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 C H A R A C T E R I Z A T I O N OF RETINAL DISEASE LINKED A B C R MUTANTS by AZIEN  SAFARPOUR  B.Sc. (Hons.), The University o f British C o l u m b i a , 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 THE REQUIREMENTS FORTHE DEGREE OF M A S T E R OF SCIENCE in T H E F A C U L T Y OF G R A D U A T E STUDIES (Department o f Biochemistry and M o l e c u l a r Biology)  We accept this thesis as conforming to the required standard  THE UNIVERSITY  OF BRITISH C O L U M B I A  September 2001 © A z i e n Safarpour, 2001  In  presenting this  degree at the  thesis  in  partial  fulfilment  of  the  requirements  University  of  British  Columbia,  I agree that the  freely available for reference and study. I further agree that copying  of  department  this thesis for scholarly or  by  his  or  her  for  an  Library shall make it  permission for extensive  purposes may be granted by the  representatives.  It  is  understood  that  publication of this thesis for financial gain shall not be allowed without permission.  Department of  OcWYvnST'j  The University of British Columbia Vancouver, Canada  Date  DE-6 (2788)  QcToUgr  2-^  i  ^  advanced  T£> ^T] vc)  head of my copying  or  my written  ABSTRACT  The  retinal A B C transporter, A B C R , is a 220 k D a glycoprotein localized in the outer  segments o f the photoreceptor cells. organized into two  tandem halves  nucleotide binding domain.  A s a member o f the A B C superfamily, A B C R  is  each containing a transmembrane domain and a  Studies have suggested that A B C R may either act as an a l l -  trans-retinal extruder or a retinylidene-phosphatidylethanolamine flippase.  Eighty-nine mutations in  ABCR gene have been identified and associated with a number o f  retinal degenerative diseases including Stargardt's disease.  In this study the effects o f  Stargardt's disease causing mutations ( D 8 4 6 H , T 1 5 2 6 M , R 2 0 3 8 W , R 2 0 7 7 W and R 2 1 0 6 C ) on the structure and function o f 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 o f the proteins were compared, it was found that the R 2 0 3 8 W and  R 2 0 7 7 W mutants were expressed at 57-59% o f the w i l d type while R 2 1 0 6 C and  T 1 5 2 6 M expressed at 71% and 81% o f the w i l d type protein, respectively.  The expression  level o f the D 8 4 6 H mutant varied between 31-68% o f the w i l d type, depending on the transfection  method  used.  The recombinant  proteins  were  purified from  CHAPS  solubilized C O S cell by immunoaffinity chromatography.  Immunofluorescence microscopy o f the transfected C O S cells revealed that the expressed wild type A B C R was localized in the vesicles, whereas almost 100% o f both D 8 4 6 H and T 1 5 2 6 M , 77% o f R 2 0 3 8 W , 80% o f R 2 0 7 7 W and 34% o f R 2 1 0 6 C showed an endoplasmic reticulum/Golgi labeling pattern.  C a l n e x i n interaction studies showed that all o f the  mutants examined co-purified with calnexin to a greater extent than the w i l d type A B C R .  The  ATPase  activity  of  purified  proteins  phosphatidylethanolamine was measured.  reconstituted  in  liposomes  rich  in  In the presence o f 50 u M all-frvms-retinal, the  basal activity o f the wild type A B C R was increased by 1.6 fold; however, no stimulation was seen in T 1 5 2 6 M and R 2 0 3 8 W variants.  The A T P a s e activities o f both D 8 4 6 H and  R 2 0 7 7 W were impaired, whereas the R 2 1 0 6 C mutant showed a retinal-stimulated activity  ii  similar to the wild type.  The a z i d o - A T P labeling  study showed that the w i l d type,  T 1 5 2 6 M and R 2 1 0 6 C bound A T P , whereas the D 8 4 6 H and R 2 0 7 7 W did not.  iii  TABLE OF CONTENTS  Abstract  ii  Table o f Contents  iv  List o f Tables  vii  List of Figures  viii  List of Abbreviations  x  Acknowledgements  xiii  Dedication  xiv  1. I N T R O D U C T I O N  1  1.1  The H u m a n E y e  1  1.2  The Retina  1  1.3  The Photoreceptor Cells  4  1.4  The R o d Outer Segment and the D i s k  6  1.5  Phototransduction  7  1.5.1 1.5.2 1.5.3  The Dark Current  7  Photoexcitation and Recovery The V i s u a l (Retinoid) C y c l e  10 14  1.6  Retinal Pigment Epithelium  14  1.7  Macular Degeneration and Stargardt's Disease  20  1.8  The Superfamily o f A B C Proteins  22  1.9 1.10  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  Possible Functions of A B C R  29  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 B i o l o g y Techniques  35  2.3  Generation o f 3F4 C o u p l e d Sepharose Beads  37  2.4  Generation o f the Constructs  37 iv  2.5 Expression o f 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 o f 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 o f 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 o f 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 A z i d o - A T P Labeling o f C O S Membranes  50  2.13.1 Membrane Preparation  50  2.13.2 A z i d o - A T P 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 o f the A B C R Mutants  54  3.2 Generation o f Constructs  56  3.3 Transfection and Expression o f W i l d Type and Mutant A B C R  56  3.4 Purification o f W i l d Type and Mutant A B C R from Solubilized C O S Cells 3.5 Immunofluorescence  58 Microscopy  62  3.6 Calnexin Association  65  3.7  65  Reconstitution o f A B C R s into Liposomes  3.8 The A T P a s e Assay  67  3.9  74  A z i d o - A T P Labeling v  4. D I S C U S S I O N  77  4.1  Expression o f W i l d Type and Mutant A B C R s in C O S Cells  77  4.2  Purification and Reconstitution o f A B C R Variants  78  4.3  The A T P a s e Assay  79  4.4  Analysis of the Expressed W i l d Type and Mutant A B C R s  81  4.5  4.4.1  W i l d Type A B C R  81  4.4.2  D846H  83  4.4.3  T1526M  84  4.4.4  R2038W  85  4.4.5  R2077W  86  4.4.6  R2106C  87  Summary  88  5. R E F E R E N C E S  90  vi  LIST 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 o f mutant A B C R s relative to w i l d type A B C R in C O S cells  60  vii  LIST OF FIGURES  F i g 1 Anatomy o f the human eye  2  F i g 2 Organization o f the retina  3  F i g 3 The rod and cone photoreceptor cells  5  F i g 4 Dark current in the rod photoreceptor cells  9  F i g 5 The process o f phototransduction in the rod outer segment  11  F i g 6 Schematic diagram depicting the phagocytosis o f the rod outer segment by the retinal pigment epithelium  15  F i g 7 The visual cycle and the role o f retinal pigment epithelium  18  F i g 8 N o r m a l and Stargardt's disease affected macula  21  F i g 9 Prototype domain arrangements in A B C proteins  23  F i g 10 Putative topological models o f A B C R  27  F i g 11 Possible functions o f A B C R  31  F i g 12 Strategy used to generate D 8 4 6 H construct  40  F i g 13 Putative model o f A B C R showing the five Stargardt's disease mutations  55  F i g 14 Generation o f the selected constructs  57  F i g 15 Expression o f the w i l d type and the five mutant A B C R constructs in C O S cells  59  F i g 16 Purification o f wild type and mutant A B C R s from solubilized C O S cells  61  F i g 17 Immunofluorescence labeling o f C O S cells transfected with w i l d type and mutant A B C R constructs  63  F i g 18 Co-purification o f calnexin with mutant A B C R s  66  F i g 19 A B C R proteins reconstituted in liposomes  68  F i g 20 The effect o f all-frans-retinal on the A T P hydrolysis o f reconstituted ROS A B C R F i g 21  69  Basal and retinal stimulated A T P hydrolysis o f R O S A B C R reconstituted in two different lipid mixtures  F i g 22 The effect o f  70  a\\-trans-ret'ma\ concentration on the A T P a s e activity o f  reconstituted R O S A B C R  71  viii  F i g 23 Basal and retinal stimulated A T P a s e activities o f R O S A B C R and wild type A B C R from transfected C O S cells F i g 24 A T P hydrolysis o f reconstituted wild type and mutant A B C R s purified from transfected C O S cells F i g 25 A z i d o - A T P labeling and yield o f wild type and mutant A B C R s  ix  LIST O F A B B R E V I A T I O N S  ABC  A T P binding cassette  ABCR  retinal A B C transporter  ADP  adenosine 5'-diphosphate  ALDP  transporter protein responsible for adrenoleukodystrophy  ATP  adenosine 5'-triphosphate  BCA  bicinchroninic acid  BES  N,N-bis(2-hydroxyethyl)-2-aminoethanesulfonic acid  BHT  butylated hydroxytoluene  bp  base pairs  BSA  bovine serum albumin  cAMP  adenosine 3',5'-cyclic monophosphate  cDNA  D N A reverse-transcribed from an m R N A template (copy D N A )  CFTR  cystic fibrosis transmembrane regulator  cGMP  guanosine 3',5'-cyclic monophosphate  CHAPS  3 - [(cholamidopropyl)-dimethylammonio] -1 -propanesulfonate  CRBP  cellular retinal-binding protein  CY3  fluorescent cyanine dye  DMEM  Dulbecco's modified Eagle medium  dNTP  deoxyribonucleoside triphosphate  DOPE  1,2-dioleoylphosphatidylethanolamine  DTT  dithothrieitol  ECL  enhanced chemiluminescence  EDTA  ethylaminediamine tetraacetic acid  FCS  fetal calf serum  GAP  GTPase-activating protein  GCAP  guanylate cyclase-activating protein  GDP  guanosine 5'-diphosphate  GTP  guanosine 5'-triphosphate  HEPES  N-2-hydroxyethylpiperazine-N'-2-ethanesulphonic acid x  HH1  highly hydrophobic sequence  Ig  immunoglobulin  IRBP  interphotoreceptor retinal-binding protein  kb  kilobase  Km  Michaelis constant  LB  Luria-Bertani broth  LRAT  lecithin-retinol acyltransferase  MDR  multi-drug resistance  MHC  major histocompatibility complex  MRP  multidrug resistance associated protein  NBD  nucleotide binding domain  PAGE  polyacrylamide gel electrophoresis  PBS  phosphate buffered saline  PBS-T  phosphate buffered saline with 0.05% Tween 20  PC  phosphatidylcholine  PCR  polymerase chain reaction  PDE  phosphodiesterase  PE  phosphatidylethanolamine  PI  phosphatidylinositol  PS  phosphatidylserine  R*  metarhodopsin II  rds  retinal degeneration slow  RGS  regulator o f G protein signaling  ROS  rod outer segment  RPE  retinal pigment epithelium  SDPE  1 -stearoyl-2-docosahexaenoylphosphatidyl  SDS  sodium dodecyl sulfate  SUR  sulfonylurea receptor  T py  transduction complex  TAP  transporter associated with peptide presentation  TBS  Tris-buffered saline  TMD  transmembrane domain  a  XI  ethanolamine  Tris [hydroxymethyl] aminoethane volume per volume weight per volume  xii  ACKNOWLEDGEMENTS  I would like to thank D r . Robert M o l d a y 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 D r . Ross M a c G i l l i v r a y and D r . M i c h a e l Murphy for being on my committee.  I cherish the experience that I gained as a graduate student and owe it to every member o f the M o l d a y lab, past and present.  I would like to thank Laurie M o l d a y 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 o f the lab, Stephanie Bungert for being a good friend full o f energy, and for her encouragement, Jason W o n g for his statistical hockey discussions, D a n 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 W i n c o W u , for their help and A n d r e w H o 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  THE HUMAN EYE  The human eye (Fig 1) is an elongated sphere o f about 1 inch in diameter that transduces light in the range o f 400 nm to 700 nm into neural discharges, allowing visual perception. The eye consists o f three layers or coats and two chambers. The white fibrous outer layer known as the sclera forms part o f the supporting wall o f the eyeball. Near the front o f 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.  T o w a r d the front, the  choroid, first, thickens and forms the ciliary body which controls the shape o f the lens and, finally becomes a thin, muscular diaphragm called the iris.  T h i s colored circular muscle  controls the amount o f light entering the eye by altering the size o f the pupil.  The inner  layer of the eye is the retina. It consists o f a thin layer o f neural tissue lining the back o f 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 m m thick (http://retina.umh.es/Webvision) and lines the back of the eye. A l l vertebrate retinas are composed o f three layers o f 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 o f 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. two layers o f synapses.  The retina is composed o f three layers o f cells and  The closest layer to the choroid contains the photoreceptor cells,  the rods and the cones. The photoreceptor cells communicate by way o f electrical synapse with each other and with the second layer o f cells.  The horizontal cells, bipolar cells and  the amacrine cells make up the second cell layer o f the retina. consists o f ganglion cells whose fibers become the optic nerve. http://retina.umh.es/Webvision.  3  The innermost layer Figure obtained from  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 o f 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 o f 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 o f an area known as the macula.  1.3  THE PHOTORECEPTOR CELLS  There are two types o f 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 o f 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). array o f flattened membranous disks.  The outer segment contains an  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 G o l g i apparatus,  mitochondria and endoplasmic reticulum ( E R ) . The cell body contains the nucleus and the synaptic terminus has synaptic vesicles containing the neurotransmitter glutamate.  The  release o f neurotransmitter is high in the dark and is reduced in a graded fashion in the light.  4  Plasma membrane Outer segment  Disks  Connecting cilium  Inner segment Cell body  Nucleus  Synaptic Terminus  Rod  Cone  Fig 3. The rod and cone photoreceptor cells. There are two types o f photoreceptor cells in the retina: the rods and the cones. E a c h photoreceptor cell consists o f four regions: the outer segment, the inner segment, cell body and the synaptic terminus. The outer segment of the rods contains an array o f light sensitive membranous disks that is surrounded by a plasma membrane. the disk membrane.  The outer segment plasma membrane o f the cones is continuous with The outer segment o f both rods and cones is connected to the inner  segment by a thin connecting cilium. terminus releases glutamate.  The cell body contains the nucleus and the synaptic  Figure adapted from Dose (1995).  5  1.4  T H E ROD O U T E R SEGMENT AND T H E DISK  Phototransduction in the rod cells occurs specifically in the outer segment, which consists of  hundreds o f membranous disks.  Disks are formed at the connecting c i l i u m by  successive evagination o f the plasma membrane such that each newly made disks become stacked one on top o f 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. T o maintain the outer segment at a steady length, old disks are pushed up toward the tip o f the outer segment, where they are eventually released from the outer segment and engulfed by the retinal pigment epithelium (reviewed in B o k , 1985).  The  lipids and proteins that make up the disk membrane are synthesized in the inner segment and transported to the base o f the outer segment, where they are incorporated into the basal folding o f the outer segment ( Y o u n g and D r o z , 1968; Y o u n g , 1968).  Although the disk membrane arises from the folding o f the plasma membrane, the protein and phospholipid compositions o f the two membranes are somewhat different.  Boesze-  Battaglia and Albert (1992) have shown that phosphatidylethanolamine (PE) accounts for 11% o f the total phospholipids in the plasma membrane, whereas it accounts for 42% in the disk membranes.  The amount o f phosphatidylserine (PS) in the plasma membrane  (24%) is twice that in the disk membrane (14%). Phosphatidylcholine ( P C ) in the plasma membrane represents 65% o f the total phospholipids but 45% o f total disk phospholipids. Phosphatidylinositol (PI) represents a very small percentage o f 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 o f the membrane phsopholipids between the disks and  the surrounding plasma membranes.  