UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Calmodulin modulation of the cGMP-gated channel of photoreceptor rod cell Hsu, Yi-te 1994

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

Item Metadata

Download

Media
831-ubc_1994-953497.pdf [ 3.44MB ]
Metadata
JSON: 831-1.0088090.json
JSON-LD: 831-1.0088090-ld.json
RDF/XML (Pretty): 831-1.0088090-rdf.xml
RDF/JSON: 831-1.0088090-rdf.json
Turtle: 831-1.0088090-turtle.txt
N-Triples: 831-1.0088090-rdf-ntriples.txt
Original Record: 831-1.0088090-source.json
Full Text
831-1.0088090-fulltext.txt
Citation
831-1.0088090.ris

Full Text

CALMODULIN MODULATION OF THE cGMP-GATED CHANNEL OFPHOTORECEPTOR ROD CELLbyYl-TE HSUB.Sc., University of British Columbia, 1989A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDEPARTMENT OF BIOCHEMISTRY AND MOLECULAR BIOLOGYWe accept this thesis as conformingto the required standardTHE UNWERSITY OF BRITISH COLUMBIAJune 1994©Yi-Te Hsu, 1994In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(SignatureDepartment of E’ IoThe University of British ColumbiaVancouver, CanadaDate J,(- 30, i99/—DE.6 (2/88)iiABSTRACTDuring the past decade, great advances have been made towards deciphering themolecular mechanism of vision. Findings from various laboratories have led to theestablishment of the visual cascade and the identification of the main members of thispathway. However, little is known so far about the intricate regulation of the variousenzymes involved, in particular, the roles of Ca2 and Ca2-binding proteins. This thesisdescribes the finding of a novel regulation of the visual transduction process through theCa2-calmodulin dependent modulation of the cGIVIP-gated channels.An unknown 17/20 kDa protein found in the EDTA extract of bovine rod outersegments (ROS) was purified and identified as calmodulin by a variety of biochemical andimmunochemical studies. The cGMP-gated channel complex, which plays an importantrole in phototransduction, was found to be the major calmodulin binding protein in ROSmembranes by calmodulin affinity chromatography and Western blotting analysis withiodinated calmodulin. This association between calmodulin and the channel complex,which is mediated through a high molecular weight channel associated protein, was foundto be conserved among various species.The effect of calmodulin on the ROS cGIVIP-gated channel activity wasinvestigated by a Ca2 influx assay using ROS membrane vesicles pre-loaded withArsenazo III dye. In the absence of calmodulin, the channel displayed an apparent Km of19 ± 0.4 iiM for cGMP, while in the presence of calmodulin, its apparent Km increased to33 ± 2 iiM. Similar shifts in the Km of the channel for cO1VIP were observed usingextruded ROS membrane vesicles containing either dichiorophosphonazo III or neutralred. This effect is specific for calmodulin since otherCa2+binding proteins such as bovinerecoverin and brain S-100 caused no change on channel activity. The calmodulinmediated shift in the affinity of the channel for cGMP could be reversed by mastoparan, apeptide inhibitor of calmodulin. In addition to Ca2, calmodulin also appeared to affectthe translocation of various monovalent and divalent cations across the ROS membrane.iiiCalmodulin modulation of the channel was observed to occur within a physiological Ca2+range of 20-300 nM.The calmodulin effect was shown to be mediated through the 240 kDa channelassociated protein as determined by a Ca2+ efflux assay using the immunoaffinity purifiedchannel complex reconstituted into lipid vesicles. In the absence of calmodulin, thereconstituted channel complex displayed a Km of 33 p.M for cGMP, while in the presenceof calmodulin, the Km increased to 44 p.M. N-terminal sequence analysis of a 105 kDacalmodulin-binding fragment of the 240 kDa protein revealed a sequence that matchedwith the recently cloned f3-subunit of the human rod cGMP-gated channel.In summary, the presence of calmodulin in ROS has been detected. This proteinbinds preferentially to the 13-subunit of the cGMP-gated channel complex and canmodulate the affinity of the channel for cGMP. A model of the calmodulin binding to thechannel complex is presented and the importance of the calmodulin mediated regulation ofthe channel activity is discussed.ivTABLE OF CONTENTSPageABSTRACT.iiTABLE OF CONTENTS ivLIST OF TABLES ixLIST OF FIGURES xLIST OF ABBREVIATIONS xiiiACKNOWLEDGEMENT xvCHAPTER 1INTRODUCTION1.1. PHOTORECEPTOR ROD CELL 11.1.1. Structure 11.1.2. Rod outer segments 31.2. PHOTOTRANSDUCTION 51.2.1. Ca2 vs. cGMP as the primary messenger for photoexcitation 71.3. VISUAL TRANSDUCTION CASCADE 101.3.1. Photoactivation of rhodopsin 101.3.2. Activation of transducin 121.3.3. Hydrolysis of cGMP byPDE 121.3.4. Closure of the cGIvlP-gated channel 131.3.5. Deactivation and regeneration of rhodop sin 131.3.6. Deactivation of transducin and phosphodiesterase 141.3.7. Resynthesis of cGMP 141.3.8. Roles of calcium in visual transduction 151.4. THE cGMP-GATED CHANNEL COMPLEX 171.4.1. Properties of the cGMP-gated channel 17V1.4.2. Purification and localization of the cGMP-gated channel 191.4.3. Molecular cloning of the cGMP-gated channel complex 201.4.4. Structural analysis and topography of the channel subunits 211.5. CALMODULIN 261.5.1. Molecular cloning of calmodulin 261.5.2. Structure of calmodulin 281.5.3. Post-translational modification 281.5.4. Calcium binding to calmodulin 301.5.5. Calmodulin binding sequences 311.6. FUNCTIONAL ROLES OF CALMODULIN 311.6.1. Cyclic nucleotide metabolism 341.6.2. Protein phosphorylation 341.6.3. Protein dephosphorylation 351.6.4. Cation transport 361.6.5. Cytoskeletal organization 361.6.6. Control of cell proliferation 371.7. THESIS INVESTIGATION 38CHAPTER 2IDENTIFICATION AND CHARACTERIZATION OF A 20 kDa PROTEIN,CALMODULIN, FROM THE EDTA EXTRACT OF PHOTORECEPTOR RODOUTER SEGMENTS2.1. MATERIALS 402.2. METHODS 402.2.1. Preparation of ROS membranes 402.2.2. Purification of the 20 kDa protein from ROS membranes 412.2.3. Gel filtration chromatography 41vi2.2.4. Preparation of ROS disk and plasma membranes 422.2.5. Extraction of calmodulin from ROS disk and plasma membranes... 432.2.6. Immunofluorescence localization of calmodulin 432.2.7. Radioiodination of calmodulin 432.2.8. SDS polyacrylamide gel electrophoresis, Western blottinganalysis, and protein concentration determination 442.3. RESULTS 452.3.1. Extraction and purification of a 20 kDa protein from ROS 452.3.2. Identification of the 20 kDa protein as calmodulin 452.3.3. Immunofluorescence labelling 472.3.4. Extraction of calmodulin from ROS disk and plasma membranes 532.3.5. Calmodulin binding proteins in ROS 532.4. DISCUSSION 53CHAPTER 3IDENTIFICATION OF THE cGMP-GATED CHANNEL COMPLEX AS TIlEMAJOR CALMODULIN BINDING PROTEIN IN BOVINE ROS MEMBRANES3.1. MATERIALS 593.2. METHODS 593.2.1. Calmodulin affinity chromatography 593.2.2. PMc 6E7 antibody affinity chromatography 603.2.3. DEAE purification of the cGMP-gated channel complex 613.2.4. Calcium efflux assay on reconstituted channels 613.2.5. Ca2-dependent association of calmodulin with the channelcomplex 623.2.6. SDS polyacrylamide gel electrophoresis, Western blottinganalysis, and protein concentration determination 62vii3.3. RESULTS.633.3.1. Identification of the cGMP-gated channel complex as the majorcalmodulin binding protein 633.3.2. Calmodulin affinity chromatography 633.3.3. Channel complexes from ROS of other species 693.3.4. Extraction of calmodulin by the immobilized channel complex 733.4. DISCUSSION 73CHAPTER 4MODULATION OF THE cGMP-GATED CHANNEL OF PHOTORECEPTORROS BY CALMODULIN4.1. MATERIALS 764.2. METHODS 764.2.1. Calcium influx assay using Arsenazo III and dichlorophosphona.zoIII dyes 764.2.2. Inbibition of the calmodulin effect by mastoparan 774.2.3. Cation influx assay using neutral red dye 774.2.4. Purification of calmodulin and recoverin from bovinephotoreceptor ROS 784.3. RESULTS 794.3.1. Effect of calmodulin on the affinity of the channel for cGMP 794.3.2. Specificity of the calmodulin effect 814.3.3. Inhibition of the calmodulin effect by mastoparan 844.3.4. Cation selectivity and calcium dependence of the calmodulineffect 884.4. DISCUSSION 88CHAPTER 5viiiCHARACTERIZATION OF THE 240 kDa CHANNEL ASSOCIATED PROTEII%15.1. MATERIALS 945.2. METHODS 945.2.1. Purification of the channel complex by PMc 6E7 antibodyaffinity and DEAE anion exchange chromatography 945.2.2. Calcium effiux assay of the immunoaffinity purified channelcomplex reconstituted into lipid vesicles 955.2.3. Mild proteolytic digestion ofROS membranes with tlypsin,chymotrypsin, and kallikrein 955.2.4. Purification of the kallikrein-treated channel complex by PMc 6E7antibody affinity chromatography 965.2.5. Protein sample preparation for N-terminal sequence analysis 965.2.6. SDS polyacrylamide gel electrophoresis and Westernblotting analysis 965.3. RESULTS 975.3.1. Mediation of the calmodulin effect through the 240 kDa channelassociated protein 975.3.2. Effectiveness of anti-240 kDa protein monoclonal antibodies ininhibiting calmodulin binding to the channel complex 995.3.3. Effect of proteolysis on the cGMP-gated channel complex 995.3.4. Identification of a calmodulin binding fragment of the 240 kDaprotein 1085.4. DISCUSSION 108SUIVIIVIARY 115REFERENCES 123ixLIST OF TABLESTable PaneI. Rod outer segment proteins 4II. Gel filtration molecular weight sizing of the 17/20 kDa protein 48Ill. Comparison of the amino acid composition of the 17/20 kDa protein withthose of otherCa2binding proteins 52IV. Comparative purification of the channel complex by DEAR anion exchangeand calmodulin affinity chromatography 68V. Summary of the calmodulin modulation of the cGMP-gated channelcomplex as determined by various assay systems 116xLIST OF FIGURESFigure Page1. Schematic representation of a vertebrate photoreceptor rod cell 22. Hyperpolarization of the rod cell in response to light 63. Calcium hypothesis for visual excitation 84. cGMP hypothesis for visual excitation 95. A diagram showing the basic reactions of the visual transduction pathway 116. Calcium-dependent modulation of guanylate cyclase 167. Sequence alignment of the cyclic nucleotide binding domains of channelsfrom different species 228. Sequence alignment of the N-terminal cleavage sites of the cGMP-gatedchannels from different species 249. Sequence alignment of the putative voltage sensor motifs (S4) and poreregions of the cyclic nucleotide gated channels 2510. Working model for the organization of the cL-subunit of bovine rodcGMP-gated channel within the lipid bilayer 2711. Structure and sequence of calmodulin 2912. Helical wheel plot of a model amphiphilic calmodulin binding peptide 3213. Amino acid sequences of the calmodulin binding domains of varioustarget proteins 3314. Extraction and purification of calmodulin from bovine ROS membranes.... 4615. Calcium-45 binding analysis of the 20 kDa protein 4916. Calcium-dependent mobility shift analysis of the 17/20 kDa protein bySDS polyacrylamide gel electrophoresis 5017. Western blotting analysis with an anti-calmodulin monoclonal antibody 5118. Immunofluorescence localization of calmodulin 5419. Extraction of calmodulin from ROS disk and plasma membranes 5520. Labelling of the ROS proteins with iodinated calmodulin 56xi21. Purification of the cGMP-gated channel complex by calmodulin andimmunoaffinity chromatography and identification of the 240 kDa channel-associated protein as a major calmodulin binding polypeptide of ROSmembranes 6422. Purification of the bovine ROS channel complex by calmodulin affinitychromatography 6523. Activation of the calmodulin affinity purified channel complex bycGMP 6724. Comparative purification of the channel complex by DEAE and calmodulinaffinity chromatography 7025. Purification of the channel complex from ROS of various mammals 7126. Calmodulin affinity purification and Western blotting analysis of the cGMPgated channel complexes from bovine and frog ROS membranes 7227. Extraction of calmodulin by the immobilized cGMP-gated channelcomplex 7428. Effect of calmodulin on the activation of the cGMP-gated channel 8029. Effect of calmodulin on the activation of the rod channel by cGMP asdetermined by the dichiorophosphonazo Ill assay system 8230. Effect of calmodulin on the activation of the rod channel by cGIvIP asdetermined by the neutral red assay system 8331. Purification of recoverin from bovine ROS 8532. Effect of variousCa2-binding proteins on the cGMP-gated channelactivity 8633. Inhibition of the calmodulin effect by mastoparan 8734. Effect of calmodulin on the influx of various cations 8935. Calcium dependence of the calmodulin effect on the cGMP-gated channelactivity 9036. Effect of calmodulin on the cGMP-dependent activation of the reconstitutedchannel complex 98xii37. Inability of anti-240 kDa monoclonal antibodies to inhibitcalmodulin binding to the 240 kDa protein 10038. Chymotrypsin digestion of ROS membranes 10139. Trypsin digestion of ROS membranes 10340. Reconstitution of trypsinized ROS membrane proteins 10441. Effect of trypsin treatment of ROS membranes on the calmodulinmodulation of the channel activity 10542. Purification of the trypsinized channel complex by calmodulin affinitychromatography 10643. Digestion ofROS membranes with kallikrein 10744. Reconstitution of the kallikrein-treated ROS membrane proteins 10945. Purification of the kallikrein-treated cGIvJP-gated channel complex bythe anti-cx-subunit PMc 6E7 monoclonal antibody column 11046. N-terminal sequence alignment of the 105 kDa fragment with thecGMP-gated channel 13-subunit 11147. Possible role for calmodulin modulation of the channel during the visualtransduction process 11848. Schematic model of the cGMP-gated channel complex 121xiiiLIST OF ABBREVIATIONSATP adenosine 5’-triphosphateBCA bicinchoninic acidBSA bovine serum albumincAMP adenosine 3’ 5’-cyclic monophosphatecGMP guanosine Y 5’-cyclic monophosphateCHAPS (3 -[(cholamidopropyl)-dimethylammonio]- 1 -propanesulfonate)DEAE diethylaminoethylDFP diisopropylfluorophosphateDTT dithiothreitolECL enhanced chemiluminescenceEDTA ethylenediamine tetraacetic acidEGTA ethylene glycol bis-(f3-aminoethyl ether) N,N,N’,N’,-tetraacetic acidFCCP carbonyl cyanide p-trifluoromethoxyphenyl-hydrazoneFITC fluorescein isothiocyanateGDP guanosine diphosphateGMP guanosine monophosphateGTP guanosine triphosphateh hourHEPES (N-[2-hydroxyethyl]piperazine-N’-[2-ethanesulfonic acid])Ig immunoglobulinkDa kilodaltonmm minuteMr relative molecular weightPBS phosphate buffered salinePDE phosphodiesteraseROS rod outer segmentsxivSDS.sodium dodecyl sulphateTris (Tris[hydroxymethyl]aminomethane)xvACKNOWLEDGEMENTI would like to thank my supervisor Dr. Robert Molday for his superb guidancethroughout the course of my thesis investigation. I am also very grateful to my parentsand sister for their support and encouragement during the last four years. Also, I wouldlike to extend my appreciation to my committee members Dr. Ken Baimbridge and Dr.Peter Candido for their helpful advice and letters of reference. In addition I would like tothank other members of the laboratory, Laurie Molday, Dr. Delyth Wong, Dr. AndyGoldberg, Michelle filing, Andrea Dose, Orson Moritz, and Dr. Greg Connell for technicaladvice and helpful discussions. Finally, I would like to express my gratitude to othermembers of the Department of Biochemistry and Molecular Biology, Wendy Rodrigueza,Dr. Richard Barton, Dr. Surhail Ahmad, Dr. Archie Chonn, and Rachel Zhande for theirinvaluable technical assistance.1CHAPTER 1INTRODUCTION1.1. PHOTORECEPTOR ROD CELL1.1.1. StructureThe photoreceptor rod cell is a highly specialized neuron found in the peripheralregion of the retina. It functions to mediate vision under dim light conditions. In thehuman retina, the rod cell is generally 1-3 jim in diameter and 40-60 pm in length (Shichi,1983). It is divided into 4 major compartments: the outer segment, the inner segment, thecell body, and the synaptic terminal (Fig. 1). The outer segment is the primary site forvisual transduction. This compartment is joined to the inner segment via a nonmotilecilium containing microtubule doublets and actin and myosin filaments. This contractilesystem is believed to mediate the formation of new disks at the base of the outer segment(Williams, 1991). The inner segment contains the metabolic machinery of the cellincluding the endoplasmic reticulum, golgi apparatus, and mitochondria. Newlysynthesized ROS proteins are believed to be transported from Golgi apparatus in the innersegment via vesicles (Bird et at., 1988). These vesicles fuse with the ciliary plasmamembrane and the proteins are sorted to either disk or plasma membranes by an unknownmechanism. Mitochondria supply the majority of the energy, in the form of ATP, requiredby the various metabolic processes of the rod cell. The cell body is the compartment thatencloses the nucleus. Messenger RNAs from the nucleus direct the synthesis of variousrod cell proteins. The synaptic terminal of the rod photoreceptor cell contains numerousneurotransmitter vesicles, In the dark, neurotransmitters are continuously released toactivate the connecting bipolar and horizontal cells. The release of these neurotransmittersare inhibited in response to light.2inner SegmentCell BodyFig. 1: Schematic representation of a vertebrate photoreceptor rod cell.The photoreceptor rod cell is divided into four major compartments: the outersegment, the inner segment, the cell body, and the synaptic terminal (diagram provided byDr. R. S. Molday).Ftasnia IOuter SegmentGd ComØexNucleusSynaptic Vestcles Synaptic Terminal31.1.2. Rod outer segmentsEach outer segment is made up of a stack of hundreds of disks surrounded by aplasma membrane. At the base of the outer segment, evagination of the ciliary plasmamembrane produces continuously folded nascent disk membranes from which mature disksand the plasma membrane form (Steinberg el aL, 1980). A disk consists of two closelyspaced lamellar membranes joined together by a hairpin loop structure known as the rimregion. The disks have been shown to interact with the plasma membrane at these rimregions through filamentous structures (Roof and Heuser, 1982). In the outer segment,approximately 95 % of the membranes are disk membranes while only 5 % of themembranes are plasma membranes (Molday and Molday, 1987).Intact rod outer segments can be prepared free of other retinal cells and organellesby gently shaking the retinas in a sucrose solution followed by density gradientcentriftigation (Godchaux and Zimmerman, 1979; Molday and Molday, 1987). Theprotein components of the outer segments can be separated into soluble and membranefractions (Godchaux and Zimmerman, 1979; KUhn, 1980). Using a ricin-gold-dextranmembrane perturbation method, Molday and Molday (1987) have shown that themembrane fraction can be further separated into disk and plasma membranes. Many of theproteins found in the soluble and membrane fractions as listed in Table I (Molday andMolday, 1993), have been identified and characterized by various laboratories. Thesoluble fraction includes enzymes required for the visual cascade, glucose metabolism, andcalcium-dependent pathways. As for the membrane fraction, studies have shown that thedisk and plasma membranes, with the exception of rhodopsin, have quite distinct proteincompositions.In terms of lipids, the two membrane systems also appear to have distinctcompositions. Plasma membranes have been reported to have a high content ofcholesterol, unsaturated C18 (18:2; 18:3) fatty acids and the saturated C14 fatty acids,whereas the disk membranes are high in the saturated C18 fatty acid and unsaturated C224Table IRod outer segment proteinsProtein Mr. (daltons)Soluble and membrane-associated proteinsPhosphodiesterasece-subunit 88,00013-subunit 84,000y-subunit 11,000Protein kinase C 85,000Rhodopsin kinase 68,000Arrestin 48,000Creatine kinase 43,000Transducince-subunt 39,00013-subunit 37,000y-subunit 8,000Pyrophosphatase 44,000Glyceraldehyde-3-phosphate dehydrogenase 38,000Phosducin 33,000Recoverin 26,000Calmodulin 16,700Phosphatase 2A 38,000Other proteinsGlycolytic enzymesNucleotide diphosphokinasePentose shunt enzymesGuanylate kinaseTubulin 58,000Disk membrane proteinsRim protein 220,000Rhodopsin 38,000Peripherinlrds 35,000ROM-i 33,000Retinol dehydrogenase 37,000Plasma membrane proteinscGMP-gated channel-associated protein 240,000NaiK-Ca2exchanger 230,000cGMP-gated channel 63,000Glut-i glucose transporter 50,000Rhodopsin 38,000Guanylate cyclase? ii 2,0005(22:6) docosahexaenoic acid (Boesze-Battaglia and Albert, 1989). Thus, specific sortingmechanisms are likely to exist in the rod cell to direct the lipids and proteins to theirrespective membrane systems.1.2. PHOTOTRANSDUCTIONPhototransduction is the process by which the photoreceptors convert absorbedlight into electrical signals. In the dark, there is a continuous flow of current across thephotoreceptor plasma membrane. The direction of flow is inward at the outer segmentand outward at the inner segment (Fig. 2). By suction electrode recording studies, it wasshown that this circulating dark current reaches a value of 20-50 pA per vertebrate rodcell (Baylor et aL, 1979a, b). The inward current, mainly in the form of Na ions, iselectrically balanced by an outward K+ current at the inner segment. This is brought aboutby the maintenance of Na+ and K+ gradients across the inner segment plasma membraneby the electrogenic action of Na+/K+ ATPase. Sodium ions from the inward current arepumped out of the rod inner segment into the extracellular medium where the [Naj ismuch higher, in exchange for the translocation of K+ ions into the inner segment from theextracellular medium where the [K+] is lower. Potassium ions can then exit the innersegment down their concentration gradient via voltage sensitive K+channels in the plasmamembrane of the inner segment to complete the circuit.The Na current entering the outer segment is suppressible by light. Thissuppression is referred to as the photocurrent. Generally, it has been found that theisomerization of a single rhodopsin will give rise to a peak photocurrent of 0.2-1.5 pA. Inother words, the dark current will be suppressed by 0.2-1.5 pA. This photocurrent hasthree main features. First, the response of photocurrent is linear for 5-10 isomerization ofrhodopsin per rod cell (Penn and Hagins, 1972; Baylor et aL, 1979 a, b). Second,individual photons will cause the suppression of the dark current only over a restrictedlength of the outer segment due to compartmentalization of the photoresponse (Lamb and6Dark LightNaNaKFig. 2: Ryperpolarization of the rod cell in response to light.In the dark, Na+ enters the ROS through the plasma membrane specific cationchannels. Na+ then flows into the inner segment and is continuously extruded by Na+/K+ATPase in exchange for the influx of K (3 Na extruded for the influx of 2 Kj. The Ktranslocated into the inner segment by Na+JK+ ATPase then exits the inner segmentthrough the voltage-gated potassium channels. The continuous flow of this dark currentenables the release of neurotransmitter vesicles from the synaptic terminal. In the light,the photoactivated process causes the closure of the ROS cation channels. Thecontinuous action of N&1K+ ATPase to pump out Na+ from the inner segment will thenresult in the hyperpolarization of the ROS plasma membrane. This hyperpolarizationdiminishes the release of neurotransmitter vesicles from the synaptic terminal.Na7Yau, 1981). Thirdly, the single photon response is multistage in nature due the underlyingcomplex pathways of the visual cascade system (Fuortes and Hodgkin, 1964).1.2.1. Ca2vs cGMP as the primary messenger for photoexcitationIn the early 1970s, various laboratories directed much of their efforts towardidentif’ing the primary messenger which relays the photoabsorption event on the diskmembrane to a change in plasma membrane conductance. Two distinct hypotheses wereproposed at the time. The first hypothesis, proposed by Yoshikami and Hagins (1970),suggested that the divalent cation Ca2, is the primary messenger of photoexcitation. Intheir model (Fig. 3), it was suggested that in the dark, intracellular Ca2 concentration isrelatively low and there is a high accumulation of Ca2 within the disks. The lightactivated process causes the release of Ca2+ from the disks and an increase in theintracellular Ca2+ level. This increase in Ca2+ suppresses the dark current by directlyblocking the cation channels in the plasma membrane. The second hypothesis, also knownas the cGMP hypothesis, stemmed from the findings of several laboratories (Bitensky etal., 1971; Cohen et a!., 1978; Kilbride and Ebrey, 1979; Woodruff and Bownds, 1979).This hypothesis suggests that the light activated process induces a change in theintracellular cGIvlP concentration which ultimately affects the cation conductance of theplasma membrane (Fig. 4). The first hypothesis was eventually disproved by a number ofexperimental results. First, it was shown that Ca2+ closed the channels more slowly thandid light (Yau and Nakatani, 1985). Second, ROS depleted of Ca2 could still respond tolight (Matthews et a!., 1985; Cote et a!., 1985). Third, the cation channels on excisedROS plasma membrane patches are opened by the addition of cGMP, but not by theaddition of Ca2 (Fesenko et a!., 1985). The cGMP hypothesis, together with thediscovery of the various visual cascade enzymes, formulated the present model of visualtransduction.8LightFig. 