M filter Ultrafree-MC spin columns. Before the supernatant was mixed with the R i m 3F4-Sepharose coupled beads, the 3F4-beads were pre-equilibrated with buffer A (50 m M N a H E P E S , p H 7.5, 100 m M NaCI, 1 mg/ml sonicated soybean phospholipid, 10 m M C H A P S , 10% glycerol, 1 m M D T T and 3 m M M g C l 2 ) . The 3F4-beads and the cell lysates were incubated for 1 hour at 4 ° C , after which the unbound fraction was removed by low speed centrifugation. The beads were then washed 6x with 0.4 ml of buffer A in spin columns. To elute the wild type or the mutant A B C R s , 30 ul of buffer A containing 0.2 mg/ml 3F4 peptide ( Y D L P L H P R T G ) was added to the column and allowed to incubate for 15 min at 4 ° C . The eluted fraction was obtained by low speed centrifugation, and the elution was repeated a second time. A third elution step using only 30 ul of buffer A was also carried out to yield a final volume of 90 u.1 of eluate. The eluted fractions (30-45 ul) were removed and added to 10-15 ul of 4x SDS buffer (8% S D S , 200 m M T r i s - H C l p H 6.8, 40% glycerol, 0.4% bromophenol blue and 16% (3-mercaptoethanol) and loaded on a 6.5 or 7.5% polyacrylamide gel. 2.5.6 Immunofluorescence labeling of COS cells For immunofluorescence labeling, the C O S cells were grown and transfected by either one of the transfection methods outlined in sections 2.5.2-2.5.3. A sterile glass cover slide, however, was added to each 10 cm dish when the cells were re-plated the day before 43 transfection. Forty-eight hours post-transfection, before the cells were scraped off the plate, each glass cover slide covered with cells was removed and placed in a well of a 6-well plate. While the rest of the C O S cells were harvested as before (section 2.5.4), the glass cover slides were gently washed twice with P B S containing 0.1 m M C a C l 2 and 1 m M M g C l 2 . The cells on the glass cover slides were fixed by incubating the slide for 15 min in 4% paraformaldehyde made in P B S . Each well containing a glass slide was washed three times (5 min per wash) with PBS supplemented with 0.1 m M C a C l 2 , 1 m M M g C l 2 and 10 m M glycine. Next, the cells were both blocked and permeabilized for 15-30 min in 100 m M sodium phosphate buffer, p H 7.4, containing 0.3% Triton-X-100 and 10% goat serum. When the blocking/permeabilizing buffer was removed, the primary antibody 3F4, diluted 1:3 in 100 m M sodium phosphate buffer containing 0.1%) Triton-X-100 and 2.5% goat serum, was added to the cells. The primary antibody was incubated with the cells for 1 hour at room temperature before being washed away with three 10 min washes of 100 m M sodium phosphate containing 0.05% Tween 20. The secondary antibody, goat anti-mouse antibody conjugated to C Y 3 , was diluted 1:800 in 100 m M sodium phosphate buffer supplemented with 0.1% Triton-X-100 and 2.5% goat serum. Finally the cells were washed twice in sodium phosphate buffer and mounted by adding a few drops of Mowio l mounting solution ( C a l b i o c h e m ® ) onto the glass cover slips. Each cover slip was then placed face down on a microscope slide and fixed in position by adding transparent nail polish around the periphery of the glass cover slip. After the nail polish was dried, each microscope slide was viewed using a Zeiss Axioplan 2 microscope equipped with a digital imager. 2.6 COS C E L L MEMBRANE PREPARATION To prepare C O S cell membranes from transfected cells, ten 10 cm dishes were transfected per plasmid D N A construct using the calcium phosphate method of transfection (section 2.5.2). Two days following transfection, the cells were scraped in 2 ml of hypotonic buffer (10 m M T r i s - H C l , 1 m M E D T A , p H 7.4) per dish, pooled together to a final volume 44 of 20 ml and centrifuged at 2800 g for 10 min at 4 ° C . The cells were washed in 12 ml of the same buffer and re-centrifuged again at 2800 g for 10 min, 4 ° C . To lyse the cells, the pellet was resuspended in 12 ml of the hypotonic buffer containing l x Complete Protease Inhibitor Cocktail (Roche Diaganostics, Laval , Quebec) and left on ice for 1 hour. The cells were further disrupted by both using a Duall glass homogenizer 15 times and by passing the homogenate 3x through a 26-gauge needle. The solubilized C O S cells (6ml) were then layered on top of an ice cold step sucrose gradient made up of 5 ml of 5% sucrose (w/v) and 6ml of 60% sucrose in gradient buffers (10 m M T r i s - H C l , 150 m M NaCI, 1 m M M g C l 2 , 1 m M C a C l 2 , 0.1 m M E D T A ) . The samples were centrifuged in a Beckman SW27.1 rotor at 20 000 rpm for 1 hour at 4 ° C . The C O S membranes formed a band at the interface between the two sucrose layers and were collected with the aid of a syringe equipped with a 16-gauge needle. The membranes were washed twice with 10 m M T r i s - H C l , 1 m M E D T A , p H 7.4 and collected by centrifugation at 50 000 rpm for 20 min at 4 ° C in a TLA100 .4 rotor (Beckman). The membrane pellet was resuspended in 500 ul of 5% gradient buffer, 10% glycerol and lx Complete Protease Inhibitor Cocktail . The membrane fractions were stored at - 8 0 ° C . 2.7 BOVINE ROD OUTER SEGMENT PREPARATION Bovine R O S were prepared under dim red light as follows. One hundred retinas were gently inverted in 40 ml of homogenizing buffer containing 20% (w/v) sucrose, 10 m M taurine, 10 m M p-D-glucose, 0.25 m M M g C l 2 , and 20 m M T r i s - H C l , p H 7.4. The retinas were passed through a Teflon filter (300 um mesh). The resulting filtrate was divided and layered on top of six 28-50% (w/v) sucrose gradients. The samples were centrifuged in a SW28 rotor (Beckman Instruments) at 26 000 rpm for 1 hour at 4 ° C . The R O S band located at around 33% sucrose was removed from each gradient and washed with 20 ml of homogenizing buffer by centrifuging in a Sorvall SS-34 rotor at 10 000 rpm for 10 min. The washed R O S were resuspended in 1 ml of homogenizing buffer and pooled together, and an extra 2 ml was used to rinse the centrifuge tubes. The resulting 8 ml of R O S in homogenizing buffer was divided into 1 ml aliquots and stored at - 8 0 ° C . The protein 45 concentration was determined by bicinchoninic acid ( B C A ) assay and was approximately 5-8 mg/ml. 2.8 IMMUNOAFFINITY PURIFICATION OF ABCR F R O M ROS A B C R was purified from 0.5-5 mg of R O S membranes, depending on the experiment. A l l washing steps were carried under the dim light. R O S membranes were first diluted in 10 m M H E P E S , p H 7.4 and centrifuged at 86 000 g for 10 min in a T L A 100.4 rotor. After two more washes in 10 m M H E P E S , the R O S membrane suspension was solubilized in buffer A containing 18 m M C H A P S and lx Complete Protease Inhibitor Cocktail for 20-30 min on ice. The mixture was centrifuged as above. The soluble fraction was removed and incubated for 1 hour at 4 ° C with 100 or 150 ul of a 50% suspension of R i m 3F4-Sepharose 2B beads, pre-equilibrated with buffer A (50 m M N a H E P E S , p H 7.5, 100 m M NaCI, 1 mg/ml sonicated soybean phospholipid, 10 m M C H A P S , 10% glycerol, 1 m M D T T and 3 m M M g C b ) . The unbound fraction was removed by low speed centrifugation and the bound A B C R was eluted by using 3F4 peptide, as outlined in section 2.5.5. 2.9 EXTRACTION OF PHOSPHOLIPIDS F R O M ROS Phospholipids were extracted from R O S by following the method of Folch et al. (1957) and when feasible, steps were performed under nitrogen gas. Organic solvents such as chloroform and methanol contained 50 |J,g/ml of butylated hydroxytoluene ( B H T ) and nitrogen gas was bubbled through them prior to use. R O S membranes (5 mg) were washed three times with 10 m M potassium phosphate (pH 7.0) by centrifugation at 40 000 rpm in a T L A 100.4 rotor for 10 min. The membranes were resuspended to a final volume of 200 ul with 10 m M potassium phosphate. Approximately 200 ul of 1 M NH2OH (adjusted to a p H 7.0 by the addition of 1 M N a H C 0 3 ) and 930 u.1 of methanol (containing B H T ) were added to the resuspended R O S membrane. After 10 min of incubation on ice, 1 ml of water and 1.87 ml of chloroform (containing B H T ) were added and the resulting mix was vortexed thoroughly. The organic and the aqueous phase were separated by centrifugation 46 in a table top centrifuge for - 1 0 min at ~ 400 g. The organic phase was removed and washed with 1.4 ml of 0.3 M N a C l in water and 930 ul of methanol, before being dried under nitrogen gas. After the organic solvent was evaporated, the lipids were resuspended in 150 ul of 1:1 solution of methanol/chloroform containing 50 ug/ml of B H T and applied to a 0.5 mm Silicagel G , thin layer chromatography plate (Merck) under nitrogen gas. The plate was developed in 1:1 hexane/ether solution, in the presence of nitrogen gas. The phospholipids, estimated to remain at the origin, were scrapped off the plate along with the Silicagel. The phospholipids were recovered from the Silicagel by adding 3 ml of 1:1 methanol/chloroform solution (containing 50 |J.g/ml of B H T ) , vortexing, and separating the phases by centrifugation at low speed in a table top centrifuge. These steps were repeated once more and the resulting R O S extracted phopholipids were analyzed for their phosphorus content by using the method of Zhou and Arthur (1992). 2.10 DETERMINATION OF LIPID PHOSPHORUS CONTENT The yield of phospholipids extracted from R O S (section 2.9) was determined by the malachite green method of Zhou and Arthur, 1992. Briefly, different volumes (10, 20, 30, etc) of the R O S extracted lipids were added to 13 x 100 mm glass tubes and dried under N 2 gas. In the fumehood, 100 jul of perchloric acid was added to all tubes including empty standard tubes. Each tube was sealed with aluminum foil and heated to ~ 1 4 0 ° C for 1-3 hours or until the yellowish color of the samples turned colorless. Phosphorus standards, containing different nmole amounts of phosphorus, were prepared in duplicate by adding the required amount of 1 m M potassium phosphate to the glass tubes containing 100 ul perchloric acid. Water was added to each standard tube to bring the final volume to 400 ul, while 300 ul of water was added to the sample tubes. The working solution (2 ml), prepared from solutions I, II and III (below), was added to each tube and mixed. After 30 min incubation at room temperature, the absorbance of each solution at 660 nm was measured and the amount of phosphorus (nmol/ul) in the R O S extract was determined by a standard curve. The concentration of phopholipids in the R O S extract (mg/ml) was then determined by taking into account the volume of lipids extracted and by using the molecular weight of phospholipids (775 g/mol). 47 Working solution: 3 volumes of solution I, 1 volume of solution II and 0.016 volume of solution III Solution I: 0.3% malchite green base in 1.5 M H C I Solution II: 4.2% ammonium molybdate in 1.5 M H C I Solution III: 4% Tween 20 2.11 RECONSTITUTION OF ABCR PROTEINS INTO LIPID VESICLES Purified wild type, mutant and R O S A B C R s were reconstitued into l ipid vesicles consisting of R O S extracted lipids or a 50:50 mixture of brain polar lipids with either D O P E or S D P E . To produce the 20 mg/ml R O S lipid solution, a known amount of R O S extracted lipids were dried under N 2 gas and resuspended in the required volume of buffer C (25 m M H E P E S , p H 7.4, 140 m M N a C l , and 10% glycerol). To produce the 20 mg/ml of a mixture of 50:50 brain polar lipid and D O P E or S D P E , first 8 mg of each of brain polar lipids and D O P E or S D P E were dissolved in 0.4 ml of 1:1 mixture of chloroform:methanol. Then, 0.2 ml of each of the two lipids were mixed together and dried under N 2 gas. The lipid mixture was dissolved in 0.4 ml of buffer C (25 m M H E P E S , p H 7.4, 140 m M N a C l , and 10% glycerol) by extensive vortexing and sonicating. Next, 9 u l of the 20 mg/ml lipid solution was mixed with 3 ul of 15% octyl-P-D-glucopyranoside (w/v) in buffer C . Then, 24 u l o f either the purified wild type or mutant or R O S A B C R was added to the lipid-detergent mix and incubated for 30 min on ice. Similarly, 24 u l of the eluate from untransfected cells and 24 u l of reconstitution buffer (termed buffer blank) were added to the lipid-detergent mix, acting as negative controls. After the incubation, 200 ul of reconstitution buffer (25 m M H E P E S , p H 7.5, 0.14 M N a C l , 10% glycerol, 1 m M D T T , 1 m M E D T A ) at room temperature was added and the resulting mixture was incubated for a further 2 min. Immediately following the two minutes of incubation, the mixture was passed through 200 ul of packed Extracti-Gel D resin in a Mobicol mini-column equipped with a 10 um pore-size filter (MoBiTec , Gottingen, Germany), previously equilibrated in reconstitution buffer. The reconstituted proteins in lipid vesicles were collected on ice by applying gentle pressure with a 10 ml 48 syringe. Before carrying out the ATPase assay, M g C l 2 was added to the reconstituted samples to a final concentration of 5 m M . 