W h e n 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 o f the outer segment, the ratio o f cholesterol phospholipids decreases  (Boesze-Battaglia  et  al,  1990); however,  to  the phospholipid  headgroup and fatty acyl composition o f the disk membranes do not change.  Although the protein compositions o f the two membranes differ ( M o l d a y and M o l d a y , 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 -K +  2+  +  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 , C a +  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 -K +  2+  +  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 C a  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  M o l e c u l a r weight , D a  Function  Rhodopsin  36 000 (38 000)  Phototransduction  c G M P - c h a n n e l a subunit  63 000 (79 600)  Phototransduction  c G M P - c h a n n e l P subunit  240 000 (155  000)  Na /Ca  230 000 (130  000)  Plasma membrane proteins  +  2 +  -K  +  exchanger  Glucose transporter  Cation, exchange  50 000  Glucose transport  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  ABCR  220 000 (257  Retinol dehydrogenase  33 500  Disk membrane proteins  000)  Retinoid transport Retinal reduction  a  Table reproduced from M o l d a y , 1998.  b  Molecular weight values were determined by S D S - P A G E ; values in parentheses were calculated from sequence.  8  Na7Ca -K 2+  +  exchanger 4Na  •  Fig 4. Dark current in the rod photoreceptor cells.  Dark current consists o f the  entrance o f cations mainly through c G M P - g a t e d channel and their exits through different channels or pumps. and the N a / C a +  segment.  Ca  2 +  2 +  -K  Na +  +  enters the rod outer segment through the c G M P - g a t e d channels  exchanger and is pumped out by a N a / K +  +  Ca  A T P a s e in the inner  also enter the outer segment through the c G M P - g a t e d channel but is  extruded through the N a / C a 2 +  +  2 +  -K  is pumped out through a C a  +  2 +  exchanger in the outer segment. ATPase. K  +  In the inner segment  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 Na /K +  +  A T P a s e . Although M g  2 +  is transported through the c G M P - g a t e d channel, its efflux  mechanism needs to be determined.  A s shown, in the dark there is a flow o f cations that  enters the rod photoreceptor cell and results in the polarization o f 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 - K +  2 +  +  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 - K +  (Cervetto et al,  2 +  +  exchanger in the outer segment  1989; K i m et al., 1998). The rod inner segment plasma membrane does j 2~1~ ~F 2"i" 2"i"  not contain the Na / C a - 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\\-transretinal. 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 p ) is a multisubunit peripheral a  y  membrane protein consisting of three subunits: a or T (39 kDa), (3 or Tp (36kDa) and y or a  T (8kDa) (Hamm and Gilchrist, 1996). The a subunit ( T ) contains both a binding site y  a  for G T P / G D P and a catalytic site for the hydrolysis of bound GTP. The P and y subunits form the Tp which remain membrane bound. T associates with Tp when GDP is bound, T  a  Y  but may separate and be released into the space between disks when G T P is attached.  10  Fig 5. The process of phototransduction in the rod outer segment. A photon o f light isomerizes  1 l-cis-retinal  attached  to  rhodopsin  to  all-rrarcs-retinal  to  produce  metarhodopsin II, the activated form o f rhodopsin. Metarhodopsin II activates transducin by catalyzing the exchange o f G D P for G T P on the a subunit of transducin ( T ) . a  The  activated transducin ( T - G T P ) dissociates from its py subunits and in turn activates a  cGMP-phosphodiesterase ( P D E ) by removing P D E ' s inhibitory y subunits. Activated P D E catalyzes the hydrolysis o f c G M P to 5 ' - G M P .  The decrease in the c G M P concentration  causes the c G M P - g a t e d channels to close, preventing N a causing the cell to hyperpolarize. The C a extrusion o f C a  2 +  by the N a / C a +  2 +  -K  +  2 +  +  and C a  2 +  to enter the cell and  concentration decreases as a result o f constant  exchanger.  11  Figure reproduced from M a h (1999).  U p o n binding to the activated rhodopsin (R*), transducin exchanges its G D P for G T P on its T , producing the active form T - G T P which dissociates from T p and R * . The T - G T P a  a  Y  a  diffuses and binds to a photoreceptor specific phosphodiesterase ( P D E ) , while the freed R * catalyzes another round o f G T P / G D P exchange on a second transducin molecule.  Thus, a  single R * may activate hundreds o f molecules o f transducin.  Phosphodiesterase  ( P D E ) is a peripheral membrane protein consisting o f 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 - G T P interact and activate the aPy2 holoenzyme, by a  removing its inhibitory constraint (Fung and Griswold-Prenner, 1989; Fung et al, The activated P D E complex hydrolyzes c G M P to 5 ' - G M P and effectively  1990).  reduces the  concentration o f c G M P in R O S . The decrease in the intracellular c G M P concentration causes the c G M P - g a t e d channels on the plasma membrane, that were maintained open in the dark, to close (Molday, 1998; Zagotta and Siegelbaum, 1996). channel prevents the flow o f N a , C a a n d M g +  membrane to become hyperpolarized. extrude C a , the C a 2 +  2 +  2 +  2 +  The closure o f the  into the R O S , causing the rod cell  Since the N a / C a +  2 +  -K  +  exchanger continues  to  concentration decreases from ~ 500-700 n M in dark to - 3 0 - 5 0 n M  in light (Sampath et al,  1998).  Hyperpolarization o f 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 o f  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 c G M P - g a t e d channel. R * is inactivated by two sequential processes: rhodopsin kinase (Udovichenko et al,  the phosphorylation o f its C-terminus by  1997) followed by the binding o f arrestin. Arrestin  is a 48 k D a cytosolic protein that prevents phosphorylated rhodopsin from interacting with transducin (Krupnick et al,  1997; Zhang et al,  1997).  T - G T P is inactivated through its a  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 ( G A P ) , the most important being R G S 9 (He.er al,  12  1998). P D E activity is inhibited as T a  G D P dissociates and re-associates with T p . Alternative mechanisms of P D E inactivation y  have also been suggested including phosphorylation o f 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 o f the c G M P  to its original level in order to re-open the c G M P - g a t e d channel. c G M P is re-synthesized by guanylate cyclase present in the outer segment. c G M P - g a t e d channel re-opens and N a , C a +  2 +  A s the level o f c G M P increases, the  and M g  2 +  enter, bringing the membrane  potential back to ~ - 40 m A .  It has been shown that the changes in C a  concentration may trigger several negative  2 +  feedback pathways  (Gray-Keller and Detwiler, 1994).  synthesizes c G M P  is inhibited by high C a  Following  illumination however,  guanylate cyclase.  low  2 +  Ca  First guanylate cyclase  which  concentration (Lolley and Racz, 1982). concentration  leads to the  activation  This process seems to be mediated through a specific family o f C a  of 2 +  -  binding proteins termed guanylate cyclase activating proteins or G C A P s ( K o c h and Stryer, 1988; Palczewski et al,  1994; D i z h o o r et ah, 1995). C a  2 +  concentrations also regulate the  affinity o f the c G M P - g a t e d channel for c G M P through calmodulin, another C a - b i n d i n g 2+  protein (Hsu and M o l d a y , 1993; Chen et al,  1994).  F o r example a fall in the  concentration increases the affinity o f the channel for c G M P . recoverin, Kawamura,  affects 1993).  the  phosphorylation  A t low C a  2 +  of  activated  Similarly C a  rhodopsin  levels, recoverin, a C a  2 +  (Chen et  2 +  Ca  2 +  , through  al,  1995;  binding protein, seems to  interfere with the inhibition o f rhodopsin kinase, which may lead to shorten the lifetime o f 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 o f the protein moiety  (opsin) and the conversion o f the chromophore all-frans-retinal to 11-czs-retinal.  The transduction mechanism o f vertebrate cone was not discussed here; however it very much resembles that o f 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,  13  1986).  1.5.3  The visual (retinoid) cycle  F o l l o w i n g 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 o f retinoids between the photoreceptor cell and the surrounding retinal pigment epithelium ( R P E ) , in a process termed the visual or retinoid cycle (Crouch et 1996).  al,  In the cones, the cone pigment regeneration occurs in less than 10 m i n (Rushton  and Henry, 1968), whereas in rods, it lasts 20-30 m i n (Ripps and Weale, 1969). F o l l o w i n g the photoexcitation o f rhodopsin and the isomerization o f 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, Flannery  et al, 1990).  (section  1.6).  Within the R P E ,  1 l-cw-retinal  photoreceptor cells via I R B P .  1982; L i o u et al,  1982;  \\-trans-xe\mo\ is converted to 11-c/s-retinal  is then released  from R P E cells and returned to  the  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 o f both the rods and the cones are closely associated with a monolayer of cuboidal cells called retinal pigment epithelium ( R P E ) .  R P E is located between the  capillary bed o f the choroid and the photoreceptor cells.  The cells o f R P E contain  pseudopodial  attachments  that envelop  the  outer  segments.  T w o o f the  important  functions o f R P E include: i) the phagocytosis o f rod and cone outer segment fragments that are shed and ii) the uptake, processing and release o f retinoids involved in the visual cycle (reviewed in B o k , 1993).  The  phagocytosis  function  o f R P E in the  renewal  process  o f outer  segments  was  discovered by autoradiography. Y o u n g and B o k (1969) showed that when radioactive frog outer segment disks reached the apex o f the outer segment, they became detached (a process called disk shedding) and were rapidly phagocytized by R P E (Fig 6).  14  Although  Fig 6. Schematic diagram depicting the phagocytosis of the rod outer segment by the retinal pigment epithelium. A monolayer o f retinal pigment epithelium ( R P E ) is adjacent to the outer segment o f the photoreceptor cells (only rods are shown). A l t h o u g h the exact mechanism is still unknown, the phagocytosis process begins as peusodopodial attachments o f R P E cells envelop and internalize the old discarded disks. Inside the R P E , the discarded disks form phagosomes. T o break down and digest the contents o f the phagosomes, lysosomes fuse with them, forming larger structures called phagolysosomes.  15  <Tmnnnnnnnnnnnnt?nnnnnmiiiiiiii/inl.1...Ti  E  J5  <D <u  |E£ Q)  or:  D 3 Q _  Q_ <D  co  CO  O  a: 16  the mechanism underlying disk shedding and internalization by R P E is still obscure, some investigators suggest that R P E 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 R P E (Young, 1971).  Once  internalized in the R P E 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 R P E (Fig 7). As explained earlier, all-rrans-retinol is transported to R P E 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 ( L R A T ) (Saari and Bredberg, 1989).  The  resulting retinyl ester serves as a stable storage form of retinoids in the R P E 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 R P E 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 R P E and may compromise the function of R P E . 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 R P E 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 R P E where it is esterified to all-/>aw,s-retinyl ester by lecithin-retinol acyltransferase ( L R A T ) (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/sretinal 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 o f retinal-specific genes have been identified that, when mutated, result in retinal degeneration (the RetNet website summarizes the current genetic knowledge o f the various retinal diseases).  This should not be o f great  surprise when one considers the large number o f proteins that are involved in the visual cascade and visual cycle alone.  For example mutations in rhodopsin and catalytic a and (3  subunits o f P D E result in autosomal recessive retinitis pigmentosa (Rosenfeld et al, M c L a u g h l i n et al,  1993; Huang et al,  peripherin/rds and A B C R  1995).  1992;  Similarly, mutations in guanylate cyclase,  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. , 1 9 9 3 ; A l l i k m e t s et al. ,1997a).  Stargardt's disease, which was first described by K a r l Stargardt in 1909, is a recessive form of macular degeneration with a juvenile to young-adult onset.  T h i s disease is  characterized by the loss o f central vision, progressive atrophy o f 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 o f lysosomal material similar to lipofuscin within the aging R P E cells.  The abnormal and premature accumulation o f lipofuscin may be the result o f any  one or more o f the following defects: i) increased turnover o f photoreceptor cells, ii) abnormal phagocytosis  o f R O S , iii) abnormal degradative activity in the R P E and iv)  defective visual cycle.  In 1997, a study by A l l i k m e t s Stargardt's disease.  et al. (1997a) identified ABCR as the gene responsible for  Immediately, this gene was cloned and the resulting protein, called  A B C R , was identified as an A B C transporter. T o date mutations in the A B C R gene have been linked to a number o f retinal diseases such as: Stargardt's disease, recessive  cone-rod  dystrophy,  retinitis  pigmentosa,  fundus  controversial age-related macular degeneration (Cremers et al, 1998; Allikmets et al,  1997a,b).  20  flavimaculatus  autosomal and  the  1998; Martinezmir et  al,  A  ~~  — ^  Fig 8 . Normal and Stargardt's disease affected macula. (A) The macular region o f a normal individual. (B) features  The macula o f a Stragardt's disease patient.  of Stargardt's disease is the  appearance  One o f the clinical  o f yellowish/orange  deposition  lipofuscin material around the macula, as shown. Diagram from Yannuzzi et al,  21  1995.  of  1.8  T H E SUPERFAMILY OF A B C PROTEINS  The superfamily of ATP-binding cassette ( A B C ) proteins which includes P-glycoprotein, multidrug resistance  associated protein ( M R P ) 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 C F T R .  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 A B C 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. A s transporters, ABC  proteins  contain  a  six membrane  spanning  segment,  denoted  T M D for  transmembrane domain. The T M D s and N B D s 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 ( T M D - N B D ) or ( N B D - T M D ) configuration. Some examples of half size A B C transporters are TAP1 and 22  Fig 9. Prototype domain arrangements in A B C proteins. (A) Linear representation o f the nucleotide binding domain ( N B D ) .  Shown are the Walker A and B motifs and the  signature sequence (motif C ) o f A B C transporters. The amino acids are designated by one letter code,  ' h ' stands for hydrophobic amino acids and 'x' can be varied.  reproduced from Schneider and Hunke, 1998.  (B)  Figure  T h e different organizations o f the  nucleotide binding domains ( N B D ) 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 -  NBD, (TMD-NBD)  2  and T M D 0 ( T M D - N B D ) . 2  23  Figure adapted from K l e i n et al,  1999.  Walker A GxxGxGKT/S  Walker B hhhhD  I  NH 2  COOH  Motif C LSGGQQQR R K  B  NBD  NBD^J  TMD  7  \ ~ f ^^^^^ ~^f-"^^^^^^  7  TMDO  nnnr TMD1  24  TMD2  _^ & ^^^^^^ '"1  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 . either  duplicated  forward  (TMD-NBD)  The domain arrangements in these proteins are or  2  duplicated  reverse  (NBD-TMD) .  