3. Calcium hypothesis for visual excitation.In the dark, calcium ions are believed to be sequestered within the ROS disks andthis maintains a relatively low level of Ca2 within the cytoplasm. Under this condition,Na+ flows into the outer segment unhindered through the plasma membrane specificcation channels. In the light, the light activated process will cause the release of Ca2from the disks. The elevated level of cytoplasmic Ca2 will then prevent further entry ofNa by directly blocking the cation channels.DarkNa+NaNaNa• •••• •.•9.Fig. 4. cGMP hypothesis for visual excitation.In the dark, the elevated level of cGMP maintains a significant number of cationchannels in their open state and this allows the influx of Na+ into the outer segment. Inthe light, the photobleaching of rhodopsin activates the visual cascade resulting in adecrease in the level of cGMP. This decrease in the level of cGIvlP will then cause theclosure of the cation channels and prevent further influx ofNa+ into the outer segment.DarkI.cNaI••Na+ IL•CicNa?— NaLight=-==101.3. VISUAL TRANSDUCTION CASCADEThe visual transduction process is divided into two major stages: the initial lightstimulated excitation step and the subsequent recovery to the dark resting state. Briefly,the visual excitation cascade (Fig. 5) is initiated when light induces the photoisomerizationof the 1 1-cis-retinal chromophore of rhodopsin to its all trans-form. This photoexcitedform of rhodopsin then activates transducin by catalyzing the exchange of GTP for GDPon the cx-subunit of transducin. Activated transducin turns on the cGMP-dependentphosphodiesterase which catalyzes the hydrolysis of cGMP to 5-GIvW. In the dark, anelevated level of cGMP maintains the cGMP-gated channels in their open state and allowsthe influx of Na and Ca2 into the outer segment. The light-stimulated decrease in thecGMP level causes the closure of these cGMP-gated cation channels. Closure of thesecation channels results in the hyperpolarization of the rod outer segment plasma membraneand initiates the transmission of electrical signals. Recovery of the rod cell to its darkresting state involves the deactivation of rhodopsin, transducin, phosphodiesterase, andthe stimulation of guanylate cyclase to resynthesize cGMP required for the reopening ofthe cGMP-gated channels.1.3.1. Photoactivation of rhodopsinVertebrate rhodopsin, a membrane protein of 38 kDa, is present in high abundancein ROS disk membranes. In the dark, it is conjugated to an 1 1-cis-retinal chromophore viaa Schiff base to a specific lysine residue. In the case of bovine rhodopsin, the linkage is atlysine 296, in the middle of the seventh transmembrane helix (Thomas and Stryer, 1982).Upon absorption of a photon, 1 1-cis-retinal isomerizes to its all trans isomer within 200femtoseconds (Schoenlein et al., 1991). This isomerization of retinal leads to theformation of various rhodopsin photolysis intermediates beginning with bathorhodopsinand proceeding on to form lumirhodopsin, metarhodopsin I and finally to metarhodopsinII (Lamola et al., 1974). Formation of metarhodopsin II then leads to changes in the11cGMPPhosphodesterose / “,c1GW5’GMP’Fig. 5: A diagram showing the basic reactions of the visual transduction pathway.In the dark, elevated cGIvIP concentrations maintain a number of cGIVIP-gatedchannels in their open conformation. This allows the influx ofNa+ and Ca2+ into the outersegment and maintains the outer segment in a partially depolarized state. Photobleachingof rhodopsin, involving the isomerization of 11 -cis retinal to its all-trans isomer, results inthe formation of activated (meta II) rhodopsin which catalyzes the exchange of boundGDP for GTP on transducin. Transducin will then activate phosphodiesterase whichcatalyzes the hydrolysis of cGMP to GMP. The decrease in free cGMP concentrationwithin the outer segment will cause a closure of the cGMP-gated channels to the influx ofNa+ and Ca2+ and a transient hyperpolarization of the outer segment. The recovery of theROS to its dark resting state occurs through 1) the inactivation of rhodopsin by arhodopsin kinase (RK) catalyzed phosphorylation reaction and the binding of arrestin(Ar); 2) inactivation of transducin by hydrolysis of bound GTP to GDP; 3) inhibition ofphosphodiesterase by rebinding of the inhibitory subunits to the catalytic subunits; 4)resynthesis of cGMP from GTP by gunaylate cyclase (GC); and 5) the reopening of thecGMP-gated channel as cGMP levels increase.Exchanger4 NaNaCa’NaDisk Membrane++44+44++Plasma Membrane12environment of various rhodopsin amino acids (Chabre, 1985; Chen and Hubbell, 1978)and the dissociation of retinal from the opsin (Bownds, 1967; Cooper eta!., 1987).1.3.2. Activation of transducinPhotoreceptor ROS transducin is a member of the heterotrimeric GTP bindingprotein family. Each transducin consists of an ce-, f3-, and y-subunit. In its inactive form,transducin is peripherally associated with the ROS disk membrane and its ce-subunitcontains a bound GDP molecule. Upon photoactivation, metarhodopsin II associates withthe cL-subunit of transducin via rhodopsins cytoplasmic loops (KUhn and Hargrave, 1981;Konig et a!., 1989). Metarhodopsin II catalyzes the exchange of GTP for GDP on the csubunit of transducin (KUhn, 1980). The binding of GTP to transducin results in aconformational change which causes the cL-subunit of transducin to dissociate from boththe f3, y-subunits and metarhodopsin II (KUhn, 1980). The free metarhodopsin II can theninteract with another transducin complex. A single metarhodopsin II can activate severalhundred transducin molecules. This represents the first step of signal amplification invisual activation.1.3.3. Hydrolysis of cGMP by PDEPDE in ROS is composed of one x-, one f3-, and two ‘y-subunits. The cGMPhydrolysis function of the enzyme resides in the c- and f3- subunits while the ‘y-subunits actas inhibitory subunits (Hurley and Stryer, 1982). In addition to its catalytic site, PDE alsohas two additional high affinity noncatalytic cGMP binding sites (Yamazaki et al., 1980).The phosphodiesterase complex becomes activated when the two y inhibitory subunits areinactivated by binding to two GTP-Ta subunits (Wensel and Stryer, 1990). Removal ofthe inhibitory restraint of these subunits increases the phosphodiesterase activity by 50 to100 fold (Baehr et a!., 1979; Hurley and Stryer, 1982). Approximately, one hundred13phosphodiesterase molecules are activated from a single photon absorption. Each PDEmolecule can lead to the hydrolysis of up to io cGIVIP molecules.1.3.4. Closure of the cGMP-gated channelThe free cGMP concentration in the ROS is approximately 4-10 p.M. Thismaintains a significant number of the cGMP-gated channels in their open state (Nakataniand Yau, 1988a). The opening of these channels allows the influx of Na and Ca2 intothe outer segment. The activation of phosphodiesterase leads to a rapid hydrolysis ofcGMP to 5’-GIYIP. This results in a local depletion of cGMP and causes the closure of thenearby plasma membrane cGMP-gated channels.1.3.5. Deactivation and regeneration of rhodopsinTo turn off the visual activation cascade, rhodopsin, transducin, andphosphodiesterase have to be inactivated and cGMP has to be regenerated. Inactivationof photoactivated rhodopsin involves the phosphorylation of rhodopsin followed by thebinding of arrestin to phosphorylated rhodopsin. Phosphorylation of rhodopsin has beenshown to be a light-dependent process (KUhn and Dreyer, 1972; Bownds et a!., 1972).This phosphorylation reaction occurs at serine and threonine residues and is universalamong various species. More than 85% of phosphate incorporation is found at the C-terminal region of rhodopsin (Hargrave et a!., 1980). Under conditions of high level ofrhodopsin bleaching, the extent of phosphorylation is proportional to the fraction ofrhodopsin being bleached with a maximum of 9 moles of phosphate incorporated into eachmole of rhodopsin (Wilden and KUhn, 1982).Rhodopsin undergoes a conformational change upon photobleaching. However,rhodopsin cannot be phosphorylated until the formation of metarhodopsin II (Yamamotoand Shichi, 1983; Seckler and Rando, 1989). This conformational change is necessary toallow the cytoplasmic surface of the opsin to interact with rhodopsin kinase. Rhodopsin14kinase is a serine and threonine kinase that is dependent on Mg2+. It is a member of a newclass of receptor kinases that phosphorylate the G-protein coupled receptors (Palczewskiand Benovic, 1991).Phosphorylation of rhodopsin is insufficient to completely stop the activation oftransducin by metarhodopsin II (Wilden et at., 1986), although a graded decrease intransducin activation is associated with an increasing level of rhodopsin phosphorylation(Miller et at., 1986). A complete inhibition of metarhodopsin’s ability to activatetransducin requires the further binding of arrestin to the phosphorylated form ofmetarhodopsin II (KUhn et at., 1984). Arrestin dissociates from metarhodopsin II whenthe all-trans-retinal chromophore is reduced to all-trans-retinol by retinol dehydrogenase(Hofmann et at., 1992, Ishiguro et aL, 1991). Upon arrestin dissociation, metarhodopsinII is dephosphorylated by phosphatase 2A (Palczewski et a!., 1989). All-trans-retinol istransported to the retinal pigment epithelial cells and converted back to 11 -cis-retinal (Bokand Heller, 1976). The newly synthesized 11 -cis-retinal is then transported back into theROS where it reassociates with opsin to regenerate rhodopsin.1.3.6. Deactivation of transducin and phosphodiesteraseThe ci-subunit of transducin has intrinsic GTPase activity. It is believed that GTPhydrolysis deactivates GTP-Tcx (Dratz et a!., 1987; Vuong and Chabre, 1990). Theresulting inactive GDP-Tc complex dissociates from the y-subunit of phosphodiesterase.This free GDP-Tc subunit can then reassociate with the 13- and y- heterodimers oftransducin. However, some studies have suggested that the rate of GTP hydrolysis is tooslow and that Tcc inactivation may require an alternate pathway (Erickson et a!., 1992).The catalytic activity of phosphodiesterase is inactivated when the y inhibitory dimersreassociates with the ci- and f3-. catalytic subunits of phosphodiesterase.1.3.7. Resynthesis of cGMP15Guanylate cyclase is a key enzyme responsible for visual recovery. This enzymecatalyzes the formation of cGIvIP from GTP. The activity of this enzyme is inhibited bypyrophosphate, a by-product of the cGMP synthesis (Hakki and Sitaramayya, 1990), andis dependent on the level of intracellular Ca2 (Koch and Stryer, 1988). The resynthesisof cGMP by guanylate cyclase results in an increase in the level of cGMP. This elevatedlevel of cGMP reopens the cGMP-gated channels to allow the influx ofNa and Ca2 intothe outer segment.1.3.8. Roles of calcium in visual transductionIn the dark, cytoplasmic Ca2 in the photoreceptor ROS is kept at a constant levelby two transporting systems. Influx of calcium ions entering the outer segment via thecGMP-gated channels are counterbalanced by their constant extrusion through the plasmamembrane specific Na+/Ca2+K+ exchanger. This maintains the cytoplasmic Ca2+ at aconcentration of between 0.2-0.6 .tM (McNaughton et at., 1986; Ratto et at., 1988).When light triggers the visual activation cascade, a decrease in cGMP concentrationresults in the closure of the cGMP-gated channels. The closure of the cGIVIP-gatedchannels prevents the influx of Ca2 into the outer segment. Since the activity of theNa+/Ca2+.K+ exchanger is not affected by the light activated process, Ca2+ continues tobe extruded and the Ca2+ level within the outer segment decreases to below 0.1 tM.Koch and Stryer (1988) proposed that the main target of this Ca2 flux betweenlight and dark conditions is guanylate cyclase. The activity of this enzyme was shown toincrease 5-20 fold when Ca2 is lowered from 300 nM to 20 nM. This effect appeared tobe mediated through a soluble factor. It was proposed that in the dark when the Ca2level is high, this factor binds to Ca2 and is maintained in its inactive state. Guanylatecyclase is in turn maintained at its basal level of activity (Fig. 6). When thephotobleaching of rhodopsin results in a decrease in the level of Ca2+ through the closureof the cGIvlP-gated channels, Ca2 dissociates from this soluble factor. This activated16light ..—, Rho —. oc CGMP 4Ca4MIGC (basal) —. GC (acilve)IGW GMPtGC (basal) 4— GC (ocltve)tCaFig. 6: Calcium-dependent modulation of guanylate cyclase.In the dark, elevated levels of cGMP maintain a relatively high concentration ofCa2 within the outer segment through the opening of the cGMP-gated channels. Themodulator (M) of guanylate cyclase binds Ca2 and is in its inactive state. Under theseconditions, guanylate cyclase is kept at its basal level of activity. When the light activatedprocess causes a decrease in the level of cGMP ‘D and the closure of the cGMP-gatedchannels, the continuous action ofNa+/Ca2+K+ exchanger will then result in a decrease inthe level of Ca2 within the outer segment . The dissociation of Ca2 from thismodulator will activate it and this protein will then stimulate the resynthesis of cGMPfrom GTP . As the level of cGMP rises within the outer segment, the cGMP-gatedchannels will reopen and replenish the level of Ca2 within the outer segment to its darkresting state ©. The rebinding of Ca2 to the modulator will once again deactivate it andminimize the stimulation of guanylate cyclase activity .INo-aCoNo17factor then stimulates guanylate cyclase to resynthesize cGMP. Once the cGMP-gatedchannels reopen in response to an elevated level of cGMP, the Ca2+ level within the outersegment rises. This allows Ca2+ to rebind to the guanylate cyclase activating factor andrestores the guanylate cyclase to its basal level of activity.A 26 kDa Ca2 binding protein, recoverin, was originally believed to be this Ca2-dependent modulator of guanylate cyclase (Dizhoor et al., 1991; Lambrecht and Koch,1991). However, subsequent studies indicated that recoverin is unlikely to be themodulator of guanylate cyclase (Hurley et a!., 1993). Recombinant recoverin, expressedin E. coli., failed to stimulate rod guanylate cyclase. In addition, at the low Ca2concentrations required to stimulate guanylate cyclase, recoverin dissociates from theROS membranes. More recently, it has been shown that recoverin affects instead,phosphodiesterase activation in a Ca2+dependent manner, through regulating rhodopsinphosphorylation (Kawamura, 1993). With increasing Ca2 concentrations, this proteinwas found to decrease the extent of rhodopsin phosphorylation and thus have a directeffect on the rate of phosphodiesterase activation. This modulation is thought to play arole in visual adaptation, a phenomenom characterized by a decrease in sensitivity of thephotoreceptor cell in the presence of background illumination (Fain and Matthews, 1990;Pugh Jr. and Lamb, 1990; Shapley and Enroth-Cugell, 1984).1.4. THE cGMP-GATED CHANNEL COMPLEX1.4.1. Properties of the cGMP-gated channelThe cGMP-gated channel of ROS is a cation selective channel that does notdiscriminate well between monovalent cations (Fesenko et a!., 1985; Hodgkin et a!.,1985). The relative permeabilities of various monovalent cations have been measured andare shown in the decreasing order: NH+4 > Li+ Na+ > K+ > Rb+> Cs+. The cGMPgated channel is also permeable to divalent cations such as Ca2+ and Mg2+. In the dark,the inward current is mainly carried by Na+ and to a lesser degree by Ca2+ and Mg2+18(Hodgkin et al., 1985; Nakatani and Yau, 1988a; Cervetto et al., 1988). Suppression ofthe Na component of this current results in the hyperpolarization of the rod cell. TheCa2+ component plays a role in visual recovery, while the Mg2+ component has no knownfunction in phototransduction so far.Voltage clamping analyses of isolated ROS indicate that the current-voltagerelation shows an outward rectification (Bader et a!., 1979; Bodoia and Detwiler, 1985;Baylor and Nunn, 1986). At negative voltages, the current-voltage relation is flat,implying that the rod dark current is insensitive to physiological changes in membranepotential (between -40 mV in dark to -80 mV in light, Cervetto and Fuortes, 1978). Thisrectification of the cGMP-gated channels together with the absence of leakage channelsallows the effective propagation of the light-induced hyperpolarization through the outersegment (Yau and Baylor, 1989).Single-channel current recordings of the rod plasma membrane patches at lowcGMP concentrations (Haynes et a!., 1986; Zimmerman and Baylor, 1986) indicate thatthe channel has a maximum single channel conductance of 25 pS in the absence of divalentcations. The channel openings last for a few milliseconds and display a characteristicflickery nature. This suggests that the cGMP molecules may be loosely bound to thechannel. This fast gating kinetics of the channel improves the signal to noise ratio forphoton detection and also allows the rapid propagation of the photoresponse. Singlechannel activity is suppressed when a low concentration of Ca2+ or Mg2+ is added.Bursts of high-frequency flickering events are observed in place of the single channelopening events, suggestive of the intermittent blocking of the open channel by divalentcations. At physiological potentials, the unit conductance is lowered from 25 to 0.1 pSand this translates into a decrease of unitary current from a maximum of 1 pA to 4 fA.Thus, a large number of channels must be open to maintain the dark current at a fixed size.This effectively reduces the level of background noise.19Several pharmacological agents have been shown to block the cGMP-gatedchannel conductance. One of these inhibitors is the Ca2+ channel antagonist 1-cis-diltiazem which blocks the conductance at micromolar concentrations in both the ROSmembrane preparations and excised ROS plasma membrane patches (Koch and Kaupp,1985; Stern eta!., 1986). However, it does not inhibit the activity of the purified channelsthat have been reconstituted into lipid vesicles (Cook et a!., 1987). The compound, 3’,4’-dichlorobenzamil, which is a derivative of the Na+ channel blocker amiloride has also beenreported to block channel activity at micromolar range (Nicol et at., 1987). Morerecently, Nicol (1993) has shown that pimozide, a Ca2 channel and calmodulinantagonist, is also an effective blocker of the rod cGMP-gated channel in the micomolarrange.1.4.2. Purification and localization of the cGMP-gated channelThe rod cGMP-gated channel was first purified by Cook et a!. (1987) fromCHAPS solubilized ROS membranes by DEAE anion exchange and TSK AF-red affinitychromatography. The purified channel preparation, as visualized on SDS polyacrylamidegels, consisted of a prominent 63 kDa and a fainter 240 kDa band. This preparationexhibited cGIvIP-dependent calcium flux activity when reconstituted into lipid vesicles,Further studies by Molday et a!. (1990) revealed that these two proteins are tightlyassociated to one another. The channel complex, consisting of the 63 kDa and the 240kDa proteins, could be copurified by immunoaflinity chromatography using either an anti-63 kDa cGMP-gated channel monoclonal antibody PMc 1D1 or an anti-240 kDa proteinmonoclonal antibody PMs 4B2. More recently, Brown et a!. (1993) demonstrated thatboth the 63 kDa and the 240 kDa proteins can be photoaffinity labelled by 8-p-azidophenacylthio-cGMP, an analog of cGIvIP. This suggests that the cGMP-gatedchannel complex may consist of at least two distinct subunits, each with the capability tobind cGMP.20Various biochemical and immunocytochemical analyses using the anti-63 kDacGMP-gated channel monoclonal antibody PMc ID1 and the anti-240 kDa proteinmonoclonal antibody PMs 4B2 have been carried out to study the distribution of thechannel in rod outer segments. The results indicate that the cGMP-gated channel complexis exclusively localized to the ROS plasma membrane and its density is approximately 300tm2 (Cook eta!., 1989; Molday eta?., 1990).1.4.3. Molecular cloning of the cGMP-gated channel complexThe sequence of the 63 kDa channel subunit was determined by Kaupp et aL(1989) by screening a retinal cDNA library using degenerate oligonucleotide probesgenerated from the partial amino acid sequences of the tryptic channel fragments. Thefull length cDNA encodes a membrane protein of 690 amino acids (Mr. 79,601).Homologues of this channel subunit have been cloned and sequenced from several otherspecies. The mouse and human rod channel sequences are 85% identical to the bovinesequence (Pittler et a!., 1992; Dhallan et a?., 1992) while the chicken rod and conesubunits share 76% and 65% identity with the bovine sequence, respectively. When thefull length eDNA was transcribed into mRNA and microinjected into Xenopus oocytes, theexpressed channel displayed many of the electrophysiological properties of the ROSchannel (Kaupp eta!., 1989). It was observed to be activated by cGMP with a K112 of 52p.M and a cooperativity of 1.75. However, it was less sensitive to the channel blocker, lcis-diltiazem, and it did not display rapid flickering of the channel conductance.Until recently, it was believed that the channel was composed of a homooligomeric complex. Chen et aL (1993), however, demonstrated the presence of a secondsubunit (designated as the 13-subunit) of the channel by homology cloning from a humanretinal library. The human 13-subunit shares 30% homology with the original bovinechannel (now designated as the x-subunit). The full length cDNA of the [3-subunitencodes a membrane protein of 909 amino acids (Mr 102,330). Alternate splicing21produces a shorter protein product that has its first 286 amino acids at its N-terminusmissing. The relative distribution of these two forms of the 13-subunit in ROS, however, isnot known. When this f3-subunit was transiently expressed in human kidney 293 cells, theexpressed channel was inactive as determined by patch clamping study. But, when this f3-subunit was coexpressed with the a-subunit, the resulting complex displayed ionconductance properties that were more similar to the native ROS channel. In particular,the coexpressed a- and 13-subunits displayed rapid flickering of the channel conductanceand increased sensitivity to inhibition by l-cis-diltiazem.1.4.4. Structural analysis and topography of the cGMP-gated channel subunitsSequencing analysis of the a- and 13-subunits of the cGMP-gated channels ofvarious species revealed several interesting features. Stretches of 80-100 amino acidsclose to the C-termini of the channels have been identified as the cGMP binding sites(Kaupp et at., 1989; Chen et aL, 1993; also see Fig. 7), on the basis of their similarity tothe putative cGIvlP binding domain of the bovine lung cGMP-dependent protein kinase(Takio et at., 1984). Hydropathy plots of the channel sequences suggest the presence of 5hydrophobic segments, designated as Hi - H5, which are of sufficient length to span thelipid bilayer. In the bovine cL-subunit, there