2.12 ATP HYDROLYSIS OF RECONSTITUTED PROTEINS 2.12.1 ATPase assay The ATPase assay for each sample was generally carried out in triplicate. The reconstituted samples (8 ul) were pipetted into 0.5 ml microcentrifuge tubes and kept on ice while the lOx solutions of 0.5 m M A T P and 0.5 m M all-fr^ms-retinal (in ethanol) were being prepared fresh. The 0.5 m M A T P solution contained: 0.5 m M A T P and either 0.1 u C i (for overnight exposures) or 0.2 u C i of [ a - 3 2 P ] A T P (for 3 hours exposures) in reconstitution buffer. To the 8 ul of reconstitued samples, 1 ul of the 0.5 m M a\\-trans-retinal or 1 ul of the ethanol blank (produced along side the 0.5 m M retinal solution, but instead of retinal contained ethanol), was added and the resulting mixture was incubated on ice for 10 min. The reaction was initiated by the addition of 1 ul of the 0.5 m M labeled A T P solution and was carried out for 30 min at 3 7 ° C before being stopped by 4 ul of 10% SDS. In order to separate A D P from A T P , 1 ul from each reaction mixture was deposited onto a polyethyleneimine cellulose plate (Aldrich) and separated by thin layer chromatography in 0.5 M LiCl /1 M formic acid. After the plate was dried, it was exposed to a storage phosphor screen for either 3 hours or overnight depending on the amount of hot A T P used (0.2 u C i versus 0.1 uCi) . The screen was then scanned in a Phosphorlmager SI (Molecular Dynamics) and the data analyzed (below). 2.12.2 Activity calculations The information obtained by scanning the phosphor screen was used to calculate the A T P hydrolysis of the reconstituted wild type, mutant and R O S A B C R proteins. After scanning the phosphor screen, IPLab Gel analysis Software (Signal Analytics Corp. , Vienna, V A ) was used to quantify the spots corresponding to A D P and A T P . The ratio of A D P to the original A T P 0 ( A T P 0 = A D P + A T P ) was first calculated for all of the samples, then the buffer blank was subtracted from all samples. Since each sample in the assay was carried 49 out in triplicate, an average value for each sample was obtained. The total nmole of A T P hydrolyzed for each sample was then calculated (below) by taking into account a number of dilutions carried over from the solubilization, purification and reconstitution steps: Total nmoles of A T P hydrolyzed = (average A D P / A T P o ) x (0.5 m M A T P x 1 ul A T P solution) x (volume of reconstituted samples / 8 ul) x (volume of eluate / 24 ul) x (volume of total cell lysate / 500ul) Next the total nmoles of A T P hydrolyzed per mg of total proteins in cell lysate (determined by B C A , section 2.15) was determined for each sample. The values obtained for the untransfected cells, which acted as background values, were subtracted from the rest of the wild type and mutant samples. The nmoles of A T P hydrolyzed in the reconstituted sample was then calculated as follows: nmoles A T P hydrolyzed per reconstituted sample = (total nmoles of A T P hydrolyzed / mg of total protein) x total mg of protein x ( 500 ul / total cell lysate) x ( 24 ul / eluate volume) Finally, the total nmoles of A T P hydrolyzed per minute in the reconstituted sample was determined by dividing the nmole of A T P hydrolyzed in the reconstituted samples by the 30 min during which the assay was carried out. The resulting values divided by the amount of protein present in each of the reconstituted samples represented the specific activities (i.e. the nmole of A T P hydrolyzed per minute per mg of protein). 2.13 AZIDO-ATP LABELING OF COS MEMBRANES 2.13.1 Membrane preparation A z i d o - A T P labeling was carried out on both R O S membranes (section 2.7) and membranes obtained from C O S cells transfected with either the wild type or mutant constructs (section 2.6). These membranes were first washed as follows. R O S membranes (0.5 mg) were washed twice with 10 m M Tris, 1 m M E D T A , p H 7.4 by centrifuging at 20 000 rpm for 8 min in a T L A 100.4 rotor (Optima T L Ultracentrifuge, Beckman). The pellet was resuspended in 0.5 ml of the same buffer, allowed to incubate for 20 min (to lyse plasma membrane) and re-centrifuged for 12 min at 30 000 rpm. The 50 pellet was resuspended in 120 ul of assay buffer (20 m M Tris, 0.15 M NaCI, 5 m M M g C l 2 , p H 7.4) and 60 ul was saved for running on a gel. Approximately 3-4 mg of C O S cell membranes were washed once in 10 m M Tris, 1 m M E D T A , p H 7.4 containing lx Complete Protease Inhibitor Cocktail , by centrifuging in T L A 100.4 rotor for 12 min at 30 000 rpm. The membrane pellet was resuspended in 60 ul of assay buffer (20 m M Tris, 0.15 M NaCI, 5 m M M g C l 2 , p H 7.4). 2.13.2 Azido-ATP assay The 8-azidoadenosine-5'-triphosphate [a- P] (20 Ci /mmol solution), at 1.5 ul per sample of C O S or R O S membranes, was dried under nitrogen gas in a darkened fume hood and resuspended in 6 ul of assay buffer (20 m M Tris, 0.15 M NaCI, 5 m M M g C l 2 , p H 7.4) per sample. Then, 6 ul of the azido A T P solution was added to each sample and the crosslinking reactions were initiated by irradiating the samples with ultraviolet light (mineral light U V Lamp; 254) at a distance of 11 cm above the samples for 10 min. The reactions were stopped by turning off the light. The samples were then washed in 500 ul of 20 m M Tris, by centrifuging them in a TL45 rotor at 30 000 rpm for 15 min. The samples were resuspended in 30 ul of 20 m M Tris before being solubilized for 30 min on ice in 200 ul of 2% Triton-X-100, 20 m M Tris, 0.15 M NaCI and lx Complete Protease Inhibitor Cocktail. The solubilized membranes were centrifuged in a T L A 100.4 rotor at 40 000 rpm for 10 min, 4 ° C . The resulting supernatant was removed and incubated for 1 hour at 4 ° C with 150 ul of a 50% suspension of R i m 3F4-Sepharose 2B beads, pre-equilibrated with 0.2% Triton-X-100, 20 m M Tris, 0.15 M NaCI. The immunoaffinity chromatography procedure employed to purify the A B C R s , was similar to the one outlined in section 2.5.5. Briefly, following the 1 hour incubation, the R i m 3F4-beads were washed six times with 0.4 ml of 0.2% Triton-X-100, 20 m M Tris, 0.15 M NaCI. The bound proteins were eluted twice in 40 ul of 0.2% Triton-X-100, 20 m M Tris, 0.15 M NaCI containing 4% SDS and lx Complete Protease Inhibitor Cocktail . O f the 80 ul of eluates, 40 ul was mixed with 13 ul of 4x SDS sample buffer and loaded on one 6.5% polyacrylamide gel, while the other 40 ul was loaded on a second gel. S D S - P A G E was 51 carried out as outlined in section 2.14. One gel was destined for western blotting and thus was transferred onto an Immobilon membrane (detailed explanation in section 2.14) while the other gel was stained in Coomassie blue stain for 20 min and destained in 10% acetic acid for 15 min, before being dried in a gel dryer (Bio-Rad). The dried gel was then exposed to either an X-ray film for 2-4 days or a phosphor storage screen overnight and scanned in a Phosphorlmager (Molecular Dynamics). 2.14 PROTEIN ELECTROPHORESIS AND WESTERN BLOTTING Sodium dodecyl sulfate polyacrylamide gel electrophoresis ( S D S - P A G E ) was routinely carried out using the gel apparatus from Hoefer. Samples were prepared by mixing with the appropriate amount of 4x SDS sample buffer (8% SDS, 200 m M T r i s - H C l p H 6.8, 40% glycerol, 0.4% bromophenol blue and 16% P-mercaptoethanol). Typically, 10-45 ul of samples were applied to each lane of a 6.5% or a 7.5% polyacrylamide gel and electrophoresis was carried out using the Laemmli buffer system (1970) at 150 V , 27 mA/gel . Gels were stained for 1 hour in Coomassie Blue stain (25% isopropanol, 10% acetic acid, 0.025% Coomassie Brilliant Blue R-250) and subsequently destained in 10% acetic acid for 4 hours. Gels were dried between two sheets of cellophane (Bio-Rad) or using the gel drying apparatus. Western blotting was performed following gel electrophoresis as follows. The unstained gel was soaked in the transfer buffer for 10 min (25 m M Tris, 195 m M glycine, 10%) methanol, p H 8.3). The proteins on the gel were then transferred onto the Immobilon P membrane (Millipore) at 300 m A for 40 min in a semidry transfer apparatus (Biorad) using the transfer buffer. After the Immobilon was soaked in methanol followed by 5 min in P B S , it was blocked in 1 % skim mi lk/PBS containing 0.05 % Tween 20 (PBS-T) for 30 min. The membrane then was incubated with the primary antibody diluted in 0.1% milk/PBS for 1 hour. The primary antibodies used were either N-terminus R i m 5B4 monoclonal (1:20 dilution) or anti-calnexin Ig (1:1000). The blots were washed three times for 10 min each with P B S - T . The secondary antibody, a goat anti-mouse Ig or a goat anti-rabbit Ig linked to horseradish peroxidase, was diluted 1:5000 in 0.1 % mi lk /PBS-52 T. After 1 hour of incubation the blot was washed three times with P B S - T as before. Proteins on the blot were detected by the E C L Western blotting detection system (Amersham) according to the manufacturer's instructions. To strip a blot of antibody, the membrane was incubated at 5 0 ° C in stripping buffer (100 m M (3-mercaptoethanol, 2% SDS, 62.5 m M T r i s - H C l , p H 6.7) for 30 min, while rotating. The membrane was washed twice in P B S - T for 10 min and blocked in 4% skim mi lk /PBS for lhour followed by a second incubation in 1% skim milk/PBS at room temperature. The primary antibody was then added and the western blotting procedure was continued as outlined above. 2.15 PROTEIN DETERMINATIONS The amount of proteins in R O S membrane preparations was determined by B C A method (Pierce), using different concentrations of a stock B S A solution (2mg/ml). The protein concentrations in samples termed "total cell lysates" in section 2.5.4 (obtained after solubilizing the transfected C O S cells) were also determined by B C A . The amounts of A B C R s purified from solubilized C O S cells were determined by running a polyacrylamide gel, staining it with Coomassie Brilliant Blue stain and then scanning the gel with an Ultroscan X L laser densitometer. By comparing the intensities of Coomassie stain of the eluted A B C R bands with those of B S A standards, the amounts of purified proteins were determined. The amounts of reconstituted proteins were calculated from the amount of purified proteins present in the eluates, while taking into account a number of dilutions. Furthermore, to quantitate the amount of mutant A B C R expressed in the C O S cells, western blots of total cell lysates were scanned with the Ultroscan X L laser densitometer and expressed as a percentage of the wild type protein. 53 RESULTS 3.1 SELECTION OF ABCR MUTANTS O f the 89 disease causing A B C R mutations, the following five were chosen (Fig 13). i) D846H: This is a missense mutation in the putative sixth transmembrane segment, in which a negatively charged residue (aspartate) is replaced with a neutral/positively charged histidine. This mutation is of interest since aspartate is the only charged residue in the middle of the putative sixth transmembrane segment, a relatively hydrophobic region. Thus this substitution may play a significant role in the folding or function of A B C R . ii) T1526M: This is a missense mutation in which two polar residues are exchanged. This mutation was chosen due to its location. According to the current A B C R model of Illing et al. (1997) and Bungert et al. (2001) (Fig 10, A), this position is located inside the disk lumen, on the second large intradiskal loop between hydrophobic segments 7 and 8. The model proposed by Azarian and Travis (1997) and Sun et al. (2000), however, places this mutation on the cytoplasmic side of the disk membrane (Fig 10, B). Thus, this mutation may help in supporting one model versus the other. iii) R2038W and iv) R2077W: These missense mutations involve the replacement of positively charged arginines by tryptophans (an aromatic residue). Both of these mutants are located in the second nucleotide binding domain (NBD) of A B C R . These mutations may provide some information regarding the ability of A B C R to bind and/or hydrolyze A T P . v) R2106C: This missense mutation is located in the second N B D following the Walker B sequence. The positively charged arginine residue is replaced with a polar, uncharged cysteine residue. This mutation is predicted to have a less dramatic effect on the A T P binding or hydrolysis, as it is not located in a conserved sequence in the N B D . 54 T1526M R2077W Fig 13. Putative model of ABCR showing the five Stargardt's disease mutations. The location of the five missense mutations resulting in Stargardt's disease (Lewis et al, 1999) selected for study are shown. D846H missense mutation occurs in the putative sixth transmembrane segment. T1526M is located in the second intradiskal loop in this model. R2038W, R2077W and R2106C are located in the second nucleotide binding domain. (NBD) . 