glycoprotein, C F T R and A B C R are three examples o f full size transporters.  2  P-  Some A B C  proteins such as S U R , and some o f M R P s contain the configuration o f 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 o f 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). M a n y o f these subfamilies have a structural counterpart in yeast. F o r example, the M D R , M R P and ALD  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 o f 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 o f this study.  1.8.2  A B C A subfamily and A B C R  The A B C A subfamily includes members such as A B C A 1 , A B C A 2 , A B C A 3 and A B C A 4 or A B C R . NBD)2.  These proteins are full size transporters with the forward configuration ( T M D -  Further more, they contain a stretch o f highly hydrophobic amino acids ( H H 1 )  linking the two tandem halves (Luciani et al, 66% identity along their entire sequence,  1994). Members o f this subfamily show 4555-61%) along their first N B D and 57-69%)  identity along their second N B D (Broccardo et al, A B C R is the focus o f this study.  25  1999). A m o n g the A B C A subfamily,  A B C R , which was named the R i m protein, was first reported by Papermaster et al. in 1978 as a 290 k D a glycoprotein localized in the rim region of the frog outer segment disks. Two  decades later, in 1997 the bovine homologue o f the R i m protein was cloned and  identified as a novel member o f 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.  F r o m 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 o f this protein, putative models have been proposed based on hydropathy profiles and sequence similarities to members of the A B C superfamily.  other  Fig 10 shows putative models o f the A B C R proposed  by Illing et al. (1997) and supported by Bungert et al. (2001) and that o f A z a r i a n and Travis (1997) and Sun et al.  (2000).  In both models  ABCR  consists o f a single  polypeptide chain arranged in two tandem halves, each containing a N B D proceeded by a TMD.  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 o f A B C R (Fig 10, B), the highly hydrophobic segment ( H H 1 ) is thought to partially enter the membrane and form a hairpin loop. This H H 1 segment in the model of Illing et al. and Bungert et al. is regarded as the seventh transmembrane segment. The result o f H H 1 spanning the membrane instead of looping in the membrane, is that the second intradiskal loop in Fig 10, A is located inside o f the disk (in the lumen) whereas in the other model it is in the cytoplasmic side o f the disk. Bungert et al. support their model by  locating four /V-glycosylation sites on each o f the putative hydrophilic intradiskal  loops.  In addition, their study also provided evidence that the two tandem halves o f the  A B C R interact through disulfide bonding.  26  Fig 10.  Putative topological models of ABCR.  ( A ) A representation o f the model  proposed by Illing et al. (1997) and supported by Bungert et al. (2001).  T h i s model is  characterized by two large intradiskal loops that are located in the disk lumen. The highly hydrophobic segment ( H H 1 ,  in white) between the first T M D and the second T M D is  envisioned as the seventh transmembrane segment.  (B) T o p o l o g i c a l model o f  proposed by A z a r i a n and Travis (1997) and Sun et al. (2000).  ABCR  In this model the H H 1  loops in and out o f 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,  27  2001.  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 o f retinoids. The accumulation o f lipofuscin materials seen in Stargardt's disease may then be the result of a nonfuctional protein.  A l t h o u g h the exact function o f A B C R  is currently under  investigation, different studies (below) have shed light on the putative function and the environmental requirements o f this protein.  In 1999, Sun et al. investigated the A T P a s e activities o f purified and reconstituted A B C R in the presence o f a number o f compounds, including various geometric isomers o f retinal. The basis o f this experiment was that i f A B C R (in the presence o f A T P ) encountered its substrate or its allosteric activator, its A T P a s e activity would be stimulated as it transports its  substrate  across  the  membrane.  Similar  studies  conducted  for  CFTR  glycoprotein had been successful in identifying their substrates (Bear et al, and L i n g , 1994).  A s a result o f the study by Sun et al  and P-  1992; Shapiro  (1997) retinoids, especially all-  /rarcs-retinal, were identified as possible substrates o f A B C R .  Shortly after, W e n g  et al. (1999) reported the generation o f 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 o f both  phosphatidylethanolamine and protonated /V-retinylidene-PE (a complex  of  aW-trans-  retinal and P E ) in R O S , and accumulation o f lipofuscin materials in R P E . observations  along with the  clinical features  of  Stargardt's disease,  accumulation o f lipofuscin in R P E , led the authors to suggest A B C R  These  especially  the  as a putative  transporter o f all-^nms-retinal or /V-retinylidene-PE.  In a third study, A h n et al. (2000) measured the A T P a s e and G T P a s e activities o f purified bovine A B C R reconstituted in different lipid environments and in the presence o f different retinoids. A s a result o f 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 A T P a s e activity o f purified and reconstituted bovine A B C R .  Following these studies two models for the function o f A B C R have been suggested (Fig  11, A and B). B o t h A and B o f Fig 11 show the release o f all-trans-retinal following photoexcitation o f rhodopsin.  In A , A B C R acts as a transporter and actively extrudes  retinal from the disk membrane, making it available to Panel  all-trans-retmol dehydrogenase.  B shows that some o f 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 o f 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 o f 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 o f lipofuscin.  The death o f R P E and  photoreceptors would shortly follow, resulting in Stargardt's disease.  1.10  THESIS INVESTIGTION  A t the start o f this project, Lewis  et al. (1999) published a list o f 89 variations in ABCR  gene that were considered disease causing. The goal o f this project was to study the effect of five o f these A B C R variants on the structure and possibly the function o f A B C R .  The  first step toward this goal was the selection o f the mutations, followed by the  generation o f 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 o f the mutants relative to the w i l d type protein were revealed by western blotting. Furthermore, the localization patterns o f the expressed proteins (mutants and w i l d type) to different  compartments  in  the  C O S cells  were  examined  by  immunofluorescence  microscopy. Finally, the degree o f interactions between the w i l d type or mutant  30  ABCRs  Fig 11. Possible functions of ABCR. extruder.  ( A ) Function o f A B C R as an all-trans-retinal  After being released from rhodopsin (not shown) all-zra/w-retinal is actively  extruded from the disk membrane by A B C R dehydrogenase.  and becomes exposed to all-frtws-retinol  Once reduced to a l l - / r a / M - r e t i n o l , it enters the visual cycle. (B) Function  of A B C R as a retinylidene-phosphatidylethanolamine flippase.  After the release o f all-  trans-xttina\ from rhodopsin (not shown), some o f 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 dehydrogenase.  side o f the  disk membrane  is  not  accessible to  all-/ra/M-retinol  A B C R acting as a flippase, flips the VV-retinylidene-PE that forms on the  inner leaflet o f the disk to the outer leaflet where it can become accessible to retinol dehydrogenase.  all-trans-  The flipping action o f 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 W e n g et al,  31  1999.  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 o f mutant proteins in comparison to the wild type A B C R .  The next part o f the study was to investigate the effect o f the selected mutations on the A T P hydrolysis activity o f 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 ROS  lipids or brain polar lipids and D O P E  vesicles.  The A T P a s e  activity o f the  reconstituted proteins (wild type and mutants) were measured in the presence and absence of  a\\-trans-ret'ma\, the possible substrate o f A B C R .  T h e nucleotide binding abilities o f  four o f the mutants were also briefly investigated by 8 - a z i d o - A T P labeling.  33  M A T E R I A L AND METHODS  2.1  M A T E R I A L S 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. W i l d type A B C R c D N A was previously inserted into pRK5 plasmid by J. Ahn. The primers designed for both P C R and sequencing were synthesized by Gibco B R L . The Pfu Turbo D N A polymerase and the QuickChange Site-Directed Mutagenesis Kit 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 P C R purification kit, QIAprep Spin Miniprep K i t and the Q I A G E N Plasmid Maxiprep K i t were all from Qiagen. AW-trans-xeXm&X, soybean phospholipid (CHAPS)  and  were  3-[(3-cholamidopropyl)dimethylammonio]-l-propanesulsonic purchased  from  Aldrich,  while  brain  polar  lipids,  acid 1,2-  dioleoylphosphatidylethanolamine (DOPE) and l-stearoyl-2-docosahexaenoylphosphatidyl ethanolamine (SDPE) were from Avanti Polar Lipids. 32  8-azido-[a- P]ATP and [a32  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 C Y 3 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 , lx 2  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 o f 1 u.1 o f 20 m M d N T P s , 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 o f 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 Q I A q u i c k G e l Extraction K i t and eluted in 50 u.1 water. The D N A concentrations were either determined spectophotmetrically or by running on agarose gel along side L a m d a / H i n d l l l ladder ( M B I Fermentas).  During ligation reactions the vector and the insert were added in the molar ratio o f 1:2 to l x T ligation buffer and 1 ul o f bacteriophage (T ) D N A ligase. 4  4  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 ( D H 5 a ) cells were prepared using CaCl2 method outlined in Sambrook et al. (1989). T o transform the competent cells, 45 ul o f 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 o f 0.2 m l o f L B , the cells were allowed to incubate for 1 hour at 3 7 ° C while shaking (300 rpm). F o l l o w i n g the incubation, 250 ul o f the cells was spread on agar L B plates containing 50 u.g/ml o f ampicillin. The plates were left overnight at 3 7 ° C in an incubator. T o screen the cells for the desired construct, single colonies from agar plates were inoculated in 2 m l o f L B / a m p i c i l l i n 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 K i t . T o obtain larger amount o f D N A (1-2 jag) a maxiprep was done on 500 m l o f E. coli culture using the manufacturer's instructions. T o 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 o f the E. coli cells containing the desired recombinant plasmid were prepared by mixing 0.85 m l o f cultured cells with 0.15 m l o f sterile glycerol. glycerol stocks were stored at - 8 0 ° C .  36  The  2.3  G E N E R A T I O N OF 3F4 C O U P L E D SEPHAROSE BEADS  The R i m 3F4 antibody against an epitope near the C-terminus of bovine A B C R generated and described by Illing et al,  1997.  was  Approximately 1 mg o f antibody per 1 m l  o f beads was dialysed in three changes o f 20 m M borate buffer p H 8.4, at 4 ° C over a three day period. The Sepharose 2 B 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). activated beads were washed four times with cold borate buffer (20 m M , p H 8.4)  The by  centrifugation in a table top centrifuge for 3-5 m i n , and subsequently incubated with the 3F4 antibody at 1 mg/ml beads.  After 2-4 hours o f incubation at 4 ° C on a rotating wheel,  the 3F4-beads were washed twice with Tris Buffered Saline ( T B S : 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). 0.01% N a N  2.4  3  The 3F4-beads were stored in an equal volume o f T B S and  at4°C.  G E N E R A T I O N OF T H E CONSTRUCTS  W i l d 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 ABCR/pRK5.  The R 2 0 3 8 W , R 2 0 7 7 W and R 2 1 0 6 C 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 A B C R / b l u e s c r i p t 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 w i l d type A B C R / p R K 5 constructs, also cut with and Aflll.  Hindlll  Once inserted, the w i l d type and the three mutant A B C R D N A s were cut out o f  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 o f the two restriction sites in the final constructs.  37  T 1 5 2 6 M mutant was generated by using the QuickChange Site-Directed Mutagenesis K i t (Stratagene) and by following the manufacturer's guidelines. T w o 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 ( P C R s ) were carried out using 10, 20 and 50 ng o f template w i l d type A B C R / p C D N A 3 .  E a c h reaction was carried in 50 ul volume and  contained: 250 u M d N T P s , 125 ng o f each o f the primers, 5 pi o f lOx Pfu buffer and 2.5 units o f  Pfu Turbo. A negative control reaction was also carried out in which one o f the  primers was missing.  T h e 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 m i n (to denature the D N A double helix). 2- 9 5 ° C for 30 seconds (to denature the helix at the beginning o f each cycle), 5 7 ° C for 1 m i n (to allow annealing o f the primers) and 7 2 ° C for 30 m i n (to allow extension o f the new strand by the polymerase). 10 minute extension at 7 2 ° C .  T h e cycle was repeated 15 times followed by a final  While the samples were stored at 4 ° C , usually 10 u.1 was  removed and analyzed on a 1% agarose gel for the presence o f the amplified plasmids. T o differentiate between the o l d 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 o f  Dpnl restriction enzyme (10 units). This enzyme digests the methylated template  D N A s and thus allows the direct use o f the resulting P C R products in the transformation o f competent  To  E. coli cells.  generate D 8 4 6 H , 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 o f 6 2 ° C ) and A Z 2 A - (reverse primer having a predicted melting temperature o f 6 7 ° C ) .  T h e A Z 2 A - primer contained the nucleotide change G  instead o f 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 o f the mutation, just 5' to an EcoRV site (Fig 12).  The P C R reaction contained: 10 ng o f template, 250 ng o f  both A Z 1 + and A Z 2 A - , 250 u M d N T P s , l x  Pfu buffer and 2.5 units o f Pfu Turbo. After a  first heat-denaturation at 95 ° C for 1 m i n , the following P C R cycle was repeated 25 times: heat-denaturation at 9 5 ° C for 1 m i n , annealing o f primers for 1 m i n at 5 7 ° C , extension by the D N A polymerase for 2.5 m i n at 7 2 ° C . After the 25 rounds o f P C R , the amplified  38  iJ fi  %  t t t  o  o  <  o o <  fi c3  Si < u a  < O  o <  <  a  H H <  <  <  a o  o < o U o u u o < a u < H  cr I CO  s  1  u <  o o < a u o  c  t—  o u o < o <  o  6  U U H O  <  0  a HI  < o u  < < H H  9  Co  o  H O H H U H  o o <n u H  <l U  u  H O H O H H O H  < o < o o u  fi  l-fi c o  en <U  o, di o fi <u fi cr 0) C/3 >H  1)  CCJ  0)  0>  o  I  < 01  N  +  cn N  ro N  << <  c o CCJ +-» fi  cn CS  <  o  u u < o < a  +  U U H H H <  fi c  I-a fi fi CD  "Cf  00 Q  uo  l-fi H  39  rat  5  -  >  AZ1 +  M  |»  m  GH  1 1  5 ' — — 3'  m  C-l  EcoRV  M  3' 5'  Kpnl  (546)  (2649)  1- P C R 2- Cut with EcoRV and Kpnl  c HI EcoRV  Q—j  2103 bp  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 o f the A B C R  upstream o f EcoRV site. fragment was cut with  cDNA,  After 25 rounds o f polymerase chain reactions, the amplified  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 o f  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 Q I A q u i c k P C R Purification K i t . 