are 5 consensus sequences for N-linkedglycosylation at Asn residues 90, 91, 177, 327, and 423. However, by lectin binding,endoglycosidase treatment, and immunochemical studies, the channel a-subunit is likely tobe N-linked glycosylated only at Asn position 327 (Wohlfart et at., 1989, 1992). Incontrast, there is no consensus sequence for a N-linked glycosylation site in the human 13-subunit.The full length cDNA of the bovine a-subunit encodes an 80 kDa protein.However, the purified ROS membrane channel subunit generally displays a molecularweight of 63 kDa on the SDS polyacrylamide gels. N-terminal sequence analysis of thepurified a-subunit revealed the absence of the first 92 amino acids from the N-terminus of22CYCLIC NUCLEOTIDE BINDING DOMAINBovine Rod (a) Y5PGDYICKKGDtGREKTflKzOKLkVVAD-DGrtQPVL$DGSypG!ZBIU(txGSr3.RTJiMxEsxGysDLrcLsw (499.577):::::::: ::Human Rod (a) YSPGDYICKKGDIGREMYIIKEGKLkVD-DGV7QPVVL8DGsypGEXSILNxxQ$zkGa5RT)yxEgxGyBDLycLs (496-575):::::::::::::s:::::::::::::::: :::::::::::::::::::::s:::s:s:::;::::,,:::::::::::sMouse Rod (a) YBPODYICKKGDIGREMYXXK!GxLVVaD—DGvrQPVVLSDGSypGflXaILNIKa5KkQpRT7ijlxz5xGygDLrCx.s (491470):1:::::: ::::::::::::::::;::t: :::::::::::::;::::::;::::::::z :::i:::::::::Chicken Rod (cx) YBPGDYICRKGDIGREMYIXK!GKIJ.VVAD-DGVTQPVVLBDGBYPGBXBILNXKGBKkGRR?PJItRSXGTBDLPCX.SKD (452431):::::::::::::::::::.::::::z: :::::::::::::::::::s::::: ::::::: :S:t:::::::::Chicken Cone (cx) (545424)::::::: ::::g :::::::::::::: :: :: :: :H;::::::::: ::::::::Bovine Olfactory FBPODYICRKGDxGKEMYIxzEGKLAvD-DGvTQThLLskoscFonxsxzaizxasFJomTaxRsLGTsoLrcLsID (475.554):: :::: :::::::s : : :: :: :s::: : ::::::: : ::Human Rod () YLPNDYVCKKGEXGREMYIIQAGQVQVLGGPDGKsVLVTIQBVpGEXSLLAvGG---oRRTNIvGPJ.WXLDKI (365.432)Fig. 7: Sequence alignment of the cyclic nucleotide binding domains of channelsfrom different species.Alignment of the potential cyclic nucleotide binding sites of rod and cone csubunits, the rod 3 subunit, and the olfactory subunit. Identical amino acids are indicatedby ( : ). A high degree of sequence identity within the nucleotide binding domain isobserved.23the channel. The ct-subunits of other mammalian species also display an apparent Mr. 63K on the SDS polyacrylamide gels, thus suggesting that this type of N-terminal truncationmay be universal for mammalian photoreceptor channels (Molday et a!., 1991). Analysisof the chicken rod and cone cL-subunits also indicate the occurrence of some type ofcleavage reactions (BOnigk et aL, 1993). The cleavage site between Ser 92 and 93 of thebovine sequence, N-N-S-S-N-K-E, is conserved in other mammals (Fig. 8). Related butnonidentical sequences are also found in the ct-subunits of chicken photoreceptors.A modified voltage sensor motif S4, as found in the voltage-gated shaker Kchannel, is present in both the photoreceptor and olfactory nucleotide gated channelsubunits (Jan and Jan, 1990; Kaupp et a!., 1989). In voltage gated channels, this motifgenerally consists of up to 7 repeats of a positively charged arginine or lysine residueseparated by two hydrophobic amino acids. This motif is thought to span the lipid bilayer.For the photoreceptor and olfactory channel subunits, there are only four of these repeatsand the motif also contains several negatively charged residues (Fig. 9). It is likely thatthis divergence of the S4 motifs of the photoreceptor and olfactory channels from itsvoltage gated channel counterpart is responsible for their insensitivity to physiologicalchanges in membrane potential (Kaupp, 1991).In addition, the photoreceptor and olfactory channel subunits also contain asegment (Fig, 9) which shows considerable similarity to the pore region of the voltagegated K channel (Goulding et a!., 1992; Heginbotham et a!., 1992; Bonigk et a!., 1993).The pore region has been suggested to consist of two antiparallel 13 strands connected by ahairpin turn (Yellen et a!., 1991; Durell and Guy, 1992). The pore regions of theindividual subunits are likely to line the internal cavity of the channel complex in order totranslocate ions across the membrane.Based on immunocytochemical labelling, biochemical, and sequence analysis, amodel showing the structural features of the ct-subunit of the cGMP-gated channel hasbeen proposed (Bonigk et aL, 1993) as shown in Figure 10. The H1-H5 hydrophobic24N-TERMINAL CLEAVAGE SITEBovine Rod (a)Human Rod (a)Mouse Rod (a)Chicken Rod (a)Chicken Cone (a)NNSSNKENNSSNKDNNNSNKDSNNTNED(90-96)(89-95)(84-90)(53-59)(145-151)Fig. 8: Sequence alignment of the N-terminal cleavage sites of the cGMP-gatedchannels from different species.Alignment of the N-terminal cleavage sites from bovine, human, mouse, andchicken rod and cone cGMP-gated channels.25VOLTAGE SENSOR MOTIF (S4)Shaker KBovine Rod (a)Human Rod (a)Mouse Rod (a)Chicken Rod (a)Chicken Cone (a)Bovine OlfactoryHuman Rod (3)(360-380)(267-287)(265-285)(260-280)(221-241)(314434)(244-264)(132-152) *Fig. 9: Sequence alignment of the putative voltage sensor motifs (S4) and poreregions of the cyclic nucleotide gated channels.The voltage sensor motifs (S4) and pore regions of the bovine rod channel werecompared with those of the voltage-gated Shaker K channel and other cyclic nucleotidegated channels.ILRVIRLVRVFRIPKLSRH SR: : : :EIRLNRLLRISRMPEFPQRTB: : : : :EIRLNRLLRPSRMPEFYQRBs:::::EIRLNRLLRISRKPBFFQRTEELRINRLLRVARMYEPFQRTES ::E LRPNRILRIkRLfE PPDR TB: : :::: :::::: 5::EVR?NRLLRFARM?EFPDRTB:: : S :: : SLLRLPRCLRYMAFPBFNSRLEPUTATIVE POREShakerlc’ DAFWWAVVTMTTVGYODMTPV (431.451)BovineRod(a) YSLYWSTLTLTTIG--ETPPP (349-367):: :s : : S S S : : : S S :Humanflod(a) YSLYWSTLTLTTIG--ETPPP (347465): : : :: : : :: S : : : : ::MouseRod(a) YSLYWSTLTLTTIO--ETPPP (342-360): : : S S S : S : :: : S :S S : SChickenRod(a) YSLYWSTLTLT?XGETPPP (303421)S S : :: S : : : : : S S S S S : : SChickenCone(a) YSLYW8TLTLTTIG--ETPPP (390-414)BovirieOlfactory YCLYW8TLTLTTIGETPPP (326-344)HumanRod(á3) RCYYFkVKTLITIGOLPDP (206-224)*26segments and the S4 motif are likely to span the membrane while the putative pore isviewed as extending into the membrane in an extended 13 conformation. The N-terminaland C-terminal segments of the channel are both oriented to the cytoplasmic side. TheAsn 327 glycosylation site is exposed to extracellular environment. The cGMP bindingdomain of the channel, located near the C-terminus, is likely to interact with the pore tocontrol cation translocation. The 13-subunit, which displays sequence homology with thect-subunit, probably has similar structural features.1.5. CALMODULINCalmodulin was first discovered as a component required to activate cAMPphosphodiesterase from bovine brain (Cheung, 1970). Subsequently, it was shown by Teoand Wang (1973) to be a Ca2-binding protein that confers Ca2 sensitivity to thisenzyme. Calmodulin has been purified from bovine brain by either the conventionalchromatographic methods such as DEAE anion exchange, hydroxyapatite, and gelfiltration chromatography or by more specific affinity chromatographic methods usingligands such as troponin I and melittin (Grand et at., 1979; Kincaid, 1987) or antipsychoticdrugs such as phenothiazine, trifluoroperazine, and chloropromazine (Charbonneau andCormier, 1979; Jamieson and Vanaman, 1979) immobilized onto agarose.1.5.1. Molecular cloning of calmodulinThe first calmodulin cDNA sequence was obtained from the electroplax tissue ofthe electric eel (Munjaal et at., 1981). Subsequently, the calmodulin cDNA sequenceswere determined from a variety of species such as chicken (Putkey et a!., 1983), toad(Chien and Dawid, 1984), rat (Nojima and Sokabe, 1987), human (Wawrzynczak andPerham, 1984), fruit fly (Smith et aL, 1987), yeast (Davis et a!., 1986), slime mold(Goidhagen and Clarke, 1986), and trypanosome (Tschudi et at., 1985). The amino acidsequences of these calmodulins show a high degree of homology with one another.27Fig. 10: Working model for the organization of the tx-subunit of bovine rod cGMPgated channel within the lipid bilayer.The first 92 amino acids (dashed line) as predicted from cDNA sequence analysisare absent in the channel in ROS membranes possibly by a posttranslational cleavagereaction. Segments labeled Hi- H5 and the S4 voltage sensor motif are possibletransmembrane ct-helix segments and the segment between H4 and H5 is the putative poreregion. Immunogold labelling studies have established the N and C terminus on thecytoplasmic side and Asn-327 containing an N-linked oligosaccharide on the extracellularside. The cGMP binding domain is located near the C-terminus.PoreExtrocellularIntraceflular281.5.2. Structure of calmodulinCalmodulin is a protein of 148 amino acids. It is dumbbell shaped with two of itsglobular lobes joined by an 8-turn central os-helix. It has 4 structurally and functionallysimilar domains (Fig. 11) which are related to the Ca2-binding motif of parvalbumin(Vanaman et at., 1977). Each Ca2-binding domain, also known as an EF-hand, consistsof a loop of 12 amino acids flanked by two (x-helices that are oriented at approximately900 to one another (Kretsinger and Nockolds, 1973). Each lobe has two EF-hands andwithin each EF-hand, the negatively charged residues and a main-chain carbonyl oxygen ofthe loop form a pentagonal bipyramid with six coordinating residues to facilitate Ca2binding. Five of these residues coordinate a Ca2 or Mg2 ion while the sixth coordinatesa water molecule (Kretsinger et at., 1988).Calmodulin has 7 a-helices spanning residues 7-19, 29-40, 46-55, 65-92, 102-1 12,119-128, and 13 8-148. The helices are separated by non helical stretches containing 6-9amino acids. These helices are involved in the Ca2-binding helix-loop-helix structure.The central x-helix is shared by domains II and III (a. a. 41-64 and 93-118 respectively).Beta sheet structures occur between the pair of domains in each lobe. In addition,calmodulin has four reverse turns within eachCa2-binding loop (Babu et al., 1985).Within each half of calmodulin, there is a hydrophobic cleft. The aromatic rings ofPhe 19 and 68 are situated at the surface of the cleft in the N-terminal half while Phe 92and 114 are found at the cleft surface in the C-terminal half The entrance leading to theseclefts are blocked from the solvent by lysine residues. It is believed that the binding ofCa2 by calmodulin exposes these hydrophobic clefts to allow them to interact with thetarget sites (Babu etaL, 1985).1.5.3. Post-translational modificationCalmodulin undergoes various post-translational modifications. The N-terminus ofcalmodulin has been shown to be acetylated (Watterson et al., 1980). In addition, most29N-terminusaçX Ca’h/Ca2C-terminusCCa2Ca2b1 ADQLTEEQIAEFKEAFSLFDKDGDG26 TITTKELGTVMRSLGQNPTEAELQD51 MINEVDADGNGTIDFPEFLTMMARK76 MKDTDSEEEXREAFRVFDKDGNGYI*101 SAkE LRHVNNLGEKLTDEEVDEMI126 R S A D I D G D G 0 V N Y S S F V Q N N T A KFig. 11: Structure and sequence of calmodulin.a. A schematic diagram of calmodulin, Calmodulin is dumbbell shaped with 2globular lobes, each containing two Ca2-binding sites (Babu et al., 1985). b. Bovinecalmodulin is a protein of 148 amino acids. The Ca2-binding loops of the 4 EF-hands(amino acids 20-31, 56-67, 93-104, and 129-140) are underlined. The first amino acid,alanine, is acetylated and the protein is trimethylated at lys-1 15 (*).30calmodulins are trimethylated at lysine position 115. This N-methylation reaction iscatalyzed by S-adenosylmethionine:calmodulin N-methyltransferase (Sitaramayya et a!.,1980). Although this modification does not affect its modulation of the cAMPphosphodiesterase activity, this methylation has been implicated in its modulation of NADkinase activity (Rowe et a!., 1986; Roberts et a!., 1986). Calmodulin is also a substratefor protein carboxylmethyltransferase at either aspartic or glutamic acid residues (Freitagand Clarke, 1981). Finally, the carboxyl terminal lysine of calmodulin can be cleaved by acalmodulin converting enzyme, resulting in a change in mobility on SDS polyacrylamidegels (Murtaugh eta!., 1983).1.5.4. Calcium binding to calmodulinCalcium binding to calmodulin has been examined by a variety of methodsincluding equilibrium and flow dialysis and spectroscopic analysis (Yoshida et a!., 1983;Minowa and Yagi, 1984; Burger et a!., 1984; lida and Potter, 1986; Ogawa andTanokura, 1984). At physiological concentration of salts, the Ca2-binding curvesindicate that there are four apparent binding sites with Kd in the range of 5 x 10-6 to io-M. The binding of Ca2 to calmodulin appears to be a sequential event leading to astepwise Ca2+induced conformational transition of calmodulin (Klee and Vanaman,1982). The apo form of calmodulin, Ca0aM, is believed to form a Ca2aM intermediatefirst upon Ca2 addition. This form then goes on to become Ca4aM upon the binding oftwo additional Ca2 ions.In addition to Ca2+, calmodulin can also bind many other divalent cations andtrivalent lanthanides. Some of these cations (ie. Cd2,Zn2,Pb2, and Tb3j, like Ca2,will induce a conformational change leading to an activation of calmodulin dependentenzymes. Others, such as Mg2+, Be2+, Ni2+, and Co2+ will bind to calmodulin and causea different conformational change which will not activate the target enzymes (Klee, 1988).311.5.5. Calmodulin binding sequencesThe mechanisms of calmodulin interaction with its target sites were studied by theutilization of various calmodulin binding peptide hormones and toxins (DeGrado, 1988;Anderson and Malencik, 1986). The Ca2-dependent association of these peptides tocalmodulin appears to correlate with their abilities to form positively charged amphiphilichelices (Giedroc et a!., 1983; McDowell et a!., 1985; DeGrado et al., 1985; Cox et a!.,1985) which are characterized by clusters of hydrophobic residues adjacent to clusters ofbasic residues (Fig. 12).The primary sequences of a number of calmodulin binding proteins have beendetermined and their precise calmodulin binding domains have been mapped (Fig. 13).The calmodulin binding sequences are generally 25 residues or less. These sequencescontain a large number of positively charged residues and little or no negatively chargedresidues. Hydrophobic residues generally repeat with a 3 to 4 residue period to match thea-helix.Studies of calmodulin binding peptides with circular dichroism, proton NMR, andfluorescence emission indicate that the peptide binding to calmodulin leads to an increasein helicity of the complex. The hydrophobic helical faces of the peptides comes in directcontact with calmodulin while their hydrophilic faces are exposed to the surroundingenvironment (O’Neil and DeGrado, 1990). Photoaffinity labelling analyses indicate thatthe hydrophobic patches on the two halves of calmodulin must come together to form asingle binding site for either the peptide or the target enzyme binding (Kauer et a!., 1986;O’Neil etal., 1989).1.6. FUNCTIONAL ROLES OF CALMODULINCalmodulin is present in virtually all eukaryotic cells (Klee and Vanaman, 1982).It acts as a regulatory protein to modulate the activities of a large number of proteinsinvolved in cellular Ca2 signalling pathways. These include cAMP metabolism, protein32LysLysFig. 12: Helical wheel plot of a model amphiphilic calmodulin binding peptide.Calmodulin binds with high affinity to the basic amphiphilic ct-helical peptide (AcWKKLLKLLKKLLKL-CONH)in which one face of the helix is hydrophobic while theother face is positively charged (O’Neil and DeGrado, 1990).LysLys33Fig. 13: Amino acid sequences of the calmodulin binding domains of various targetproteins.The amino acid sequences of the calmodulin binding sites of various knownproteins and their dissociation constants for calmodulin are listed (O’Neil and DeGrado,1990). These sequences generally form amphiphilic x-helices.NARRKL KGAINFIAVSAANRFKKI SSSGAi mMT OHAV RAI OR L S S SSKMLCK KRRWKKSMMLCK aaxwoxType II kinaseCaMkinase ARRKLKAPhosphorytase b kinase ‘. a a I. I 0Phosphorylase b kinase G K 6 KPFkinase F MN N WECa2 pump a a 0 I I. W F aSpectrin K T A S P W K SNeuromodulin A H KAdenyate cyclase I 0 I. I. * K I ACaldesmon a A S F I. KCalspermin a a K L K A ACalcineurin K S v a N K I R AModelpeptide LKwKKLLIT MLATRNF SAVKAVVA$$RLGA V A FR YOU WV I. 6000 N RI CLTVLA$VRI YYQYRRVV K I. L AHI a PP A P KS 65 YrOLNRIQTWI KVVNFSSSAR I. MV NT V AT F N SI K SAA V K AVVAS SaL OS1 flM1 flM3—10 flM20 nMK 6.5nMV ll.4flM600 n<100 nun.d.580 flun.d.n.d.n.d.0.2 fluEAGARS AVGSAOKSGMKPANVXAVV AS SR LOSI OK MAR V F S V LRI.KLLKKLLKKLLKLG34phosphorylation and dephosphorylation, cation transport, cytoskeletal organization, andcell cycle control.1.6.1. Cyclic nucleotide metabolismCyclic AIVIP acts as a second messenger in the action of various hormones. It issynthesized by adenylate cyclase from ATP and is degraded by a phosphodiesterase to 5-AMP. Calmodulin regulates both the synthesis and hydrolysis of cAIVIP throughmodulating the activities of adenylate cyclase and phosphodiesterase. Several forms ofboth enzymes have been shown to be activated by calmodulin (Brostrom et a!., 1975;Cheung et a!., 1975; Klee and Vanaman, 1982). Thus, calmodulin has the effect ofcausing a transient change in cAMP level in response to an increase in free cytoplasmicCa2,Cyclic GMP is a key messenger in the visual cascade and in the cGMP-dependentprotein kinase activation. This compound is synthesized by guanylate cyclase from GTPand hydrolyzed by cGMP phosphodiesterase. In the protozoan Tetrahymena pyrformis,guanylate cyclase has been reported to be activated by calmodulin (Nagao et a!., 1979).However, in photoreceptor rod outer segments, the activities of both cGMPphosphodiesterase and guanylate cyclase have been shown to be unaffected by calmodulin(Del Priore and Lewis, 1983; Koch and Stryer, 1988).1.6.2. Protein phosphorylationCalmodulin is involved in the phosphorylation of a large number of proteinsthrough the activation of various kinases. In smooth muscle, it stimulates the myosin lightchain kinase to phosphorylate myosin light chain which activates the actomyosin ATPase.This subsequently results in ATP hydrolysis and the contraction of myosin, leading to thecontraction of smooth muscle (Hoar eta!., 1979; Sherry et a!., 1978). Calmodulin is also35involved in glycogen metabolism by stimulating the activity of phosphorylase kinase,leading to glycogen breakdown (Cohen eta!., 1978; Shenolikar, eta!., 1979).Calmodulin dependent multiprotein kinase, also known as CaM-PK II, is a kinasewith a broad substrate specificity (Kennedy et a?., 1983a, b; Yamauchi and Fujisawa,1983). In brain, this kinase phosphorylates synapsin-1 to control neurotransmitter release(Llinas et a?., 1985). It also phosphorylates tyrosine and tryptophan hydroxylases toregulate catecholamine and serotonin synthesis (Vuffiet et a!., 1984; Yamauchi et a?.,1981). In skeletal muscle, CaM-PK II inhibits glycogen synthesis by the phosphorylationof glycogen synthase (Cohen, 1986). In cardiac muscle, CaM-PK II phosphorylatesphospholamban and increases the Ca2-ATP se activity to accumulate Ca2 within thesarcoplasmic reticulum (Simmerman et a?., 1986). In liver, in addition to modulatingglycogen synthase, this kinase is also involved in the x-adrenergic agonist response bymodulating the activity of pyruvate kinase and phenylalanine hydroxylase (Schworer et a?.,1985a, b; Doskeland eta?., 1984). Phosphorylation of pyruvate kinase leads to a decreasein its affinity for its substrate while phosphorylation of phenylalanine hydroxylase leads tothe activation of this enzyme.1.6.3. Protein dephosphorylationCalcineurin, also known as protein phosphatase 2B, is present in various tissues(Ingebritsen et a?., 1983). This serine and threonine specific phosphatase is a. heterodimercomposed of a 61 kDa and a 19 kDa subunit and is stimulated by Ca2 and calmodulin(Stewart et a?., 1983). The 61 kDa catalytic subunit binds calmodulin while the 19 kDasubunit binds Ca2 (Merat et aL, 1985; Klee et a!., 1979; Aitken et aL, 1984).Calcineurin can be found in both the soluble and membrane fractions, with the membranebound species likely to be associated with cytoskeletal elements (Pallen et a?., 1985;Tallant and Cheung, 1983).36Calcineurin plays an important role in the regulation of glycogen metabolism. Itdephosphorylates the o-subunit of phosphorylase kinase (Antoniw and Cohen, 1976) toprevent glycogen breakdown. It is also involved in the cAMP-dependent pathway bydephosphorylating the regulatory subunit of cAMP-dependent protein kinase, leading tothe deactivation of the enzyme. Calcineurin stimulates phosphatase- 1 activity bydephosphorylating inhibitor-i (Cohen, 1982; Hemmings et al., 1984). In addition, thisenzyme dephosphorylates various microtubule associated proteins including Tau factorand microtubule associated protein-2 (MAP-2). This promotes microtubule assembly(Goto etat., 1985).1.6.4. Cation transportIntracellular cation levels are regulated by a variety of pumps, exchangers, andchannels. Many of these ion translocators are regulated by calmodulin. The red blood cellCa2-ATP se pump, which extrudes Ca2 from the cytoplasm, is activated by calmodulin(Carafoli and Zurini, 1982; Penniston, 1982; Gopinath and Vincenzi, 1977). In heartmyocyte plasma membranes, calmodulin has been suggested to play a role in modulatingthe Na/Ca2exchanger activity through phosphorylation by calmodulin dependent kinase(Caroni and Carafoli, 1983). In addition, calmodulin has been shown to increase theactivity of aCa2-dependent Na channel of Paramecium (Saimi and Ling, 1990) and todecrease the activity of the Ca2+rele se channels of the sarcoplasmic reticulum in cardiacand skeletal muscle (Smith et at., 1989) and the unglycanated M1P26 gap junctionchannels in lens (Swamy and Abraham, 1992).1.6.5. Cytoskeletal organizationCalmodulin interacts with various cytoskeletal structures. Calmoclulin has beenshown to regulate the microtubule assembly and disassembly process. In the presence ofCa2+calmodulin, brain microtubules were shown to undergo dissociation with a37concurrent inhibition of its assembly process (Nishida and Sakai, 1980; Keller eta!., 1982;Job eta!., 1981). Similarly, calmodulin has been shown to stimulate the polymerization ofG-actin (Piazza and Wallace, 1985) and the dissociation of caldesmon from F-actin (Sobueez’ aL, 1981). In red blood cells, calmodulin binds to spectrin and other cytoskeletalelements (Berglund et a!., 1984; Burns and Gratzer, 1985). In neurons and other celltypes, calmodulin associates with various structural proteins including fodrin, ahomologue of red blood cell spectrin (Glenney Jr. eta!., 1982).1.6.6. Control of cell proliferationCa2-calmodulin is believed to play a role in mitosis by regulating the progressionof cells from the Gi phase into the S phase (Hazelton et a!., 1979). Both Ca2 andcalmodulin are found in abundance in the centrosomal region of the mitotic spindle (Welsheta!., 1978; Wolniak eta!., 1980). Transient increases in Ca2 have been associated withthe breakdown of the nuclear envelope, chromatin condensation, and chromosomalmovement during anaphase (Keith eta!., 1985; Poenie eta!., 1985; Baitinger eta!., 1990).381.7. THESIS INVESTIGATIONIn photoreceptor cells, intracellular calcium plays important roles in modulating thevisual recovery process. These modulations are mediated through various Ca2+bindingproteins. Prior to the initiation of this thesis work, little was known about the identities ofthe Ca2-binding proteins in ROS and their involvement in the visual pathway. In anattempt to learn more about the biochemical mechanisms underlying Ca2+ feedback inROS, functional characterization of an unknown 17/20 kDa Ca2-binding protein,extracted from the bovine ROS membranes by EDTA, was carried out in this thesis. Thischaracterization was accomplished through identification of this Ca2-binding protein andits main target in ROS membranes and elucidation of the modulatory role of this Ca2-binding protein on its target.This thesis investigation is divided into four major chapters. Chapter 2 describesthe extraction, purification, identification, and localization of a 17/20 kDa protein frombovine ROS membranes. The identification of thisCa2-binding protein as calmodulin ledto the studies in Chapter 3 which describe the purification and identification of calmodulinbinding proteins on ROS membranes. Using calmodulin affinity chromatography andWestern blotting analysis with radioiodinated calmodulin and specific monoclonalantibodies, a 240 kDa protein which forms a complex with the c&subunit of the cGMPgated channel was found to be the major calmodulin binding protein on the ROS plasmamembranes. This association between calmodulin and the channel complex was alsoexamined among photoreceptor ROS of various vertebrates.Chapter 4 describes studies which investigate the potential modulation of thecGIvlP-gated channel by calmodulin. A series of cation influx assays carried out onextruded ROS membrane vesicles either in the presence or absence of calmodulin revealedthat calmodulin regulates the channel activity by changing its affinity for cGMP. Thespecificity of this effect was examined in terms of its cation specificity,Ca2-dependency,and inhibition by calmodulin antagonists.39Chapter 5 looks into the characterization of the 240 kDa channel associatedprotein. The importance of this 240 kDa protein in mediating the calmodulin effect wasstudied by reconstitution assays and proteolytic analysis. In addition, a 105 kDa fragmentof the 240 kDa protein was found to contain the calmodulin binding site.40CHAPTER 2IDENTIFICATION AND CHARACTERIZATION OF A 20 kflaPROTEIN, CALMODULIN, FROM THE EDTA EXTRACT OFPHOTORECEPTOR ROD OUTER SEGMENTS2.1. MATERIALSBovine eyes were obtained from J & L Meats (Surrey, British Columbia). Murineanti-calmodulin monoclonal antibody was from Upstate Biotechnology Incorporated.DEAE Sephacel and Ultrogel AcA-54 were purchased from Pharmacia. 