55 3.2 GENERATION OF CONSTRUCTS The wild type A B C R was inserted in pRK5 plasmid by J. Ahn and denoted A B C R / p R K 5 . As explained in section 2.4, the R2038W, R2077W and R2106C were first placed in the wild type A B C R / p R K 5 and then the entire mutated A B C R s were individually subcloned into pCDNA3. Possible clones containing the correct constructs were screened by restriction enzyme digestion. One enzyme often used in the screening process was Apal. Digestion of A B C R / p C D N A 3 with this enzyme resulted in four fragments of approximately 6.8, 3.3, 2.0 and 0.15 kb in size (Fig 14, A ) . Positive clones were sequenced for the presence of the desired mutation. To generate the T1526M mutant, the QuickChange Site Directed Mutagenesis Ki t was used. As shown in Fig 14, B the entire 12.3 kb plasmids containing the possible nucleotide change were generated from all three PCR reactions. Sequencing revealed the presence of the correct nucleotide change in all three amplified constructs. D846H was generated by amplifying a 2.1 kb fragment bearing the mutation via PCR (Fig 14, C). The fragment was subsequently inserted in modified A B C R / p C D N A 3 , after the corresponding wild type fragment was removed. 3.3 TRANSFECTION AND EXPRESSION OF WILD TYPE AND MUTANT ABCR Two methods of transient transfections were employed separately: calcium phosphate (Chen and Okayama, 1987) or the SuperFect (Qiagen). The calcium phosphate method required the transfection of 3-4 dishes of COS cells to get a similar expression level as one plate transfected by SuperFect, for the wild type A B C R / p C D N A 3 . This is due in part to the fact that a single dish used for calcium phosphate was 30-40% confluent whereas the dish used for SuperFect was 70-90% confluent. However, the calcium phosphate method proved to be more economical. The COS cells were examined 48 hours post-transfection for the level of protein expression. This was achieved by solubilizing the cells in 18 m M C H A P S and centrifuging the solubilized cells to pellet the cell membranes. The resulting total cell 56 A B C 1 2 1 2 3 4 5 1 2 Fig 14. Generation of the selected constructs. (A) Restriction enzyme digestion of A B C R / p C D N A 3 with Apal. Two ug of A B C R / p C D N A 3 was digested with 10 units of Apal for 1 hour at 3 7 ° C , and the resulting fragments were separated by 1% agarose gel electrophoresis (lane 2). The size of the fragments following Apal digestion of A B C R / p C D N A 3 are: 6.8 kb, 3.3 kb, 2.0 kb and 0.15 kb (not visible). Lamda/Hindlll ladder was used in lane 1. (B) Generation of T 1 5 2 6 M / p C D N A 3 construct by site directed mutatgenesis. Different amounts of template A B C R / p C D N A 3 were used in three P C R reaction mix. After 16 rounds of P C R , 10 ul from each mixture was removed, loaded on a 1% agarose gel and subjected to electrophoresis; lane 1, Lamda/Hindlll ladder; lane 2, amplified plasmid from 10 ng of template D N A ; lane 3, amplified plasmid from 20 ng of D N A ; lane 4, amplified plasmid from 50 ng of template D N A ; lane 5, control reaction: missing one primer. (C) Amplification of the 2.1 kb fragment bearing the D846H mutation. Ten ng of A B C R / p C D N A 3 was subjected to 25 rounds of P C R . Ten (al of the resulting product was loaded on 1% agarose gel; lane 1, Lamda/Hindlll ladder; lane 2, D846H fragment of 2.1 kb. 57 lysates were examined for the presence of the recombinant proteins by western blotting. Untransfected COS cells were used as negative controls. The Coomassie Brilliant Blue stained gel of the total cell lysates showed numerous proteins present in the cell extract (Fig 15, A). The recombinant A B C R s (wild type and mutants) were not distinguished among other COS cell proteins on this gel. However, the western blot, labeled with the Rim 5B4 antibody, revealed the presence of a band corresponding to A B C R in all lanes except the untransfected control lane (Fig 15, B). As the figure shows the wild type A B C R along with the five mutants were expressed in COS cells. Generally 30% of the transfected COS cells expressed the recombinant proteins. The level of expression was quantified by scanning the western blot bands with a densitometer, and expressing the results as a percentage of the wild type. The levels of A B C R s expressed were comparable between calcium phosphate and SuperFect except in the D846H mutant, which expressed at a higher level using the calcium phosphate method. Table 3 shows the expressed levels of mutant A B C R s as a percentage of the wild type. 3.4 PURIFICATION OF WILD TYPE AND MUTANT ABCR F R O M SOLUBILIZED COS C E L L S Once the A B C R variants were shown to be expressed in COS cells, the proteins were purified from total cell lysates using immunoaffinity chromatography. A B C R was previously purified from ROS using the Rim 3F4 antibody coupled to beads (Illing et al, 1997). A similar purification scheme was employed here. The Coomassie blue stained gel and western blots of immunoaffinity purified A B C R are shown in Fig 16 A and B. Both the wild type and the mutant A B C R s were detected as the principal protein in each lane. No protein band corresponding to A B C R was present in the untransfected lane, as expected (Fig 16, A). To quantify the amount of proteins purified, varying amounts of B S A were loaded along side the eluted samples on the polyacrylamide gel. Following Coomassie blue staining and destaining, the purified A B C R s along with the B S A standards were scanned with a 58 1 2 3 4 5 6 7 2 0 5 -116_ 9 7 -6 6 -4 5 -2 9 -B 1 2 3 4 5 6 7 Fig 15. Expression of the wild type and the five mutant ABCR constructs in COS cells. The untransfected C O S cells (Janes 1) and the cells transfected with wild type A B C R (lanes 2), D846H (lanes 3), R2038W (lanes 4), R2077W (lanes 5), R2106C (lanes 6), T1526M (lanes 7) were solubilized in 18 m M C H A P S . The insoluble fractions were removed and the supernatants were loaded onto 7.5% S D S - P A G E gels. (A) Proteins stained with Coomassie Brilliant Blue; (B) Proteins transferred to an Immobilon membrane for western blotting using R i m 5B4 antibody for detection of A B C R . Thirty ul of the supernatant was loaded in (A), 10 ul of the supernatant in (B). The size of the molecular weight markers are shown in kDa. 59 Table 3 Expression levels of mutant ABCRs relative to wild type ABCR in COS cells A B C R variant Expression level Wi ld type 100 % D846H C 31 ± 1 % D 8 4 6 H d 68 ± 16 % R2038W 59 ± 21 % R2077W 57 ± 20 % R2106C 71 + 17 % T1526M 81 ± 9 % a With the exception of the D846H variant, values compiled are the averages of both methods of transfection (SuperFect and CaPC^j) obtained from nine experiments. b Expression levels of the proteins are cacluated as the percentage (%) of wild type. c The expression level of D846H obtained from four experiments during which SuperFect was used to transfect the cells. d The expression level of D846H obtained from six experiments when CaPCM method of transfection was used. 60 1 2 3 4 5 6 7 8 9 10 11 12 - A B C R r - BSA B 1 2 3 4 5 6 Fig 16. Purification of wild type and mutant ABCRs from solubilized COS cells. The supernatant obtained from solubilized C O S cells (0.5 ml) were mixed with 150 u.1 of 3F4-beads for 1 hour at 4 ° C . After the column was washed and the A B C R proteins were eluted using the 3F4 peptide, 30 ul from each eluate was loaded on a 7.5% gel and stained with Coomassie Brilliant Blue (A) while another 10 ul was loaded on a gel used for western blotting (B). In both panels lanes 1, wild type A B C R ; lanes 2, D846H; lanes 3, R2038W; lanes 4, R2077W; lanes 5, R2106C; lanes 6, T1526M. In panel A lane 7, untransfected; lanes 8-12, B S A standards: 0.1, 0.2, 0.3, 0.4, 0.5 ug. 61 densitometer. From a standard curve, it was determined that between 0.3-0.9 u.g of proteins could be purified from two 30-40% confluent C O S cell plates. These values varied from one experiment to another and strongly depended on the transfection efficiency and the purification scheme used. The immunoaffinity purification scheme employed here and explained in section 2.5.5 allowed the recovery of 40-50%) of the A B C R proteins (wild type and mutants). The rest was either present in the unbound fraction or remained attached to the column after peptide elution. Increasing the bead volume (from 50 to 70 ul) and increasing the incubation time with the peptide during elution, slightly increased the recovery of the purified proteins. Improving the transfection efficiencies, in order to increase the yield of expressed proteins was difficult and often fruitless despite extensive human efforts. 3.5 IMMUNOFLUORESCENCE MICROSCOPY Immunofluorescence labeling of the transfected C O S cells (using the R i m 3F4 as the primary antibody and goat anti-mouse Ig-CY3 as the secondary antibody) further confirmed that all five mutant constructs expressed. The labeling patterns, however, were somewhat different between the wild type and the mutants. The wild type proteins localized mostly in vesicles (red dots in the figure) that were scattered throughout the cells (Fig 17, A). The D846H variant and T 1 5 2 6 M showed different labeling patterns than the wild type. These variants showed intense labeling around the nucleus, a region corresponding to the E R / G o l g i compartment (Fig 17, B and F). The two mutations located in the second N B D , R2038W and R2077W did not show the wild type labeling pattern, and in addition, their E R / G o l g i labeling seemed slightly different from those of D846H and T1526M. That is, the labeling pattern of these mutants showed filamentous extensions from the E R / G o l g i compartment (Fig 17, C and D). The last mutant, R2106C, showed both the vesicular and E R / G o l g i labeling pattern; however, the extent of vesicular labeling was less than the wild type (Fig 17, E). To quantify these results, 100 cells were counted under the microscope and they were classified as showing either the vesicular or E R / G o l g i labeling pattern. D846H and T 1 5 2 6 M both showed 90-100% E R / G o l g i whereas 77% of R2038W, 80% of R2077W and only 34% of R2106C were classified as E R / G o l g i . 62 Fig 17. Immunofluorescence labeling of COS cells transfected with wild type and mutant ABCR constructs. The cells were fixed with 4% paraformaldehyde, permeabilized in 0.3% Triton-X-100 and labeled with Rim 3F4 antibody. Anti-mouse antibody conjugated to C Y 3 was used to view the labeled cells by fluorescence microscopy. Cells in panel A were transfected with wild type constructs. Cells in panel B with D846H, panel C with R2038W, panel D with R2077W, panel E with R2106C and panel F with T1526M construct. 63 64 3.6 CALNEXIN ASSOCIATION The amount of time a mutant or wild type protein/polypeptide spends in the E R of a cell may be an indication of its folding. To support the immunofluorescence results, regarding the E R / G o l g i labeling of the mutant proteins, blots containing total cell lysates and purified proteins were stripped and labeled with anti-calnexin antibody (Fig 18). The basis of this experiment was as follows. If a mutation causes a slight change in the folding of a protein, the expressed mutant protein may be retained in the E R more often than the wild type. A s a result, calnexin may also be associated with the mutant protein more often. Fig 18 shows that in the total cell lysate fractions calnexin was present in relatively equal amounts. This is to be expected since calnexin is present in the cells. In the purified protein fractions, however, calnexin did not co-purify with the A B C R purified from R O S and the level of calnexin that co-purified with the wild type A B C R was very low, indicating that wild type A B C R may not interact with calnexin. With the mutants, on the other hand, more calnexin was co-purified (Fig 18, B) even when the amounts of purified mutant A B C R s were comparable to the wild type (Fig 18, A). This suggests that the mutants are misfolded and retained in the E R / G o l g i . It should be mentioned that the amount of T1526M purified was larger than the other mutants. This may be the reason why more calnexin co-purified with this mutant. R2106C was not included in this study. 