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 D 8 4 6 H mutation back into an equivalent fragment in the w i l d 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 o f this construct needed to be removed. A B C R / p C D N A 3 both with  T h i s was done by cutting  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.  C O S - 1 or C O S - 7 cells were grown in 5% CO2 and  3 7 ° C , in Dulbecco's M o d i f i e d Eagle M e d i u m ( 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"). passaged almost twice a week as follows.  T o maintain the cell line, the cells were  First the old medium was removed and the cells  were washed in 2 m l o f t r y p s i n - E D T A . Then, to detach the cells from the 10 c m dish, 1ml of  trypsin-EDTA  was  added and the  cells were incubated for 5 minutes  at  37°C.  Sometimes the C O S cells remained attached to the dish, and a gentle tapping o f the dish was required to dislodge them.  T o stop trypsinization, 9 m l o f new complete D M E M was  added. The cells were re-plated at a dilution o f 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 sites o f the p C D N A 3 mammalian C O S cells.  (Invitrogen)  or the modified p C D N A 3 ,  One o f the transfection methods  NotI blunted Xbal  were used to  transfect  employed was the calcium  phosphate method of Chen and Okayama (1987). The cells were plated at a concentration  41  of 6x10 cells per 10 c m dish the night before the transfection. The next day, 30 jj,g o f the 5  ABCR/pCDNA3  or mutant A B C R / p C D N A 3  final volume o f 372 ul.  construct was diluted with sterile water to a  Then, 123 ul o f I M C a C l  mixed. While gently vortexing the D N A - C a C l bis-(2-hydroxyethyl)-2-aminoethanesulfonic  Na2HP04,  2  2  was added to the diluted D N A and  mixture, 495 ul o f 2x B B S (50 m M N , N -  acid  (BES),  280  mM  NaCl,  1.4  mM  p H 6.95) was added dropwisely. The mixture was incubated for 20 m i n at room  temperature and then added to the C O S cells while gently swirling the plate. were incubated at 3 7 ° C and 5% C 0  2  overnight.  The cells  The following morning the cells were  washed once with 10 ml o f filtered phosphate buffered saline ( P B S ) containing 5 m M E D T A , and incubated for another 24 hours at 3 7 ° C , 5% C 0  2  in 10 m l o f 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 o f 50 i l l . T o further dilute the D N A , 500 ul o f D M E M (without serum and antibiotics) was added.  Then  60  ul  of  SuperFect transfection  reagent  (Qiagen)  was  the  DNA/DMEM  dropwisely and the mixture was  temperature.  Meanwhile, the adherent C O S cells at 70-90% confluency per one 10 c m  dish were washed with 10 m l o f sterile P B S . DMEM  was  added  to  the  allowed to sit for 5-10  added to  m i n at room  After the incubation, 3 m l o f complete  DNA/SuperFect  mixture  and  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  resulting  The cells were  allowed to incubate for 2 hours at 3 7 ° C and 5% C 0 , before being washed with P B S . 2  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 o f 5% C 0 . 2  2.5.4  Harvesting the COS cells  Forty-eight hours post-transfection, each 10 c m C O S cell dish was first washed in P B S 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 m i n at  4 ° C and the pellet resuspended in 50-100 ul o f P B S .  42  The resuspended cells were then  added dropewisely to 0.5 m l o f solubilization buffer (50 m M N a H E P E S , p H 7.5, 100 m M N a C I , 1 mg/ml sonicated soybean phospholipid, 18 m M C H A P S , D T T and 3 m M M g C l  2  10% glycerol, 1 m M  and l x complete protein inhibitor cocktail).  carried for 30-40 m i n on ice with frequent, yet gentle vortexing.  Solubilization was  The solubilized cells  were finally centrifuged at 40 000 rpm in T L A 100.4 rotor (Optima T L Ultracentrifuge, Beckman) for 10 m i n at 4 ° C . Approximately 50 ul o f 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 U l t r a f r e e - M C spin columns.  o f R i m 3F4-Sepharose 2 B beads, in 0.4  u M filter  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 NaHEPES,  p H 7.5,  100 m M N a C I , 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. columns.  The beads were then washed 6x with 0.4 m l o f buffer A in spin  T o elute the w i l d type or the mutant A B C R s , 30 ul o f 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 o f buffer A was  also carried out to yield a final volume o f 90 u.1 o f eluate. were removed and added to 10-15  The eluted fractions (30-45 ul)  ul o f 4x S D S 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 o f a 6well plate.  While the rest o f 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 m i n in 4% paraformaldehyde made in P B S .  E a c h well containing a glass slide was washed three  times (5 m i n per wash) with P B S supplemented with 0.1 m M C a C l , 1 m M M g C l 2  m M glycine.  2  and 10  Next, the cells were both blocked and permeabilized for 15-30 m i n in 100  m M sodium phosphate buffer, p H 7.4, containing 0.3% T r i t o n - X - 1 0 0 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%) T r i t o n - X - 1 0 0 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 m i n 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  supplemented with 0.1% T r i t o n - X - 1 0 0 and 2.5% goat serum.  buffer  Finally the cells were  washed twice in sodium phosphate buffer and mounted by adding a few drops o f M o w i o l mounting solution ( C a l b i o c h e m ® ) onto the glass cover slips.  E a c h cover slip was then  placed face down on a microscope slide and fixed in position by adding transparent nail polish around the periphery o f the glass cover slip. After the nail polish was dried, each microscope slide was viewed using a Zeiss A x i o p l a n 2 microscope equipped with a digital imager.  2.6  COS C E L L M E M B R A N E PREPARATION  T o prepare C O S cell membranes from transfected cells, ten 10 c m dishes were transfected per plasmid D N A construct using the calcium phosphate method o f transfection (section 2.5.2).  T w o days following transfection, the cells were scraped in 2 m l o f 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 m l and centrifuged at 2800 g for 10 m i n at 4 ° C .  The cells were washed in 12 m l o f  the same buffer and re-centrifuged again at 2800 g for 10 m i n , 4 ° C .  T o lyse the cells, the  pellet was resuspended in 12 m l o f the hypotonic buffer containing l x Complete Protease Inhibitor Cocktail (Roche Diaganostics, L a v a l , Quebec) and left on ice for 1 hour.  The  cells were further disrupted by both using a D u a l l glass homogenizer 15 times and by passing the homogenate 3x through a 26-gauge needle.  T h e solubilized C O S cells (6ml)  were then layered on top o f an ice cold step sucrose gradient made up o f 5 m l o f 5% sucrose (w/v) and 6ml o f 60% sucrose in gradient buffers (10 m M T r i s - H C l , NaCI, 1 m M M g C l , 2  150 m M  1 m M C a C l , 0.1 m M E D T A ) . The samples were centrifuged in a 2  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 o f 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 T L A 1 0 0 . 4 rotor (Beckman). The membrane pellet was resuspended i n 500 ul o f 5% gradient buffer, 10% glycerol and l x Complete Protease Inhibitor Cocktail. T h e membrane fractions were stored at - 8 0 ° C .  2.7  BOVINE ROD O U T E R SEGMENT PREPARATION  Bovine R O S were prepared under d i m red light as follows.  One hundred retinas were  gently inverted in 40 m l o f 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 , and 20 m M T r i s - H C l , p H 7.4. 2  were passed through a Teflon filter (300 um mesh).  The retinas  The resulting filtrate was divided and  layered on top o f six 28-50% (w/v) sucrose gradients.  T h e samples were centrifuged in a  S W 2 8 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 m l o f 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 m l o f homogenizing buffer and pooled together, and an extra 2 m l was used to rinse the centrifuge tubes.  The resulting 8 m l o f R O S in  homogenizing buffer was divided into 1 m l aliquots and stored at - 8 0 ° C .  45  The protein  concentration was determined by bicinchoninic acid ( B C A ) assay and was approximately 5-8 mg/ml.  2.8  IMMUNOAFFINITY PURIFICATION OF A B C R 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. washing steps were carried under the d i m light.  All  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 m i n 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 l x Complete Protease Inhibitor Cocktail for 2030 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 o f a 50% suspension o f R i m 3F4Sepharose 2 B beads, pre-equilibrated with buffer A (50 m M N a H E P E S , p H 7.5, 100 m M N a C I , 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  E X T R A C T I O N OF PHOSPHOLIPIDS F R O M ROS  Phospholipids were extracted from R O S by following the method o f F o l c h et al. (1957) and when feasible, steps were performed under nitrogen gas.  Organic solvents such as  chloroform and methanol contained 50 |J,g/ml o f 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 ( p H 7.0) by centrifugation at 40 000 rpm in a T L A 100.4 rotor for 10 m i n . The membranes were resuspended to a final volume o f 200 ul with 10 m M potassium phosphate.  Approximately 200 ul o f 1 M NH2OH (adjusted to a  p H 7.0 by the addition o f 1 M N a H C 0 ) and 930 u.1 o f methanol (containing B H T ) were 3  added to the resuspended R O S membrane.  After 10 m i n o f incubation on ice, 1 m l o f  water and 1.87 m l o f 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  m i n at ~ 400 g.  The organic phase was removed and  washed with 1.4 m l o f 0.3 M N a C l in water and 930 ul o f methanol, before being dried under nitrogen gas.  After the organic solvent was evaporated, the lipids were resuspended  in 150 ul o f 1:1 solution o f methanol/chloroform containing 50 ug/ml o f B H T and applied to a 0.5 m m Silicagel G , thin layer chromatography plate (Merck) under nitrogen gas.  The  plate was developed in 1:1 hexane/ether solution, i n the presence o f 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 m l o f 1:1  methanol/chloroform solution (containing 50 |J.g/ml o f B H T ) , vortexing, and separating the phases by centrifugation at low speed in a table top centrifuge.  These steps were repeated  once  were  more  and the  resulting  R O S extracted  phopholipids  analyzed  for  their  phosphorus content by using the method o f Z h o u and Arthur (1992).  2.10  DETERMINATION OF LIPID PHOSPHORUS C O N T E N T  The yield o f phospholipids extracted from R O S (section 2.9) malachite green method o f Z h o u and Arthur, 1992.  was determined by the  Briefly, different volumes (10, 20, 30,  etc) o f the R O S extracted lipids were added to 13 x 100 m m glass tubes and dried under N gas.  In the fumehood, 100  standard tubes.  2  jul o f perchloric acid was added to all tubes including empty  E a c h tube was sealed with aluminum foil and heated to ~ 1 4 0 ° C for 1-3  hours or until the yellowish color o f the samples turned colorless.  Phosphorus standards,  containing different nmole amounts o f phosphorus, were prepared in duplicate by adding the required amount o f 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 u l , while 300 u l o f water was added to the sample tubes. T h e working solution (2 ml), prepared from solutions I, II and III (below), was added to each tube and mixed. After 30 m i n incubation at room temperature, the absorbance o f each solution at 660 nm was measured and the amount o f phosphorus (nmol/ul) in the R O S extract was determined by a standard curve.  T h e concentration o f phopholipids in the R O S extract (mg/ml) was then  determined by taking into account the volume molecular weight o f phospholipids (775 g/mol).  47  o f lipids extracted  and by using  the  W o r k i n g solution: 3 volumes of solution I, 1 volume o f solution II and 0.016 volume o f 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 A B C R PROTEINS INTO LIPID VESICLES  Purified  wild type,  mutant  and R O S A B C R s  were reconstitued  into  lipid  vesicles  consisting of R O S extracted lipids or a 50:50 mixture o f brain polar lipids with either D O P E or S D P E .  T o produce the 20 mg/ml R O S lipid solution, a known amount o f R O S  extracted lipids were dried under N  2  gas and resuspended i n the required volume o f buffer  C (25 m M H E P E S , p H 7.4, 140 m M N a C l , and 10% glycerol).  T o produce the 20 m g / m l  of a mixture o f 50:50 brain polar lipid and D O P E or S D P E , first 8 mg o f each o f brain polar  lipids  and  DOPE  or  SDPE  were  dissolved  in  0.4  ml  of  1:1  mixture  of  chloroform:methanol. Then, 0.2 m l o f each o f the two lipids were mixed together and dried under N  2  gas.  The lipid mixture was dissolved in 0.4 m l o f 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 o f the 20 m g / m l lipid solution was  mixed with 3 ul o f 15% o c t y l - P - D -  glucopyranoside (w/v) in buffer C . Then, 24 u l o f either the purified w i l d type or mutant or R O S A B C R was added to the lipid-detergent mix and incubated for 30 m i n on ice. Similarly, 24 u l o f the eluate from untransfected cells and 24 u l o f reconstitution buffer (termed buffer blank) were added to the lipid-detergent mix, acting as negative controls. After the incubation, 200 ul o f 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 r o o m temperature was added and the resulting mixture was incubated for a further 2 m i n .  Immediately following the  two  minutes of incubation, the mixture was passed through 200 ul o f packed Extracti-Gel D resin in a M o b i c o l 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 m l  48  syringe.  Before carrying out the A T P a s e assay, M g C l  2  was added to the reconstituted  samples to a final concentration o f 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 m l microcentrifuge tubes and kept on ice while the lOx solutions o f 0.5 m M A T P and 0.5 m M all-fr^ms-retinal (in ethanol) were being prepared fresh. uCi  The 0.5 m M A T P solution contained: 0.5 m M A T P and either 0.1  (for overnight exposures) or 0.2  reconstitution buffer.  u C i o f [ a - P ] A T P (for 3 hours exposures) in 3 2  T o the 8 ul o f reconstitued samples, 1 ul o f the 0.5 m M  a\\-trans-  retinal or 1 ul o f the ethanol blank (produced along side the 0.5 m M retinal solution, but instead o f retinal contained ethanol), was added and the resulting mixture was incubated on ice for 10 min. The reaction was initiated by the addition o f 1 ul o f the 0.5 m M labeled A T P solution and was carried out for 30 m i n at 3 7 ° C before being stopped by 4 ul o f 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 L i C l / 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 o f hot A T P used (0.2 u C i versus 0.1 u C i ) . 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 o f the reconstituted wild type, mutant and R O S A B C R proteins. After scanning the phosphor screen, IPLab G e l analysis Software (Signal Analytics C o r p . , V i e n n a , V A ) was used to quantify the spots corresponding to A D P and A T P . original A T P  0  (ATP  0  The ratio o f A D P to the  = A D P + A T P ) was first calculated for all o f the samples, then the  buffer blank was subtracted from all samples.  49  Since each sample in the assay was carried  out in triplicate, an average value for each sample was obtained. The total nmole o f 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 o f 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 o f reconstituted samples / 8 ul) x (volume o f eluate / 24 ul) x (volume o f total cell lysate / 500ul) Next  the total nmoles  o f A T P hydrolyzed per mg o f total proteins  (determined by B C A , section 2.15) was determined for each sample.  in cell  lysate  The values obtained  for the untransfected cells, which acted as background values, were subtracted from the rest o f the w i l d type and mutant samples.  