45CaCl2 and[1251] Bolton-Hunter reagent were bought from New England Nuclear. Arthrobacterureafaciens neuraminidase was acquired from Boehringer-Mannheim. Paraformaldehydeand Tissue-Tek were bought from J. B. EM Services and Miles, respectively. Horseradishperoxidase conjugated sheep anti-mouse F’ab fragment and the ECL kit were purchasedfrom Amersham. The BCA protein assay kit was from Pierce and nitrocellulosemembranes were from Schleicher & Schuell. All other chemicals were obtained fromSigma, Fisher Scientific, or Calbiochem.2.2. METHODS2.2.1. Preparation of ROS membranesROS membranes were prepared according to the method described by Molday andMolday (1987). Briefly, retinas dissected from 100 bovine eyes were gently shaken inhomogenizing solution (20 mM TrisHCl pH 7.2, 20 % sucrose, 10 mM 3-D-glucose, 10mM taurine, and 0.25 mM MgCl2). The retinal mixture was filtered through a Teflonscreen and the filtrate was loaded onto six 20 ml 30% - 50% w/v continuous sucrosegradients (same composition as the homogenizing solution with the exception of thesucrose concentration). The gradients were then placed in a SW28 rotor and centrifugedat 25,000 rpm for 1 h at 4°C using a Beckman ultracentrifuge. The ROS band was41collected and diluted 1:5 with the homogenizing solution. The collected ROS werepelleted by centrifuging at 12,000 rpm using a Sorvall SS-34 rotor. The pellet was thenresuspended in 8 ml of homogenizing solution. Typically, 70-80 mg of ROS proteins wereobtained from 100 bovine retinas.2.2.2. Purification of the 20 kDa protein from ROS membranesEighty milligrams of ROS proteins were washed twice in 10 mM TrisHCl pH 7.2and 0.5 mM GTP and the soluble proteins were removed by pelleting the membranes at15,000 rpm using a Sorvall SS-34 rotor. The ROS membrane pellet was washed once in10 mM TrisHCl pH 7.2 and 150 mM NaCI. EDTA extraction was then carried out bywashing the membranes twice with 10 mM TrisHCl pH 7.2 containing 1 mM EDTA.This pooled EDTA extract was loaded onto a DEAE Sephacel column (1.5 ml bedvolume) equilibrated at 4°C in 10 mM TrisHCl pH 7.2, 1 mM EDTA, and 0.2 M NaC1.The 20 kDa protein was eluted off the column using a linear NaC1 gradient from 0.2-0.7M. Fractions containing the 20 kDa protein as analyzed by SDS polyacrylamide gelelectrophoresis were dialyzed against 10 mM TrisHCl pH 7.2 and concentrated to 0.6 mlusing either a Savant Speed Vac or Aquacide IT-A. The sample was then loaded onto a1.5 x 42 cm Ultrogel AcA-54 gel filtration column equilibrated in 10 mM TrisHCl pH 7.2,0.1 M NaCI, and 1 mM EDTA. Fractions containing the 20 kDa protein were dialyzedagainst distilled water and concentrated by lyophilization.2.2.3. Gel filtration chromatographyThe relative molecular weights of the 20 kDa protein and calmodulin werecompared by gel filtration chromatography. Half a ml of the purified 20 kDa protein orbovine brain calmodulin (0.5 mg/mi) was loaded onto an 1.5 X 48 cm of Ultrogel AcA-54gel filtration column. The column was equilibrated at 4°C in the buffer consisting of either10 mM HEPES pH 7.4, 100 mM NaCl, and 1 mM CaC12 or 10 mM FIEPES pH 7.4, 10042mlvi NaC1, and 1 mM EGTA. One ml fractions were collected and analyzed by SDSpolyacrylamide gel electrophoresis. The peak fractions of the 20 kDa protein andcalmodulin were assigned as the elution volumes (Ve) of these two proteins for thepurpose of calculating their apparent molecular weights. Blue dextran (—‘2 x l0 kDa),egg ovalbumin (45 kDa), carbonic anhydrase (29 kDa), and cytochrome C (12.4 kDa)were run separately as molecular weight standards.2.2.4. Preparation of ROS disk and plasma membranesA nontrypsinized preparation of ROS disk and plasma membranes was carried outaccording to the method of Molday and Molday (1987). Briefly, unbleached ROS (80 mgof proteins) in 12 ml of homogenizing solution were treated with 0.1 unit of Arthrobacterureafaciens neuraminidase for 2 h at 4 °C. The neuraminidase treated ROS were washedwith homogenizing solution by centrifI.igation at 10,000 rpm for 15 mm using a SorvallSS-34 rotor. The ROS pellet was gently resuspended in 6 ml of homogenizing solution,mixed with 1.5 ml of ricin-gold-dextran (A520 4), and allowed to incubate for 14 h at 40C. The ricin-gold-dextran labelled ROS were washed once in homogenizing solution and3 times in 10 mlvi TrisHCl pH 7.2. The ROS pellet was resuspended in 1 mM TrisHClpH 7.2 and incubated for 2 h at 4 °C. The hypotonically lysed ROS membranes wererepeatedly extruded through a syringe needle (gauge 14) for 20 mm to separate the disksfrom the plasma membranes. The membrane mixture was pelleted by centrifugation at17,000 rpm for 30 mm. The membrane pellet was then resuspended in 8 ml of 10 mMTrisHCl pH 7.2 and loaded onto six continuous gradients consisting of 9 ml of 25% -50% w/v sucrose (containing 10 mM TrisHCl pH 7.2) overlayering a 1 ml of 60 % w/vsucrose. The gradients were centrifuged at 37,000 rpm for 3 h in a SW41 rotor. Theband containing the disk membranes was collected, diluted with 5 volumes of 10 mMTrisHCl pH 7.2, and spun at 17,000 rpm for 30 mm. The disk membrane pellet wasresuspended in 3 ml of 10 mM Tris•HC1 pH 7.2. The ROS plasma membranes, which43pellet to the bottom of the gradients, were collected, washed with 10 mM TrisHCl pH7.2, and resuspended in 0.5 ml of 10 mM TrisHCl pH 7.2.2.2.5. Extraction of calmodulin from ROS disk and plasma membranesROS disk and plasma membranes were prepared as described above. One mg ofdisk membrane proteins and 0.6 mg of plasma membrane proteins were washed with 10mM TrisHCl pH 7.2 containing 0.15 M NaC1 to remove glyceraldehyde-3-phosphatedehydrogenase. The pelleted membrane samples were then incubated with 100 p.1 of theEDTA buffer to extract calmodulin. The extracted samples were then analyzed by SDSpolyacrylamide gel electrophoresis.2.2.6. Immunofluorescence localization of calmodulinA bovine retina was fixed in 3 % paraformaldehyde in phosphate buffered salinesolution for 3 h, embedded in acrylamide, and frozen in Tissue-Tek according to themethod described by Johnson and Blanks (1984). Retinal cryostat sections, 10 p.m inthickness, were cut and blocked in 3 % bovine serum albumin (BSA) and 1.5 % Triton X100 in PBS for 2 h. Anti-calmodulin monoclonal antibody diluted 10 X with 3 % BSA inPBS was added to the blocked retinal sections and allowed to incubate at 4 °C overnight.The sections were washed to remove unbound primary antibody and incubated withfluorescein isothiocyanate conjugated goat anti-mouse Ig for 1 h at 4 °C.Immunofluorescence microscopy was carried out using a Zeiss Axiophotphotomicroscope.2.2.7. Radioiodination of calmodulinBovine brain calmodulin (40 p.g) was radioiodinated by [1251] Bolton Hunterreagent (Bolton and Hunter, 1973). The iodinated calmodulin was isolated by gel44filtration. Typically, the radioiodinated calmodulin had a specific activity of 4 j.tCi I p.g ofprotein.2.2.8. SDS polyacrylamide gel electrophoresis, Western blotting analysis, andprotein concentration determinationSDS polyacrylamide gel electrophoresis was carried out using the buffer system ofLaemmli (1970). For theCa2-dependent mobility shift analysis, 1 mM CaC12 or 1 mMEGTA was added to the running gel buffer. For Western blotting analysis with the anticalmodulin monoclonal antibody, protein samples were run on a 12 % SDSpolyacrylamide gel. The proteins were electrophoretically transferred at 0.2 A for 20 mmfrom the SDS polyacrylamide gel onto an Immobilon membrane in the presence of thetransfer buffer (12.5 mM TrisHC1, 96 mM glycine, and 10 % methanol). The blot wasblocked in 0.5 % Tween in PBS for 1 h at room temperature, incubated in anti-calmodulinantibody (1 tg/ml) for 1 h, washed with 0.05 % Tween in PBS, and then incubated inhorseradish peroxidase conjugated goat anti-mouse Ig for 1 h. Visualization of theantibody labelling was carried out by ECL according to the manufacturer’s direction. Forthe 45Ca2labelling study, the purified 20 kDa protein and calmodulin were run on a 12 %SDS polyacrylamide gel and transfered onto nitrocellulose membrane. The blot wasequilibrated in 10 mM imidazole pH 6.8, 60 mM KC1, and 5 mM MgC12 for 1 h at roomtemperature. The blot was then labelled with 1 tCiIml of 45CaC12 for 10 mm, washedwith the equilibration buffer, air dried, and visualized by autoradiography. For theiodinated calmodulin labelling study, the labelling was carried out according to the methodof Flanagan and Yost (1984). Protein samples were run on an 8 % SDS polyacrylamidegel and transfered onto an Immobilon membrane. The blot was blocked in 50 mMTris•HC1 pH 7.4, 100 mM NaC1, 1mM CaC12, and 0.05% Tween 20 for 1 h and labelledwith iodinated calmodulin (2 pCi/mi) in the blocking buffer for 90 mm at room45temperature. The blot was washed with the blocking buffer and air dried. Labelledprotein bands were visualized by autoradiography.The protein content of ROS membranes was determined by the BCA protein assaykit using bovine serum albumin as a standard.2.3. RESULTS2.3.1. Extraction and purification of a 20 kDa protein from ROSSelective extractions of soluble proteins from ROS were carried out by firstwashing the intact ROS with a hypotonic solution (10 mM TrisHCl pH 7.2 and 0.5 mMGTP), following by an isotonic solution (10 mM TrisHCl pH 7.2 and 150 mM NaC1)prior to treatment with EDTA. The initial hypotonic wash lysed the ROS and removedmajority of the ROS soluble proteins, including phosphodiesterase, transducin, and Santigen (data not shown). The subsequent isotonic wash removed glyceraldehyde-3-phosphate dehydrogenase from ROS membranes (data not shown) as reported by Hsu andMolday (1990). A 20 kDa protein (Fig. 14, lane e) was eluted by EDTA from the ROSmembranes together with several other proteins with molecular weights ranging from 26kDa to 160 kDa (Fig. 14, lane b). The 20 kDa protein which comigrated with bovinebrain calmodulin by SDS polyacrylamide gel electrophoresis was subsequently purified byDEAE anion exchange and gel filtration chromatography (Fig. 14, lanes c and d). The gelfiltration step was utilized to remove a 90 kDa protein which sometimes cofractionatedwith the 20 kDa protein. A 26 kDa protein which was coeluted with the 20 kDa proteinby EDTA was later identified as recoverin.2.3.2. Identification of the 20 kfla protein as calmodulinThe purified 20 kDa protein was compared with bovine brain calmodulin by gelfiltration molecular weight sizing, 45Ca labelling, Ca2-dependent mobility shift by SDSpolyacrylamide gel electrophoresis, Western blotting with anti-calmodulin monoclonal46kDa97-66-45-31-20-14-abc d eFig. 14: Extraction and purification of calmodulin from bovine ROS membranes.Bovine ROS calmodulin was selectively extracted with an EDTA solution after theROS membranes had been washed with hypotonic and isotonic solutions. Calmodulin waspurified from the EDTA extract by DEAR anion exchange and gel filtrationchromatography. The protein samples were analyzed on a 12 % SDS polyacrylamide gelby staining with Coomassie blue. Lane a, bovine ROS membranes (30 jig); lane b, EDTAextract of ROS membranes; lane c, DEAR column purified calmodulin; lane d, gelfiltration column purified calmodulin; lane e, Sigma bovine brain calmodulin (1 jig).I47antibody, and amino acid composition analysis. When the 20 kDa protein and calmodulinwere chromatographed on an Ultrogel AcA-54 gel filtration column either in the presenceof Ca2 or EGTA, they displayed apparent molecular weights in the range of 35 to 37 kDa(Table II). The higher than expected values for the molecular weights of both proteinssuggest that the 20 kDa protein is probably axially asymmetric, a recognized characteristicof calmodulin.The potential Ca2-binding property of the 20 kDa protein was investigated by45Ca2labelling. Figure 15 shows that the 20 kDa protein (lane a), like calmodulin (laneb), can bind Ca2. Another unique property exhibited by calmodulin is its ability toundergo aCa-dependent mobility shift when electrophoresed on a SDS polyacrylamidegel. In the absence of Ca2 (+EGTA), the 20 kDa protein and calmodulin displayed anapparent molecular weight of 20 kDa, In the presence of Ca2, both proteins migratedwith an apparent molecular weight of 17 kDa (Fig. 16). Because of this ambiguity in themolecular weight assignment, the 20 kDa protein is referred to as the 17/20 kDa protein.Western blotting analysis with an anti-calmodulin monoclonal antibody showedthat the antibody labelled the 17/20 kDa protein in intact ROS (Fig. 17, lane a) and in theROS EDTA extract (Fig. 17, lane b), as well as the purified bovine brain calmodulin (Fig.17, lane c). The amino acid analysis of the purified 17/20 kDa protein (Table III)indicated that the amino acid composition of the 17/20 kDa protein is very similar to thatof calmodulin but not to otherCa2+binding proteins of similar molecular weights, such astroponin C, calcineurin B, S-l00 ct, and S-100 3.2.3.3. Immunofluorescence labellingImmunofluorescence labelling of retinal sections by the anti-calmodulinmonoclonal antibody suggested that calmodulin is ubiquitously distributed among allretinal cell layers, including the outer segments of the photoreceptor rod cells (Fig. 18). In48Table IIGel filtration molecular weight sizing of the 17/20 kfla protein17/20 kDa protein CalmodulinCa2 35kDa 36kDaEGTA 37kDa 37kDa49.4Fig. 15: Calcium-45 binding analysis of the 20 kDa protein.The purified 20 kDa protein (lane a, 0.5 jig) and bovine brain calmodulin (lane b,0.5 j.ig) were run on a 12% SDS polyacrylamide gel and the proteins were transferredonto a nitrocellulose membrane. The blot was then incubated in a 45CaC12 solution. The45Ca2binding was visualized by autoradiography.ab5097—6645—Fig. 16: Calcium-dependent mobility shift analysis of the 17/20 kfla protein by SDSpolyacrylamide gel electrophoresis.The purified 20 kDa protein (lanes a, 1 g) and bovine brain calmodulin (lanes b, 1tg) were run on a 12% SDS polyacrylamide gel either in the presence of Ca2 or EGTA.The two proteins displayed an apparent Mr of 20 K in the presence of EGTA and anapparent Mr of 17 K in the presence ofCa2.kDa EGTA Ca2a b a b51kDa CB CaM66-45-31-20-abcFig. 17: Western blotting analysis with an anti-calmodulin monoclonal antibody.Whole ROS proteins (lanes a, 30 .tg), EDTA extract of ROS membranes (lanes b),and bovine brain calmodulin (lanes c, 0.5 p.g) were run on a 12 % SDS polyacrylamidegel. The gel slices were either stained with Coomassie blue (CB) or transferred onto anInunobilon membrane and subjected to Western blotting analysis with an anti-calmodulinmonoclonal antibody (CaM).abc52Table IIIComparison of the amino acid composition of the 17/20 kDa protein with those ofotherCa2-binding proteins17/20 kDa CaM Troponin Calcineurin S-100 S-100Residue protein C B a. 13Lys 6 8 9 15 9 8His 1 1 1 2 2 5Arg 6 6 7 6 0 1Asx 23 23 22 28 13 9Thr 10 12 6 3 4 3Ser 4 4 7 11 5 5Gix 26 27 31 22 15 19Pro 3 2 1 3 0 0Gly 13 11 13 14 6 4Ala 11 11 13 5 8 5Cys 0 1 0 1 2Val 8 7 7 13 8 7Met 7 9 10 6 2 3Iso 8 8 10 11 1 4Leu 10 9 9 14 11 8Tyr 2 2 2 3 2 1Phe 8 8 10 12 5 7Trp D’ 0 0 0 1 0a The content of cysteine and tryptophan residues cannot be determined by amino acidcomposition analysis.53a control study, preadsorption of the anti-calmodulin antibody by Protein-G beads greatlydiminished antibody labelling.2.3.4. Extraction of calmodulin from ROS disk and plasma membranesIn order to determine whether calmodulin binding proteins in ROS are localized tothe disk or plasma membrane, ROS disk and plasma membranes were separated by thericin-gold-dextran affinity perturbation method under nontrypsinizing conditions (Fig. 19,lanes a and c). A subsequent extraction of these two membrane preparations by EDTAindicated that calmodulin binds preferentially to the ROS plasma membranes (Fig. 19,lanes b and d). When ROS membranes were lightly trypsinized to allow a better disk andplasma membrane separation, similar results were also obtained (data not shown).2.3.5. Calmodulin binding proteins in ROSCalmodulin binding proteins in ROS were determined by labelling with iodinatedcalmodulin. ROS proteins were first separated into soluble and membrane proteinfractions (Fig. 20 (left), lanes b and c) and the membrane fraction was further divided intodisk and plasma membranes (Fig. 20 (left), lanes d and e). These protein samples werethen run on a SDS polyacrylamide gel and transferred onto an Immobilon PVDFmembrane. The membrane was subjected to Western blotting analysis with iodinatedcalmodulin. Calmodulin binding components in the ROS soluble protein fraction consistmainly of a protein of 67 kDa (Fig. 20 (right), lane b). The membrane associatedcalmodulin binding proteins have molecular weight ranging from 35 kDa to 240 k.Da.These proteins are preferentially localized to the ROS plasma membranes (Fig. 20 (right),lanes c, d, and e).2.4. DISCUSSION54Fig. 18: Immunofluorescence localization of calmodulin.Fluorescent micrographs of retinal sections labelled with either (b) anti-rhodopsinmonoclonal antibody rho 1D4, (c) anti-calmodulin monoclonal antibody, or (d)flowthrough from a Protein G column used to adsorb out the anti-calmodulin antibody.The photoreceptor layers, including the outer segment (OS), inner segment (IS), and cellbody (CB), visualized under phase contrast microscopy are shown on (a).55kDaa4e v97_66___-45— :31—20—14—abc d eFig. 19: Extraction of calmodulin from ROS disk and plasma membranes.ROS disk and plasma membranes were prepared by the ricin-gold-dextranperturbation method. The isolated disk (lane a, 30 jig) and plasma membranes (lane c, 40jig) were then subjected to an EDTA wash, and their respective supernatants aftercentrifligation of the membranes are shown in lanes b and d. Lane e is bovine braincalmodulin (1 jig). The protein samples were analyzed on a 12% SDS polyacrylamide geland stained with Coomassie blue.56kDaCB CaMabcdeFig. 20: Labelling of the ROS proteins with iodinated calmodulin.ROS proteins were divided into soluble and membrane fractions. The membraneproteins were further separated into disk and plasma membranes. The protein sampleswere run on an 8% SDS polyacrylamide gel. The gel slices were either stained withCoomassie blue or transferred onto an Immobilon membrane and subjected to iodinatedcalmodulin labelling. Lanes a, whole ROS proteins (40 jig); lanes b, soluble ROS extract(7 jig); lanes c, stripped ROS membrane proteins (30 j.tg); lanes d, ROS disk membranes(25 fig); lanes e, ROS plasma membranes (30 .tg).IFabcde57Selective extraction of ROS membranes under various conditions led to thedetection of several proteins that appeared to interact with the ROS membranes in adivalent cation dependent manner. A 20 kDa protein found in the EDTA extract of ROSmembranes became the focus of this study. This protein was purified by DEAE anionexchange and gel filtration chromatography. Subsequent studies by gel filtrationmolecular weight sizing, 45Ca labelling, Ca2-dependent mobility shift on SDSpolyacrylamide gel, amino acid composition analysis, and Western blotting analysis withanti-calmodulin monoclonal antibody identified this protein as calmodulin.Immunofluorescence labelling of retinal sections with an anti-calmodulin monoclonalantibody showed that calmodulin is found in rod photoreceptor ROS as well as otherretinal layers and is not a contaminant from other retinal layers.Recoverin, a 26 kDa protein originally hypothesized to modulate ROS guanylatecyclase activity (Lambrecht and Koch, 1991; Dizhoor et a!., 1991) but more recentlyfound to regulate rhodopsin phosphorylation in ROS (Kawamura, 1993), was also presentin this EDTA extract. However, this protein does not appear to interact with themembranes as tightly as calmodulin, since only small amounts of this protein remainedbound to the ROS membranes after the initial hypotonic and isotonic washes. Unlikeiodinated calmodulin which readily bound to distinct bands on the Western blots, iodinatedrecoverin failed to label any ROS proteins. In addition, a recoverin affinity column alsofailed to extract proteins from the solubilized ROS membranes in a Ca2-dependentmanner (data not shown).Calmodulin has also been detected in photoreceptor outer segments from varioussources, including bovine, frog, squid, and teleost (Kohnken et a!., 1981; Morelli et al.,1989; Nagao et a!., 1987; de Couet et a!., 1986; Asai et a!., 1989; Nagle and Burnside,1983). This suggests that calmodulin may play an important functional role in thephotoreceptor rod cell. No definitive identification of the calmodulin binding proteins inROS, however, had been made. As a first step in the characterization of the calmodulin58binding proteins in ROS, the ROS proteins were separated into soluble and membranefractions, and the membrane fraction was further divided into disk and plasma membranes.Analysis of the ROS proteins by either selective EDTA extraction or iodinated calmodulinlabelling on Western blots suggests that the calmodulin binding proteins are preferentiallylocalized to ROS plasma membranes.ROS disk and plasma membranes have been previously shown to have quitedistinct protein compositions (Molday and Molday, 1987). Proteins such as the cGMPgated channel complex and the Na+/Ca2+K+ exchanger are found exclusively in theplasma membranes (Cook et aL, 1989; Molday et al., 1990; Reid et al., 1990). Thepreferential association of calmodulin to ROS plasma membranes raises the possibility thatcalmodulin may be interacting with one of these plasma membrane specific proteinsinvolved in phototransduction. Thus, it may have a role in regulating the visualtransduction process. Identification of a major calmodulin binding protein in ROS plasmamembranes is described in Chapter 3.59CHAPTER 3IDENTIFICATION OF THE cGMP-GATED CHANNEL COMPLEXAS THE MAJOR CALMODULIN BINDING PROTEIN IN BOVINEROS MEMBRANES3.1. MATERIALSCalmodulin Sepharose, DEAE Sephacel and Sepharose 2B beads were obtainedfrom Pharmacia. DEAE-650 S gel was purchased from Supelco and cyanogen bromidewas bought from Fisher. [1251] Bolton-Hunter reagent was from New England Nuclear.Immobilon membranes were acquired from Millipore and the protein assay kit wasobtained from Bio-Rad. Arsenazo III, asolectin, proteolytic inhibitors, cGMP, and allother chemicals were purchased from Sigma.3.2. METHODS3.2.1. Calmodulin affinity chromatographyROS membranes were prepared from freshly dissected bovine retinas as previouslydescribed (Molday et at., 1987). Proteolytic inhibitors: 0.1 mM DFP, 5 ig/ml aprotinin, 2tg/ml leupeptin, and 2 jig/ml pepstatin were present in all subsequent chromatographicsteps. The ROS membranes were washed three times in 10 mM HEPES pH 7.4containing 1 mM EDTA and solubilized in solubilization