3.7 RECONSTITUTION OF ABCRS INTO LIPOSOMES Previously, a study by A h n et al. (2000) showed that P E and a reducing environment are both required for the ATPase activity of A B C R . The liposomes used here were made from either R O S extracted phospholipids or a 50:50 mixture of both brain polar lipids and D O P E or S D P E . R O S extracted lipids contained ~ 40% P E , 40% P C , 10% PS and 10% other lipids, while the brain polar extract constituted of - 33% P E , 13% P C , 18% PS and 36%) other phospholipids (Ahn et al., 2000). These lipids were used in reconstitution experiments in order to create an environment rich in P E . To produce the reducing environment needed for functional proteins, 1 m M of D T T was included at all times in all buffers used. The purified proteins were reconstituted into liposomes by the removal of 65 1 2 3 4 5 6 7 8 9 10 11 12 13 A - A B C R B wm mm m^mm \- Calnexin Fig 18. Co-purification of calnexin with mutant ABCRs. A B C R s from C O S cell lysates (lanes 1-6), R O S membranes (lane 7) and A B C R proteins purified by immuoaffinity chromatography (lanes 8-13) were resolved by S D S - P A G E , transferred onto Immobilon membrane and labeled with R i m 5B4 antibody (A). The same blot was stripped and labeled with anti-calnexin antibody (B). Ten ul from each sample was loaded on the gel. Lane 1, wild type A B C R ; lane 2, D846H; lane 3, R2038W; lane 4, R2077W; lane 5, R2106C; lane 6, T1526M; lane 7, R O S membrane; lane 8, A B C R purified from R O S membrane; lane 9, purified wild type; lane 10, purified D846H; lane 11, purified R2038W; lane 12, purified R2077W; lane 13, purified T1526M. 66 the detergent ( C H A P S ) in the presence of excess lipids. The presence of proteins (wild type and mutants) in the vesicles was determined by western blotting using R i m 5B4 antibody (Fig 19). This figure shows that A B C R s were present in the reconstituted samples in relatively equal amounts. The amount of reconstituted proteins in each sample was quantified by densitometry from the standard curve generated from B S A . Approximately 30-50 ng of A B C R were calculated to be present in the reconstituted samples. These values varied between experiments as the transfection efficiencies varied. 3.8 T H E ATPASE ASSAY The ATPase assay was carried out as outlined in section 2.12. In this study, ATPase activity and specific activity were defined as the nmoles of A T P hydrolyzed per min per reconstituted sample and the nmoles of A T P hydrolyzed per min per mg of protein, respectively. To optimize the assay conditions, the ATPase assay was first performed for reconstituted samples containing purified A B C R from R O S . It was found that upon incubation with 50 u M all-zrafts-retinal, the basal activity of reconstituted R O S A B C R , in a 50:50 mixture of brain polar l i p i d : D O P E , was stimulated 1.8 ± 0.3 fold as shown in Fig 20. The specific activities were determined to be 47 nmole of A T P hydrolyzed/min/mg (basal) and 80 nmole of A T P hydrolyzed/min/mg (stimulated). When R O S A B C R was reconstituted in a 50:50 mixture of brain polar lipids and S D P E , the basal activity was increased ~ 4 fold (Fig 21, A). However, when the absolute values for A T P hydrolysis of R O S A B C R reconstituted in brain polar /SDPE mixture were compared to those of A B C R reconstituted in brain po lar /DOPE lipids, they were significantly smaller (Fig 21, B). As a result, brain polar l ip ids /DOPE mixture was used in reconstitution experiments. The assay was also performed with increasing concentrations of all'/r^ms-retinal in the presence of 50 u M A T P (Fig 22). A s shown in the figure, the activity was increased with increasing retinal concentration. 67 1 2 3 4 5 6 7 Fig 19. ABCR proteins reconstituted in liposomes. The purified A B C R proteins (24 ul) were mixed with 12 ul of 20mg/ml of solubilized lipids in 15% octyl-glucoside, and allowed to incubate for 30 min on ice. Two ml of the reconstitution buffer was added to the proteolipid mixture, and after 2 min of incubation, the resulting mixture was passed through 200 ul o f Extracti-gel D resin in a mini-column. Thirty ul of the eluates from wild type A B C R (lane 1), D846H (lane 2), R2038W (lane 3), R2077W (lane 4), R2106C (lane 5), T1526M (lane 6) and R O S (lane 7) were loaded on a 7.5% polyacrylamide gel. Following electrophoresis, the proteins were transferred onto an Immobilon membrane for western blotting and detected with R i m 5B4 antibody. 68 Fig 20. The effect of all-fra/is-retinal on the ATP hydrolysis of reconstituted ROS ABCR. A B C R was solubilized and purified from bovine R O S membranes by immunoaffinity chromatography. The purified protein was reconstitued in 20 mg/ml of a 50:50 mixture of brain polar lipid extract and D O P E . The ATPase assay was performed in the absence (basal) and presence (retinal) of 50 m M all-frows-retinal. The results represent the mean ± S.D. of five experiments. 69 Fig 21. Basal and retinal stimulated ATP hydrolysis of ROS ABCR reconstituted in two different lipid mixtures. Purified A B C R from R O S membrane was reconstituted in a 50:50 mixture of brain polar lipids and D O P E or a 50:50 mixture of brain polar lipids and S D P E . (A) Retinal stimulation expressed as % of basal A T P hydrolysis. (B) Absolute values of basal and retinal stimulation expressed as nmoles of A T P hydrolyzed/min/reconstituted sample. Each value is the mean of two determinations. 70 0.06 -a 0 20 40 60 80 100 120 retinal uM Fig 22. The effect of all-fra/ts-retinal concentration on the ATPase activity of reconstituted ROS ABCR. Purified bovine A B C R from R O S membranes was reconstituted in a 50:50 mixture of brain polar l ip ids /DOPE and assayed for the ATPase activity in the presence of 10, 20, 50 and 100 u M of all-trans-retinal. Each value is the mean of two determinations. 71 A B C R s reconstituted in lipid vesicles retained their ATPase activities over a three day period on ice. Purified A B C R s in column buffer (10 m M C H A P S ) on the other hand, were degraded when kept on ice overnight. This was apparent by the absence of A B C R bands on western blots. Furthermore, the presence of 5 m M M g C l 2 was required in the reconstituted samples in order to measure A T P hydrolysis. The ATPase activity of recombinant wild type A B C R purified from transfected C O S cells was also determined. In the presence of 50 u M all-trans-retinal, the basal activity of wild type A B C R reconstituted in a mixture of brain polar l ip ids /DOPE, increased by 1.6 ± 0.4 fold. It should be mentioned, however, that the absolute values for ATPase activity differed from one experiment to the next, and were substantially smaller than the values obtained from R O S A B C R (Fig 23). This was because approximately ten times less recombinant wild type protein could be purified and reconstituted from C O S cells compared to R O S A B C R . The basal and stimulated specific activities of wild type A B C R were measured to be 13 nmoles. of A T P hydrolyzed/min/mg and 22 nmoles of A T P hydrolyzed/min/mg, respectively. Most of the ATPase activity measurements for the wild type and the mutants, however, were expressed as nmoles of A T P hydrolyzed/min/reconstituted sample. This was because the amounts of purified A B C R s were small and scanning the Coomassie blue stained gel by a densitometer did not result in accurate measurements of protein concentrations. Thus the specific activity values were not reliable. Although the basal and retinal stimulated ATPase activities varied between experiments, no significant change or increase was seen when R O S extracted lipids were used instead of brain po lar /DOPE lipids to reconstitute the wild type A B C R . Different mutant A B C R s reconstituted in either R O S extracted lipids or a 50:50 mixture of brain polar lipid and D O P E were assayed along side wild type A B C R , untransfected and buffer blank samples. Measuring the ATPase activities of the mutants proved to be very difficult, as the basal activity values for the mutants were smaller than the basal activity of the wild type. Furthermore, as mentioned earlier with the wild type samples, the ATPase values of the mutants differed between preparations and depended on the amounts of proteins that purified and reconstituted into the vesicles. Since the amount of purified protein differed between experiments and, in turn, depended on the efficiency of 72 Fig 23. Basal and retinal stimulated ATPase activities of ROS ABCR and wild type ABCR from transfected COS cells. Purified A B C R from R O S membrane (ROS) and purified wild type A B C R obtained from solubilized C O S cells (WT) were both reconstituted in 50:50 mixture of brain polar lipids and D O P E . Both the basal (open columns) and retinal stimulated (solid columns) ATPase activities were measured and expressed as absolute values for comparison. Each value is the mean of two determinations. 73 the transfection, an experiment during which relatively high amounts of proteins were purified was chosen as a representative experiment. A s shown in Fig 24, the basal ATPase activities of R2038W and T1526M resembled the recombinant wild type A B C R . Upon the addition of 50 u M retinal, however, they did not display significant retinal stimulation activities, whereas the activity of wild type A B C R increased 1.6 fold. A s illustrated in this figure, the D846H and R2077W variants displayed both basal and retinal stimulated activities lower than basal level of the wild type. Although the basal activity of the R2106C variant was lower than that of the wild type, it was stimulated in the presence of 50 u M retinal. When R O S extracted phospholipids were used, the basal and retinal stimulated ATPase activities of the wild type and the mutants remain similar to when brain po lar /DPOE lipids were used. 3.9 AZIDO-ATP LABELING To complement the A T P hydrolysis measurements, photoaffinity labeling experiments 3 2 using a- P az ido-ATP were carried out (section 2.13). The az ido-ATP labeling assays were performed on C O S cell membrane fractions. The labeled A B C R s were purified by immunoaffinity chromatography and analyzed on SDS gels. One gel was exposed to a film in order to measure the A T P binding abilities of R O S A B C R , the recombinant wild type A B C R and four of the five mutants (Fig 25, A ) . The other gel was destined for western blotting to determine the amounts of purified proteins (Fig 25, B). Fig 25, A shows that A B C R from R O S was strongly labeled, as expected. The recombinant wild type, R2106C and T1526M proteins were also labeled, indicating that they bind A T P . The D846H and R2077W variant, however, showed no az ido-ATP labeling. The R2038W variant was not investigated. A s shown in Fig 25, B the amount of purified protein in each lane were comparable. Therefore, the lack of az ido-ATP labeling by D846H and R2077W was not due to the low levels of protein present, but to their inability to bind A T P . 74 a 0.006 WT D846H R2038W R2077W R2106C T1526M Fig 24. ATP hydrolysis of reconstituted wild type and mutant ABCRs purified from transfected COS cells. A B C R proteins were solubilized and purified from transfected C O S cells and reconstituted in lipid vesicles. A T P hydrolysis was measured in the absence (open columns) and presence of 50 u M all-zrafM-retinal (solid columns). Each data represents the mean ± of S.D. of triplicate values from a single representative experiments. The name of the samples are shown in the figure, WT: wild type. 75 1 2 3 4 5 6 7 A B C R 1 2 3 4 5 6 7 B A B C R Fig 25. Azido-ATP labeling and yield of wild type and mutant ABCRs. R O S membrane and C O S cell membranes expressing the recombinant wild type and mutant 32 proteins were labeled with 8-azido-[a P ] - A T P . The labeled proteins were purified by immunoaffinity chromatography and the eluates were loaded on two 6.5% SDS polyacrylamide gels and subjected to electrophoresis. (A) The labeled A B C R s were detected by autoradiography and (B) the relative amounts of proteins were detected by western blotting using R i m 5B4 antibody. Lane 1, A B C R from R O S ; lane 2, recombinant wild type A B C R ; lane 3, D846H; lane 4, R2077W; lane 5, R2106C; lane 6, T1526M and lane 7, R O S membrane. 76 DISCUSSION The next step following the identification of A B C R as a member of the A B C transporter superfamily and its involvement in macular dystrophy, was to find a suitable expression system in order to produce high quantities of this protein. Previously, attempts were made to express both full length and N-truncated A B C R in Pichia pastoris (Safarpour, 1999). The study however, concluded that this protein could not be constitutively expressed in Pichia pastoris. A s a follow up to that study, a mammalian system using monkey kidney C O S cells were used in this project. 