The nmoles o f A T P hydrolyzed in the  reconstituted sample was then calculated as follows: nmoles A T P hydrolyzed per reconstituted sample = (total nmoles o f A T P hydrolyzed / mg of total protein) x total mg o f protein x ( 500 ul / total cell lysate) x ( 24 ul / eluate volume) Finally, the total nmoles o f A T P hydrolyzed per minute in the reconstituted sample was determined by dividing the nmole o f 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 o f protein present in each o f the reconstituted samples represented the specific activities (i.e. the nmole o f A T P hydrolyzed per minute per mg o f protein).  2.13  AZIDO-ATP L A B E L I N G OF COS M E M B R A N E S  2.13.1 Membrane preparation Azido-ATP  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 w i l d 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 m i n in a T L A 100.4 rotor (Optima T L Ultracentrifuge, Beckman).  The pellet was resuspended in 0.5 m l o f the same buffer, allowed to incubate  for 20 m i n (to lyse plasma membrane) and re-centrifuged for 12 m i n at 30 000 rpm. T h e  50  pellet was resuspended in 120 ul o f assay buffer (20 m M Tris, 0.15 M N a C I , 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 o f 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 l x Complete Protease Inhibitor Cocktail, by centrifuging in T L A 100.4 rotor for 12 m i n at 30 000 rpm. The membrane pellet was resuspended in 60 ul o f assay buffer (20 m M Tris, 0.15 M N a C I , 5 m M M g C l , p H 7.4). 2  2.13.2 Azido-ATP assay The 8-azidoadenosine-5'-triphosphate  [a-  P] (20 C i / m m o l solution), at 1.5 ul per sample  o f 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 N a C I , 5 m M M g C l , p H 7.4) per 2  sample.  Then, 6 ul o f 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 L a m p ; 254) at a distance o f 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  o f 20 m M Tris, by centrifuging them in a T L 4 5 rotor at 30 000 rpm for 15 min.  The  samples were resuspended in 30 ul o f 20 m M Tris before being solubilized for 30 m i n on ice in 200 ul o f 2% T r i t o n - X - 1 0 0 , 20 m M Tris, 0.15 M N a C I and l x Complete Protease Inhibitor Cocktail.  The solubilized membranes were centrifuged in a T L A 100.4 rotor at  40 000 rpm for 10 m i n , 4 ° C . The resulting supernatant was removed and incubated for 1 hour at 4 ° C with 150 ul o f a 50% suspension o f R i m 3F4-Sepharose 2 B beads, preequilibrated with 0.2% T r i t o n - X - 1 0 0 , 20 m M Tris, 0.15 M N a C I .  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 m l o f 0.2% T r i t o n - X - 1 0 0 , 20 m M Tris, 0.15 M N a C I .  The bound  proteins were eluted twice in 40 ul o f 0.2% T r i t o n - X - 1 0 0 , 20 m M Tris, 0.15 M N a C I containing 4% S D S and l x Complete Protease Inhibitor Cocktail. O f the 80 ul o f eluates, 40  ul was  mixed with  13  ul o f 4x  S D S sample  buffer and loaded on one  polyacrylamide gel, while the other 40 ul was loaded on a second gel.  51  6.5%  S D S - P A G E was  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 m i n 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 - r a y film for 2-4 days or a phosphor storage screen overnight and scanned in a Phosphorlmager (Molecular Dynamics).  2.14  PROTEIN ELECTROPHORESIS AND W E S T E R N BLOTTING  Sodium dodecyl sulfate polyacrylamide gel electrophoresis carried out using the gel apparatus from Hoefer.  ( S D S - P A G E ) was routinely  Samples were prepared by m i x i n g with  the appropriate amount o f 4x S D S sample 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% P-mercaptoethanol). samples were applied to electrophoresis mA/gel.  each  lane  o f a 6.5%  or a 7.5%  Typically, 10-45 polyacrylamide gel  ul o f and  was carried out using the L a e m m l i buffer system (1970) at 150 V , 27  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 m i n (25 m M Tris, 195 m M glycine, methanol, p H 8.3).  10%)  The proteins on the gel were then transferred onto the Immobilon P  membrane (Millipore) at 300 m A for 40 m i n in a semidry transfer apparatus (Biorad) using the transfer buffer.  After the Immobilon was soaked in methanol followed by 5 m i n in  P B S , it was blocked in 1 % skim m i l k / P B S containing 0.05 % Tween 20 ( P B S - T ) for 30 min.  The membrane then was incubated with the primary antibody diluted in  m i l k / P B S for 1 hour. monoclonal (1:20  0.1%  The primary antibodies used were either N-terminus R i m 5B4  dilution) or anti-calnexin Ig (1:1000).  times for 10 m i n each with P B S - T .  The blots were washed three  T h e secondary antibody, a goat anti-mouse Ig or a  goat anti-rabbit Ig linked to horseradish peroxidase, was diluted 1:5000 in 0.1 % m i l k / P B S -  52  T.  After 1 hour o f 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.  T o strip a blot o f antibody, the membrane was incubated at 5 0 ° C in stripping buffer (100 m M (3-mercaptoethanol, 2% S D S , 62.5 m M T r i s - H C l , p H 6.7) for 30 m i n , while rotating. The membrane was washed twice in P B S - T for 10 m i n and blocked in 4% skim m i l k / P B S for  lhour followed by a second incubation in 1% skim m i l k / P B S 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 o f proteins in R O S membrane preparations was determined by B C A method (Pierce), using different concentrations o f a stock B S A solution (2mg/ml). concentrations in samples termed "total cell lysates" in section 2.5.4 solubilizing the transfected C O S cells) were also determined by B C A .  The protein  (obtained after The amounts o f  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.  B y comparing the intensities o f Coomassie stain o f the  eluted A B C R bands with those o f B S A standards, the amounts o f purified proteins were determined. The amounts o f reconstituted proteins were calculated from the amount o f purified proteins present in the eluates, while taking into account a number o f dilutions. Furthermore, to quantitate the amount o f mutant A B C R expressed in the C O S cells, western blots o f total cell lysates were scanned with the Ultroscan X L laser densitometer and expressed as a percentage o f the w i l d 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) D 8 4 6 H : This is a missense mutation in the putative sixth transmembrane segment, in which a negatively charged histidine.  charged residue  (aspartate) is replaced with a  neutral/positively  This mutation is o f interest since aspartate is the only charged residue  in the middle o f the putative sixth transmembrane segment, a relatively hydrophobic region.  Thus this substitution may play a significant role in the folding or function o f  ABCR.  ii) T 1 5 2 6 M : This is a missense mutation in which two polar residues are exchanged. mutation was chosen due to its location.  This  A c c o r d i n g to the current A B C R model o f 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 A z a r i a n and Travis (1997) and Sun et al. (2000), however, places this mutation on the cytoplasmic side o f the disk membrane (Fig 10, B).  Thus, this mutation  may help in supporting one model versus the other.  iii) R 2 0 3 8 W and iv) R 2 0 7 7 W :  These missense mutations involve the replacement  positively charged arginines by tryptophans (an aromatic residue).  of  B o t h o f these mutants  are located in the second nucleotide binding domain ( N B D ) o f A B C R .  These mutations  may provide some information regarding the ability o f A B C R to bind and/or hydrolyze ATP.  v) R 2 1 0 6 C : 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. binding  This mutation is predicted to have a less dramatic effect on the A T P  or hydrolysis,  as  it is  not  located  54  in a conserved  sequence  in the N B D .  T1526M  R2077W  Fig 13. Putative model of ABCR showing the five Stargardt's disease mutations. The location o f the five missense mutations resulting in Stargardt's disease (Lewis et al, 1999) selected for study are shown. D 8 4 6 H missense mutation occurs in the putative sixth transmembrane segment. T 1 5 2 6 M is located in the second intradiskal loop in this model. R 2 0 3 8 W , R 2 0 7 7 W and R 2 1 0 6 C 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. A h n 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 p C D N A 3 .  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 K i t was used. A s shown in Fig 14, B the entire 12.3 kb plasmids containing the possible nucleotide change were generated from all three P C R 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 P C R (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 T Y P E AND M U T A N T  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  C  B 1  Fig 14.  2  1  2  3  4  Generation of the selected constructs.  A B C R / p C D N A 3 with Apal.  5  (A)  1  2  Restriction enzyme digestion o f  T w o ug o f A B C R / p C D N A 3 was digested with 10 units o f  Apal for 1 hour at 3 7 ° C , and the resulting fragments were separated by 1% agarose gel electrophoresis (lane 2). The size o f the fragments following Apal digestion o f ABCR/pCDNA3  are: 6.8 kb, 3.3 kb, 2.0 kb and 0.15  ladder was used in mutatgenesis.  kb (not visible).  Lamda/Hindlll  lane 1. (B) Generation o f T 1 5 2 6 M / p C D N A 3 construct by site directed  Different amounts o f template A B C R / p C D N A 3  were used in three P C R  reaction mix. After 16 rounds o f P C R , 10 ul from each mixture was removed, loaded on a  lane 1, Lamda/Hindlll ladder; lane 2, amplified plasmid from 10 ng o f template D N A ; lane 3, amplified plasmid from 20 ng o f D N A ; lane 4, amplified plasmid from 50 ng o f template D N A ; lane 5, control reaction: 1% agarose gel and subjected to electrophoresis;  missing  one primer.  mutation.  (C)  A m p l i f i c a t i o n o f the 2.1  kb fragment  bearing the  T e n ng o f A B C R / p C D N A 3 was subjected to 25 rounds of P C R .  resulting product was loaded on 1% agarose gel; D 8 4 6 H fragment o f 2.1 kb.  57  D846H  T e n (al o f the  lane 1, Lamda/Hindlll ladder; lane 2,  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  FROM  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 R i m 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  205116_ 97664529-  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), D 8 4 6 H (lanes 3), R 2 0 3 8 W (lanes 4), R 2 0 7 7 W (lanes 5), R 2 1 0 6 C (lanes 6), T 1 5 2 6 M (lanes 7) were solubilized in 18 m M C H A P S .  T h e insoluble fractions were  removed and the supernatants were loaded onto 7.5% S D S - P A G E stained  with  Coomassie  Brilliant  Blue; (B) Proteins  gels.  ( A ) Proteins  transferred to an Immobilon  membrane for western blotting using R i m 5B4 antibody for detection o f A B C R .  Thirty ul  o f the supernatant was loaded in ( A ) , 10 u l o f the supernatant in (B). The size o f the molecular weight markers are shown in k D a .  59  Table 3 Expression levels of mutant ABCRs relative to wild type ABCR in COS cells A B C R variant  Expression level  W i l d type  100 %  D846H  C  31 ± 1 %  D846H  d  68 ± 16 %  R2038W  59 ± 21 %  R2077W  57 ± 20 %  R2106C  71 + 17 %  T1526M  81 ± 9 %  a  With the exception o f the D 8 4 6 H variant, values compiled are the averages o f both methods o f transfection (SuperFect and CaPC^j) obtained from nine experiments.  b  Expression levels o f the proteins are cacluated as the percentage (%) o f w i l d type.  c  The expression level o f D 8 4 6 H obtained from four experiments during which SuperFect was used to transfect the cells.  d  The expression level o f D 8 4 6 H obtained from six experiments when CaPCM method o f transfection was used.  60  1  2  3 4  5 6 7  8  9 10 11 12 -ABCR  r-  B  1  2  3  4  BSA  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 o f 3F4beads 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, w i l d type A B C R ; lanes 2, D 8 4 6 H ; lanes 3, R 2 0 3 8 W ;  lanes 4, R 2 0 7 7 W ; lanes 5, R 2 1 0 6 C ; lanes 6, T 1 5 2 6 M . 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.  F r o m a standard curve, it was determined that between 0.3-0.9 u.g o f  proteins could be purified from two 30-40% confluent C O S cell plates. varied from  one  experiment  to  another and strongly depended  efficiency and the purification scheme used. employed here and explained in section 2.5.5 A B C R proteins (wild type and mutants).  These values  on the  transfection  The immunoaffinity purification scheme allowed the recovery o f 40-50%) o f the  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 o f expressed proteins was difficult and often fruitless despite extensive human efforts.  3.5  I M M U N O F L U O R E S C E N C E MICROSCOPY  Immunofluorescence primary  antibody  labeling o f the transfected C O S cells (using the R i m 3F4 as the  and goat  anti-mouse  Ig-CY3  confirmed that all five mutant constructs expressed. somewhat  as  the  secondary  antibody)  further  The labeling patterns, however, were  different between the wild type and the mutants.  The w i l d type proteins  localized mostly in vesicles (red dots in the figure) that were scattered throughout the cells (Fig 17, A). wild  type.  The D 8 4 6 H variant and T 1 5 2 6 M showed different labeling patterns than the 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 , R 2 0 3 8 W and R 2 0 7 7 W did not show the w i l d type labeling pattern, and in addition, their E R / G o l g i labeling seemed slightly different from those o f D 8 4 6 H and T 1 5 2 6 M . That is, the labeling pattern o f these mutants showed filamentous extensions from the E R / G o l g i compartment (Fig 17, C and D ) .  The last mutant, R 2 1 0 6 C , showed  both the vesicular and E R / G o l g i labeling pattern; however, the extent o f vesicular labeling was less than the w i l d type (Fig 17, E).  T o 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.  D 8 4 6 H and T 1 5 2 6 M both showed 90-100% E R / G o l g i whereas 77% o f  R 2 0 3 8 W , 80% o f R 2 0 7 7 W and only 34% o f R 2 1 0 6 C 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. T h e cells were fixed with 4% paraformaldehyde, permeabilized in 0.3% T r i t o n - X - 1 0 0 and labeled with R i m 3F4 antibody. antibody  conjugated to C Y 3 was used  to view  Anti-mouse  the labeled cells by fluorescence  microscopy. Cells in panel A were transfected with w i l d type constructs. Cells in panel B with D 8 4 6 H , panel C with R 2 0 3 8 W , panel D with R 2 0 7 7 W , panel E with R 2 1 0 6 C and panel F with T 1 5 2 6 M construct.  63  64  3.6  C A L N E X I N ASSOCIATION  The amount o f time a mutant or w i l d type protein/polypeptide spends in the E R o f a cell may be an indication o f its folding. T o support the immunofluorescence results, regarding the E R / G o l g i  labeling o f the mutant proteins, blots containing total cell lysates and  purified proteins were stripped and labeled with anti-calnexin antibody (Fig basis o f this experiment was as follows.  18).  The  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 o f calnexin that co-purified with the w i l d type A B C R was very low,  indicating that w i l d type A B C R may not interact with calnexin. W i t h the mutants, on the other hand, more calnexin was co-purified (Fig 18, B) even when the amounts o f purified mutant A B C R s were comparable to the w i l d type (Fig mutants are misfolded and retained in the E R / G o l g i .  18, A).  This suggests that the  It should be mentioned that the  amount o f T 1 5 2 6 M purified was larger than the other mutants.  This may be the reason  why more calnexin co-purified with this mutant. R 2 1 0 6 C 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 A T P a s e activity o f A B C R . The liposomes used here were made from either R O S extracted phospholipids or a 50:50 mixture o f 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 o f - 33% P E , 13% P C , 18% PS and 36%) other phospholipids ( A h n et al., 2000).  These lipids were used in reconstitution  experiments in order to create an environment rich in P E .  