buffer (10 mM HEPES pH 7.4,150 mM KC1, 10 mM CaC12, 1 mM DTT, 18 mM CHAPS, and 0.2 % asolectin) at aprotein concentration of 1.5-2 mg/ml. Forty three mg of solubilized ROS proteins wereloaded onto 2 calmodulin columns (2.5 ml each) equilibrated in the solubilization buffer at4 °C. After washing the columns with 10 column volumes of washing buffer (10 mMHEPES pH 7.4, 15 mM CHAPS, 150 mM KC1, 1 mM CaC12, 1 mM DTT, and 0.18 %asolectin), the channel complex was eluted off the columns with washing buffer in which 1mM CaC12 was replaced by 1.5 mM EDTA. Fractions (0.75 ml) were collected and their60absorbance was measured at 280 nm. Peak fractions containing the channel complex wereadjusted to 2 mM CaC12 and mixed 1:1 with 18 mg/mi of asolectin and 2 mM CHAPS inthe dialyzing buffer (10 mM HEPES pH 7.4, 0.1 M KC1, 0.1 mM DTT, and 2 mM CaC12).The sample was first dialyzed against 5 changes of the dialyzing buffer for 2 days and thendialyzed against 3 changes of the same buffer without Ca2 over a period of 8 h. Thereconstituted channel was used for Ca2+ efflux assays (Cook et al., 1987).For the purification of channel complex from various species, 0.5-0.7 mg ofbovine, pig, and rat and 3 mg of frog ROS membranes were first washed with the EDTAsolution to remove endogenous calmodulin, and then solubilized and mixed with 100 tl ofcalmodulin-Sepharose beads. A.fter washing the beads with washing buffer, the respectivechannel complexes were eluted off the beads with elution buffer containing 2 mM EDTA.Asolectin was included with CHAPS in order to maximize the interaction betweencalmodulin and the channel complex. In the absence of asolectin, the CHAPSconcentration had to be decreased below 15 mM. Other detergents (in the absence ofasolectin) such as 1 % Triton X-100 can be used to substitute CHAPS while 1 % sodiumcholate cannot be used. This detergent appeared to interfere with the interaction betweencalmodulin and the channel complex.3.2.2. PMc 6E7 antibody affinity chromatographyPMc 6E7 monoclonal antibody was purified by ammonium sulphate precipitationand DEAE anion exchange chromatography. The antibody was precipitated from mouseascites fluid with 50 % ammonium sulphate at 4°C. The precipitate was pelleted at 10,000rpm using a Sorvall SS-34 rotor. The pellet was resuspended in 10 mI\4 TrisHCl pH 7.2containing 20 mM NaC1 and dialyzed against 3 changes of the same buffer. The antibodysolution was then loaded onto a DEAE Sephacel column equilibrated at room temperaturein 10 mM Tris•HCI pH 7.2. The column was washed with 5 column volumes of theequilibration buffer, and the antibody was eluted with a 0 - 0.3 M NaC1 gradient. One ml61fractions were collected and analyzed by both A280 measurements and SDSpolyacrylamide gel electrophoresis. The purified antibody was dialyzed against 10 mMborate pH 8.4 and coupled to CNBr activated Sepharose 2B beads (Molday et al., 1990).Purification of the cGMP-gated channel complex using the PMc 6E7 antibodycolumn was carried out essentially by the same procedure used for calmodulin affinitychromatography except DTT was omitted and 0.9 mg/mi of synthetic peptidecorresponding to the N-terminal of the 63 kDa ROS channel (Ser-Asn-Lys-Glu-Gln-GluPro-Lys-Glu-Lys-Lys-Lys-Lys-Lys) was used to elute the channel complex (Molday et a!.,1991).3.2.3. DEAE purification of the cGMP-gated channel complexROS were washed three times with 10 mM HEPES pH 7.4, 1 mM DTT, and 1mM EDTA. The ROS membranes were solubilized at a final protein concentration of 1.6mg/mi in 10 mM HEPES pH 7.4, 150 mM KC1, 18 mM CHAPS, 0.2 % asolectin, 2 mMCaC12, and 0.1 mM DFP. The DEAE fractionation of ROS proteins was carried outaccording to the method of Cook et a!. (1987). Briefly, the solubilized ROS membraneproteins were loaded onto 4 ml of DEAE-650 S column equilibrated at 4 °C in thesolubilization buffer. The column was washed with 10 column volumes of the washingbuffer (10 mM HEPES pH 7.4, 150 mM KCI, 15 mM CHAPS, 0.18 % asolectin, 2 mMCaCl2, and 0.1 mM DFP). The bound proteins were eluted off the column with 0.75 MKC1 in the washing buffer, and 1.5 ml fractions were collected. Fractions containing thechannel complex were pooled and reconstituted into lipid vesicles according to the methodof Cook eta!. (1987) as described above.3.2.4. Calcium efflux assay on reconstituted channelsThe dialyzed lipid vesicles containing the reconstituted channel complex (0.3 ml)were mixed with 1.7 ml of 55 jiM Arsenazo III dye in the vesicle dialysis buffer. The62calcium efflux assays were induced by the addition of 4 p1 of varying concencentrations ofcGMP. The absorbance change of the Arsenazo III dye induced by Ca2 efflux from thevesicles was monitored spectrophotometrically at 650-730 rim using a SLM Aminco DW2000 dual wavelength spectrophotometer.3.2.5.Ca+dependent association of calmodulin with the channel complexAnti-channel a*subunit monoclonal antibody PMc 1D1 was purified from mouseascites fluid and coupled to Sepharose 2B beads as described above. The channelcomplex was immobilized onto the antibody colunm using the method described aboveexcept 1 % Triton X-100 was used in place of CHAPS and NaC1 was used in place ofKC1. One hundred p1 of the antibody beads containing the immobilized channel complexwas washed with 2 column volumes of the washing buffer containing 0.1 % Trition X100. Three hundred p.1 of bovine brain calmodulin (50 jig) were adjusted to 1 mM CaC12and 0.1 % Triton X-100 and allowed to incubate with the beads for 2 h at 4 °C. Thebeads were then washed with 20 column volumes of the washing buffer (10 mM HEPESpH 7.4, 150 mM NaC1, 0.1 % Tx-100, and 1 mlvi CaC12). The bound proteins wereeluted off the beads with the washing buffer containing 2 mM EDTA in place of 1 m]VICaC12.3.2.6. SDS polyacrylamide gel electrophoresis, Western blotting analysis, andprotein concentration determinationSDS polyacrylamide gel electrophoresis was carried out as previously described.Calmodulin was iodinated by the Bolton-Hunter reagent and 125-calmodulin labelling wascarried out as described in section 2.2.8.. For immunoblotting analysis, the blot was firstlabelled with PMc 6E7 and PMs SEll and then relabelled with 125-goat antimouse Ig forvisualization by autoradiography. The protein concentration of various solubilizedsamples were determined by the Bio-Rad protein assay kit using y-globulin as a standard.633.3. RESULTS3.3.1. Identification of the cGMP-gated channel complex as a major calmodulinbinding proteinIn order to identify the calmodulin binding proteins, ROS membranes weresolubilized in detergent and passed through a calmodulin column. Two polypeptides ofapparent Mr 63 K and 240 K were isolated as the major constituents (Fig. 21 (left), laneb). These polypeptides comigrated with the 63 kDa ct-subunit of the channel and the 240kDa channel associated polypeptide of the cGMP-gated channel complex isolated on ananti-channel immunoaffinity column (Molday et a!., 1990; Molday et a!., 1991; Fig. 21(left), lane c). Western blots labelled with antibodies PMc 1D1 and PMs 5E1 1 to thechannel ct-subunit and to the 240 kDa protein, respectively, confirmed that the calmodulinaffinity column extracted the 63 kDaJ24O kDa cGMP-gated channel complex from ROSmembranes (Fig. 21 (middle)).A Western blot labelled with 125-calmodulin in the presence of Ca2 was used toidentify the calmodulin binding proteins. Whereas calmodulin labelled severalpolypeptides of apparent Mr 240 K, 140 K, 70 K, 67 K, and 35 K in both the ROSmembranes and the calmodulin affinity purified fraction (Fig. 21 (right), lanes a and b),only the 240 kDa polypeptide was labelled with calmodulin in the immunoaffinity purifiedchannel complex (Fig. 21 (right), lane c). This labelling is specific since in the presence ofEGTA, no labelling by iodinated calmodulin was observed. The intense labelling of the 67kDa and 70 kDa proteins in the calmodulin column eluant as compared to the 240 kDaprotein is probably due to the low transfer efficiency of the 240 kDa protein.3.3.2. Calmodulin affinity chromatographyPurification of the channel complex from ROS membranes was carried out usingcalmodulin affinity chromatography. Figure 22a shows the SDS polyacrylamide gel64Fig. 21: Purification of the cGMP-gated channel complex by calmodulin andimmunoaffinity chromatography and identification of the 240 kfla channel-associated protein as a major calmodulin binding polypeptide of ROS membranes.ROS were subjected to an EDTA wash to remove soluble proteins. Themembranes were then solubilized and passed through either a calmodulin or antibodyaffinity column. Lanes a, ROS membranes after the extraction of soluble proteins (30 p.g);lanes b, EDTA eluant from a calmodulin affinity column; and lanes c, N-terminal peptideeluant from a N-terminal specific anti-channel c&subunit monoclonal (PMc 6E7) column.Left: SDS polyacrylamide gel stained with Coomassie Blue (CB); Middle: Western blotlabelled with both an anti-63 kDa channel cL-subunit monoclonal antibody (PMc 1D1) andan anti-240 kDa monoclonal antibody (PMs 5E1 1). Right: Western blot labelled with125-calmodulin (CaM).kDaPMs 5E11CB PMc1D1240 —___.___CaMabc a b C6590c1ab0.120.080.0400 2 4 6 8Fraction NumberkDaFig. 22: Purification of the bovine ROS channel complex by calmodulin affinitychromatography.a. SDS polyacrylamide gel electrophoretic analysis of the calmodulin columnfractions. Fifteen p.1 of sample from each column fraction was run on an 8 % SDSpolyacrylamide gel and the gel was stained with Coomassie blue (CB). The resultsindicated that the 63/240 kDa channel complex is the predominant component eluted offthe calmodulin column. b. Elution profile of the channel complex from a calmodulinaffinity column. The 63/240 kDa channel complex was eluted off the calmodulin columnby EDTA and the absorbance at 280 nm of each fraction (0.75 ml) was monitored. Peakfractions (3-7) were pooled and reconstituted into lipid vesicles.12 34 56 789I I I1066electrophoretic analysis of the column fractions and Figure 22b shows thechromatographic profile. The pooled peak fractions from the calmodulin columns werereconstituted into lipid vesicles for calcium efflux assays (Cook et at., 1987). Figure 23ais a plot of the channel activation by varying concentrations of cGMP. The channeldisplayed a Km of 30 p.M for cGMP and a Hill coefficient (n) of 3.9 (Fig. 23b). Thesevalues are comparable to those of the channel purified by DEAE anion exchange and TSKAF-red affinity chromatography (Km= 11 p.M and n 3.1; Cook et at., 1987).Purification of the channel complex by calmodulin affinity chromatographyrecovered approximately 33% of the net activity from the pooled fractions (3-7). Thepurified channel displayed a specific activity of 163 units/mg protein, representing a 36fold purification. For the purpose of comparison, purification of the channel complex byDEAE anion exchange column (Cook et at., 1987) was carried out as well. The DEAEchromatography method recovered 50 % of the net activity. The fractionated channelcomplex was purified 14 fold and had a specific activity of 65.7 units/mg protein (TableIV).Calmodulin affinity chromatography, thus, not only gave a channel complex ofcomparable yield as the DEAR column but also of higher purity and specific activity. The36 fold purification of the channel complex obtained by calmodulin affinitychromatography is lower than the 110 fold purification reported by Cook et al. (1987)using the combined DEAR anion exchange and the TSK AF-red affinity chromatography.However, the 240 kDa protein was reported to be either absent or present in minutequantity in their channel preparations.SDS polyacrylamide gel electrophoretic analysis was also carried out to comparethe purity of the channel complex obtained from these two chromatographic procedures.Both columns extracted the majority of the channel complex from the solubilized ROS asevident from Western blotting analysis with anti-channel complex monoclonal antibodiesPMs 5E1 1 and PMc 1D1 (Fig. 24 (right), lanes b and e). However, whereas the DEAR67a1.00.8Vo 0.6Vmax0.40.200b10010-2[cGMPI (pM)Fig. 23: Activation of the calmodulin affinity purified channel complex by cGMP.The channel complex purified by calmodulin affinity chromatography wasreconstituted into lipid vesicles. 0.3 ml of vesicles containing 8 p.g of the channel complexwere assayed at varying concentrations of cGMP. a. The relative initial velocities of thechannel complex were plotted as a function of cGIvlP concentration. The solid curved linewas drawn from the sigmoidal isotherm (Vo/Vmax = 1- (1+ lOn(P1< pCcGMP)y.1 using aKm of 30 p.M and a Hill coefficient (n) of 3.9. b. Hill plot analysis of the channelactivation as a function of cGMP concentration. The slope of the straight line is 3.9.20 40 60 80[cGMPJ (pM)10 5068Table IVComparative purification of the channel complex by DEAE anion exchange andcalmodulin affinity chromatographyChannel activityVolume Protein Umts’ Total Units/mg Recovery Purification(ml) (mg/mi) /ml units of protein % foldSolubilized ROS 27 1.60 7.6 205 4.75 100 1DEAE (unbound) 27 1.18 0 0 0 0DEAEpoo1 9.0 0.17 11.3 102 65.7 49.6 14Solubilized ROS 28 1.52 6.8 190 4.47 100 1CaM (unbound) 28 1.20 0.6 16.8 0.50 8.8CaM Pool 7.5 0.05 8.5 63.8 163 33b 36a One unit of channel activity is defined as the amount of Ca2+ released by saturatingconcentration of cGMP that is equivalent to 1 % of one ml of lipid vesicles.b Determined by pooling fractions 3-7.69fractionated channel complex was contaminated with Na+/Ca2+K+ exchanger (Mr= 230K), guanylate cyclase (Mr= 112 K), and rhodopsin (Mr= 38 K), the calmodulin columneluant consisted of mainly the 240 kDa and 63 kDa components of the channel complex(Fig. 24 (left), lanes c and f).3.3.3. Channel complexes from ROS of other speciesIn addition to bovine ROS membranes, calmodulin affinity chromatography wasused to purify the channel complexes from ROS membranes of various mammals. Figure25 (left) shows a Coomassie blue stained SDS polyacrylamide gel of the channelcomplexes purified from bovine, pig, and rat ROS membranes. Identification of the 63kDa polypeptides as the cL-subunit of the cGMP-gated channels was confirmed by Westernblotting analysis with the anti-cL-subunit monoclonal antibody PMc 1D1 (Fig. 25, right).In frog ROS membranes, 3 bands corresponding to proteins of apparent Mr 240 K,112 K, and 74 K were observed in the EDTA eluted fractions from a calmodulin affinitycolumn (Fig. 26 (left), lane d). The 74 kDa protein stained only faintly with Coomassieblue. Anti-cL-subunit monoclonal antibody PMc 1F6, which binds to a highly conservedsegment of this subunit in various species (BOnigk et al., 1993), was used to identify thefrog cL-subunit of the cGIvIP-gated channel. As shown in Figure 26 (right), PMc 1F6labelled the 63 kDa channel a.-subunit in bovine preparations and the 74 kDa polypeptidein frog preparations. The anti-240 kDa channel associated protein monoclonal antibodyPMs 4B2 (Molday et al., 1990) did not cross-react with any protein in frog ROSmembrane preparations. A Western blot labelled with iodinated calmodulin revealed thatcalmodulin intensely labelled the 112 kDa protein in addition to a 63 kDa protein (Fig. 26(middle), lanes c and d). The 63 kDa protein is likely a homolog of the bovine 67 kDaprotein which has been shown not to be a component of the cGMP-gated channelcomplex.70.4Fig. 24: Comparative purification of the channel complex by DEAE and calmodulinaffinity chromatography.Solubilized ROS membrane proteins (lanes a and d) were loaded onto either aDEAE or a calmodulin column. Lanes b and e are the flowthrough from the columns,respectively. Lanes c and f are the respective eluants from the two columns. The proteinsamples were analyzed on an 8 % SDS polyacrylamide gel. The gel slices were eitherstained with Coomassie blue (left) or transferred onto an Immobilon membrane andsubjected to Western blotting analysis with both the anti-cGIvIP-gated channel ct-subunitmonoclonal antibody PMc 1D1 and the anti-240 kDa protein monoclonal antibody PMs5E1 1 (right).CBkDa2O5-’’- .4116-.97-66-—45- .PMs5E11PMc ID1— .4.4abcdefab C def71Fig. 25: Purification of the channel complex from ROS of various mammals.Calmodulin affinity chromatography was used to purif,’ the cGMP-gated channelcomplexes (lanes b, d, and 0 from bovine (lane a), pig (lane c), and rat (lane e) ROSmembranes. The protein samples were analyzed on an 8% SDS polyacrylamide gel andstained with Coomassie blue (left). The presence of the cGIvIP-gated channel cc-subunitfrom each species was confirmed by Western blotting analysis with the anti-cGMP-gatedchannel cc-subunit monoclonal antibody PMc 1D1 (right). The relative positions of the cx-subunit of the cGMP-gated channel and the high molecular weight channel associatedprotein are marked with arrows.PMc 1D1CBa b Cd e31La b Cd e I72CB24O-t. ‘—L-63-—.abcdCaMFig. 26: CaM affinity purification and Western blot analysis of the cGMP-gatedchannel complexes from bovine and frog ROS membranes.Solubilized bovine and frog ROS membranes (lanes a and c, 30 p.g) were loadedonto calmodulin-Sepharose columns in the presence of Ca2+ and bound proteins wereeluted with EDTA (lanes b and d). SDS polyacrylamide gels were either stained withCoomassie blue (CB) or transferred onto Immobilon membranes and labelled with eitherradioiodinated CaM or anti-channel a-subunit monoclonal antibody PMc 1F6 and anti-240kDa monoclonal antibody PMs 4B2. The relative positions of the a-subunit of frogcGMP-gated channel as detected with PMc 1F6 monoclonal antibody are marked witharrows.PMs4B2PMc 1F6.4abcd abc d733.3.4. Extraction of calmodulin by immobilized channel complexIn order to further demonstrate theCa2-dependent interaction of calmodulin withthe cGMP-gated channel complex, this complex was immobilized onto the anti-ct-subunitmonoclonal antibody PMc 1D1 Sepharose beads. SDS polyacrylamide gel electrophoreticanalysis confirmed the binding of the channel complex to the antibody beads (Fig. 27, lanec). Bovine brain calmodulin was then incubated with the antibody beads containingimmobilized channel complex in the presence ofCa2+. The bound proteins were eluted offthe beads by EDTA. As shown in Figure 27 (lane e), bovine brain calmodulin wasextracted by the channel complex. When a similar binding experiment was carried Outwith the EDTA extract of ROS membranes, a 90 kDa protein, in addition to calmodulin,was found to bind to the immobilized channel complex in aCa2+dependent manner (datanot shown). This suggests that another Ca2 binding protein can also interact with thechannel complex.3.4. DISCUSSIONIn order to determine the role of calmodulin in theCa2+dependent regulation ofROS processes, identification of the calmodulin binding proteins was first carried out.Using calmodulin affinity chromatography, the cGMP-gated channel complex consisting ofthe 63 kDa ct-subunit and the 240 kDa channel associated polypeptide, was found to bethe major calmodulin binding protein on ROS membranes. By Western blotting analysisusing iodinated calmodulin, this association between calmodulin and the channel complexwas found to be mediated through the 240 kDa protein. Identification of the plasmamembrane specific cGMP-gated channel complex as the major calmodulin binding proteinin ROS membranes is in agreement with the findings in Chapter 2 in which calmodulin wasfound to be preferentially associated with the ROS plasma membranes.Calmodulin affinity chromatography is a useful method for purification of thechannel complex. In a single step, the channel complex can be isolated in high yield and74.4—.4—Fig. 27: Extraction of calmodulin by immobilized cGMP-gated channel complex.The bovine cGMP-gated channel complex consisting of the 63 kDa ct-subunit andthe 240 kDa channel associated protein was immobilized onto anti-ct-subunit PMc 1D1monoclonal antibody Sepharose beads. The association of the channel complex to theantibody beads was confirmed on a 10% SDS polyacrylamide gel. Lane a is the ROSmembranes (30 pg). Lane b is the PMc 1D1 beads boiled in the gel loading buffer andlane c is the PMc 1D1 antibody beads with immobilized channel complex. The positionsof the channel complex components and antibody heavy and light chains are marked byarrows. Bovine brain calmodulin (lane d, 0.5 .tg) was adjusted to 1 mM CaC12 andincubated with the PMc 1D1 antibody beads immobilized with the channel complex. Thebeads were washed and the bound calmodulin was eluted by EDTA (lane e).kDa—97-6645—31-24063Ab H.C.Ab L.C.abc de75with a greater than 90 % purity as assessed by SDS polyacrylamide gel electrophoresis.The channel complex purified by calmodulin affinity chromatography is functionally activeas determined by the reconstitution assays. The Km and Hill coefficient of thereconstituted channel complex are comparable to those of the channel complex preparedby DEAE anion exchange and TSK ÀY-red affinity chromatography as described by Cooket al. (1987). Calmodulin affinity chromatography is also a valuable method for thepurification of the channel complexes from various species. This channel complex canalso be purified from frog ROS membranes, suggesting that the association between thechannel ct-subunit and a calmodulin binding subunit may be universal among vertebrates.In frog ROS membranes, the 74 kDa protein is most likely the ct-subunit of the channelsince it binds an anti-ct-subunit antibody. The 112 kDa calmodulin binding protein mayitself constitute the calmodulin binding component of the channel complex, oralternatively, it may be a proteolytic fragment of the 240 kDa protein which is moreefficiently transferred and renatured than the 240 kDa protein.The association between calmodulin and the channel complex was also shown bythe Ca2-dependent binding of calmodulin to the channel complex immobilized onto ananti-channel ct-subunit antibody-Sepharose support. The 90 kDa protein which was foundto co-elute with calmodulin from the column by EDTA may be another Ca2-bindingprotein that associates with the channel. It will be interesting to determine whether thisprotein binds to the ct-subunit or to the 240 kDa channel associated protein. Based on theabove studies, this interaction between calmodulin and the channel complex may play arole in regulating the function of the channel. Investigation of the effect of calmodulin onthe channel activity is described in Chapter 4.76CHAPTER 4MODULATION OF THE cGMP-GATED CHANNEL OFPHOTORECEPTOR ROS BY CALMODULIN4.1. MATERIALSPolycarbonate membranes were obtained from Nuclepore. DichiorophosphonazoIII was purchased from Fluka. DEAR Sephacel and Ultrogel AcA 54 were fromPharmacia. Centricon-lO concentrator was acquired from Amicon. Bovine brain S-100protein, bovine brain calmodulin, Arsenazo Ill, Sephadex G50, phenyl Sepharose, neutralred, FCCP, and all other chemicals were obtained from Sigma.4.2. METHODS4.2.1. Calcium influx assay using Arsenazo Ill and dichiorophosphonazo ifi dyesROS membranes were prepared as previously described. ROS membrane vesicleswere loaded with Arsenazo III dye as follows: ROS membranes were washed twice inbuffer A (10 mM HEPES pH 7.4, 1 mM DTT, 1 mM EDTA, and 0.1 mM DFP) and oncein buffer B (2 mM HEPES pH 7.4, 0.2 mM DTT, and 0.02 mM DFP) to remove solubleproteins. The membrane pellet was resuspended in buffer B at a protein concentration of12 mglml and bleached by continuous white light. After one cycle of freezing andthawing, the membranes were suspended in 10 mM HEPES pH 7.4, 100 mM KC1, 1 mMDTT, and 2 mM Arsenazo ifi and sonicated for 1 mm in a beaker of water using a broadtip sonicator probe (Heat Systems-Ultrasonics, Inc.) at setting 7. The sonicated ROSmembranes were then extruded 3 times using a Lipid Extruder (Lipex Biomembranes,Vancouver, B.C.) through two layers of 800 nm pore size Nuclepore polycarbonate filters(Hope et al., 1985). The extrusion of ROS membranes was repeated using 400 nm and200 nm pore size polycarbonate filters. The membrane vesicles were passed through aSephadex G50 column (1.5 X 28 cm) equilibrated at 4 °C in buffer C (10 mlvi HEPES pH777.4, 0.1 mM DTT, and 100 mM KC1) to remove untrapped dye. The eluted vesicles werefrirther dialyzed against buffer C for 3 h at 4 °C. ROS membrane vesicles