4.1 EXPRESSION OF WILD TYPE AND MUTANT ABCRS IN COS C E L L S First, the wild type A B C R , cloned in the desired plasmid p C D N A 3 , was used to transfect COS-1 or 7 cells using the SuperFect reagent or calcium phosphate method of transfection. Since the expressed proteins were destined for functional studies, a non-denaturing detergent with a high critical micelle concentration ( C M C ) was required during the solubilization step. A high C M C would allow for the rapid removal of the detergent in the subsequent steps. Previously, Illing et al. (1997) had used 18 m M C H A P S ( C M C = 8 m M ) to solubilize R O S membrane and purify A B C R . Similarly, 18 m M C H A P S was used in this study to solubilize the expressed proteins from C O S cell membranes. It should be mentioned that the A B C R proteins were both solubilized and purified in the presence of both soybean phospholipids (-40% PC) and D T T . A n excess of phospholipids and a reducing condition were previously shown to preserve the function of P-glycoprotein (Callaghan et al. 1997). Glycerol (10%) was also included to stabilize the proteins. When the cells were solubilized and the insoluble fraction removed, the total cell lysate was analyzed for the expression of wild type A B C R . Since the wild type A B C R was successfully expressed in C O S cells, the same system and transfection methods were used to express the five mutant constructs: D846H, T 1 5 2 6 M , R2038W, R2077W and R2106C. Following transfection, analysis of the soluble fractions 77 obtained from detergent solubilized COS cells revealed that all five mutants could be expressed. These proteins had the same molecular weight as the A B C R from ROS, indicating that they were glycosylated to the same extent as the native A B C R . The expression levels of these recombinant proteins, however, differed from the wild type, and also from one another. Furthermore, the transfection methods used had some effects on the expression level of D846H. The D846H variant was expressed twice as much when C a P 0 4 was used instead of the SuperFect. The expression levels of the other four proteins varied between 50-80% of the wild type (Table 3). Approximately 30-40% of adherent COS cells were transiently transfected by either methods. 4.2 PURIFICATION AND RECONSTITUTION OF ABCR VARIANTS Once the wild type and mutant proteins were shown to be expressed in COS cells, there was a need for a purification scheme that could preserve the native conformation of the proteins. Rim 3F4-Sepharose 2B beads were previously shown to be successful in purifying A B C R from ROS (Illing et al, 1997). Since none of the missense mutations were located on or near the Rim 3F4 binding site of the A B C R variants, Rim 3F4 antibody coupled to Sepharose beads was also employed in this study. The column buffer contained all of the components present in the solubilization buffer, with the exception of a lower concentration of CHAPS (10 mM). The harvesting of cells, purification and reconstitution steps were carried out on the same day to minimize degradation of A B C R . Elution was carried out using the 3F4 peptide in order to preserve the native conformation of the proteins. Approximately 40-50 % of the wild type and mutant proteins were recovered by this method. More proteins could have been purified i f longer incubations were allowed. To reconstitute the A B C R proteins into vesicles, several steps were required. First the lipids, either ROS extracted phospholipids or a mixture of brain polar/DOPE, were solubilized in a mild detergent with high C M C value such as octyl-P-D-glucopyranoside. Next, the purified A B C R proteins in a buffer containing C H A P S were mixed with the solubilized lipids. The detergents were then removed, leaving the proteins reconstituted in lipid vesicles. Detergents may be removed by dialysis, dilution or gel filtration 78 chromatography (Doige et al, 1992; Urbatsch et al, 1994; Shapiro and Ling , 1994). The method of detergent removal used in this study was essentially that of Sun et al. (1999), where Extracti-Gel resin was used. Detergents bind to this resin allowing the proteins reconstituted in the lipid vesicles to be collected. This method was favorable as it was both time efficient and effective in removing the detergents. Approximately 30-50 ng of proteins could be reconstituted by this method. 4.3 T H E ATPASE ASSAY In order to investigate the effects of the selected mutations on the function of A B C R , a functional assay was needed. Previously, substrate-dependent A T P hydrolysis studies were performed on purified and reconstituted P-glycoprotein (Shapiro and Ling , 1994; Urbatsch et al, 1994). Recently, two studies by Sun et al. (1999) and A h n et al. (2000) took a similar approach for A B C R . They identified a number of possible substrates, including all-/r<ms-retinal and Af-retinylidene-phosphatidylethanolamine (JV-retinylidene-PE) , that stimulated the ATPase activity of reconstituted A B C R purified from R O S . In the present study the efficiency of the ATPase assay was first assessed, as a number of experiments were performed with A B C R purified from R O S and reconstituted in lipid vesicles. Since studies by Sun et al. (1999) and A h n et al. (2000) had revealed a K m value of 25-33 u M for A T P , 50 u M of A T P was used here to insure that there were enough A T P in the assay. Similarly, since half maximal stimulation of reconstituted A B C R , was calculated to occurred at -10-15 u M retinal, 50 u M retinal was used in the assays, unless specified otherwise. The same studies had also shown the importance of P E on the ATPase activity of R O S A B C R , therefore, experiments in this study were carried out mostly in a mixture of brain polar lipids (33% P E , 18% PS, 13% P C , 36% other phospholipids) and D O P E to ensure an environment rich in P E . When R O S A B C R was reconstituted in a 50:50 mixture of brain polar lipids and D O P E , its basal activity was increased by an average of 1.8 ± 0.3 fold. A h n et al. (2000) had seen a 2-3 fold increase when the same lipids were used to reconstitute the protein. The specific activities, defined as the nmoles of A T P hydrolyzed per min per mg of protein, were determined to be 47 (basal) and 80 (retinal) in the presence of 50 u M of A T P . Increasing 79 the retinal concentrations while keeping A T P constant, resulted in an increase in the A T P hydrolysis of R O S A B C R . This illustrated that a\\-trans-retma\ may be a possible substrate or modulator of A B C R . When S D P E was substituted for D O P E in the 50:50 mixture with brain lipid, the basal A T P hydrolysis was stimulated ~ 4 fold, a result also reported by A h n et al. (2000). A s shown in Fig 21, B however, the absolute values of basal and stimulated activities in brain polar /SDPE vesicles were smaller than the basal value for A B C R in brain p o l a r / D O P E vesicles. This may complicate the interpretation of the data. For instance, one may consider the effect of S D P E to be inhibitory since the basal activity was lower than when D O P E was used. Increasing the retinal concentration then would stimulate the activity and remove the inhibitory effects of S D P E , by bringing the activity to the basal level. When the recombinant wild type protein was reconstituted in 50:50 mixture of brain polar l ip ids /DOPE, the presence of 50 u M alWrarcs-refinal caused an average increase of 1.6 ± 0.4 fold from its basal activity. In some experiments a 2.5 fold increase in stimulation was seen. Although the basal A T P hydrolysis of wild type A B C R was found to increase in the presence of all-rnms-retinal, the absolute values for ATPase activities varied between experiments and were significantly smaller than those calculated for R O S A B C R . A h n et al. (2000) had observed an increase in the retinal stimulation when phospholipids extracted from R O S were used to reconstitute the R O S purified A B C R s . Creating an environment similar to the disk membrane, in terms of the major phospholipid constituents, seemed very prominent. N o significant increase in A T P hydrolysis (basal and retinal stimulated), however, was observed when R O S lipids were used in this study. One reason may be the presence of possible contaminants during the extraction of phospholipids from R O S membranes. When mutant A B C R s obtained from solubilized C O S cells were used in the ATPase assay, the data was even more difficult to interpret. In some experiments, very little protein 80 could be purified and reconstituted, as the transfection efficiency was low. In these experiments distinguishing between basal activity due to a functional A B C R and non-enzymatic A T P hydrolysis was very difficult. Similarly, due to the high background of A T P hydrolysis, it was hard to attribute slight stimulation or slight inhibition to the presence of alWra/w-retinal. Another concern with the assay was the variation in the ATPase activities between experiments. To make the ATPase data more reliable, one solution was to obtain and purify more proteins. This was achieved by transfecting three C O S cell dishes instead of one per D N A sample. In order to prevent variations between basal activities from one experiment to the next, each experiment which included transfection of C O S cells, solubilization of the cells, purification and reconstitution of the proteins, was performed for the wild type and the five mutants at the same time. In this case, the ATPase data obtained for the mutants was interpreted relative to the wild type. Fig 24 was a representative experiment in which 30-50 ng of proteins were reconstituted and each assay was performed in triplicates. A l l of the mutants showed a basal ATPase activity lower or essentially equal to the wild type. ATPase activities of both D846H and R2077W were diminished regardless of retinal. R2106C, however, was stimulated by retinal, while R2038W and T 1 5 2 6 M remained unaffected by it. A more detailed interpretation of the results from Fig 24 is presented in section 4.4. Specific activities were not calculated for the mutants, since the amount of purified proteins in the reconstituted samples was very small, and would result in large and inaccurate specific activity values. Instead, results were reported as nmoles of A T P hydrolyzed per min per reconstituted sample. 4.4 ANALYSIS OF T H E EXPRESSED WILD TYPE AND MUTANT ABCRS 4.4.1 Wild type ABCR When the transfected C O S cells were analyzed by immunofluorescence microscopy, the wild type A B C R was mainly localized in the vesicles, although some endoplasmic reticulum (ER)/Golgi labeling was also present. W i l d type A B C R did not seem to localize in the plasma membrane, as there was no labeling along the perimeter of the cells and the 81 labeled vesicles were scattered throughout the C O S cells. A study by Sun et al. (2000) however, suggested that wild type A B C R was localized within the E R of the transfected 293 cells. The discrepancy regarding wild type A B C R ' s location between Sun et al. and the present study may be due to the different cell types used. The expressed wild type A B C R was also examined for its association with calnexin, one of the E R chaperones involved in primary quality control in the secretory pathway. The primary quality control retains proteins that contain features characteristic of incompletely folded proteins. Calnexin and calreticulin (soluble homologue of calnexin) are lectins that recognize monoglucosylated misfolded proteins, and in conjunction with other proteins and enzymes mediate the retention and the proper folding of misfolded glycoproteins in the E R . The proposed model for quality control of glycoproteins in mammalian cells (reviewed by Ellgaard et al., 1999) is as follows. A s a nascent polypeptide is synthesized in the E R , N-linked oligosaccharides in the form of a 14-saccharide core unit (GIC3 Mang GlcNAc2; Glc: glucose, Man: mannose, G l c N A c : JV-acetylglucosamine) is covalently attached to an asparagine residue by an oligosaccharyltransferase. In the absence of the added sugars the proteins misfold, aggregate and become degraded without being transported to the Golgi complex. Two of the three glucose residues on the core unit are rapidly trimmed by the actions of glucosidase I and II to produce a monoglucosylated protein. Calnexin and calreticulin recognize the monogucosylated protein and associate with it. The removal of the last glucose molecule by glucosidase II results in the release of the glycoprotein from calnexin/calreticulin and allows the correctly folded glycoprotein to continue its journey to the Golgi complex. However, i f the protein remains incompletely folded an enzyme called UDP-Glucose:glycoprotein glucosyltransferase (GT) reglucosylates the protein, causing it to bind to calnexin/calreticulin again. This glucosylation and deglucosylation cycle continues until glycoproteins are completely folded, and as a result misfolded proteins are retained in the E R . Looking for an interaction between the expressed A B C R and calnexin, therefore, may support the localization patterns revealed by immunoflurorescence microscopy and may further suggest the effect of a given missense mutation on the conformation of the A B C R . 