T o produce the reducing  environment needed for functional proteins, 1 m M o f D T T was included at all times in all buffers used. The purified proteins were reconstituted into liposomes by the removal o f  65  1 2 3 4 5  6 7 8 9 10 11 12 13  A  -  B  Fig 18. lysates  wm mm m^mm \- Calnexin  Co-purification of calnexin with mutant ABCRs.  (lanes 1-6),  immuoaffinity  ABCR  ROS  ABCRs  from C O S cell  (lane 7) and A B C R proteins purified by (lanes 8-13) were resolved by S D S - P A G E , transferred  membranes  chromatography  onto Immobilon membrane and labeled with R i m 5B4 antibody ( A ) .  The same blot was  stripped and labeled with anti-calnexin antibody (B). T e n ul from each sample was loaded  Lane 1, wild type A B C R ; lane 2, D 8 4 6 H ; lane 3, R2038W; lane 4, R 2 0 7 7 W ; lane 6, T 1 5 2 6 M ; 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 D 8 4 6 H ; lane 11, purified R 2 0 3 8 W ; lane 12, purified R 2 0 7 7 W ; lane 13, purified T 1 5 2 6 M . on the gel.  lane 5,  R2106C;  66  the detergent ( C H A P S ) in the presence o f excess lipids.  The presence o f 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 o f reconstituted proteins in each sample was quantified by densitometry from the standard curve generated  from B S A .  Approximately 30-50 ng o f A B C R  calculated to be present in the reconstituted samples.  were  These values varied between  experiments as the transfection efficiencies varied.  3.8  T H E ATPASE ASSAY  The A T P a s e assay was carried out as outlined in section 2.12. In this study, A T P a s e activity and specific activity were defined as the nmoles o f A T P hydrolyzed per m i n per reconstituted sample and the nmoles o f A T P hydrolyzed per m i n per mg o f protein, respectively.  T o optimize the assay conditions, the A T P a s e 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 o f reconstituted R O S A B C R , in a 50:50 mixture o f 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 o f A T P hydrolyzed/min/mg  (basal) and 80 nmole o f A T P hydrolyzed/min/mg (stimulated).  When R O S A B C R was reconstituted in a 50:50 mixture o f brain polar lipids and S D P E , the basal activity was increased ~ 4 fold (Fig 21, A). for A T P hydrolysis o f R O S A B C R compared to  those o f A B C R  However, when the absolute values  reconstituted in brain p o l a r / S D P E mixture were  reconstituted  in brain p o l a r / D O P E  lipids, they  were  significantly smaller (Fig 21, B).  A s a result, brain polar l i p i d s / D O P E mixture was used  in  The  reconstitution  experiments.  assay  was  also  performed  with  concentrations o f all'/r^ms-retinal in the presence o f 50 u M A T P (Fig 22). the figure, the activity was increased with increasing retinal concentration.  67  increasing  A s shown in  1 2 3 4 5  Fig 19.  6 7  A B C R proteins reconstituted in liposomes.  The purified A B C R proteins (24  ul) were mixed with 12 ul o f 20mg/ml o f solubilized lipids in 15% octyl-glucoside, and allowed to incubate for 30 m i n on ice.  T w o m l o f the reconstitution buffer was added to  the proteolipid mixture, and after 2 min o f incubation, the resulting mixture was passed through 200 ul o f Extracti-gel D resin in a mini-column.  Thirty ul o f the eluates from  wild type A B C R (lane 1), D 8 4 6 H (lane 2), R 2 0 3 8 W (lane 3), R 2 0 7 7 W (lane 4), R2106C (lane 5), T 1 5 2 6 M (lane 6) and R O S (lane 7) were loaded on a 7.5% polyacrylamide gel. F o l l o w i n g 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 o f a  50:50 mixture o f brain polar lipid extract and D O P E . the absence  The A T P a s e assay was performed in  (basal) and presence (retinal) o f 50 m M all-frows-retinal.  represent the mean ± S . D . o f five experiments.  69  The results  Fig 21. Basal and retinal stimulated ATP hydrolysis of ROS A B C R 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 o f brain polar lipids and SDPE. values  (A) Retinal stimulation expressed as % o f basal A T P hydrolysis. (B) Absolute of  basal  and  retinal  hydrolyzed/min/reconstituted sample.  stimulation  expressed  as  nmoles  of  E a c h value is the mean o f two determinations.  70  ATP  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 o f brain polar l i p i d s / D O P E and assayed for the A T P a s e activity in the presence o f 10, 20, 50 and 100 u M o f all-trans-retinal. E a c h value is the mean o f two determinations.  71  ABCRs  reconstituted in lipid vesicles retained their A T P a s e 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. on  western blots.  This was apparent by the absence o f A B C R bands  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 A T P a s e activity o f recombinant w i l d type A B C R purified from transfected C O S cells was also determined. In the presence o f 50 u M  all-trans-retinal, the basal activity o f w i l d  type A B C R reconstituted in a mixture o f brain polar l i p i d s / D O P E , increased by 1.6 ± 0.4 fold.  It should be mentioned, however, that the absolute values for A T P a s e 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 w i l d type protein could be purified compared to R O S A B C R . were measured to be  and  the  from C O S cells  The basal and stimulated specific activities o f w i l d type A B C R  13 nmoles. o f A T P hydrolyzed/min/mg and 22 nmoles o f A T P  hydrolyzed/min/mg, respectively. type  and reconstituted  mutants,  M o s t o f the A T P a s e activity measurements for the w i l d  however,  hydrolyzed/min/reconstituted sample.  were  expressed  as  nmoles  of  This was because the amounts o f purified  ATP ABCRs  were small and scanning the Coomassie blue stained gel by a densitometer did not result in accurate measurements o f protein concentrations. not reliable.  Thus the specific activity values were  A l t h o u g h the basal and retinal stimulated A T P a s e activities varied between  experiments, no significant change or increase was seen when R O S extracted lipids were used instead o f brain p o l a r / D O P E lipids to reconstitute the w i l d 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 o f brain polar lipid and D O P E were assayed along side w i l d type A B C R , untransfected and buffer blank samples.  Measuring the A T P a s e activities o f the mutants proved to be very  difficult, as the basal activity values for the mutants were smaller than the basal activity o f the w i l d type.  Furthermore, as mentioned earlier with the w i l d type samples, the A T P a s e  values o f the mutants differed between preparations and depended on the amounts o f proteins that purified and reconstituted into the vesicles.  Since the amount o f purified  protein differed between experiments and, in turn, depended on the efficiency o f  72  Fig 23. Basal and retinal stimulated ATPase activities of ROS ABCR and wild type A B C R from transfected COS cells.  Purified A B C R from R O S membrane (ROS) and  purified  from  wild  type  ABCR  obtained  solubilized  C O S cells  reconstituted in 50:50 mixture o f brain polar lipids and D O P E .  (WT) were  both  Both the basal (open  columns) and retinal stimulated (solid columns) A T P a s e activities were measured and expressed  as  absolute  values  for comparison.  determinations.  73  Each  value  is  the mean  o f two  the transfection, an experiment during which relatively high amounts o f proteins were purified was chosen as a representative experiment. A s shown in Fig 24, the basal A T P a s e activities of R 2 0 3 8 W and T 1 5 2 6 M resembled the recombinant w i l d type A B C R .  U p o n the  addition of 50 u M retinal, however, they did not display significant retinal stimulation activities, whereas the activity o f w i l d type A B C R increased 1.6 fold.  A s illustrated in  this figure, the D 8 4 6 H and R 2 0 7 7 W variants displayed both basal and retinal stimulated activities lower than basal level of the w i l d type.  A l t h o u g h the basal activity o f the  R 2 1 0 6 C variant was lower than that o f the w i l d type, it was stimulated in the presence o f 50 u M retinal.  W h e n R O S extracted phospholipids were used, the basal and retinal  stimulated A T P a s e activities of the w i l d type and the mutants remain similar to when brain p o l a r / D P O E lipids were used.  3.9  To  AZIDO-ATP L A B E L I N G  complement the A T P hydrolysis measurements,  photoaffinity labeling  experiments  32 using a -  P a z i d o - A T P were carried out (section 2.13).  The a z i d o - A T P 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 S D S 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 w i l d 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 o f purified proteins (Fig 25, B). shows that A B C R from R O S was strongly labeled, as expected.  Fig 25,  A  The recombinant w i l d  type, R 2 1 0 6 C and T 1 5 2 6 M proteins were also labeled, indicating that they bind A T P . The D 8 4 6 H and R 2 0 7 7 W variant, however, showed no a z i d o - A T P labeling. variant was not investigated.  The R 2 0 3 8 W  A s shown in Fig 25, B the amount of purified protein in each  lane were comparable. Therefore, the lack o f a z i d o - A T P labeling by D 8 4 6 H and R 2 0 7 7 W was not due to the low levels o f 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. COS  A B C R proteins were solubilized and purified from transfected  cells and reconstituted i n lipid vesicles.  absence  A T P hydrolysis was measured i n the  (open columns) and presence o f 50 u M all-zrafM-retinal (solid columns). Each  data represents the mean ± o f S . D . o f triplicate values from a single  representative  experiments. The name o f the samples are shown in the figure, WT: w i l d type.  75  1 2 3 4  5  6  7  ABCR  1 2  B  3  4  5  6  7  ABCR  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]-ATP.  immunoaffinity  the  chromatography  and  The labeled proteins were purified by  eluates  were  polyacrylamide gels and subjected to electrophoresis.  loaded  (A)  on  two  6.5%  SDS  The labeled A B C R s were  detected by autoradiography and (B) the relative amounts o f proteins were detected by  Lane 1, A B C R from R O S ; lane 2, recombinant lane 3, D 8 4 6 H ; lane 4, R 2 0 7 7 W ; lane 5, R 2 1 0 6 C ; lane 6, T 1 5 2 6 M and  western blotting using R i m 5B4 antibody. wild type A B C R ;  lane 7, R O S membrane. 76  DISCUSSION  The next step following the identification o f A B C R as a member o f 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 o f 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 T Y P E AND M U T A N T ABCRS IN COS C E L L S  First, the w i l d type A B C R , cloned in the desired plasmid p C D N A 3 , was used to transfect C O S - 1 or 7 cells using the SuperFect reagent or calcium phosphate method o f transfection. Since the expressed detergent  proteins were destined  for functional studies,  a non-denaturing  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 o f 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 o f both soybean phospholipids (-40%  P C ) and D T T . A n excess o f phospholipids and a  reducing condition were previously shown to preserve the function o f P-glycoprotein (Callaghan et al. 1997). Glycerol (10%) was also included to stabilize the proteins. W h e n the cells were solubilized and the insoluble fraction removed, the total cell lysate was analyzed for the expression o f w i l d type A B C R .  Since the w i l d 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: D 8 4 6 H , T 1 5 2 6 M , R 2 0 3 8 W , R 2 0 7 7 W and R 2 1 0 6 C .  F o l l o w i n g transfection, analysis o f the soluble fractions  77  obtained from detergent solubilized C O S cells revealed that all five mutants could be expressed.  These proteins had the same molecular weight as the A B C R from R O S ,  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 CaP04  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.  R i m 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 R i m 3F4 binding site of the A B C R variants, R i m 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 C H A P S (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 L i n g , 1994). The  method o f detergent removal used in this study was essentially that o f 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.  T h i s method was favorable as it was  both time efficient and effective in removing the detergents.  Approximately 30-50 ng o f  proteins could be reconstituted by this method.  4.3  T H E ATPASE ASSAY  In order to investigate the effects o f the selected mutations on the function o f 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 L i n g , Urbatsch et al,  1994).  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 o f possible substrates,  including all-/r<ms-retinal and Af-retinylidene-phosphatidylethanolamine (JV-retinylideneP E ) , that stimulated the A T P a s e activity o f reconstituted A B C R purified from R O S .  In the  present study the efficiency o f the A T P a s e assay was first assessed, as a number o f 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 o f A T P was used here to insure that there were enough A T P in the assay.  Similarly, since half maximal stimulation o f reconstituted A B C R ,  was  calculated to occurred at -10-15 u M retinal, 50 u M retinal was used i n the assays, unless specified otherwise.  The same studies had also shown the importance o f P E on the  A T P a s e activity o f R O S A B C R , therefore, experiments in this study were carried out mostly in a mixture o f brain polar lipids (33% P E , 18% P S , 13% P C , 36% other phospholipids) and D O P E to ensure an environment rich in P E .  W h e n R O S A B C R was reconstituted in a 50:50 mixture o f brain polar lipids and D O P E , its basal activity was increased by an average o f 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. T h e specific activities, defined as the nmoles o f A T P hydrolyzed per m i n per mg o f protein, were determined to be 47 (basal) and 80 (retinal) in the presence o f 50 u M o f A T P .  79  Increasing  the retinal concentrations while keeping A T P constant, resulted in an increase in the A T P hydrolysis o f R O S A B C R .  T h i s illustrated that a\\-trans-retma\ may be a possible  substrate or modulator o f A B C R .  W h e n S D P E was substituted for D O P E i n 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 o f basal and stimulated activities in brain p o l a r / S D P E vesicles were smaller than the basal value for A B C R i n brain p o l a r / D O P E vesicles.  This may complicate the interpretation o f the data.  For instance, one may  consider the effect o f 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 o f S D P E , by bringing the activity to the basal level.  W h e n the recombinant w i l d type protein was reconstituted i n 50:50 mixture o f brain polar l i p i d s / D O P E , the presence o f 50 u M alWrarcs-refinal caused an average increase o f 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 o f w i l d type A B C R was found to increase i n the presence o f all-rnms-retinal, the absolute values for A T P a s e 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 . environment  similar  to the disk  membrane,  constituents, seemed very prominent.  i n terms  Creating an  o f the major phospholipid  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  o f possible  contaminants  during the extraction o f  phospholipids from R O S membranes.  When mutant A B C R s obtained from solubilized C O S cells were used in the A T P a s e assay, the data was even more difficult to interpret.  80  In some experiments, very little protein  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 nonenzymatic A T P hydrolysis was very difficult.  Similarly, due to the high background o f  A T P hydrolysis, it was hard to attribute slight stimulation or slight inhibition to the presence o f alWra/w-retinal. Another concern with the assay was the variation in the A T P a s e activities between experiments.  