containingtrapped Arsenazo III dye were then added to buffer C in a cuvette to a final volume of 2ml and at a final protein concentration of 0.29 mg/mI either in the absence or presence of 5jig/mi (235 nM) calmodulin. An aliquot (4 p.1) of stock CaC12 was then added to give afinal concentration of 100 p.M CaC12. One mm after the addition of calcium, the calciuminflux assay was initiated by the injection of 4 p.1 of varying concentrations of cGMP. Thechange in Arsenazo III dye absorbance upon binding to Ca2 was monitored at 650-73 0nm using a SLM Aminco DW 2000 dual wavelength spectrophotometer.For dichiorophosphonazo ifi (Wohlfart et a!., 1990), a similar dye trappingprocedure was employed. The assays were carried out in the presence of CaC12 or MgC12(200 jiM). For assays using Mg2 as the main influx ion, 1 p.M CaCl2 was also included.The divalent cation influx assays were carried out either in the presence or absence of 235nM bovine brain calmodulin and the assays were initiated by the addition of varyingconcentrations of cGMP. The change in dye absorbance was monitored at 600-672 nmusing the SLM Aminco DW 2000 dual wavelength spectrophotometer.4.2.2. Inhibition of the calmodulin effect by mastoparanROS membrane vesicles containing Arsenazo III dye were prepared as describedabove. ROS membrane vesicles were added to buffer C (10 mM HEPES pH 7.4, 0.1 mMDTT, and 100 mM KC1) to a final volume of 2 ml at a final protein concentration of 0.3mg/mi. Varying concentrations of calmodulin (0.8-23 5 nM) either in the presence orabsence of 430 nM mastoparan were added. The calcium influx assays were carried outby the addition of CaC12 (100 p.M) followed by the injection of 12.5 p.M cGMP 1 mmlater. The change in dye absorbance was monitored as described above.4.2.3. Cation influx assay using neutral red dye78ROS were washed once in buffer D (10 mM HEPES/arginine pH 7.4, 0.5 mMDTT, and 1 mM EDTA) and once in buffer E (10 mM HEPES/arginine pH 7.4 and 0.5mM DTT) to remove soluble proteins. The membrane pellet was then resuspended inbuffer E at a protein concentration of 14 mg/mi and subjected to one cycle of freezing andthawing. The membranes were mixed with an equal volume of the resuspension buffer (20mM HEPES/arginine pH 7.4, 10 % sucrose, 2.5 % Ficoll, and 0.5 mM DTT) and extruded4 times through one layer of 800 nm pore size Nuclepore polycarbonate filter using theLipid Extruder. The extrusion of ROS membranes was repeated using the 400 nm poresize polycarbonate filter. The neutral red assay was carried out according to a modifiedmethod of Schnetkamp (1990) as follows: The membrane vesicles were added to the assaybuffer containing 20 mM HEPES/arginine pH 7.4, 5 % sucrose, 0.5 mM DTT, 30 1iMneutral red, 1 p.M FCCP, and with or without 400 nM calmodulin. The final assay volumewas 2 ml and with a protein concentration of 0.5 mg/mI. The assay was initiated by theaddition of 50 p.1 of 0.5 M CaCI2,MgCl2,BaCI2,MnCl2,NaCl, KC1, LiC1, or CsC1 (12.2mM final concentration) and 0.4 mM CaC12 (10 p.M final concentration). Four p1 of 10mM cGMP (20 p.M final concentration) was added 1 mm later. The change in absorbancewas monitored at 540-650 nm using a SLM Aminco DW 2000 dual wavelengthspectrophotometer. For measuring the Ca2+ dependence of calmodulin effect, Na+ (12.2mM) was used as the main influx ion and the assay mixture contained varyingconcentrations of Ca2 ranging from 10 nM to 1 p.M. Buffers of varying Ca2concentrations were prepared according to the program of Fabiato (1988). The followingratios of Ca2 to 1 mM EGTA were used (estimated free Ca2 in parentheses): 0.134 (10nM), 0.279 (25 nM), 0.436 (50 nM), 0.607 (100 nM), 0.755 (200 nM), 0.830 (316 nM),0.860 (398 nM), 0.908 (631 nM), and 0.940 (1000 nM).4.2.4. Purification of calmodulin and recoverin from bovine photoreceptor ROS79Bovine ROS calmodulin was purified as described in section 2.2.2.. For thepurification of recoverin, ROS membranes from 200 bovine retinas were washed with 100ml of 10 mM HEPES pH 7.4 and 1 mM EDTA. The supematant was loaded onto a 4 mlDEAE Sephacel column equilibrated at 4 °C in 10 mM HEPES pH 7.4, 0.1 mM DTT, and1 mM EDTA. The column was washed with 5 column volumes of equilibration buffer andrecoverin was eluted off the column with 0.125 M NaCl in the same buffer. The DEAEfractions were dialyzed at 4 °C against 10 mM HEPES pH 7.4, 0.1 mM CaC12, and 0.1mM DTT for 10 h. The dialyzed sample was adjusted to 0.5 M NaC1 and loaded onto a 4ml phenyl Sepharose column. The column was washed first with 5 column volumes of 10mM HEPES pH 7.4, 0.5 M NaCl, 0.1 mM DTT, and 0.1 mM CaCl2 and then with thesame buffer without the NaC1. Recoverin was eluted with 10 mlvi HEPES pH 7.4, 0.1mM DTT, and 5 mM EGTA. The phenyl Sepharose column eluant was concentrateddown to 1 ml with a Centricon-lO concentrator and loaded onto an 1.5 x 54 cm UltrogelAcA-54 gel filtration column equilibrated in 10 mM HEPES pH 7.4 and 100 mM NaCl.One ml fractions were collected and analyzed by SDS polyacrylamide gel electrophoresis.Fractions containing recoverin were dialyzed against 10 mM HEPES pH 7.4 andlyophilized down to 0.25 ml.4.3. RESULTS4.3.1. Effect of calmodulin on the affinity of the channel for cGMPThe effect of calmodulin on the cGMP-gated channel activity was investigated byCa2+ influx assays using ROS membrane vesicles loaded with the divalent cation sensitivedye, Arsenazo III. The rate of Ca2+ influx was measured spectrophotometrically (Kochand Kaupp, 1985) as a function of cGIVIP concentration in the presence and absence ofcalmodulin. As shown in Figure 28a and b, the apparent Km for cGIvIP increased from 19± 0.4 LM in the absence of calmodulin to 33 ± 2 iiM in the presence ofCa2-calmodulin.This change in Km translates into a relatively large change in the rate of Ca2 influx (6-80a1.00.8Vo 0.6Vmax0.40.200 10 20 30 40 50 60cGMP Concentration (iiM)0q0IIt =1 OscGMPFig. 28: Effect of calmodulin on the activation of the cGMP-gated channel.a. The relative initial velocity of channel activation by cGMP was plotted againstthe concentration of cGMP either in the presence ( • ) or absence ( • ) of calmodulin. Inthe absence of calmodulin, the channel had an apparent Km of 19 iiM and a Hillcoefficient of 3.8 for cGMP. In the presence of calmodulin, its Km increased to 33but its Hill coefficient remained the same at 3.8. These assays were carried out in threeseparate experiments using different ROS membrane vesicle preparations. In all cases, anincrease in Km (from 19 ± 0.4 tM to 33 ± 2 pM) in the presence of calmodulin (235 nM)without significant change in either the Vmax or the Hill coefficient (n= 3.7 ± 0.1 and n3.5 ± 0.6 in the absence and presence of calmodulin, respectively) was observed. Thesolid curved lines are calculated from a sigmoidal relationship using the indicated Km andHill coefficients. b. A typical trace for cGMP dependent influx of Ca2 into ROS vesiclescontaining trapped Arsenazo III. A cGMP concentration of 12.5 tM was used to initiatethe influx ofCa2+ either in the presence (+) or absence (-) of calmodulin.+81fold) at 12.5 iM cGMP. The Vmax and the cooperativity for cGMP as determined by aHill plot, however, was unaffected byCa2+calmodulin.Dichlorophosphona.zo III, a divalent cation sensitive dye, that had been previouslyused by Wohifart et a!. (1990) to measure Mg2 translocation through the reconstitutedcGMP-gated channel complex, was also used in this study to further verify the calmodulineffect. This dye was trapped within ROS membrane vesicles and assays were carried outusing either Ca2+ or Mg2+ as the main influx ion. When Ca2+ was used as the main influxion, a similar shift in the affinity of the channel for cGMP was observed, as had beenobserved with Arsenazo III loaded ROS membrane vesicles (Fig. 29a). The Km of thechannel for cGIvlP increased from 22 jiM to 33 jiM while its Hill coefficient remainedrelatively unchanged (n= 3.0). When Mg2 was used as the influx ion (Fig. 29b), thechannel displayed a Km of 37 jiM with a Hill coefficient of 2.4. In the presence ofcalmodulin, the Km of the channel for cGMP increased to 45 jiM without significantlychanging its Hill coefficient (n= 2.1). The higher Km and lower Hill coefficient observedwith Mg2 translocation may be due to the differential gating of this cation by the channelcomplex.Neutral red, a dye that is sensitive to changes in surface electrostatic potentialinduced by cation influx (Schnetkamp, 1990), was also used to examined the calmodulineffect. For Ca2+ influx rates using the neutral red assay (Fig. 30), Ca2+calmodulin wasobserved to increase the Km for cGMP from 34 jiM to 51 jiM without significantlyaffecting either the Vmax or the Hill coefficient (n= 2.4 and 2.2 respectively). The higherKm and lower Hill coefficient obtained with this assay may be due to the hydrophobicnature of this dye and its interaction with the phospholipids and/or the channel.4.3.2. Specificity of the calmodulin effectIn order to assess whether the observed calmodulin-induced change in the affinityof the cGMP-gated channel for cGMP is a unique phenomenom or is a universal property82Y2VmaxaFig. 29: Effect of calmodulin on the activation of the rod channel by cGMP asdetermined by the dichlorophonazo ifi assay system.a. Calcium influx assays were carried out using ROS membrane vesicles containingdichiorophosphonazo ifi dye either in the absence ( • ) or presence ( • ) of 235 nMcalmodulin. Calmodulin was shown to increase the Km of the channel for cGMP from 22iM to 34 iM without affecting either its maximum velocity or cooperativity for cGMP(n 3.0). The solid curved lines were drawn from a sigmoidal isotherm using the indicatedKm and Hill coefficients. b. Magnesium influx assays were carried out as above either inthe absence ( • ) or presence ( • ) of calmodulin. The Km of the channel was shown toincrease from 37 iiM in the absence of calmodulin to 45 iM in the presence ofcalmodulin. Their respective Hill coefficients, however, remained relatively unchanged(n== 2,4 and 2.1 respectively).[cGMP] (pM)b1.0[cGMPJ (pM)831.00.80.6V0Vmax 040.20-0 20 40 60 80 100 120 140 160[cGMP] (pM)Fig. 30: Effect of calmodulin on the activation of the rod channel by cGMP asdetermined by the neutral red assay system.Calcium influx assays were carried out using extruded RO S membrane vesicles inthe presence of neutral red. The assays were carried out either in the absence ( • ) orpresence ( • ) of calmodulin. The initial velocities of the channel were plotted as afunction of cGMP concentration. The Km of the channel was observed to increase from34 iM to 51 p.M in the presence of calmodulin while its Hill coefficient remainedrelatively unchanged (n= 2.4 and 2.2 respectively). The solid curved lines were drawnusing the indicated Km and Hill coefficient values.•••••I I I I I I I84shared by various Ca2-binding proteins, bovine ROS recoverin and brain S-100 proteinswere added to the Arsenazo III loaded vesicles in place of calmodulin. In addition,calmodulin isolated from bovine ROS was compared with bovine brain calmodulin for itseffect on channel activity. Recoverin was purified from the EDTA extract of ROS byDEAE anion exchange, phenyl Sepharose, and gel filtration chromatography and thepurified recoverin was analyzed by SDS polyacrylamide gel electrophoresis as shown inFigure 31. Figure 32 shows the cGMP-dependent influx of Ca2 into ROS membranevesicles in the presence of various Ca2+binding proteins. Whereas bovine braincalmodulin and ROS calmodulin decreased the channel activity equally well (Fig. 32b),bovine ROS recoverin and brain S-100 protein had no effect on channel activity undersimilar conditions (Fig. 32a).4.3.3. Inhibition of the calmodulin effect by mastoparanIn order to reverse the calmodulin effect, two anti-calmodulin agents, mastoparanand calmidazolium, were used. In the absence of mastoparan, the half maximal inhibition(IC50) of the channel by calmodulin in the presence 12.5 tM cGMP was found to be 1.85± 0.25 nM (averaged from 3 trials; Fig. 33). In the presence of 430 nM mastoparan,higher calmodulin concentrations were required to modulate the channel activity, and thehalf maximal inhibition (IC50)of the channel was found to increase over ten fold to 20.3 ±3.8 nM (averaged from 3 trials). Higher mastoparan concentration was not used since thispeptide was found to affect the channel activity at concentrations above 800 nM.Calmidazolium, a potent hydrophobic calmodulin inhibitor, was found to inhibit thechannel activity at concentrations above 560 nM. However, at or below thisconcentration, it was ineffective in reversing the calmodulin effect to any significantextent. It has been reported that the hydrophobic nature of the calmodulin inhibitors canresult in their incorporation into lipid bilayers (Tanimura et al., 1991). As a result, theseinhibitors can become unavailable for binding to calmodulin. This may explain the85kDa97-66-45-20-a b c d eFig. 31: Purification of recoverin from bovine ROS.Bovine ROS were washed with EDTA to extract soluble proteins. Recoverin waspurified from this EDTA extract by DEAE anion exchange, phenyl Sepharose, and gelfiltration chromatography. The protein samples were run on a 9% SDS polyacrylamidegel and the gel was stained with Coomassie blue. Lane a, intact ROS (30 fig); lane bEDTA extract of ROS (5 jig); lane c, pooled recoverin from the anion exchange column;lane d, pooled recoverin from the phenyl Sepharose column; lane e, gel filtration columnpurified recoverin.31-86a-CaM+ 200 nM recoverin0 V____,+200nMS—100b_CaMnM CaMt lOS+ 20 nM brain CaMnM ROS CaMFig. 32: Effect of various Ca2-binding proteins on the cGMP-gated channelactivity.a. ROS membrane vesicles were trapped with Arsenazo III dye. The Ca2+ influxassays were carried out at 12.5 i.tM cGMP in the absence (-CaM) or in the presence of200 nM recoverin, S-100, and calmoclulin. b. Ca2+ influx was measured in the absence orpresence of 20 nM bovine brain calmodulin or bovine photoreceptor calmodulin.87100 --.•... _!•80-I b.%60-•\\.\ •\40-. •\20 -..____0 iiiil i I iiiiil I I I iiiiil I1 10 100fCalmodulin] (nM)Fig. 33: Inhibition of the calmodulin effect by mastoparan.The calcium influx assays were carried out using extruded vesicles containingArsenazo III dye. The assays were carried out using varying concentrations of calmodulin(0.8-235 nM) either in the presence ( • ) or absence ( • ) of 430 nM mastoparan. AcGMP concentration of 12.5 jiM was used to initiate the assay and the relative initialvelocity ofCa2 influx in the absence of calmodulin was designated as 100 % activity.88inhibition of the channel activity by mastoparan and calmidazolium at high concentrationsand the failure of calmidazolium to inhibit the calmodulin effect.4.3.4. Cation selectivity and calcium dependence of the calmodulin effectThe cGMP-gated channel is permeable to a variety of monovalent and divalentcations. A neutral red dye assay was used to determine the effect ofCa2+_calmodulin onthe cGIvIP-dependent influx rate of different cations into ROS membrane vesicles. At 20iM cGIVIP, Ca2-calmodulin decreased the initial rate of ion influx by 2-3 fold for allcations except Mg2 (Fig. 34). The smaller decrease for Mg2 (1.4 fold) may be due tothe interference of high Mg2+ concentration on the interaction of Ca2+ with calmodulin(Ogawa and Tanokura, 1984).The calmodulin modulation of the channel activity was shown to be a Ca2+dependent process. In a similar neutral red dye assay system, Na+ was used as the maininflux ion and the assay was carried out in the presence of varying concentrations of Ca2+ranging from 10 nM to 1 pM. At a constant cGMP concentration of 20 p.M, thecalmodulin effect was observed to occur within the range of 20 to 300 nM Ca2 (Fig. 35).In the control study in which calmodulin was omitted from the assay mixture, the channelactivity was relatively unaffected by Ca2.4.4. DISCUSSIONThe role of calmodulin in modulating the activity of the cGMP-gated channel wasinvestigated using the extruded ROS membrane vesicle system containing various divalentcation sensitive dyes. In all cases, it was found that the presence of calmodulin resulted ina decrease in the affinity of the channel (ie. increase in Km) for cGMP without affectingthe number of cGMP molecules required to open the channel (ie. Hill coefficients remainconstant). This effect was shown to be specific for calmodulin since two other calcium89140120-100-f80.40-20H HCa2 MgBa2Mn2 Na K+ Li CsFig. 34: Effect of calmodulin on the influx of various cations.Cation influx assays were carried out using extruded ROS membrane vesicles inthe presence of neutral red. The initial rates of cGMP-dependent influx of differentmonovalent (Na+, K+, Li+, and Csj and divalent (Mg2+, Ba2+, and IVln2j cations intoROS membrane vesicles were measured in the presence (solid) and absence (cross-hatched) of calmodulin. A cGMP concentration of 20 p.M was used to initiate cationinflux. The initial velocity of Ca2 flux in the absence of calmodulin was set at 100%. Allions except Mg2+ showed a 2-3 fold decrease in initial velocities in the presence of Ca2+calmodulin.II90100>-I‘IC.)604020Fig. 35: Calcium dependence of calmodulin effect on the cGMP-gated channelactivity.The Na+ influx assays were carried out using extruded ROS membrane vesicles inthe presence of neutral red and at varying concentrations of Ca2 (10-1000 nM). Theassays were carried out either in the presence ( • ) or absence of calmodulin ( • ). ThecGMP-clependent (20 M) Na influx in the presence of 10 nM Ca2, but in the absenceof calmodulin, was designated as 100 % activity. Results of three sets of experimentscarried out in the presence of calmodulin are shown with standard deviations marked byerror bars.102[Ca2--] (nM)91binding proteins, bovine brain S-100 and retinal recoverin, had no effect on the channelactivity.The half maximal inhibition value of 1.85 ± 0.25 nM obtained from the dose-response curve of the calmodulin inhibition of channel activity indicates that theassociation between calmodulin and the channel complex is quite strong. Othercalmodulin-binding proteins have binding constants for calmodulin in the range of 1-600nM (O’Neil and DeGrado, 1990). Inhibition of the calmodulin effect was found to bedependent on the type of inhibitors used. Only mastoparan, the least hydrophobic of thecompounds used, was capable of substantially reversing the calmodulin effect in the lipid-rich assay system employed here.The binding of calmodulin to the channel complex affected the Km but not theVmax or cooperativity of the channel for cGMP. These parameters were found to beconsistent by a variety of assay methods. This suggests that calmodulin affects the bindingaffinity of the channel for cGMP but not the ion translocation properties of the channel.The calmodulin effect on channel activity was observed over a Ca2+ concentrationof 50-3 00 nM. This is within the range of the Ca2-dependent activation of guanylatecyclase (Koch and Stryer, 1988) and physiological Ca2 concentration in photoreceptorouter segments (McNaughton eta!., 1986; Ratto et at., 1988; Korenbrot and Miller, 1989;Kaupp and Koch, 1992). Earlier, Caretta et at. (1988) observed that Ca2 decreased thebinding affinity of a cGMP analogue to ROS membranes. It is likely that this observedCa2 effect on cGMP binding is related to the effect of Ca2 on calmodulin modulation ofthe cGMP-gated channel as reported here.The above studies suggest that changes in cytoplasmic Ca2+ concentration whichare known to occur during the photoresponse can affect the activity of the cGMP-gatedchannel. The channel is opened in response to the cooperative binding of cGMP to thechannel subunits. The affinity of the channel for cGMP is further modulated by the levelof cytoplasmic Ca2 through the binding ofCa2-calmodulin to the channel complex.92This Ca2+calmodulin effect on the channel is most pronounced at low cGIvIPconcentrations as found under physiological conditions (Nakatani and Yau, 1988a).Regulation of the channel by cGIvIP andCa2+calmodulin is in some ways analogous tothe regulation of hemoglobin by oxygen and pH (Bohr effect) in which small changes inpH alter the affinity of hemoglobin for oxygen.The key question that one could raise regarding calmodulin modulation of thechannel is “Does this effect occur in ROS under physiological conditions?” To addressthis question, experimental designs must involve the use of whole ROS. This system,however, can cause potential problems due to the complexities associated with intactsystems. Earlier studies dealing with the incorporation of Ca2 buffer into detached ROSindicated that intracellular Ca2 dictates the rate of photorecovery (Torre et a!., 1986;Nakatani and Yau, 1988b). Although these observations are in general agreement withthe idea of aCa2-calmodulin mediated shift in the affinity of the channel for cGMP, it isnot possible to distinguish the role played by calmodulin with those of other Ca2-regulated processes, such as those of guanylate cyclase and phosphodiesterase. Gray-Keller et al. (1993) recently described experiments involving measurements of the kineticsof light response following the dialysis of various Ca2-binding proteins into a detachedgecko rod outer segment. Addition of exogenous recoverin was found to prolong thephotorecovery phase possibly due to its potential inhibition on rhodopsin phosphorylationwhile calmodulin was found to have little or no effect. Results in this chapter indicate thatcalmodulin binds strongly to the channel complex (half maximal inhibition or 1C50=2 nM).With an estimated calmodulin concentration of 4 j.tM in the ROS (Kaupp and Koch,1992), the addition of exogenous calmodulin would unlikely elicit an additional effect,particulary if calmodulin remains bound to the channel complex in high Ca2. Anotherapproach to study calmodulin modulation of the channel in intact ROS is to measure theeffect of calmodulin inhibitors on the phototranscluction process. However, theapplication of these calmodulin inhibitors in membrane systems is not straightforward.93Several studies including studies on the ROS membrane vesicle system, as describedabove, indicate that the calmodulin inhibitors can directly inhibit the conductance of notonly the cGMP-gated channel but also a variety of other types of channels in a Ca2-independent manner (Nicol, 1993; Kihira et al., 1990; Klockner and Isenberg, 1987). Thissuggests that these inhibitors are not acting through calmodulin. The hydrophobic natureof the inhibitors could result in their incorporation into the lipid bilayers, therebydecreasing their overall effective concentration. In a highly membraneous system such asthe intact rod outer segment, one would have to determine the actual target site of theseinhibitors as well as their effective concentrations in the cytoplasm. Without thisinformation, interpretation of inhibition results would be ambiguous. The physiologicaleffect of calmodulin on modulation of the cGMP-gated channel thus remains an openquestion at the present time.94CHAPTER 5CHARACTERIZATION OF THE 240 kDa CHANNEL ASSOCIATEDPROTEIN5.1. MATERIALSBovine pancreatic trypsin and chymotrypsin, porcine pancreatic kallikrein, soybeantrypsin inhibitor, CHAPS, asolectin, Arsenazo III, bovine brain calmodulin, and otherchemicals were purchased from Sigma. [1251] Bolton-Hunter reagent was obtained fromNew England Nuclear and horseradish peroxidase conjugated sheep anti-mouse F’abfragment and the ECL kit were acquired from Amersham. Centricon-30 concentrator wasbought from Amicon, and Immobilon membrane was obtained from Millipore.5.2. METHODS5.2.1. Purification of the channel complex by PMc 6E7 antibody affinity and DEAEanion exchange chromatographyROS membranes (40 mg protein) were solubilized in 30 ml of 10 mM HEPES pH7.4, 150 mM KC1, 18 mM CHAPS, 0.2% asolectin, 2 mM CaC12, and 0.1 mM DFP. Thesolubilized sample was loaded onto two antibody columns (3 ml bed volume each) thathad been equilibrated at 4 °C in the same buffer. After washing the columns with 10column volumes of the washing buffer (10 mM HEPES pH 7.4, 150 mM KC1, 15 mMCHAPS, 0.18 % asolectin, 1 mM CaC12, and 0.1 mM DFP), the channel complex waseluted off the columns with washing buffer containing the synthetic