82 Furthermore, both C F T R and P-glycoprotein have been shown to bind calnexin (Pind et al, 1994; Loo and Clarke, 1994). This project showed that although calnexin was present in the cell lysate fractions, it did not co-purify at high levels with the wild type A B C R expressed in COS cells. This indicated that wild type A B C R may have not been delayed in the E R and thus may be folded correctly. This result is consistent with the retinal stimulated ATPase activity and with the observation that wild type A B C R , expressed in COS cells, was mostly present in the vesicles and not ER/Golgi compartment. 4.4.2 D846H Immunofluorescence labeling of the transiently transfected COS cells revealed that the expression patterns of the mutants were somewhat different from the wild type. The A B C R mutants were classified as either vesicular (same expression pattern was wild type) or ER/Golgi . The D846H mutant was almost entirely expressed in the ER/Golgi compartment, suggesting that this variant may be folded differently. This amino acid exchange (D-»H) in the sixth transmembrane segment may be hypothesized to impede the correct folding of this A B C R variant, causing it to be retained in the E R compartment more often that the wild type. This hypothesis was tested by examining the interaction between D846H A B C R and calnexin. As revealed by the western blot (Fig 18, in the results section), calnexin co-purified with D846H variant in a slightly larger amount than the wild type, suggesting that D846H is retained in the ER by its interaction with calnexin and thus it is misfolded. Furthermore, the azido-ATP labeling experiment showed that this protein did not bind A T P and the ATPase measurement revealed that this variant could not hydrolyze ATP in the presence or absence of retinal. Since the nucleotide binding domains (NBD) are intact, the reason why this variant failed to bind A T P is likely due to its incorrect folding. The aspartate residue may play an important role in the folding of functional A B C R in COS cells. 83 4.4.3 T1526M The T1526M A B C R mutant was expressed in relatively high amounts (81 % of wild type) in C O S cells using both calcium phosphate and SuperFect reagents. The immunofluorescence labeling pattern of this variant showed an E R / G o l g i labeling, suggesting that it is retained in the E R compartment. Furthermore calnexin associates with this mutant to a greater extent than the wild type A B C R . Interestingly, this mutation that occurs three amino acid residues amino-terminal to a N-glycosylation site (N1529), is also conserved in A B C A 1 . A z i d o - A T P labeling of this protein prepared from C O S cell membranes suggested that its N B D was capable of binding A T P . The ATPase measurement complemented the az ido-ATP labeling experiment, as a basal activity similar to the wild type was measured. A l l - / r a « s - r e t i n a l , however, did not stimulate the activity of this mutant protein. The pathogenic effect of this mutant may be speculated to be linked to its inability to interact with or transport retinal. In analyzing the above results, a discrepancy arose between the immunofluorescence and calnexin interaction studies, and those of A T P binding and hydrolysis studies. That is, this mutant, on one hand seems to be retained in the E R , suggesting a misfolded protein, and on the other hand seems capable of binding and hydrolyzing A T P , an indication of a functional protein. A similar situation has been observed for C F T R . The most common genetic mutation in C F T R is a deletion of phenylalanine 508 (AF508), this mutation is recognized as abnormal by the cellular quality control machinery and is retained within the E R . Several studies, however, have suggested that although this variant of C F T R is mislocalized, it maintains its chloride channel activity (Li et al, 1993; Pind et al., 1994; Pasyk and Foskett, 1995). Thus, the observation that T1526M or any other A B C R variants that are localized in the E R and bind calnexin, does not necessarily imply that these proteins are non-functional. The E R retention of these proteins may be due to a slight change in the conformation of the native protein that is recognized by the primary quality control mechanism. Additional functional assays need to be performed in order to determine whether a protein retained in the E R is functional or not. At the present time, immunofluorescence and calnexin association studies alone may not allow the distinction between misfolded, non-functional proteins and the differently folded, mislocalized yet functional proteins. Thus care must be taken when referring to these studies. 84 Previously, it was mentioned that two different models for A B C R had been proposed (Ming et al, 1997; Bungert et al, 2001; Azarian and Travis, 1997; Sun et al, 2000). In one model (Azarian and Travis, 1997; Sun et al, 2000) the presence of a highly hydrophobic region looping in and out of the disk bilayer resulted in a small second intradiskcal loop and a relatively large loop on the cytoplasmic side of the disk. In the second model (Ming et al, 1997; Bungert et al, 2001), the highly hydrophobic region was considered as a transmembrane segment and as a result, a longer second intradiskal loop was located in the disk lumen. The T1526 residue is located outside of the disk in the first model, and on the intradiskal loop inside the disk in the second model. If the assumption that the T1526M variant loses its retinal sensitivity, holds true and assuming that A B C R transports retinal out of the disk membrane, then one may favor positioning T1526 and the second intradiskal loop in the lumen of the disk, allowing it to interact with retinal either released by activated rhodopsin or bound to PE on the inner leaflet of the disk membrane. 4.4.4 R2038W The R2038W variant with a missense mutation in the second nucleotide binding domain was expressed in COS cells at 59% of the wild type. Immunofluorescence labeling identified 77% of this variant as having the ER/Golgi labeling pattern. However, this labeling pattern was recognized as being slightly different from those of D846H and T1526M (above) in that filamentous extension from ER/Golgi were detected. Further experiments indicated that calnexin interacted with this variant as it co-purified with the mutant protein. It may be speculated, therefore, that this variant initially is retained in the ER by calnexin, but after a number of reglucosyation and deglucosylation, it is folded correctly and translocated to the Golgi complex and vesicles. Unfortunately, azido-ATP labeling experiments were not performed for this mutant. The ATPase studies revealed that the basal A T P hydrolysis of this mutant was similar to that of the wild type, suggesting that the R2038 is not likely to be involved in the A T P hydrolysis reaction. However, no retinal stimulated ATPase activity was observed, suggesting that processes such as substrate recognition or retinal transport may be affected by this missense mutation. 85 A closer examination of the second N B D of the A B C transporters revealed that R2038 is located in a region called helical domain which is located between Walker A and B motifs (Schneider and Hunke, 1998). Mutational analysis of this region using histidine permease (HisP) from S. typhimurium, an A B C transporter, had indicated that most amino acid substitutions in this region, did not alter the nucleotide binding activity. O n the other hand, the transport ability of these mutants were affected (Shymala et al., 1991). This is consistent with the findings that R2038W variant showed basal, but not retinal stimulated A T P hydrolysis activity. 4.4.5 R2077W The R2077W missense mutation is also located in the second N B D , more specifically it is part of the consensus sequence called motif C : Y S G G x K R K (Broccardo et al, 1999). The exchange of the arginine residue in the above sequence with a tryptophan did not significantly affect the expression level of this variant, as it was expressed 57% of wild type. Immunofluorescence labeling revealed a pattern similar to R2038W (a missense mutation also in the second N B D , above). That is, 80% of the protein was expressed in the E R / G o l g i and filamentous extensions from the E R / G o l g i compartment were seen. R2077W was also associated with calnexin. A z i d o - A T P labeling of R2077W in C O S cell membrane, however, showed that A T P does not bind to this variant. The ATPase activity of this mutant was also impaired, regardless of the presence or absence of retinal. In this case distinguishing between a misfolded, non-functional protein and one that is unable to catalyze A T P was difficult. Functional assays other than A T P binding and hydrolysis are needed to assess whether this variant was unable to hydrolyze A T P due to its conformation or due to the amino acid substitution in its N B D . According to Manavalan et al. (1995), the Walker A motif and the motif C of C F T R and other G-proteins such as the a sub unit of transducin, are positioned adjacent to each other 86 in the tertiary structure of these proteins. In their study, Manavalan et al. identified specific residues in both motifs A and C responsible for binding the phosphate moiety of A T P . Although the equivalent residue of R2077 was not shown to directly interact with the phosphate group of A T P , three and five amino acids away, corresponding to residues G2072 and S2074 in A B C R , were involved in the binding of y phosphate and in the case of the serine residue, in M g 2 + binding. The proximity of these residues to R2077 and the conservation of the motif C among other members of A B C A proteins may suggest a relatively important function for this residue in A T P binding. Furthermore, most mutations in this region seemed to abolish the A T P hydrolysis, as observed in this study with the R2077W variant. 4.4.6 R2106C This missense mutation occurred in the second N B D , between Walker B and a region called the switch region (Schneider and Hunke, 1998). Although the arginine residue at 2106 was conserved in A B C A 1 , it is located outside of any defined consensus sequences. It may, therefore, be hypothesized that the R2106C mutation may not have a very profound effect on the function of A B C R . The R2106C was expressed at a relatively high level (71% of wild type) in C O S cells. Immunofluorescence microscopy revealed mostly vesicular labeling, however 34% of the labeling was E R / G o l g i . This suggested that this mutant resembled more the wild type expression pattern and had a more native conformation than the other mutants. Unfortunately, calnexin interaction was not assessed for this mutant. The az ido-ATP labeling of this mutant in C O S cell membrane fractions revealed that this variant could bind A T P . Although the basal ATPase activity of this variant was lower than wild type, all-frans-retinal stimulated its activity. Thus in terms of A T P binding and A T P hydrolysis, this mutant resembled the wild type. The pathogenic effect of this mutant may be due to a property not investigated in this study. 87 4.5 SUMMARY At the start of this project, 89 mutations in the ABCR gene that resulted in Stargardt's disease were identified. To study the effects of five of these mutations on the structure and function of A B C R , the mammalian C O S cell expression system was used. After generating the desired constructs, they were heterologously expressed in the C O S cells. Approximately 300-900 ng of the recombinant proteins were purified from detergent solubilized C O S cells by immunoaffinity chromatography and used in reconstitution experiments. Immunofluorescence microscopy of transfected C O S cells revealed that the wild type A B C R was localized mostly in vesicles whereas the mutant A B C R s displayed a range of E R / G o l g i labeling pattern. Most of the D846H and T1526M proteins were E R / G o l g i localized compared to 77%, 80% and 34% of R2038W, R2077W and R2106C, respectively. To complement the immunofluorescence studies, the wild type and four of the mutants were examined for their abilities to interact with calnexin, a resident of E R that binds misfolded or incompletely folded proteins and causes their retention in the E R . Whereas wild type A B C R did not interact with calnexin, all four mutants were shown to co-purify calnexin. This association with calnexin hinted that these mutants might be folded differently than the wild type. To assess the ATPase activity of the wild type and mutant A B C R s , the purified proteins were reconstituted in liposomes rich in P E (50:50 mixture of brain polar lipids and D O P E ) . In the presence of 50 u M A T P , both the basal and the retinal stimulated activities of these samples were measured. Although the basal activity of wild type A B C R was increased 1.6 fold in the presence of 50 u M all-/r<ms-retinal, no stimulation was seen in T 1 5 2 6 M and R2038W variants. The R2106C variant displayed a retinal-stimulation of four fold, while the ATPase activities of both D846H and R2077W were impaired. 