T o make the A T P a s e data more reliable, one  solution was to obtain and purify more proteins. C O S cell dishes instead o f one per D N A sample. basal activities  from  one  T h i s was achieved by transfecting three In order to prevent variations between  experiment to the next,  each experiment  which included  transfection o f C O S cells, solubilization o f the cells, purification and reconstitution o f the proteins, was performed for the w i l d type and the five mutants at the same time.  In this  case, the A T P a s e data obtained for the mutants was interpreted relative to the w i l d type. Fig 24 was a representative experiment in which 30-50 ng o f proteins were reconstituted and each assay was performed in triplicates.  A l l o f the mutants showed a basal A T P a s e  activity lower or essentially equal to the w i l d type. A T P a s e activities o f both D 8 4 6 H and R 2 0 7 7 W were diminished regardless o f retinal. retinal,  while  R 2 0 3 8 W and T 1 5 2 6 M  R 2 1 0 6 C , however, was stimulated by  remained unaffected  by it.  A more  detailed  interpretation o f the results from Fig 24 is presented in section 4.4.  Specific activities were not calculated for the mutants, since the amount o f 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 o f A T P  hydrolyzed per m i n per reconstituted sample.  4.4  ANALYSIS OF T H E EXPRESSED WILD T Y P E AND M U T A N T ABCRS  4.4.1  Wild type ABCR  W h e n 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 ( E R ) / G o l g i 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 o f 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 w i l d type A B C R was localized within the E R o f the transfected 293 cells.  The discrepancy regarding w i l d 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 w i l d 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 o f incompletely folded proteins. C a l n e x i n and calreticulin (soluble homologue o f calnexin) are lectins that recognize monoglucosylated misfolded proteins, and in conjunction with other proteins and enzymes mediate the retention and the proper folding o f misfolded glycoproteins in the E R .  The proposed model for quality control o f 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 - l i n k e d oligosaccharides in the form o f a 14-saccharide core unit (GIC3 Mang  GlcNAc2; G l c : glucose, M a n : mannose, G l c N A c : JV-acetylglucosamine) is covalently attached to an asparagine residue by an oligosaccharyltransferase. added  sugars  the  proteins  misfold,  transported to the G o l g i complex.  aggregate  and become  In the absence o f the  degraded without  being  T w o o f the three glucose residues on the core unit are  rapidly trimmed by the actions o f glucosidase I and II to produce a monoglucosylated protein. C a l n e x i n and calreticulin recognize the monogucosylated protein and associate with it.  The removal o f 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  incompletely (GT)  its journey to  the  folded an enzyme  Golgi  complex.  However,  i f the  protein remains  called U D P - G l u c o s e : g l y c o p r o t e i n glucosyltransferase  reglucosylates the protein, causing it to bind to calnexin/calreticulin again.  glucosylation and deglucosylation cycle continues  until glycoproteins are  This  completely  folded, and as a result misfolded proteins are retained in the E R .  L o o k i n g 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 o f a given missense mutation on the conformation o f 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 C O S 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 E R 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 A T P 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 T 1 5 2 6 M A B C R mutant was expressed in relatively high amounts (81 % o f w i l d type) in  COS  cells  using  immunofluorescence  both  calcium  phosphate  labeling pattern o f this  and  SuperFect  variant showed  reagents.  an E R / G o l g i  The labeling,  suggesting that it is retained in the E R compartment. Furthermore calnexin associates with this mutant to a greater extent than the w i l d 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 . membranes  Azido-ATP  suggested  that  its  labeling o f this protein prepared from C O S cell  N B D was  capable  of  binding A T P .  The  ATPase  measurement complemented the a z i d o - A T P labeling experiment, as a basal activity similar to the w i l d type was measured. of this mutant protein.  A l l - / r a « s - r e t i n a l , however, did not stimulate the activity  The pathogenic effect o f 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 o f 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 o f binding and hydrolyzing A T P , an indication o f a functional protein.  A similar situation has been observed for C F T R .  The most c o m m o n  genetic mutation in C F T R is a deletion o f phenylalanine 508 (AF508), this mutation is recognized as abnormal by the cellular quality control machinery and is retained within the ER.  Several studies, however, have suggested that although this variant o f C F T R is  mislocalized, it maintains its chloride channel activity ( L i et al,  1993; Pind et al.,  1994;  Pasyk and Foskett, 1995). Thus, the observation that T 1 5 2 6 M 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 o f 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.  A t 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 P E 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 E R 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 o f 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 o f 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 R 2 0 3 8 W variant showed basal, but not retinal stimulated A T P hydrolysis activity.  4.4.5  R2077W  The R 2 0 7 7 W missense mutation is also located in the second N B D , more specifically it is part o f the consensus sequence called motif C : Y S G G x K R K (Broccardo et al, exchange o f the arginine residue  in the  above  1999). The  sequence with a tryptophan did not  significantly affect the expression level o f this variant, as it was expressed 57% o f w i l d type.  Immunofluorescence labeling revealed a pattern similar to R 2 0 3 8 W (a missense mutation also in the second N B D , above). ER/Golgi  and  filamentous  That is, 80% o f the protein was expressed in the  extensions  from the  ER/Golgi  compartment  were  seen.  R 2 0 7 7 W was also associated with calnexin.  A z i d o - A T P labeling o f R 2 0 7 7 W in C O S cell membrane, however, showed that A T P does not bind to this variant. The A T P a s e activity o f this mutant was also impaired, regardless of the presence or absence o f 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 .  A c c o r d i n g to Manavalan et al. (1995), the Walker A motif and the motif C o f C F T R and other G-proteins such as the a sub unit o f transducin, are positioned adjacent to each other  86  in the tertiary structure o f these proteins.  In their study, Manavalan et al.  identified  specific residues in both motifs A and C responsible for binding the phosphate moiety o f ATP.  A l t h o u g h the equivalent residue o f R2077 was not shown to directly interact with  the phosphate group o f 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 o f y phosphate and in the case o f the serine residue, in M g  2 +  binding.  The proximity o f these residues to R2077 and the  conservation o f the motif C among other members o f 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 R 2 0 7 7 W 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).  A l t h o u g h the arginine residue at  2106 was conserved i n A B C A 1 , it is located outside o f any defined consensus sequences. It may, therefore, be hypothesized that the R 2 1 0 6 C mutation may not have a very profound effect on the function o f A B C R .  The R 2 1 0 6 C was expressed at a relatively high  level (71% o f w i l d type) in C O S cells.  Immunofluorescence microscopy revealed mostly vesicular labeling, however 34% o f the labeling was E R / G o l g i .  This suggested that this mutant resembled more the w i l d type  expression  had  pattern  and  a  more  native  conformation  than  the  other  mutants.  Unfortunately, calnexin interaction was not assessed for this mutant.  The a z i d o - A T P labeling o f this mutant in C O S cell membrane fractions revealed that this variant could bind A T P . A l t h o u g h the basal A T P a s e activity o f this variant was lower than w i l d type, all-frans-retinal stimulated its activity. Thus in terms o f A T P binding and A T P hydrolysis, this mutant resembled the w i l d type. The pathogenic effect o f this mutant may be due to a property not investigated in this study.  87  4.5  SUMMARY  A t the start o f this project, 89 mutations in the disease were identified.  ABCR gene that resulted in Stargardt's  T o study the effects o f five o f these mutations on the structure  and function o f 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 o f the recombinant proteins were purified from detergent solubilized C O S cells by immunoaffinity  chromatography and used in  reconstitution  experiments.  Immunofluorescence  microscopy o f transfected  C O S cells revealed that the w i l d type  A B C R was localized mostly in vesicles whereas the mutant A B C R s displayed a range o f E R / G o l g i labeling pattern. localized  compared  to  M o s t o f the D 8 4 6 H and T 1 5 2 6 M proteins were E R / G o l g i  77%,  80%  and  34%  of  R2038W,  R2077W  and  R2106C,  respectively.  T o complement the immunofluorescence  studies, the w i l d type and four of the mutants  were examined for their abilities to interact with calnexin, a resident o f 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.  T o assess the A T P a s e activity of the w i l d type and mutant A B C R s , the purified proteins were reconstituted in liposomes rich in P E (50:50 mixture o f brain polar lipids and D O P E ) . In the presence o f 50 u M A T P , both the basal and the retinal stimulated activities o f these samples were measured. Although the basal activity o f w i l d 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 R 2 0 3 8 W variants. The R 2 1 0 6 C variant displayed a retinal-stimulation o f four fold, while the A T P a s e activities of both D 8 4 6 H and R 2 0 7 7 W were impaired.  88  Azido-ATP  labeling  was  performed  on  complement the A T P a s e activity studies.  transfected  COS  membrane  fractions  to  Whereas the w i l d type, T 1 5 2 6 M and R 2 1 0 6 C  bound A T P to the same extent, the D 8 4 6 H and R 2 0 7 7 W did not.  B y compiling all of the results obtained, it was suggested that D 8 4 6 H 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 o f  all-trans-ret'mal, thus its pathogenic effect may be  due to its inability to interact with its substrate. revealed  that  this  variant was  not  stimulated  The A T P hydrolysis ability o f R 2 0 3 8 W in the  presence  of  all-zra«s-retinal.  Therefore, it is unlikely that this mutant is capable o f transporting its substrate.  The  R 2 0 7 7 W variant was pathogenic because it could not bind and hydrolyze A T P .  The  mutant R 2 1 0 6 C seemed to be folded correctly and both the A T P a s e activity and the azidoATP  labeling  studies  suggested  that  this  mutant  could  bind  and hydrolyze A T P .  Furthermore, as with the w i l d type, it showed a retinal stimulated A T P a s e activity.  The  R 2 1 0 6 C mutation may affect another function o f 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 o f 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 , W o n g J T , M o l d a y R S . 2000. The effect o f lipid environment and retinoids on the A T P a s e 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 J R . 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 P S , Peiffer A , Zabriskie N A , L i Y , Hutchinson A , Dean M , Lupski J R , Leppert M . 1997b. Mutation o f the Stargardt disease gene ( A B C R ) in age-related macular degeneration. Science. 277:1805-1807. A m e r S, Akhtar M . 1973. Studies on regeneration o f 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 o f retinyl ester synthetase and retinoid isomerase from bovine ocular pigment epithelium. J. Biol.  Chem. 264:9231-9238. Bascom R A , Manara S, C o l l i n s L , M o l d a y R S , Kalnins V I , Mclnnes R R . 1992. C l o n i n g of the c D N A for a novel photoreceptor membrane protein (rom-1) identifies a disk r i m protein family implicated in human retinopathies. Neuron. 8:1171-1184. Baylor D A and N u n n B . 1982.  Enzymol.  Electrical signaling in vertebrate photoreceptors.  Methods  81:403-423.  Bear C E , L i C H , Kartner N , Bridges R J , Jensen T J , Ramjeesingh M , Riordan J R . Purification  and  functional  reconstitution  conductance regulator ( C F T R ) .  Cell.  of  the  cystic  fibrosis  68:809-818.  Bernstein P S , L a w W C , Rando R. 1987. Isomerization o f all-trans retinoids to retinoids in vitro. Proc. Natl Acad. Sci. USA. 84:1849-1853. Bok  D.  1985.  Retinal  1992.  transmembrane  photoreceptor-pigment  epithelium  interactions.  ll-cis  Invest.  Ophthalmol. Visual Sci. 26:1659-1694. B o k D . 1993. 17:189-195.  The retinal pigment epithelium: a versatile partner in vision. 90  J. cell Sci.  Bok D , O n g D E , Chytil F . 1984. Immunocytochemical localization o f cellular retinol binding protein in the rat retina. Invest. Ophthalmol. Visual Sci. 25:1-7. B o s c h E , Horwitz J , B o k D . 1993. Phagocytosis o f 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 S J , Albert A D . 1990. Relationship o f cholesterol content to spatial distribution and age o f disc membranes i n retinal rod outer segments. J. Biol.  Chem. 265:18867-18870. Bridges C D . 1976. V i t a m i n A and the role o f the pigment epithelium during bleaching and regeneration o f rhodopsin in the frog eye. Exp. Eye Res. 22:435-455. Broccardo C , Luciani M - F , C h i m i n i G .  1999.  The A B C A  subclass o f mammalian  transporters. Biochimica et Biophysica Acta. 1461:395-404. Bungert S, M o l d a y L L , M o l d a y R S . 2001. Membrane topology o f 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 o f P glycoprotein is dependent on maintenance o f a lipid-protein interface. Biochimica  Biophysica Acta. 1328:109-124. Cervetto L , Lagnado L , Perry R J , Robinson D W , M c N a u g h t o n P A . 1989. Extrusion o f 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 o f mammalian cells by plasmid D N A . Mol. Cell Biol. 7:2745-2752. Chen C K , Inglese J , Lefkowitz R J , Hurley J B . 1995. Ca-dependent interaction o f recoverin with rhodopsin kinase. J. Biol. Chem. 270:18060-18066. Chen T Y , Illing M , M o l d a y L L , H s u Y T , Y a u K W , M o l d a y R S . 1994. Subunit 2 (or beta) of retinal rod c G M P - g a t e d cation channel is a component o f the 240-kDa channelassociated protein and mediates Ca(2+)-calmodulin modulation. Proc. Natl Acad. Sci. USA. 91:11757-11761. Cohen A L 1968. N e w evidence supporting the linkage to extracellular space o f outer segment saccules o f frog cones but not rods. J. Cell Biol. 37:424-444. Cole SP and Deeley R G . 1998. Multidrug resistance mediated by the A T P - b i n d i n g cassette transporter protein M R P . BioEssays. 20:931-940.  91  C o o k N J , Hanke W , and Kaupp U B . 1987. Identification, purification, and functional reconstitution o f the cyclic GMP-dependent channel from rod photoreceptors. Proc. Natl  Acad. Sci. USA. 84:585-589. Cremers F P , van de Pol D J , van Driel M , den Hollander A l , van Haren F J , Knoers N V , Tijmes N , Bergen A A , Rohrschneider K , Blankenagel A , Pinckers A J , Deutman A F , H o y n g 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. C r o u c h R K , Chader G J , Wiggert B , Pepperberg D R .  1996.  Retinoids and the visual  process. Photochem. Photobiol. 64:613-621. Cuatrecasas P. 1970. Protein purification by affinity chromatography. agarose and polyacrylamid beads. J. Biol. Chem. 245:3059-3065. Dean M , Allikmets R. 1995.  