peptide (0.9 mg/mI)corresponding to the N-terminus of the cGMP-gated channel (Molday et a!., 1991).Fractions (1 ml) collected were then loaded onto a 1.5 ml DEAE anion exchange column.After washing the column with 5 colunm volumes of washing buffer in the presence of 1mM DTT, the channel complex was eluted from the column using 0.45 M KC1 in the95washing buffer. One ml fractions were collected, pooled, and reconstituted into lipidvesicles according to the method described by Cook et al. (1987).5.2.2. Calcium efflux assay of the immunoaffinity purified channel complexreconstituted into lipid vesiclesThe calcium efflux assay was carried out according to the method described byCook et al. (1987). The assay mixture consisted of 0.3 ml vesicles (6.5 .Lg protein) and1.7 ml of 55 p.M Arsenazo III dye in dialysis buffer containing 1 p.M CaC12. The assaywas carried out either in the presence or absence of 120 nM bovine brain calmodulin. TheCa2 efflux assay was initiated by the addition of varying concentrations of cGMP.Changes in dye absorbance were monitored at 650-730 nm using a SLM Aminco DW2000 dual wavelength spectrophotometer.5.2.3. Mild proteolytic digestion of ROS membranes with trypsin, chymotrypsin,and kaflikreinROS membranes were washed twice with 10 mM HEPES pH 7.4, 1 mM DTT,and 1 mM EDTA. The membrane pellet was resuspended in 10 mM HEPES pH 7.4 at aprotein concentration of 10 mg/nil for tryptic and chymotryptic digestion and 5 mg/mi forkallikrein digestion. Solutions containing trypsin (0.8 jig/mi), chymotrypsin (1.6 p.g/ml),and kallikrein (1.3 units/mI) were added to an equal volume of the resuspended ROSmembrane mixture and digestions were allowed to proceed for 30 mm at roomtemperature. The digestions were terminated by washing the membranes twice with 10mM HEPES pH 7.4 containing 50 p.g/ml of soybean trypsin inhibitor for trypsin and 0.5mM DFP for chymotrypsin and kallikrein. The digested membranes were then solubilizedin CHAPS, reconstituted into lipid vesicles, and subjected to Ca2+ efflux assays asdescribed earlier.965.2.4. Purification of the kallikrein-treated channel complex by PMc 6E7 antibodyaffinity chromatographyROS membranes which had been treated with kallikrein (10 mg) were solubilizedin 10 mlvi HEPES pH 7.4, 150 mM NaC1, 1 mM DTT, 0.1 mM DFP, and 2 mM CaC12.The solubilized sample was loaded onto a 1 ml PMc 6E7 antibody column. The columnwashing and elution procedures were carried out as previously described in section 3.2.2..5.2.5. Protein sample preparation for N-terminal sequence analysisFor N-terminal sequence analysis, ROS membrane proteins (120 mg) weresubjected to kallikrein treatment and then passed through three 5 ml calmodulin Sepharosecolumns. The EDTA eluants from the calmodulin affinity columns were concentrated witha Centricon-30 concentrator. Polypeptides were separated on an 8% SDS polyacrylamidegel and electrotransferred onto an Immobilon membrane. The protein bands werevisualized by staining with Ponseau S. The desired protein bands were excised from themembrane, destained in distilled water, and sent to the protein microsequencing servicefacility at the University of Victoria for N-terminal sequence analysis.5.2.6. SDS polyacrylamide gel electrophoresis and Western blotting analysisSDS polyacrylamide gel electrophoresis and electrotransfer of ROS membraneproteins onto Immobilon membranes were carried out as previously described. For theinhibition of125-calmodulin binding to the 240 kDa protein by anti-240 kDa monoclonalantibodies PMs 3C9, PMs 5E1 1, and PMs 4B2, the blots were first blocked for 1 h with50 mM Tris pH 7.4, 100 mM NaC1, 0.05 % Tween 20, and 1 mM CaC12. The blots werethen incubated in the same buffer containing 10 times diluted hybridoma culture fluid foranother hour. lodinated calmodulin (2 p.Ci/ml) was then added and allowed to incubatewith blots for 1 h. After washing away the unbound calmodulin, the labelled bands on theblots were visualized by autoradiography. For blots labelled with the anti-240 kDa protein97and anti-63 kDa channel a-subunit monoclonal antibodies only, horseradish peroxidaseconjugated sheep anti-mouse Ig was added instead of125-calmodulin. Labelled proteinbands were visualized by ECL.5.3. RESULTS5.3.1. Mediation of the calmodulin effect through the 240 kDa channel associatedproteinCalmodulin-induced shift in the affinity of the channel for cGMP has beendemonstrated, in the previous chapter, using ROS membrane vesicles trapped withdivalent cation sensitive dyes. In order to verifj that this effect is mediated through thecalmodulin-binding 240 kDa protein, Ca2-effiux assays were carried out using PMc 6E7antibody column purified channel complex which had been reconstituted into lipid vesicles(Fig. 36). In the absence of calmodulin, the channel displayed a Km of 33 M and a Hillcoefficient of 3.3. In the presence of calmodulin, the Km of the channel increased to 44pM while its Hill coefficient remained relatively unchanged (n= 3.3 and 3.1, respectively).The shift in the affinity of the channel for cGMP in the reconstituted system is not as largeas that observed in the intact membrane system. This difference could be attributed topotential changes in channel conformation induced by solubilization and reconstitution.The antibody column purified channel complex was also analysed by SDSpolyacrylamide gel electrophoresis and Western blotting with 125-calmodulin. Thepurified channel complex was shown to consist of the characteristic 240 kDa and 63 kDaprotein bands. The 240 kDa protein was the only protein band labeled by iodinatedcalmodulin on the Western blot (data not shown). Thus, in the absence of othercalmodulin binding proteins, the observed shift in the affinity of the channel for cGMP ismost likely the result of calmodulin interaction with the 240 kDa channel associatedprotein.981.00.80.6Vmax0.40.20[cGMPJ (pM)Fig. 36: Effect of calmodulin on the cGMP-dependent activation of the reconstitutedchannel complex.The PMc 6E7 antibody column purified channel complex was concentrated byDEAE anion exchange chromatography and reconstituted into lipid vesicles. Calciumefflux assays were carried out either in the absence or presence of 120 nM calmodulin.The relative initial velocities of the channel were plotted as a fUnction of varyingconcentrations of cGMP. The solid curved lines represent a sigmoidal isotherm using Kmvalues of 33 and 44 jM and Hill coefficients of 3.3 and 3.1 in the absence ( • ) andpresence ( • ) of calmodulin, respectively. The purified channel complex had a specificactivity of 187 units/mg protein.0 20 40 60 80995.3.2. Effectiveness of anti-240 kfla protein monoclonal antibodies in inhibitingcalmodulin binding to the channel complexSeveral monoclonal antibodies have been generated against the 240 kDa channelassociated protein. These antibodies, PMs 3C9, PMs 5E1 1, and PMs 4B2, were used onWestern blots in an attempt to inhibit iodinated calmodulin binding to the 240 kDa protein.The blots were incubated in these antibody solutions prior to calmodulin labelling. Asshown in Figure 37, none of the antibodies inhibited calmodulin binding to the 240 kDaprotein, implying that the epitopes of these antibodies on the 240 kDa protein are distantfrom the calmodulin binding site.5.3.3. Effect of proteolysis on the cGMP-gated channel complexProteolytic digestions of ROS membranes were carried out using various enzymesin an attempt to destroy the calmodulin binding site on the 240 kDa protein, and hence,the sensitivity of the channel complex to calmodulin. The enzyme digestions were carriedout under conditions which minimized cleavage of the 63 kDa channel x-subunit. Theproteolyzed ROS membrane proteins were reconstituted into lipid vesicles for Ca2+ effluxassays. Monoclonal antibody PMc 6E7 (Molday et a!., 1991) directed against the N-terminus of the 63 kDa channel ct.-subunit and monoclonal antibodies PMc 1D1 and PMc2G1 1 directed against the C-terminal regions of the cL-subunit were used to label Westernblots to examine the intactness of the cL-subunit.Chymotrypsin, an enzyme that preferentially cleaves peptide bonds on the carboxylside of aromatic amino acids, was found to cleave a variety of ROS membrane proteinsincluding the 240 kDa protein (Fig. 38). Western blots of the chymotrypsinized ROSmembrane proteins (Fig. 38) gave 2 closely spaced bands with PMc 1 Dl labelling and 3closely spaced bands with PMc 6E7 and PMc 2G1 1 labelling. This implied that cleavageshad occurred at two different regions of the C-terminus of the a-subunit. Reconstitution100CBkDa205-11666-.ICaMFig. 37: Inability of anti-240 kDa protein monoclonal antibodies to inhibitcalmodulin binding to the 240 kfla protein.Various anti-240 kDa protein monoclonal antibodies were used in an attempt toinhibit iodinated calmodulin binding to the 240 kDa protein by Western blot analysis.Left, Coomassie blue stained SDS polyacrylamide gel of the stripped ROS membranes (30jig). Right, lane a, iodinated calmodulin labelling of ROS membrane proteins withoutprior incubation with anti-240 kDa protein monoclonal antibodies; lanes b, c, and d are theiodinated calmodulin labelling of the ROS membrane proteins in the presence of anti-240kDa protein monoclonal antibodies PMs 3C9, 4B2, and 5E1 1 respectively.a b Cd101205-116-97-66-45-Fig. 38: Chymotrypsin digestion of ROS membranes.ROS membranes were subjected to mild chymotrypsin digestion. The digestedmembranes were run on an 8% SDS polyacrylamide gel. The intactness of the channel ctsubunit was monitored by Western blotting analysis with anti-c-subunit monoclonalantibodies PMc 1D1, 6E7, and 2G1 1. Lanes a are the undigested ROS membranes (30p.g) and lanes b are the chymotrypsinized ROS membranes (30 ..tg).kDa CB PMc1D1 PMc6E7 PMc 2G1131-ab ab ab ab102of the chymotrysinized ROS membrane proteins into lipid vesicles gave no channel activityas determined by the Ca2+ efflux assay.Treatment of ROS membranes with trypsin also resulted in the cleavage of variousROS membrane proteins (Fig. 39). Western blotting analysis with monoclonal antibodiesPMc 6E7 and PMc 1D1 indicated that trypsin cleaved at the N-terminal region of the a-subunit and removed the PMc 6E7 binding site. Analysis with PMs 4B2 indicated that the240 kDa protein was broken down to various smaller fragments. The trypsinized ROSmembrane proteins were also reconstituted into lipid vesicles. Calcium efflux assays of thereconstituted vesicles indicated that trypsinization of the channel complex appeared toslightly shift the Km of the channel for cGMP to the left (Km shifted from 32 iM to 28pM; Fig. 40a and b). Addition of calmodulin to the assay mixture decreased the channelactivity for both samples (Fig. 41). However, lowering calmodulin concentration from200 to 25 nM resulted in a much more rapid reversal of the calmodulin effect for thetrypsinized ROS membrane vesicles than for the untrypsinized ROS membrane vesicles.This suggests that the calmodulin binding site was still intact but its ability to bindcalmodulin was dimished by the proteolytic cleavage. Calmodulin affinity chromatographyof the trypsinized ROS membrane proteins yielded a 59 kDa trypsinized form of thechannel a-subunit and a 70 kDa fragment of the 240 kDa protein (Fig. 42, left).However, this 70 kDa protein failed to be labelled by iodinated calmodulin on the Westernblot (Fig. 42, right).A third proteolytic enzyme, kallikrein, was also used to digest ROS membranes.As shown in the Coomassie blue stained SDS polyacrylamide gel (Fig. 43), the 240 kDaprotein appeared to be the only protein cleaved by this enzyme. Western blotting analysiswith PMc 1D1 and PMc 6E7 monoclonal antibodies indicated that the a-subunit of thechannel remained relatively intact. Analysis with PMs 4B2 monoclonal antibody indicatedthat the 240 kDa protein was cleaved into smaller fragments. The reconstituted vesicles ofthe kallikrein treated ROS membrane proteins gave a Km of 29 pM and a Hill coefficientkDa2O5-.1 16-i97-I66-___45-PMc 6E7 PMs 4B2i‘aFig. 39: Trypsin digestion of ROS membranes.ROS membranes were subjected to mild trypsin digestion. The digestedmembranes were run on an 8% SDS polyacrylamide gel and the intactness of the channelx-subunit and the 240 kDa protein was monitored by Western blotting analysis with anti-ce-subunit monoclonal antibodies PMc 1D1 and PMc 6E7 and anti-240 kDa proteinmonoclonal antibody PMs 4B2. Lanes a are the undigested ROS membranes (30 jig) andlanes b are the trypsinized ROS membranes (30 jig).103PMc 1D1HCB31-—ab ab ab ba104a1.00.80.6V0Vmax0.40.2b1.00.80.8V0Vmax040.200 10 20 30 40 50 60 70 80[cGMP) (iiM)Fig. 40: Reconstitution of trypsinized ROS membrane proteins.Ca2+ efflux assays were carried out to compare the effect of trypsinization onchannel activity. Intact or trypsinized ROS membrane proteins were solubilized andreconstituted into lipid vesicles. Aliquots of 0.3 ml of vesicles were used for the Ca2efilux assays at varying concentrations of cGMP. The relative initial velocities of thechannel were plotted as a function of the cGMP concentration. a. The intact channeldisplayed a Km of 32 p.M and a Hill coefficient of 3.5. b. The trypsinized channeldisplayed a Km of 28 p.M and a Hill coefficient of 3.3. The solid curved lines in a and bwere drawn using their respective Km and Hill coefficients.LcGMP] (pM)...105U,000Ht=lOs-CaM+ 25 nM CaM+2OOnMCaMFig. 41: Effect of trypsin treatment of ROS membranes on the calmodulinmodulation of the channel activity.Untrypsinized (a) and trypsinized (b) ROS membranes were reconstituted into lipidvesicles. The reconstituted vesicles were subjected to the Ca2+ efflux assays at 17.5 pMcGMP in the presence of 0, 25, and 200 nM calmodulin.a+ 200 nM CaM-CaMb106CB CaMkDa116_97-6645-31-Fig. 42: Purification of the trypsinized channel complex by calmodulin affinitychromatography.Trypsinized ROS membranes were solubilized and passed through a calmodulincolunm. The bound proteins were then eluted with EDTA. The protein samples wereanalyzed on an 8% SDS polyacrylamide gel (left). Western blotting analysis withiodinated calmodulin was carried out to detect calmodulin binding proteins (right). Lanesa, stripped ROS membranes (30 .ig); lanes b, calmodulin column eluant from the ROSmembranes; lanes c, trypsinized ROS membranes (30 .tg); lanes d, calmodulin columneluant from the trypsinized ROS membranes. The position of the 70 kDa fragment of the240 kDa protein is markerd by the arrow.1•...5:ab Cd abcd107Fig. 43: Digestion of ROS membranes with kallikrein.ROS membranes were digested with kallikrein. The intactness of the channelcomplex was monitored by SDS polyacrylamide gel electrophoresis and Western blottinganalysis with anti-ct-subunit monoclonal antibodies PMc 1D1 and PMc 6E7. Lanes a arethe ROS membranes (30 p.g) and lanes b are the kallikrein-treated ROS membranes (30kDa CBPMc6E7 PMc 1D1ab4108of 3.1 for cGMP (Fig. 44). No Attempts to determine whether the kallikrein treatment ofROS membranes would reverse the calmodulin effect were made since the calmodulinbinding site was still intact after this treatment as described below.5.3.4. Identification of a calmodulin binding fragment of the 240 kDa proteinWhen ROS membranes were digested with kallikrein as described above, the 240kDa protein was degraded while the 63 kDa ct-subunit remained relatively intact. Theproteolyzed ROS membranes were solubilized in CHAPS and passed through a PMc 6E7antibody column. The bound proteins were then eluted off the colunm by the competingsynthetic peptide. A 105 kDa protein was found to coelute with the 63 kDa channel ctsubunit as shown on the Coomasie blue stained SDS gel (Fig. 45 (left), lane d). This 105kDa fragment was shown to be labelled on Western blots by both iodinated calmodulinand monoclonal antibody PMs 3C9 but not by monoclonal antibodies PMs 5E1 1 and PMs4B2 (Fig. 45).In order to isolate this 105 k.Da fragment in large quantity, calmodulin affinitychromatography was employed. The 105 kDa protein was transferred onto an Immobilonmembrane and the protein band was subjected to N-terminal sequence analysis. Theamino acid sequence was found to have a high degree of homology with the N-terminalregion of the recently cloned human cGMP-gated channel f3-subunit (Chen et al., 1993).The sequence alignment is shown in Figure 46. The N-terminal sequence of the 105 kDafragment is upstream of the initiation site for the shorter form of the human j3-subunit(Chen et al., 1993). This suggests that the longer form of the f3-subunit is thepredominant species expressed in bovine ROS.5.4. DISCUSSIONCalmodulin modulates the activity of the cGI\’IP-gated channel by changing itsaffinity for cGMP. This modulation is mediated through the 240 kDa channel associated1091.00.80.6V0Vmax0.20Fig. 44: Reconstitution of the kallikrein-treated ROS membrane proteins.Kallikrein-treated ROS membranes were solubilized and reconstituted into lipidvesicles. The Ca2+efflux assays were carried out at varying concentrations of cGIvlPusing 0.3 ml of reconstituted lipid vesicles per assay. The relative initial velocities of thereconstituted channels were plotted as a thnction of cGMP concentration. The solidcurved lines were drawn using a Km of 29 p.M and a Hill coefficient of 3.1.0 10 20 30 40 50 60 70 80[cGMP] (pM)Fig. 45: Purification of the kallikrein-treated cGMP-gated channel complex by theanti-o-subunit PMc 6E7 monoclonal antibody column.ROS membranes treated with kallikrein were solubilized and passed through ananti-c-subunit PMc 6E7 antibody column. The bound proteins were eluted off the columnby synthetic peptides. The purified channel complex was analysed by SDS polyacrylamidegel electrophoresis and Western blotting analysis with either iodinated calmodulin or anti-240 kDa protein monoclonal antibodies PMs 3C9, 4B2, and 5E1 1. Lanes a, stripped ROSmembranes (30 j..tg); lanes b, intact channel complex purified by PMc 6E7 column; lanes c,kallikrein-treated ROS membranes (30 jig); lanes d, kallikrein-treated channel complexpurified by PMc 6E7 column. The positions of the 105 kDa calmodulin binding fragmentof the 240 kDa protein are indicated by arrows.110PMs 3C9 PMs 482 PMs 5E1 1 CaMkDa CB• .4.4abc dIabcd abcd abcd abc d111HRCNC2 B- MPRELSRIEEEKEDEEEEEEEEEEEEEEEVTEVLLDSCWSQVGVGQSEEDGTRP -55HRCNC2B- QSTSDQKLWEEVGEEAKKEAEEKAKEEAEEVAEEEAEKEPQDWAETKEEPEAEAE —110HRCNC2 B - AASSGVPATKQHPEVQVEDTDADSCPLMAEENPPSTVLPPPSPAKSDTLIVPSSA -165AEENPPSPVQLPLSPAKSDTLHRCNC2 B- SGTHRKKLPSEDDEAEELKALSPAESPVVAWSDPTTPKDTDGQDRAASTASTNSA -220HRCNC2B- IINDRLQELVKLFKERTEKVKEKLIDPDVTSDEESPKPSPAKKAPEPAPDTKPAE -275*HRCNC2 B - AEPVEEEHYCDMLCCKFKHRPWKKYQFPQSIDPLTNLMYVLWLFFVVMAWNWNCW -330HRCNC2 B - LI PVRWAFPYQTPDNIHHWLLMDYLCDLIYFLDITVFQTRLQFVRGGDIITDKKD -385HRCNC2B- MRNNYLKSRRFKIWLLSLLPLDFLYLKVGVNPLLRLPRCLKYMAFFEFNSRLESI -440HRCNC2 B - LSKAYVYRVIRTTAYLLYSLHLNSCLYYWASAYQGLGSTHWVYDGVGNSYIRCYY -495HRCNC2 B- FAVKTLITIGGLPDPKTLFEIVFQLLNYFTGVFAFSVMIGQMRDVVGAATAGQTY -550HRCNC2 B - YRSCMDSTVKYZ.WFYKIPKSVQNRVKTWYEYTWHSQGMLDESELMVQLPDKMRLD -605HRCNC2 B- LAIDVNYNIVSKVALFQGCDRQMIFDMLKRLRSVVYLPNDYVCKKGEIGREMYII -660HRCNC2 B - QAGQVQVLGGPDGI(SVLVTLKAGSVFGEISLLAVGGGNRRTANVVAHGFTNLFIL -715HRCNC2 B - DKKDLNEI LVHYPESQKLLRKKARRNLRSNNKPKEEKSVLILPPRAGTPKLFNAA -770HRCNC2 B - LAMTGK!4GGKGAKGGKLAHLRARLKELAALEAAAKQQELVEQAKSSQDVKGEEGS -825HRCNC2B - AAPDQHTHPKEAATDPPAPRTPPEPPGSPPSSPPPASLGRPEGEEEGPAEPEEHS -880HRCNC2B - VRICMSPGPEPGEQILSVIGIPEEREEKAE -909Fig. 46: N-terminal sequence alignment of the 105 kfla fragment with the cGMPgated channel n-subunit.The N-terminal sequence of the 105 kDa kallikrein fragment of the 240 kDaprotein was aligned with the human rod cGMP-gated channel f3-subunit (HRCN2 B). Theinitiation site of the shorter transcript of the human 13-subunit is marker by an asterisk.112protein. This was verified by Ca2+ efflux assays using reconstituted channels that werepurified by immunoaffinity chromatography. The purified channel complex which consistsof the 63 kDa (i-subunit and the 240 kDa channel associated protein is devoid of othercontaminating calmodulin binding proteins as analyzed by iodinated calmodulin binding onWestern blots. The reconstituted channel displayed a similar decrease in the affinity of thechannel for cGMP in the presence of calmodulin.Attempts were made to inhibit the interaction between calmodulin and the channelcomplex by either direct competition with anti-240 kDa protein monoclonal antibodies orby mild proteolytic cleavages to remove the calmodulin binding site. In Western blottingexperiments, none of the available anti-240 kDa protein monoclonal antibodies were ableto inhibit calmodulin binding to the 240 kDa protein. This suggests that the epitopes forthese antibodies are distant from the calmodulin binding site. Digestion of the ROSmembranes with chymotrypsin not only cleaved the 240 kDa protein, but it also removedthe C-terminal end of the cL-subunit, rendering the channel inactive. This suggests that theC-terminal segment of the channel may be important for the channel activity. Digestionwith trypsin, on the other hand, cleaved the 240 kDa protein and the N-terminal end of thechannel. The trypsinized channel complex was functionally active and showed increasedaffinity for cGMP. However, it is not known whether this increased affinity is the result ofthe cleavage of the 240 kDa protein or the N-terminus of the channel (i-subunit. Inaddition, the trypsinized channel complex still displayed calmodulin sensitivity but with adecreased affinity for calmodulin. Calmodulin affinity chromatography of the trypsinizedROS proteins yielded a 70 kDa fragment of the 240 kDa protein and the degraded 59 kDa(i-subunit. The 70 kDa fragment, however, was not labelled by iodinated calmodulin onthe Western blot. This is likely the result of the decreased affinity of the proteolyzed 70kDa fragment for calmodulin. It will be interesting in the future to determine the Nterminal sequence of this 70 kDa fragment. Digestion of the ROS membranes withkallikrein appeared to cleave only the 240 kDa protein. The digested channel complex113after reconstitution, was still functionally active. Purification of the kaliikrein-treatedchannel complex by an anti-channel antibody column yielded a polypeptide of 105 kDathat binds calmodulin and the intact 63 kDa channel a-subunit.Amino terminal sequence analysis of the 105 kDa protein gave a sequence thatshowed a high degree of homology with the recently cloned human 13-subunit of thecGMP-gated channel (Chen et al., 1993). Sequences corresponding to the human f3-subunit of the cGMP-gated channel have also been detected earlier from the peptidemapping analysis of the bovine 240 kDa protein (Tiling and Molday, unpublished data).Detection of the 13-subunit within the 240 kDa protein is in agreement with the finding ofBrown et a?. (1993), indicating that the 240 kDa protein as well as the 63 kDa a-subunitis labelled by a cGMP photoaffinity derivative.The longer transcript of the 13-subunit is likely to be the predominant formexpressed in bovine ROS as suggested by the N-terminal sequence analysis of the 105 kDafragment of the 240 kDa protein. The full length cDNA of the human 13-subunit encodes aprotein of 102 kDa. Thus, it is possible that the 13-subunit is only a component of the 240kDa protein. Analysis of the human 13-subunit revealed a stretch of 20 amino acids closeto N-terminus of the channel (a.a. 283-3 02) as a potential calmodulin binding site. Thisstretch contains a large number of positively charged residues. The hydrophobic residueswithin this stretch repeat with a 3 to 4 residue period to form an amphiphilic helix.This 13-subunit can be regarded as the regulatory component of the channelcomplex. Based on the observation that the 63 kDa a-subunit of the channel is present inmore or less similar amounts as the 240 kDa protein (Molday et a?., 1990), each channelcomplex potentially consists of a 240 kDa polypeptide (or one 13-subunit) associated with3 or more a-subunits. Upon binding toCa2-calmodulin, it is possible that the 13-subunitmay undergo a conformational change, leading to a decrease in the overall affinity of thechannel complex for cGMP. Under conditions of low intracellular Ca2+, the dissociationof calmodulin from the 13-subunit will then result in an