88 A z i d o - A T P labeling was performed on transfected C O S membrane fractions to complement the ATPase activity studies. Whereas the wild type, T 1 5 2 6 M and R2106C bound A T P to the same extent, the D846H and R2077W did not. By compiling all of the results obtained, it was suggested that D846H caused a misfolded protein, unable to bind and hydrolyze A T P . T 1 5 2 6 M bound A T P , but its basal activity was not stimulated in the presence of all-trans-ret'mal, thus its pathogenic effect may be due to its inability to interact with its substrate. The A T P hydrolysis ability of R2038W revealed that this variant was not stimulated in the presence of a l l - z r a « s - r e t i n a l . Therefore, it is unlikely that this mutant is capable of transporting its substrate. The R2077W variant was pathogenic because it could not bind and hydrolyze A T P . The mutant R2106C seemed to be folded correctly and both the ATPase activity and the azido-A T P labeling studies suggested that this mutant could bind and hydrolyze A T P . Furthermore, as with the wild type, it showed a retinal stimulated ATPase activity. The R2106C mutation may affect another function of A B C R . The functional studies employed in this project and elsewhere (Sun et al., 2000) provided an in vitro system in which the effects of mutations in ABCR gene on the function and conformation of A B C R may be studied. Understanding the biochemical defects for each A B C R variant is the first step toward designing therapeutic drugs against A B C R - l i n k e d retinal dystrophies. 89 REFERENCES A h n J , Wong JT, Molday RS . 2000. The effect of lipid environment and retinoids on the ATPase activity of A B C R , the photoreceptor A B C transporter responsible for Stragardt macular dystrophy. J.Biol. Chem. 275:20399-20405. Allikmets R, Singh N , Sun H , Shroyer N F , Hutchinson A , Chidambaram A , Gerrard B , Baird L , Stauffer D , Peiffer A , Rattner A , Smallwood P; L i Y , Anderson K L , Lewis R A , Nathans J, Leppert M , Dean M , Lupski JR. 1997a. A photoreceptor cell-specific A T P -binding transporter gene ( A B C R ) is mutated in recessive Stargardt macular dystrophy. Nature Genet. 15:236-246. Allikmets R, Shroyer N F , Singh N , Seddon J M , Lewis R A , Bernstein PS, Peiffer A , Zabriskie N A , L i Y , Hutchinson A , Dean M , Lupski JR, Leppert M . 1997b. Mutation of the Stargardt disease gene ( A B C R ) in age-related macular degeneration. Science. 277:1805-1807. Amer S, Akhtar M . 1973. Studies on regeneration of rhodopsin from all-trans retinal in isolated rat retinae. Nature. 245:221-223. Azarian S M , Travis G H . 1997. The photoreceptor rim protein is an A B C transporter encoded by the gene for recessive Stargardt's disease ( A B C R ) . FEBS Lett. 409:247-252. Barry R J , Canada F J , Rando R R . 1989. Solubilization and partial purification of retinyl ester synthetase and retinoid isomerase from bovine ocular pigment epithelium. J. Biol. Chem. 264:9231-9238. Bascom R A , Manara S, Collins L , Molday RS, Kalnins V I , Mclnnes R R . 1992. Cloning of the c D N A for a novel photoreceptor membrane protein (rom-1) identifies a disk rim protein family implicated in human retinopathies. Neuron. 8:1171-1184. Baylor D A and Nunn B . 1982. Electrical signaling in vertebrate photoreceptors. Methods Enzymol. 81:403-423. Bear C E , L i C H , Kartner N , Bridges R J , Jensen T J , Ramjeesingh M , Riordan JR. 1992. Purification and functional reconstitution of the cystic fibrosis transmembrane conductance regulator ( C F T R ) . Cell. 68:809-818. Bernstein PS, Law W C , Rando R. 1987. Isomerization of all-trans retinoids to ll-cis retinoids in vitro. Proc. Natl Acad. Sci. USA. 84:1849-1853. Bok D . 1985. Retinal photoreceptor-pigment epithelium interactions. Invest. Ophthalmol. Visual Sci. 26:1659-1694. Bok D . 1993. The retinal pigment epithelium: a versatile partner in vision. J. cell Sci. 17:189-195. 9 0 Bok D , Ong D E , Chytil F . 1984. Immunocytochemical localization of cellular retinol binding protein in the rat retina. Invest. Ophthalmol. Visual Sci. 25:1-7. Bosch E , Horwitz J , Bok D . 1993. Phagocytosis of outer segments by retinal pigment epithelium: phagosome-lysosome interaction. J. Histochem. Cytochem. 41:253-263. Boesze-Battaglia K and Albert A D . 1992. Phospholipid distribution among bovine rod outer segment plasma membrane and disk membranes. Exp. Eye Res. 54:821-823. Boesze-Battaglia K , Fliesler SJ, Albert A D . 1990. Relationship of cholesterol content to spatial distribution and age of disc membranes in retinal rod outer segments. J. Biol. Chem. 265:18867-18870. Bridges C D . 1976. Vitamin A and the role of the pigment epithelium during bleaching and regeneration of rhodopsin in the frog eye. Exp. Eye Res. 22:435-455. Broccardo C , Luciani M - F , Chimini G . 1999. The A B C A subclass of mammalian transporters. Biochimica et Biophysica Acta. 1461:395-404. Bungert S, Molday L L , Molday RS. 2001. Membrane topology of the A T P binding cassette transporter A B C R and its relationship to A B C 1 and related A B C A transporters. J Biol Chem. 276:23539-23546. Callaghan R, Berridge G , Ferry D R , Higgins C F . 1997. The functional purification of P-glycoprotein is dependent on maintenance of a lipid-protein interface. Biochimica Biophysica Acta. 1328:109-124. Cervetto L , Lagnado L , Perry R J , Robinson D W , McNaughton P A . 1989. Extrusion of calcium from rod outer segments is driven by both sodium and potassium gradients. Nature. 337:740-743. Chen C and Okayama H . 1987. High-efficiency transformation of mammalian cells by plasmid D N A . Mol. Cell Biol. 7:2745-2752. Chen C K , Inglese J , Lefkowitz R J , Hurley JB . 1995. Ca-dependent interaction of recoverin with rhodopsin kinase. J. Biol. Chem. 270:18060-18066. Chen T Y , Illing M , Molday L L , Hsu Y T , Yau K W , Molday RS . 1994. Subunit 2 (or beta) of retinal rod cGMP-gated cation channel is a component of the 240-kDa channel-associated protein and mediates Ca(2+)-calmodulin modulation. Proc. Natl Acad. Sci. USA. 91:11757-11761. Cohen A L 1968. New evidence supporting the linkage to extracellular space of outer segment saccules of frog cones but not rods. J. Cell Biol. 37:424-444. Cole SP and Deeley R G . 1998. Multidrug resistance mediated by the ATP-b ind ing cassette transporter protein M R P . BioEssays. 20:931-940. 91 Cook NJ , Hanke W , and Kaupp U B . 1987. Identification, purification, and functional reconstitution of the cyclic GMP-dependent channel from rod photoreceptors. Proc. Natl Acad. Sci. USA. 84:585-589. Cremers F P , van de Pol DJ, van Driel M , den Hollander A l , van Haren FJ, Knoers N V , Tijmes N , Bergen A A , Rohrschneider K , Blankenagel A , Pinckers A J , Deutman A F , Hoyng C B . 1998. Autosomal recessive retinitis pigmentosa and cone-rod dystrophy caused by splice site mutations in the Stargardt's disease gene A B C R . Hum. Moi. Genet. 7:355-362. Crouch R K , Chader GJ, Wiggert B , Pepperberg D R . 1996. Retinoids and the visual process. Photochem. Photobiol. 64:613-621. Cuatrecasas P. 1970. Protein purification by affinity chromatography. Derivatization of agarose and polyacrylamid beads. J. Biol. Chem. 245:3059-3065. Dean M , Allikmets R. 1995. Evolution of ATP-binding cassette transporter genes. Curr. Opin. Genet. Dev. 5:779-785. Deigner PS, Law W C , Canada FJ, Rando R R . 1989. Membranes as the energy source in the endergonic transformation of vitamin A to 11-cis-retinol. Science. 244:968-71. Deterre P, Bigay J, Forquet F , Robert M , Chabre M . 1988. c G M P phosophodiesterase of retinal rods is regulated by two inhibitory subunits. Proc. Natl Acad. Sci. USA. 85:2424-2428. Dizhoor A M , Olshevskaya E V , Henzel WJ, Wong S C , Stults JT, Ankoudinova I, Hurley JB. 1995. Cloning, sequencing and expression of a 24-kDa Ca(2+)-binding protein activating photoreceptor guanylyl cyclase. J. Biol. Chem. 270:25200-25206. Doige C A , Y u X , Sharom FJ. 1992. ATPase activity of partially purified P-glycoprotein from multidrug-resistant Chinese hamster ovary cells. Biochimica Biophysica Acta. 1109:149-160. Dose A C . 1995. Molecular characterization of the cyclic nucleotide-gated cation channel of bovine rod outer segments. Ph.D. thesis. University of British Columbia. Ellgaard L , Molinari M , Helenius A . 1999. Setting the standards: Quality control in the secretory pathway. Science. 286:1882-1888. Falk G and Fatt P. 1969. Distinctive properties of the lamellar and disk-edge structures of the rod outer segment. J. Ultrastruct. Res. 28:41-60. Farahbakhsh Z T , Hideg K , Hubbell W L . 1993. Photoactivated conformational changes in rhodopsin : a time-resolved spin label study. Science. 262:1416-1419. 92 Farber D B . 1995. From mice to man: the cyclic G M P phosphodiesterase gene in vision and disease. Invest. Ophthalmol. Visual Sci. 36:263-275. Farrar G J , Kenna P, Jardan S A , Kumar-Singh R, Humphries M M , Sharp E M , Sheils D M , Humphries P. 1991. A three-base-pair deletion in the peripherin-RDS gene in one form of retinitis pigmentosa. Nature. 354:478-480. Flannery J G , O ' Day W , Pfeffer B A , Horwitz J , Bok D . 1990. Uptake, processing and release of retinoids by cultured human retinal pigment epithelium. Exp. Eye Res. 51:717-728. Folch J , Lees M , Sloane-Stanley G H . 1957. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226:497-509. Fung B K and Griswold-Prenner I. 1989. G protein-effector coupling: binding of rod phosphodiesterase inhibitory subunit to transducin. biochemistry. 28:3133-3137. Fung B K , Young J H , Yamane H K , Griswold-Prenner I. 1990. Subunit stoichiometry of retinal rod c G M P phosphodiesterase. Biochemistry. 29:2657-64. Gray-Keller M P and Detwiler P B . 1994. The calcium feedback signal in the phototransduction cascade of vertebrate rods. Neuron. 13:849-861. Hamm H E and Gilchrist A . 1996. Heterotrimeric G proteins. Curr. Opin. Cell Biol. 8:189-196. Hayward G , Carlsen W , Siegman A , Stryer L . 1981. Retinal chromophore of rhodopsin photoisomerizes within picoseconds. Science. 211:942-944. He W , Cowan C W , Wensel T G . 1998. R G S 9 , a GTPase accelerator for phototransduction. Neuron. 20:95-102. Hicks D and Molday RS . 1986. Differential immunogold-dextran labeling of bovine and frog rod and cone cells using monoclonal antibodies against bovine rhodopsin. Exp. Eye Res. 42:55-71. Higgins C F . 1992. A B C transporters: from microorganisms to man. Annu. Rev. Cell Biol. 8:67-113. Higgins C F . 1995. The A B C of channel regulation. Cell. 82:693-696. Hsu S C and Molday RS. 1991. Glycolytic enzymes and a G L U T - 1 glucose transporter in the outer segments of rod and cone photoreceptor cells. J. Biol. Chem. 266:217'45-217'52. Hsu Y T and Molday RS. 1993. Modulation of cGMP-gated channel of rod photoreceptor cells by calmodulin. Nature. 361:76-79. 93 Huang S H , Pittler SJ, Huang X , Oliveira L , Berson E l , Dryja T P . 1995. Autosomal recesive retinitis pigmentosa caused by mutations in the a-subunit of rod c G M P phosphodiesterase. Nature Genet. 11:468-471. M i n g M , Molday L L , Molday RS . 1997. The 220-kDa rim protein of retinal rod outer segments is a member of the A B C transporter superfamily. J. Biol. Chem. 272:10303-10310. Ishiguro S, Suzuki Y , Tamai M , Mizuno K . 1991. Purification of retinol dehydrogenase from bovine retinal rod outer segments. J.Biol. Chem. 266:15520-15524. Kawamura S. 1993. Rhodopsin phosphorylation as a mechanism of cyclic G M P phosphodiesterase regulation by S-modulin. Nature. 362:855-857. Kel ly A , Powis S H , Kerr L A , Mockridge I, Elliott T , Bastin J , Uchanska-Ziegler B , Ziegler A , Trowsdale J , Townsend A . 1992. Assembly and function of the two A B C transporter proteins encoded in the human major histocompatibility complex. Nature. 355:641-644. K i m T S , Reid D M , Molday RS . 1998. Structure-function relationships and localization of the N a / C a - K exchanger in rod photoreceptors. J. Biol. Chem. 273: 16561-16567. Krizaj D and Copenhagen D R . 1998. Compartmentalization of calcium extrusion mechanisms in the outer and inner segments of photoreceptors. Neuron. 21:249-256. Kle in I, Sarkadi B , Varadi A . 1999. A n inventory of the human A B C proteins. Biochimica et Biophysica Acta. 1461:237-262. K o c h K - W , Stryer L . 1988. highly co-operative feedback control of retinal rod guanylate cyclase by calcium ions. Nature. 334:64-66. Krupnick J G , Gurevich V V , Benovic J L . 1997. Mechanism of quenching of phototransduction. Binding competition between arrestin and transducin for phosphorylation. J.Biol. Chem. 272:18125-18131. Kuchler K , Thorner J. 1992. Secretion of peptides and proteins lacking hydrophobic signal sequences: the role of adenosine triphosphate-driven membrane translocators. Endocr. Rev. 13:499-514. Lai Y - L , Wiggert B , L i u Y - P , Chader G J . 