Derivatization o f  E v o l u t i o n o f A T P - b i n d i n g cassette transporter genes. Curr.  Opin. Genet. Dev. 5:779-785. Deigner P S , L a w W C , Canada F J , Rando R R . 1989.  Membranes as the energy source in  the endergonic transformation o f 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 o f retinal rods is regulated by two inhibitory subunits. Proc. Natl Acad. Sci. USA. 85:2424-  2428. Dizhoor A M , Olshevskaya E V , Henzel W J , W o n g S C , Stults JT, A n k o u d i n o v a I, Hurley JB. 1995. C l o n i n g , 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 F J . from  multidrug-resistant  1992.  Chinese  A T P a s e activity o f partially purified P-glycoprotein hamster  ovary cells. Biochimica Biophysica Acta.  1109:149-160. Dose A C .  1995.  Molecular characterization o f the cyclic nucleotide-gated cation channel  o f bovine rod outer segments. Ph.D. thesis. University o f British C o l u m b i a . Ellgaard L , M o l i n a r i 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 o f the lamellar and disk-edge structures o f  the rod outer segment. J. Ultrastruct. Res. 28:41-60. Farahbakhsh Z T , H i d e g 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. F r o m mice to man: the cyclic G M P phosphodiesterase  gene in vision  and disease. Invest. Ophthalmol. Visual Sci. 36:263-275. Farrar G J , K e n n a P, Jardan S A , K u m a r - S i n g h 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 o f retinitis pigmentosa. Nature. 354:478-480. Flannery J G , O ' D a y W , Pfeffer B A , Horwitz J , B o k D . 1990. Uptake, processing and release o f retinoids by cultured human retinal pigment epithelium. Exp. Eye Res. 51:717728. F o l c h J , Lees M , Sloane-Stanley G H . 1957. A simple method for the isolation and purification o f total lipids from animal tissues. J. Biol. Chem. 226:497-509. Fung B K and Griswold-Prenner I. 1989. G protein-effector coupling: binding o f rod phosphodiesterase inhibitory subunit to transducin. biochemistry. 28:3133-3137. Fung B K , Y o u n g J H , Yamane H K , Griswold-Prenner I. 1990. Subunit stoichiometry o f retinal rod c G M P phosphodiesterase. Biochemistry. 29:2657-64. Gray-Keller M P and Detwiler P B . 1994. T h e calcium feedback phototransduction cascade o f vertebrate rods. Neuron. 13:849-861. H a m m H E and Gilchrist A . 8:189-196.  1996. Heterotrimeric G proteins.  signal  i n the  Curr. Opin. Cell Biol.  Hayward G , Carlsen W , Siegman A , Stryer L . 1981. Retinal chromophore o f rhodopsin photoisomerizes within picoseconds. Science. 211:942-944. He  W , Cowan  C W , Wensel  TG.  1998.  RGS9,  a  GTPase  accelerator  for  phototransduction. Neuron. 20:95-102. Hicks D and M o l d a y R S . 1986. Differential immunogold-dextran labeling o f 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 o f channel regulation. Cell. 82:693-696. Hsu S C and M o l d a y R S . 1991. Glycolytic enzymes and a G L U T - 1 glucose transporter in the outer segments o f rod and cone photoreceptor cells. J. Biol. Chem. 266:217'45-217'52. H s u Y T and M o l d a y R S . 1993. Modulation o f c G M P - g a t e d channel o f rod photoreceptor cells by calmodulin. Nature. 361:76-79.  93  Huang S H , Pittler S J , Huang X , O l i v e i r a L , Berson E l , Dryja T P . recesive  retinitis  pigmentosa  caused  by mutations  1995.  in the a-subunit  Autosomal  o f rod c G M P  phosphodiesterase. Nature Genet. 11:468-471. M i n g M , M o l d a y L L , M o l d a y R S . 1997. T h e 220-kDa rim protein o f retinal rod outer segments is a member o f the A B C transporter superfamily. J. Biol. Chem. 272:1030310310. Ishiguro S, Suzuki Y , Tamai M , M i z u n o K . 1991. Purification o f retinol dehydrogenase from bovine retinal rod outer segments. J.Biol. Chem. 266:15520-15524. Kawamura S. 1993. Rhodopsin phosphorylation as a mechanism phosphodiesterase regulation by S-modulin. Nature. 362:855-857.  o f cyclic G M P  K e l l y 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 o f the two A B C transporter proteins encoded in the human major histocompatibility complex. Nature. 355:641-644. K i m T S , R e i d D M , M o l d a y R S . 1998. Structure-function relationships and localization o f the N a / C a - K exchanger in rod photoreceptors. J. Biol. Chem. 273: 16561-16567. Krizaj D and Copenhagen D R . 1998. Compartmentalization o f calcium extrusion mechanisms in the outer and inner segments o f photoreceptors. Neuron. 21:249-256. K l e i n I, Sarkadi B , Varadi A .  1999.  A n inventory o f 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 o f retinal rod guanylate cyclase by calcium ions. Nature. 334:64-66. Krupnick  J G , G u r e v i c h V V , Benovic  phototransduction.  Binding  phosphorylation. J.Biol.  JL.  competition  1997. between  Mechanism arrestin  o f quenching  and  transducin  of for  Chem. 272:18125-18131.  Kuchler K , Thorner J . 1992. Secretion o f peptides and proteins lacking hydrophobic signal sequences: the role o f adenosine triphosphate-driven membrane translocators.  Endocr. Rev. 13:499-514. L a i Y - L , Wiggert B , L i u Y - P , Chader G J . 1982. Interphotoreceptor retinol-binding proteins: possible transport vehicles between compartments o f retina. Nature. 298:848849. L a e m m l i U K . 1970. Cleavage o f structural proteins during the assembly o f the head o f bacteriophage T 4 . Nature. 227:680-685.  94  Lewis R A , Shroyer N F , Singh N , Allikmets R , Hutchinson A , L i Y , Lupski J R , Leppert M , Dean M . 1999. Genotype/phenotype analysis o f a photoreceptor-specific A T P - b i n d i n g 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. T h e cystic fibrosis mutation (delta F508) does not influence the chloride channel activity o f C F T R . Nat. Genet. 3:311-316. L i o u G L , Bridges C D , F o n g S - L . 1982. V i t a m i n 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 o f retinal guanylate cyclase in human and monkey retinas. Exp. Eye Res. 59:761-768. Lolley R N and Racz E . 1982. C a l c i u m modulation o f cyclic G M P synthesis i n rat visual cells. Vision Res. 22:1481-1486. Loo T W and Clarke D M . 1994. Prolonged association o f temperature-sensitive mutants of human P-glycoprotein with calnexin during biogenesis. J. Biol. Chem. 269:2868328689. Luciani M F , Denizot F , Savary S, Mattei M G , C h i m i n i G . 1994. C l o n i n g o f 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 o f bovine rod photoreceptor glutamic acid rich protein. M.Sc.Jhesis. University o f British Columbia. Manavalan P, Dearborn D G , M c P h e r s o n J M , Smith A E . 1995. Sequence homologies between nucleotide binding regions o f C F T R and G-proteins suggest structural and functional similarities. FEBS Lett. 366:87-91. M a r t i n e z - M i r A , Paloma E , Allikmets R , A y u s o C , del R i o T , Dean M , V i l a g e l i u 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 , W e n g J , Travis G H . 2000. Biosynthesis o f a major lipofuscin fluorophore i n mice and humans with A B C R - m e d i a t e d retinal and macular degeneration. Proc. Natl  Acad., Sci. USA. 97:7154-7159. M c B e e J K , K u k s a V , A l v a r e z R , de Lera A R , Prezhdo O , Haeseleer F , Sokal I, Palczewski K. 2000. Isomerization o f all-trans-retinol to cis-retinols i n bovine retinal pigment epithelial cells: dependence on the specificity o f retinoid-binding proteins. Biochemistry. 39:11370-11380.  95  M c L a u g h l i n M E , Sandberg M A , Berson E L , Dryja T P . 1993. Recessive mutations in the gene encoding the P subunit o f rod phosphodiesterase in patients with retinitis pigmentosa.  Nature Genet. 4:130-134. M o l d a y L L , Rabin A R , M o l d a y R S . 2000. A B C R expression in foveal cone photoreceptors and its role in Stargardt macular dystrophy. Nature Genet. 25:257-258. Molday R S .  1998.  Photoreceptor membrane proteins, phototransduction, and retinal  degenerative diseases. The Friedenwal Lecture. Invest. Ophthalmol. Visual Sci. 39:24912513. M o l d a y R S and M o l d a y L L . 1987. Differences in the protein composition o f bovine retinal rod outer segment disk and plasma membranes isolated by a ricin-gold-dextran density perturbation method. J. Cell Biol. 105:2589-2601. M o l d a y R S , H i c k s D , M o l d a y L L . 1987. Peripherin: a rim-specific membrane protein o f 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. C a l c i u m extrusion from mammalian photoreceptor terminals. J. Neurosci. 18: 2467-2474. M o r i t z O L and M o l d a y R S . 1996. Molecular cloning, membrane topology and localization o f 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 o f 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 o f the disc membranes in the retina o f 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, G o r c z y c a W A , Helekar B S , R u i z C C , Ohguro H , Huang J , Zhao X , Crabb J W , Johnson R S , W a l s h K A , GrayKeller M P , Detwiler P B , Baehr W . 1994. Molecular cloning and characterization o f retinal photoreceptor guanylyl cyclaseactivating protein. Neuron. 13:395-404. Papermaster D S , Schneider B G , Z o r n M A , Kraehenbuhl JP. 1978. Immunocytochemical localization o f a large intrinsic membrane protein to the incisures and margins o f 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 C I - channel is functional when retained in endoplasmic reticulum o f  mammalian cells. J. Biol. Chem. 270:12347-12350. Penn R D and Hagins W A . 1969. Signal transmission along retinal rods and the origin o f 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, L e o w s k i C , Bonnemaison M , L e Paslier D , Frezal J , Dufier J L , Pittler S, M u n n i c h A , K a p l a n J. 1996. Retinal-specific guanylate cyclase gene mutations in Leber's  congenital amaurosis. Nature Genet. 14:461-464. Pind S, Riordan J R , W i l l i a m s D B . 1994. Participation o f the endoplasmic reticulum chaperone calnexin (p88, IP90) in the biogenesis o f 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 o f the visual cycle. Biochemistry. 30:595-602. Reid D M , Friedel U , M o l d a y R S , and C o o k N J . 1990. Identification o f the sodiumcalcium exchanger as the major ricin-binding glycoprotein o f 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 P J , C o w l e y G S , M c G e e 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 .  Vision Res.  1968.  Bleaching and regeneration o f cone pigments in man.  8:617-631.  Saari J C and Bredberg D L . 1982. Enzymatic reduction o f 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 o f the retinal rod photoreceptor A B C protein ( A B C R ) in yeast cells. B. Sc. honours thesis. University o f British Columbia. Sambrook J , Fritsch E F , Maniatis T . 1989. E d . C o l d Spring Harbor Press, N e w Y o r k .  Mol e c u l ar C l o n i n g : A Laboratory M a n u a l . 2  n d  Sampath A P , Matthews H R , Cornwall M C , Fain G L . 1998. Bleached pigment produces a maintained decrease in outer segment C a in salamander rods. J. Gen. Physiol. 111:5364. 2 +  97  Schneider E and Hunke S. 1998. ATP-binding-cassette ( A B C ) transporter Functional and structural aspects o f the A T P - h y d r o l y z i n g subunits/domains.  system:  FEMS  Microbiol. Rev. 22:1-20. Shapiro A B and L i n g V .  1994.  A T P a s e activity o f purified and reconstituted P-  glycoprotein from Chinese hamster ovary cells. J. Biol. Chem. 269:3745-3754. Shyamala V , Baichwal V , Beall E , A m e s G F . 1991. Structure-function analysis o f the histidine permease and comparison with cystic fibrosis mutations. J. Biol. Chem. 266:18714-18719. Spitznas M and Hogan M J .  1970. Outer segments o f photoreceptors and retinal pigment  epithelium. Arch. Ophthalmol. 84:810-819. Stecher H , G e l b M H , Saari J C , Palczewski K . 1999. Preferential release o f 11-cz's-retinol from retinal pigment epithelial cells i n the presence o f cellular retinaldehyde-binding  protein. J. Biol. Chem. 274:8577-8585. Steinberg R H , Fisher S K , Anderson D H .  1980.  D i s c morphogenesis  i n vertebrate  photoreceptors. J. Comp. Neurol. 190:501-508. Sun H , M o l d a y R S , Nathans J . 1999. Retinal stimulates A T P hydrolysis by purified and reconstituted A B C R , the photoreceptor-specific A T P - b i n d i n g 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 , W i l l i a m s T , Yamazaki A . 1994. Phosphorylation o f an inhibitory subunit o f c G M P phosphodiesterase in Rana catesbeiana  rod photoreceptors.  I. Characterization o f the phosphorylation.  J.  Biol. Chem.  269:15024-15029. Tsuboi S, Matsumoto H , Yamazaki A . 1994. Phosphorylation o f an inhibitory subunit o f cGMP  phosphodiesterase  i n Rana catesbeiana rod photoreceptors.  II. A  possible  mechanism for the turnoff o f c G M P phosphodiesterase without G T P hydrolysis. J. Biol.  Chem. 269:15016-15023. Udovichenko IP, Newton A C , W i l l i a m s D S . 1997. Contribution o f protein kinase C to the phosphorylation o f rhodopsin in intact retinas. J. Biol. Chem. 272:7952-7959. Urbatsch I L , A l - S h a w i M K , Senior A E . 1994. Characterization o f the A T P a s e activity o f purified Chinese hamster P-glycoprotein. Biochemistry. 33:7069-7076. Walker J E , Saraste M , Runswick M J , G a y N J . 1982. Distantly related sequences i n the alpha- and beta-subunits o f A T P synthase, myosin, kinases and other A T P - r e q u i r i n g enzymes and a common nucleotide binding fold. EMBO J. 1:945-951.  98  Wang J K , M c D o w e l l J H , Hargrave P A . 1980. bovine rhodopsin. Biochemistry.  Site o f attachment o f 11-cis-retinal i n  19:5111-5117.  Warren R. 1999. Insights on the regulation o f the c G M P - g a t e d channel by C a / c a l m o d u l i n and on the phosphorylation o f the beta-subunit by a CKII-like protein kinase. M. Sc. thesis. University o f British C o l u m b i a . ++  Wells J , Worblewski J , K e e n J , Inglehearn C , Jubb C , Eckstein A , Jay M , A r d e n G , Bhattacharya S, Fitzke F , B i r d A C . 1993. Mutations in the human retinal degeneration slow ( R D S ) gene can cause either retinitis pigmentosa or macular dystrophy. Nature  Genet. 3:213-218. Weng J , Mata N L , A z a r i a n S M , Tzekov R T , B i r c h D G , Travis G H . 1999. Insights into the function o f rim protein in photoreceptors and etiology o f Stargardt's disease from the phenotype in abcr knockout mice. Cell. 98:13-23. Wensel T G and Stryer L . 1990. Activation mechanism o f 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 o f isomerohydrolase activity in the visual cycle. Biochemistry. 37:2044-50. Yamagata K , Goto K , K u o C H , K o n d o H , M i k i N . 1990. V i s i n i n : a novel calcium binding protein expressed in retinal cone cells. Neuron. 4:469-476. Y a n n u z z i L A , Grayer D R , Green W R . The retina A T L A S . 1995.  S t - L o u i s : M o s b y Yearbook Inc.  Y a u K W . 1994. Phototransduction mechanism in retinal rods and cones. The Friedenwald  Lecture. Invest Ophthalmol Visual Sci. 35:9-32. Y o u n g R W . 1968. Passage o f newly formed protein through the connecting c i l i u m o f retinal rods in the frog. J. Ultrastruct. Res. 23:A62-A13. Y o u n g R W . 1971. Shedding o f disks from rod outer segments i n the rhesus monkey.  J.  Ultrastruct. Res. 34:190-203. Young  RW.  1987.  Pathophysiology  o f age-related macular degeneration.  Surv.  Ophthalmol. 31:291-306. Y o u n g R W and B o k D . 1969. Participation o f the retinal pigment epithelium in the rod outer segment renewal process. J. Cell Biol. 42:392-403. Y o u n g R W and D r o z B . 1968. The renewal o f protein in retinal rods and cones. J. cell  Biol. 39:169-184.  99  Zagotta W N and Siegelbaum S A . 1996. Structure and function o f 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 i n arrestin binding. J. Biol. Chem. 272:14762-14768. Z h o u X and Arthur G . 1992. Improved procedures for the determination phosphorus by malachite green. J. Lipid Res. 33:1233-1236.  100  o f lipid  

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:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0090157/manifest

Comment

Related Items