1982. Interphotoreceptor retinol-binding proteins: possible transport vehicles between compartments of retina. Nature. 298:848-849. Laemmli U K . 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 227:680-685. 94 Lewis R A , Shroyer N F , Singh N , Allikmets R, Hutchinson A , L i Y , Lupski JR, Leppert M , Dean M . 1999. Genotype/phenotype analysis of a photoreceptor-specific ATP-binding cassette transporter gene, ABCR, in Stargardt disease. Am. J. Hum. Genet. 64:422-434. L i C , Ramjeesingh M , Reyes E , Jensen T , Chang X , Rommens J M , Bear C E . 1993. The cystic fibrosis mutation (delta F508) does not influence the chloride channel activity of C F T R . Nat. Genet. 3:311-316. Liou G L , Bridges C D , Fong S-L. 1982. Vitamin A transport between retina and pigment epithelium - an interstitial protein carrying endogenous retinol (interstitial retinol binding protein). Vision Res. 22:1457-1468. L i u X , Seno K , Nishizawa Y , Hayashi F , Yamazaki A , Matsumoto H , Wakabayashi T , Usukura J. 1994. Ultrastructural localization of retinal guanylate cyclase in human and monkey retinas. Exp. Eye Res. 59:761-768. Lolley R N and Racz E . 1982. Calcium modulation of cyclic G M P synthesis in rat visual cells. Vision Res. 22:1481-1486. Loo T W and Clarke D M . 1994. Prolonged association of temperature-sensitive mutants of human P-glycoprotein with calnexin during biogenesis. J. Biol. Chem. 269:28683-28689. Luciani M F , Denizot F , Savary S, Mattei M G , Chimini G . 1994. Cloning of two novel A B C transporters mapping on human chromosome 9. Genomics. 21:150-159. M a h N L . 1999. Heterologous expression and purification of bovine rod photoreceptor glutamic acid rich protein. M.Sc.Jhesis. University of British Columbia. Manavalan P, Dearborn D G , McPherson J M , Smith A E . 1995. Sequence homologies between nucleotide binding regions of C F T R and G-proteins suggest structural and functional similarities. FEBS Lett. 366:87-91. Martinez-Mir A , Paloma E , Allikmets R, Ayuso C , del Rio T , Dean M , Vilageliu L , Gonzalez-Duarte R, Balcells S. 1998. Retinitis pigmentosa caused by a homozygous mutation in the Stargardt disease gene A B C R . Nature Genet. 18:11-12. Mata N L , Weng J , Travis G H . 2000. Biosynthesis of a major lipofuscin fluorophore in mice and humans with ABCR-mediated retinal and macular degeneration. Proc. Natl Acad., Sci. USA. 97:7154-7159. McBee J K , Kuksa V , Alvarez R, de Lera A R , Prezhdo O, Haeseleer F , Sokal I, Palczewski K . 2000. Isomerization of all-trans-retinol to cis-retinols in bovine retinal pigment epithelial cells: dependence on the specificity of retinoid-binding proteins. Biochemistry. 39:11370-11380. 95 McLaughl in M E , Sandberg M A , Berson E L , Dryja T P . 1993. Recessive mutations in the gene encoding the P subunit of rod phosphodiesterase in patients with retinitis pigmentosa. Nature Genet. 4:130-134. Molday L L , Rabin A R , Molday RS. 2000. A B C R expression in foveal cone photoreceptors and its role in Stargardt macular dystrophy. Nature Genet. 25:257-258. Molday RS . 1998. Photoreceptor membrane proteins, phototransduction, and retinal degenerative diseases. The Friedenwal Lecture. Invest. Ophthalmol. Visual Sci. 39:2491-2513. Molday RS and Molday L L . 1987. Differences in the protein composition of bovine retinal rod outer segment disk and plasma membranes isolated by a ricin-gold-dextran density perturbation method. J. Cell Biol. 105:2589-2601. Molday R S , Hicks D , Molday L L . 1987. Peripherin: a rim-specific membrane protein of rod outer segment disks. Invest. Ophthalmol. Visual Sci. 28:50-61. Morgans C W , E l Far O, Berntson A , Wassle H , Taylor W R . 1998. Calcium extrusion from mammalian photoreceptor terminals. J. Neurosci. 18: 2467-2474. Moritz O L and Molday RS . 1996. Molecular cloning, membrane topology and localization of bovine rom-1 in rod and cone photoreceptor cells. Invest. Ophthalmol. Visual Sci. 37:352-362. Nathans J , Thomas D , Hogness D S . 1986. Molecular genetics of human color vision: the genes encoding blue, green, and red pigments. Science. 232:193-202. Nilsson S E G . 1964. Receptor cell outer segment development and ultrastructure of the disc membranes in the retina of the tadpole (Rana pipiens). J. Ultrastruct. Res. 11:581-620. Online: Webvision < http://retina.umh.es/Webvision> (date accessed: 17 July 2001). Online: Human A B C Transporters <http://www.humanabc.org> (date accessed: 6 September 2001). Palczewiski K and Saari J C . 1997. Activation and inactivation steps in the visual transduction pathway. Curr. Opin. Neurobiol. 7:500-504. Palczewski K , Subbaraya I, Gorczyca W A , Helekar B S , Ruiz C C , Ohguro H , Huang J , Zhao X , Crabb J W , Johnson R S , Walsh K A , GrayKeller M P , Detwiler P B , Baehr W . 1994. Molecular cloning and characterization of retinal photoreceptor guanylyl cyclase-activating protein. Neuron. 13:395-404. Papermaster D S , Schneider B G , Zorn M A , Kraehenbuhl JP. 1978. Immunocytochemical localization of a large intrinsic membrane protein to the incisures and margins of frog rod outer segment disks. J. Cell. Biol. 78:415-425. 96 Pasyk E A , Foskett J K . 1995. Mutant (delta F508) cystic fibrosis transmembrane conductance regulator CI- channel is functional when retained in endoplasmic reticulum of mammalian cells. J. Biol. Chem. 270:12347-12350. Penn R D and Hagins W A . 1969. Signal transmission along retinal rods and the origin of the electroretinographic a-wave. Nature (London). 223:201-204. Perrault I, Rozet J M , Calvas P, Gerber S, Camuzat A , Dollfus H , Chatelin S, Souied E , Ghazi I, Leowski C , Bonnemaison M , Le Paslier D , Frezal J, Dufier J L , Pittler S, Munnich A , Kaplan J. 1996. Retinal-specific guanylate cyclase gene mutations in Leber's congenital amaurosis. Nature Genet. 14:461-464. Pind S, Riordan JR, Williams D B . 1994. Participation of the endoplasmic reticulum chaperone calnexin (p88, IP90) in the biogenesis of the cystic fibrosis transmembrane conductance regulator. J. Biol. Chem. 269:12784-12788. Rando R R . 1991. Membrane phospholipids as an energy source in the operation of the visual cycle. Biochemistry. 30:595-602. Reid D M , Friedel U , Molday RS , and Cook N J . 1990. Identification of the sodium-calcium exchanger as the major ricin-binding glycoprotein of bovine rod outer segments and its localization to the plasma membrane. Biochemistry. 29:1601-1607. Ripps H and Weale R A . 1969. Rhodopsin regeneration in man. Nature. 222:775-777. Rosenfeld PJ, Cowley G S , McGee T L , Sandberg M A , Berson E L , Dryja T P . 1992. A null mutation in the rhodopsin causes rod photoreceptor dysfunction and autosomal recessive retinitis pigmentosa. Nature Genet. 13:358-360. Rushton W A and Henry G H . 1968. Bleaching and regeneration of cone pigments in man. Vision Res. 8:617-631. Saari JC and Bredberg D L . 1982. Enzymatic reduction of 11-cis-retinal bound to cellular retinal-binding protein. Biochimica Biophysica Acta. 716:266-272. Saari J C and Bredberg D L . 1989. Lecithin:retinol acyltransferase in retinal pigment epithelial microsomes. / . Biol. Chem. 264:8636-40. Safarpour A . 1999. Expression of the retinal rod photoreceptor A B C protein ( A B C R ) in yeast cells. B. Sc. honours thesis. University of British Columbia. Sambrook J, Fritsch E F , Maniatis T . 1989. Molecular Cloning: A Laboratory Manual. 2 n d E d . Cold Spring Harbor Press, New York. Sampath A P , Matthews H R , Cornwall M C , Fain G L . 1998. Bleached pigment produces a maintained decrease in outer segment C a 2 + in salamander rods. J. Gen. Physiol. 111:53-64. 97 Schneider E and Hunke S. 1998. ATP-binding-cassette ( A B C ) transporter system: Functional and structural aspects of the ATP-hydrolyzing subunits/domains. FEMS Microbiol. Rev. 22:1-20. Shapiro A B and Ling V . 1994. ATPase activity of purified and reconstituted P-glycoprotein from Chinese hamster ovary cells. J. Biol. Chem. 269:3745-3754. Shyamala V , Baichwal V , Beall E , Ames G F . 1991. Structure-function analysis of the histidine permease and comparison with cystic fibrosis mutations. J. Biol. Chem. 266:18714-18719. Spitznas M and Hogan M J . 1970. Outer segments of photoreceptors and retinal pigment epithelium. Arch. Ophthalmol. 84:810-819. Stecher H , Gelb M H , Saari J C , Palczewski K . 1999. Preferential release of 11-cz's-retinol from retinal pigment epithelial cells in the presence of cellular retinaldehyde-binding protein. J. Biol. Chem. 274:8577-8585. Steinberg R H , Fisher S K , Anderson D H . 1980. Disc morphogenesis in vertebrate photoreceptors. J. Comp. Neurol. 190:501-508. Sun H , Molday R S , Nathans J. 1999. Retinal stimulates A T P hydrolysis by purified and reconstituted A B C R , the photoreceptor-specific ATP-binding cassette transporter responsible for Stargardt disease. J. Biol. Chem. 274:8269-8281. Sun H , Smallwood P M , Nathans J. 2000. Biochemical defects in A B C R protein variants associated with human retinopathies. Nature Genet. 26:242-246. Tsuboi S, Matsumoto H , Jackson K W , Tsujimoto K , Will iams T , Yamazaki A . 1994. Phosphorylation of an inhibitory subunit of c G M P phosphodiesterase in Rana catesbeiana rod photoreceptors. I. Characterization of the phosphorylation. J. Biol. Chem. 269:15024-15029. Tsuboi S, Matsumoto H , Yamazaki A . 1994. Phosphorylation of an inhibitory subunit of c G M P phosphodiesterase in Rana catesbeiana rod photoreceptors. II. A possible mechanism for the turnoff of c G M P phosphodiesterase without G T P hydrolysis. J. Biol. Chem. 269:15016-15023. Udovichenko IP, Newton A C , Will iams D S . 1997. Contribution of protein kinase C to the phosphorylation of rhodopsin in intact retinas. J. Biol. Chem. 272:7952-7959. Urbatsch IL, Al-Shawi M K , Senior A E . 1994. Characterization of the ATPase activity of purified Chinese hamster P-glycoprotein. Biochemistry. 33:7069-7076. Walker J E , Saraste M , Runswick M J , Gay N J . 1982. Distantly related sequences in the alpha- and beta-subunits of A T P synthase, myosin, kinases and other ATP-requiring enzymes and a common nucleotide binding fold. EMBO J. 1:945-951. 98 Wang J K , McDowe l l J H , Hargrave P A . 1980. Site of attachment of 11-cis-retinal in bovine rhodopsin. Biochemistry. 19:5111-5117. Warren R. 1999. Insights on the regulation of the cGMP-gated channel by Ca + + / ca lmodul in and on the phosphorylation of the beta-subunit by a CKII-l ike protein kinase. M. Sc. thesis. University of British Columbia. Wells J , Worblewski J , Keen J , Inglehearn C , Jubb C , Eckstein A , Jay M , Arden G , Bhattacharya S, Fitzke F , Bird A C . 1993. Mutations in the human retinal degeneration slow (RDS) gene can cause either retinitis pigmentosa or macular dystrophy. Nature Genet. 3:213-218. Weng J , Mata N L , Azarian S M , Tzekov R T , Birch D G , Travis G H . 1999. Insights into the function of rim protein in photoreceptors and etiology of Stargardt's disease from the phenotype in abcr knockout mice. Cell. 98:13-23. Wensel T G and Stryer L . 1990. Activation mechanism of retinal rod cyclic G M P phosphodiesterase probed by fluorescein-labeled inhibitory subunit. Biochemistry. 29:2155-61. Winston A and Rando R R . 1998. Regulation of isomerohydrolase activity in the visual cycle. Biochemistry. 37:2044-50. Yamagata K , Goto K , Kuo C H , Kondo H , M i k i N . 1990. Vis inin: a novel calcium binding protein expressed in retinal cone cells. Neuron. 4:469-476. Yannuzzi L A , Grayer D R , Green W R . The retina A T L A S . St-Louis:Mosby Yearbook Inc. 1995. Yau K W . 1994. Phototransduction mechanism in retinal rods and cones. The Friedenwald Lecture. Invest Ophthalmol Visual Sci. 35:9-32. Young R W . 1968. Passage of newly formed protein through the connecting cil ium of retinal rods in the frog. J. Ultrastruct. Res. 23:A62-A13. Young R W . 1971. Shedding of disks from rod outer segments in the rhesus monkey. J. Ultrastruct. Res. 34:190-203. Young R W . 1987. Pathophysiology of age-related macular degeneration. Surv. Ophthalmol. 31:291-306. Young R W and Bok D . 1969. Participation of the retinal pigment epithelium in the rod outer segment renewal process. J. Cell Biol. 42:392-403. Young R W and Droz B . 1968. The renewal of protein in retinal rods and cones. J. cell Biol. 39:169-184. 99 Zagotta W N and Siegelbaum S A . 1996. Structure and function of cyclic nucleotide-gated channels. Annu. Rev. Neurosci. 19:235-263. Zhang L , Sports C D , Osawa S, Weiss E R . 1997. Rhodopsin phosphorylation sites and their role in arrestin binding. J. Biol. Chem. 272:14762-14768. Zhou X and Arthur G . 1992. Improved procedures for the determination of lipid phosphorus by malachite green. J. Lipid Res. 33:1233-1236. 100 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0090157/manifest

Comment

Related Items