increase in the affinity of the114channel complex for cGMP. Thus, theCa2-dependent interaction of calmodulin with thef3-subunit has the effect of increasing the sensitivity of the channel to cGMP and allows itto readily respond to fluctuations in cGMP concentration between dark and lightconditions.115SUMMARYPhotobleaching of rhodopsin in rod photoreceptors activates the visual cascadesystem leading to a decrease in the cGIvlP concentration and the closure of cGMP-gatedchannels in the rod outer segment plasma membrane. Calcium and Ca2+binding proteinsplay important roles in the recovery of the rod cell to its dark resting ‘state by regulatingthe activation of phosphodiesterase and the resynthesis of cGMP by guanylate cyclase(Kawamura and Murakami, 1991; Kawamura, 1993; Koch and Stryer, 1988). Theregulation of phosphodiesterase is mediated through a 26 kDa protein, recoverinlSmodulin, which modulates the level of rhodopsin phosphorylation and ultimately affectsthe rate of phosphodiesterase activation (Kawamura and Murakami, 1991; Kawamura,1993). Another Ca2-binding protein, originally thought to be recoverin, is responsiblefor modulating the guanylate cyclase activity in an as yet undefined manner (Koch andStryer, 1988; Hurley et al., 1993).This thesis describes the detection and purification of calmodulin from bovinephotoreceptor ROS and the identification of the cOMP-gated channel complex as themajor calmodulin binding protein in ROS membranes. In addition, by a variety of assaymethods using extruded ROS membrane vesicles and reconstituted cGIVIP-gated channelcomplex, calmodulin was shown to increase the apparent Km of the channel for cGMP(Table V). This effect has been shown to be mediated through the 240 kDa channelassociated protein. The effect was observed over a physiological Ca2+ concentrationrange of 20-300 nM. This suggests that changes in cytoplasmic Ca2 concentrationduring the photoresponse can affect the activity of the cGMP-gated channel. The channelis opened in response to the cooperative binding of cGMP to the channel subunits. Theaffinity of the channel for cGI\4P can be further modulated by the level of cytoplasmicCa2 through the binding ofCa2-calmodulin to the channel complex.116Table VSummary of the calmodulin modulation of the cGMP-gated channel complex asdetermined by various assay systems-CaM +CaMion Kin Hill Km HillDye translocated (p.M) CoefE (iiM) CoeffArsenazo III (3 trials) Ca2 19±0.4 3.7±0.1 33±2 3.5±0.6Dichiorophosphonazo III Ca 22 3.0 33 3.0Mg2 37 2.4 45 2.1Neutral red Ca2 34 2.4 51 2.2Reconstituted channel Ca2 33 3.3 44 3.1(Arsenazo III)117TheCa2+calmodulin effect on the cGMP-gated channel can be incorporated intothe current model for photoexcitation and photorecovery (Stryer, 1986; Chabre andDeterre, 1989; Pugh Jr. and Lamb, 1990; Kaupp and Koch, 1992). In the dark, the freecGMP level in ROS, estimated to be about 4-10 .tM, maintains a small, but significantnumber of channels in their open state (Nakatani and Yau, 1988a), allowing for the influxofNa and Ca2 (Fig. 47a). Under these conditions, cytoplasmic Ca2 is maintained at arelatively high concentration (approximately 0.3 .tM) by balancing the influx of Ca2through the channel with the efflux of Ca2 through the NaICa2-Kexchanger. Underthese conditions, guanylate cyclase is maintained at its basal level of activity and thecGMP-gated channel is in its low affinity (high Km) state for cGMP through Ca2-calmodulin binding to the 240 kDa channel-associated protein. Photoexcitation leads tothe activation of phosphodiesterase and a decrease in cGMP levels (Fig. 47b). Thedecrease in total cGMP has been reported to be only 10-15% (Cote et al., 1986), but thedecrease in free cGIVIP has not yet been measured. Since the channel is in its low affinitystate, it may be positioned on the cGMP dose-response curve to be sensitive to a smalldecrease in free cGMP concentration, thus, facilitating its closure. A decrease incytoplasmic Ca2 levels (estimated to be below 100 nM) will occur as a result of theclosure of the channel and the continuous extrusion of Ca2+ through the Na+/Ca2+K+exchanger. This decrease in Ca2will activate guanylate cyclase (Koch and Stryer, 1988)and limit PDE activation (Kawamura and Murakami, 1991). Under these conditions,calmodulin will also dissociate from the channel complex and cause the channel to switchto its high affinity state for cGMP. The channel will now reopen at a lower cGIvIP levelthus facilitating the recovery of the ROS to its dark level as cGMP synthesis proceeds.The opening of the channel will in turn restore the Ca2 level to its dark level, leading tothe inactivation of guanylate cyclase and the return of the channel to its low affinity state(Fig. 47c). Thus, reversible binding of calmodulin to the channel complex in response to118G + a 4NoK____Ca2®4G.C.G1P— cGMPcGMPJ.L 4NdL -NaCd’®fG.C.G1P —* cGcC III 4WKCd’®tCo”•. Na’Ca2’G1P —, cGIWLow In —‘ IiQh KmcGMP-gotd darmeI comøexNa/Co-IC oxcIaner• Cnod.*119Fig. 47: Possible role for calmodulin modulation of the channel during the visualtransduction process.a. In the dark, an elevated level of cGMP D maintains a significant number ofcGMP-gated channels in their open state and allows the influx of Na and Ca2 into theouter segment (2). The balanced influx of Ca2 through the channel with the effiux ofCa2 through the Na/CaKexchanger results in a relatively high level of Ca2 withinthe outer segment The high level of Ca2+ results in the association of calmodulin tothe channel complex and maintains the channel in its low affinity state for cGMP ©. Thechannel, in its low affinity state, will respond to a decrease in the level of cGIVIP duringphotoexcitation of the outer segment. This high level of Ca2 will also deactivate themodulator of guanylate cyclase and minimize the stimulation of guanylate cyclase activityb. Photobleaching of rhodopsin results in the activation of phosphodiesterase and adecrease in the level of cGMP D. This causes the closure of the cGIvIP-gated channel 2)and a decrease in intracellular Ca2+ due to the continuous extrusion of Ca2+ byNa+/Ca2+K+ exchanger. This drop in Ca2+ will cause calmodulin to dissociate from thechannel complex and shift the channel from its low affinity state to its high affinity state forcGMP ®. The low level of Ca2 will also activate guanylate cyclase through a Ca2sensitive modulator ®. The combined effect of resynthesizing cOMP and making thechannel more sensitive to cGMP levels would facilitate the recovery of the outer segmentto its dark level.c. As the channels reopen due to a rise in cGMP level D c2), the Ca2+ level in theouter segment is restored . This results in the rebinding of calmodulin to the channeland conversion of the channel to its low affinity state ®. Guanylate cyclase is alsorestored to its basal level of activity ®.120changes in the level of Ca2 has the effect of increasing the sensitivity of the channel tosmall changes in cGMP which may occur during the photoresponse.Light adaptation characterized by a decrease in rod sensitivity in the presence ofbackground illumination has been shown to be mediated by Ca2 (Torre et al., 1986;Nakatani and Yau, 1988b). Reduced Ca2 induced under conditions of light adaptationwill result not only in an increase in cGMP (Koch and Stryer, 1988), but also in a higheraffinity of the channel for cGMP, thus resulting in the observed reduction in the flashsensitivity characteristic of light adaptation. On this basis, Ca2 may regulatephotorecovery and light adaptation not only by regulating guanylate cyclase activity andPDE activation (Koch and Stryer, 1988; Kawamura and Murakami, 1991), but also bymodulating the affinity of the channel for cGIVIP. In addition to the visual system, asimilar calcium mediated regulation of the cAMP-gated channels has been reported in theolfactory system (Kramer and Siegelbaum, 1992). It is possible that this calcium mediatedmodulation of the cyclic nucleotide gated channels is not limited to the visual and olfactorysignal transduction systems. This regulation may potentially be an important componentin the auditory and gustatory signal transduction pathways as well.The identity of the 240 kDa channel associated protein has been under study. Inaddition to the f3-subunit recently identified by Chen et al. (1993), the 240 kDa proteinalso contains a glutamic acid rich protein (Tiling, Williams, Colville, and Molday,unpublished data) that had been cloned and sequenced earlier by Sugimoto et al. (1991).This glutamic acid rich protein (GARP) is potentially covalently linked to the 13-subunit togive an apparent Mr 240 K by SDS polyacrylamide gel electrophoresis. Since the N-terminal sequence analysis of the 105 kDa calmodulin binding fragment of the 240 kDaprotein does not contain the sequence of this glutamic acid rich protein, this protein islikely to be binding to an N-terminal stretch of the 13-subunit upstream of the kallikreincleavage site. Thus, the cOMP-gated channel complex can be envisioned as composed ofthree x-subunits associated tightly with one or more 13-subunits (Fig. 48). This 13-subunit,121ExtracellularIntracellularNaCa 2+Fig. 48: Schematic model of the cGMP-gated channel complex.The cGMP-gated channel complex can be envisioned as a pentameric complexconsisting of 3 ct-subunits, one f3-subunit, and one glutamic acid rich protein. Theglutamic acid rich protein is associated with the 13-subunit, possibly by covalent linkage.Calmodulin binds to the cGMP-gated channel complex in a Ca2+dependent mannerthrough the 13-subunit.GARP122which has the glutamic acid rich protein attached to it, interacts with calmodulin in aCa2-dependent manner. It will be interesting to determine the type of interactionbetween the 13-subunit and the glutamic acid rich protein as well as the stoichiometricratios of individual components of the cGMP-gated channel complex.123REFERENCESAitken, A., Klee, C. B., and Cohen, P. (1984) The structure of the 3 subunit ofcalcineurin. Eur. .J Biochem. 139, 663-671.Anderson, S. R. and Malencik, D. A. (1986) Peptides recognizing calmodulin. In CalciumCell Function 6 (Ed. Cheung, W. Y.), p.1-42, Academy Press, Orlando.Antoniw, J. F. and Cohen, P. (1976) Separation of two phosphorylase kinasephosphatases from rabbit skeletal muscles. Eur. I Biochem. 68, 45-54.Asai, H., Arai, T., Fujii, T., and Matsumoto, G. (1989) Purification and characterization ofCa2/caImodulin-dependent actin-binding proteins from squid retina. FEBS Lett. 247,377-380.Babu, Y. S., Sack, J. S., Greenhough, T. J., Bugg, C. E., Means, A. R., and Cook, W. J.(1985) Three-dimensional structure of calmodulin. Nature 315, 37-40.Bader, C. R., MacLeish, P. R., and Schwartz, E. A. (1979) A voltage-clamp study of thelight response in solitary rods of the tiger salamander. i: Physiol. 296, 1-26.Baehr, W., Devlin, M. J., and Applebury, M. L. (1979) Isolation and characterization ofcGIvIP phosphodiesterase from bovine rod outer segments. I Biol. Chem. 254, 11669-11677.Baitinger, C., Alderton, J., Poenie, M., Schulman, H., and Steinhardt, R. A. (1990)Multifunctional Ca2+/calmodulindependent protein kinase is necessary for nuclearenvelope breakdown. I Cell Biol. 111, 1763-1773.Baylor, D. A., Lamb, T. D., and Yau, K. -W. (1979a) The membrane current of single rodouter segments. I Physiol. 242, 685-727.Baylor, D. A., Lamb, T. D., and Yau, K. -W. (1979b) Responses of retinal rods to singlephotons. I Physiol. 288, 6 13-634.Baylor, D. A. and Nunn, B. J. (1986) Electrical properties of the light-sensitiveconductance of salamander rods. I Physiol. 371, 115-145.Bergiund, A., Backman, L., and Shanbhag, V. P. (1984) Calmodulin binding to humanspectrin. FEBS Lett. 172, 109-113.Bird, A. C., Flannery, J. G., and Bok, D. (1988) A diurnal rhythm in opsin content ofRana pipiens rod inner segments. Invest. Ophthalmol. Vis. Sci. 29, 1028-1039.124Bitensky, M. W., Gorman, R. E., and Miller, W. H. (1971) Adenyl cyclase as a linkbetween photon capture and changes in membrane permeability of frog photoreceptors.Proc. Nati. Acad. Sd. USA 68, 561-562.Bodoia, R. D. and Detwiler, P. B. (1985) Patch-clamp recordings of the light-sensitivedark noise in retinal rods from the lizard and frog. J Physiol. 367, 183-2 16.Boesze-Battaglia, K. and Albert, A. D. (1989) Fatty acid composition of bovine rod outersegment plasma membrane. Exp. Eye Res. 49, 699-701.Bok, D. and Heller, J. (1976) Transport of retinol from the blood to the retina:autoradiographic study of the pigment epithelial cell surface receptor for plasma retinolbinding protein. Exp. Eye Res. 22, 395-402.Bolton, A. E. and Hunter, W. M. (1973) The labelling of proteins to high specificradioactivities by conjugation to a 125-containing acylating agent. Biochem. 1 133, 529-539.BOnigk, W., Altenhofen, W., Muller, F., Dose, A., Illing, M., Molday, R. S., and Kaupp,U. B. (1993) Rod and cone photoreceptor cells express distinct genes for cyclic GMPgated channels. Neuron 10, 865-877.Bownds, D. (1967) Site of attachment of retinal in rhodopsin. Nature 216, 1178-1181.Bownds, D., Dawes, J., Miller, J., and Stahiman, M. (1972) Phosphorylation of frogphotoreceptor membranes induced by light. Nature 237, 125-127.Brostrom, C. 0., Huang, Y. C., Breckenridge, B. M., and Wolfi D. J. (1975)Identification of a calcium binding protein as a calcium dependent regulator of brainadenylate cyclase. Proc. Nati. Acad. Sci. USA 72, 64-68.Brown, R. L., Gerber, W. V., and Karpen, J. W. (1993) Specific labelling and permanentactivation of the retinal rod cGMP-activated channel by the photoaffinity analog 8-p-azidophenacylthio-cGMP. Proc. NatL Acad. Sci. USA 90, 5369-5373.Burger, D., Cox, 3. A., Comte, M., and Stein, E. A. (1984) Sequential conformationalchanges in calmodulin upon binding to calcium. Biochemistry 23, 1966-1971.Burns, N. R. and Gratzer, W. B. (1985) Interaction of calmodulin with the red cell and itsmembrane skeleton and with spectrin. Biochemistry 24, 3070-3074.Carafoli, E. and Zurini, M. (1982) The Ca2-pumping ATPase of plasma membranes:purification, reconstitution and properties. Biochim. Biophys. Acta 683, 279-30 1.125Caretta, A., Cavaggioni, A., Grimaldi, R., and Sorbi, R.T. (1988) Regulation of cyclicGMP binding to retinal rod membranes by calcium. Eur. J Biochem. 177, 139-146.Caroni, P. and Carofoli, E. (1983) The regulation of the Na-Ca2 exchanger of heartsarcolemma. Eur. J Biochem. 132, 451-460.Cervetto, L. and Fuortes, M. G. F. (1978) Excitation and interaction in the retina. Annu.Rev. Biophys. Bioeng. 7, 229-251.Cervetto, L., Menini, A., Rispoli, G., and Torre, V. (1988) The modulation of the ionicselectivity of the light-sensitive current in isolated rods of the tiger salamander. I PhysioL406, 181-198.Chabre, M. (1985) Trigger and amplification mechanisms in visual transduction. Annu.Rev. Biophys. Biophys. Chem. 14, 33 1-360.Chabre, M. and Deterre, P.( 1989) Molecular mechanism of visual transduction. Eur. IBiochem. 179, 255-266.Charbonneau, H. and Cormier, M. J. (1979) Purification of plant calmodulin byfiuphenazine-sepharose affinity chromatography. Biochem. Biophys. Res. Commun. 90,1039-1047.Chen, T. -Y., Peng, Y. -W., Dhallan, R. S., Ahamed, B., Reed, R. R., and Yau, K. -W.(1993) A new subunit of the cyclic nucleotide-gated cation channel in retinal rods. Nature362, 764-767.Chen, Y. S. and Hubbell, W. L. (1978) Reactions of the sulfhydryl groups of membranebound bovine rhodopsin. Membr. Biochem. 1, 107-13 0.Cheung, W. Y. (1970) Cyclic 3’,5’-nucleotide phosphodiesterase. Demonstration of anactivator. Biochem. Biophys. Res. Commun. 38, 533-53 8.Cheung, W. Y., Bradham, L. S., Lynch, T. J., Lin, Y. M., and Tallant, E. A. (1975)Protein activator of cyclic 3’: 5’-nucleotide phosphodiesterase of bovine or rat brain alsoactivates adenylate cyclase. Biochem. Biophys. Res. Commun. 66, 105 5-1062.Chien, Y. H. and Dawid, I. B. (1984) Isolation and characterization of calmodulin genesfrom Xenopus laevis. Mo!. Cell Biol. 4, 507-513.Cohen, A. I., Hall, I. A., and Ferrendelli, J. A. (1978) Calcium and cyclic nucleotideregulation in incubated mouse retina. .1 Gen. Physiol. 71, 595-6 12.Cohen, P. (1982) The role of protein phosphorylation in neural and hormonal control ofcellular activity. Nature 296, 613-620.126Cohen, P. (1986) Muscle glycogen synthase. In The Enzymes XVII, (Eds. Boyer, P. D.and Krebs, E. G.) p461-497, Academic Press, Orlando.Cohen, P., Burchell, A., Foulkes, J. G., and Cohen, P. T. W. (1978) Identification of thea2-dependent modulator protein as the fourth subunit of rabbit skeletal musclephosphorylase kinase. FEBS Lett. 92, 287-293.Cook, N. J., Hanke, W., and Kaupp, U. B. (1987) Identification, purification, andfunctional reconstitution of the cyclic GMP-dependent channel from rod photorceptors.Proc. Nat!. Acad. Sd. USA 84, 585-589.Cook, N. J., Molday, L. L., Reid, D., Kaupp, U. B., and Molday, R. S. (1989) ThecGMP-gated channel of bovine rod photoreceptors is localized exclusively in the plasmamembrane. J Blot. Chem. 264, 6996-6999.Cooper, A., Dixon, S. F., Nutley, M. A., and Robb, J. L. (1987) Mechanism of retinalschiff base formation and hydrolysis in relation to visual pigment photolysis andregeneration. Resonance raman spectroscopy of a tetrahedral carbinolamine intermediateand oxygen- 18 labeling of retinal at the metarhodopsin stage in photoreceptor membranes.I Am. C’hem. Soc. 109, 7254-7263.Cote, R. H., Burke, S. A., Nicol, B. D., and Bownds, M. D. (1985) Involvement of cGIvIPin phototransduction by isolated frog rods in the absence of transmembrane calciumgradients. Biophys. 1 47, lOOa.Cote, R. H., Nicol, G. D., Burke, S. A., and Bownds, M. D. (1986) Changes in cGIvIPconcentration correlate with some, but not all, aspects of the light-regulated conductanceof frog rod photoreceptors. I Blot. C’hem. 261, 12965-12975.Cox, J. A., Comte, M., Fitton, J. E., and DeGrado, W. F. (1985) The interaction ofcalmodulin with amphiphilic peptides. I Biol. Chem. 260, 2527-2534.Davis, T. N., Urdea, M. S., Masiarz, F. R., and Thorner, J. (1986) Isolation of the yeastcalmodulin gene: calmodulin is an essential protein. Cell 47, 423-431.Dc Couet, H. G., Jablonski, P. P., and Perkin, 3. L. (1986) Calmodulin associated withrhabdomeral photoreceptor microviffi of arthropods and squid. Cell Tissue Res. 244, 315-319.DeGrado, W. F. (1988) Design of peptides and proteins. Adv. Protein Chem. 39, 5 1-124.DeGrado, W. F., Prendergast, F. G., Wolfe Jr., H. R., and Cox, J. A. (1985) The design,synthesis, and characterization of tight binding inhibitors of calmodulin. I Cell. Biochem.29, 83-93.127Del Priore, L. V. and Lewis, A. (1983) Calcium-dependent activation and deactivation ofrod outer segment phosphodiesterase is calmodulin-independent. Biochem. Biophys. Res.Commun. 113, 3 17-324.Dhallan, R. S., Macke, J. P., Eddy, R. L., Shows, T. B., Reed, R. R., Yau, K. -W., andNathans, 3. (1992) Human rod photoreceptor cGMP-gated channel: amino acid sequence,gene structure, and ftjnctional expression. I Neurosci. 12, 3248-3256.Dizhoor, A. M., Ray, S., Kumar, S., Niemi, 0., Spencer, M., Brolley, D., Walsh, K. A.,Philipov, P. P., Hurley, J. B., and Stryer, L. (1991) Calcium sensitive activator of retinalrod guanylate cyclase. Science 251, 915-918.Doskeland, A. P., Schworer, C. M., Doskeland, 5. 0., Chrisman, T. D., Soderling, T. R.,Corbin, J. D., and Flatmark, T. (1984) Some aspects of the phosphorylation ofphenylalanine 4-monooxygenase by a calcium-dependent and calmodulin-dependentprotein kinase. Eur. I Biochem. 145, 31-37.Dratz, E. A., Lewis, 3. W., Schaechter, L. E., Parker, K. R., and Kiiger, D. S. (1987)Retinal rod GTPase turnover rate increases with concentration: A key to the control ofvisual excitaion? Biochern. Biophys. Res. ommun. 146, 379-386.Durell, S. R. and Guy, R. (1992) Atomic scale structure and functional models of voltagegated potassium channels. Biophys. .1 62, 23 8-250.Erickson, M. A., Robinson, P., and Lisman, J. (1992) Deactivation of visual transductionwithout guanosine triphosphate hydrolysis by G protein. Science 257, 1255-1258.Fabiato, A. (1988) Computer programs for calculating total from specified free or freefrom specified total ionic concentrations in aqueous solutions containing multiple metalsand ligands. Methods Enzynzol. 157, 378-417.Fain, G. L. and Matthews, H. R. (1990) Calcium and the mechanism of light adaptation invertebrate photoreceptors. Trends Neurosci. 13, 378-384.Fesenko, E. E., Kolesnikov, S. S., and Lyubarsky, A. L. (1985) Induction by cyclic GMPof cationic conductance in plasma membrane of retinal rod outer segment. Nature 313,310-313.Flanagan, S.D. and Yost, B. (1984) Calmodulin-binding proteins: visualization by 125j.calmodulin overlay on blots quenched with Tween 20 or bovine serum albumin andpoly(ethylene oxide). Anal. Biochem. 140, 510-519.Freitag, C. and Clarke, S. (1981) Reversible methylation of cytoskeletal and membraneproteins in intact human erythrocytes. I Biol. Chem. 256, 6102-6108.128Fuortes, M. 0. G. and Hodgkin, A. L. (1964) Changes in time scale and sensitivity in theommatidia of Limulus. I PhysioL 172, 239-263.Giedroc, D. P., Ling, N., and Puett, D. (1983) Identification of beta-endorphin residues14-25 as a region involved in the inhibition of calmodulin-stimulated phosphodiesteraseactivity. Biochemistry 22, 5584-5591.Glenney Jr., J. R., Glenney, P., and Weber, K. (1982 ) Erythroid spectrin, brain fodrin, andintestinal brush border proteins (TW-260/240) are related molecules containing a commoncalmodulin-binding subunit bound to a variant cell type-specific subunit. Proc. Nail. Acad.Sd. USA 79, 4002-4005.Godchaux, W. and Zimmerman, W. F. (1979) Soluble proteins of intact bovine rod cellouter segments. Exp. Eye Res. 28, 483-500.Goidhagen, H. and Clarke, M. (1986) Identification of the single gene for calmodulin inDictyostelium discoideum. Mo!. Cell BioL 6, 1851-1854.Gopinath, R. M. and Vincenzi, F. F. (1977) Phosphodiesterase protein activator mimicsred blood cell cytoplasmic activator of (Ca2-MgjATPase. Biochem. Biophys. Res.Commun. 77, 1203-1209.Goto, S., Yamamoto, H., Fukunaga, K., Iwasa, T., Matsukado, Y., and Miyamoto, E.(1985) Dephosphorylation of microtubule-associated protein 2, tau factor, and tubulin bycalcineurin. I Neurochem. 45, 276-283.Goulding, E. H., Ngai, J., Kramer, R. H., Colicos, S., Axel, R., Siegelbaum, S. A., andChess, A. (1992) Molecular cloning and single channel properties of the cyclic nucleotidegated channel from catfish olfactory neurons. Neuron 8, 45-58.Grand, R. J. A., Perry, S. V., and Weeks, R. A. (1979) Troponin C-like proteins(calmodulins) from mammalian smooth muscle and other tissues. Biochem. 1 177, 521-529.Gray-Keller, M. P., Polans, A. S., Palczewski, K., and Detwiler, P. B. (1993) The effect ofrecoverin-like calcium-binding proteins on the photoresponse of retinal rods. Neuron 10,523-531.Hakki, S. and Sitaramayya, A. (1990) Guanylate cyclase from bovine rod outer segments:solubilization, partial purification, and regulation by inorganic pyrophosphate.Biochemistry 29, 1088-1094.129Hargrave, P. A., Fong, S. L., McDowell, J. H., Mas, M. T., Curtis, D. R., Wang, J. K.,Juszczak, E., and Smith, D. P. (1980) The partial primary structure of bovine rhodopsinand its topography in the retinal rod cell disc membrane. Neurochem. mt. 1, 231-244.Haynes, L. W., Kay, A. R., and Yau, K. -W. (1986) Single cyclic GIVIP-activated channelactivity in excised patches of rod outer segment membrane. Nature 321, 66-70.Hazelton, B., Mitchell, B., and Tupper, J. (1979) Calcium, magnesium and growth controlin the W1-38 fibroblast cell. I Cell Blot. 83, 487-498.Heginbotham, L., Abramson, T., and MacKinnon, R. (1992) A functional connectionbetween the pores of distantly-related ion channels as revealed by mutant K+ channels.Science 258, 1152-1155.Hemmings Jr., H. C., Greengard, P., Tung, H. Y. L., and Cohen, P. (1984) DARPP-32, adopamine-regulated neuronal phosphoprotein, is a potent inhibitor of protein phosphatase1. Nature 310, 503-505.Hoar, P. E., Kerrick, W. G. L., and Cassidy, P. S. (1979) Chicken gizzard:relationbetween calcium activated phosphorylation and contraction. Science 204, 503-506,Hodgkin, A. L., McNaughton, P. A., and Nunn, B. J. (1985) The ionic selectivity andcalcium dependence of the light-sensitive pathway in toad rods. I PhysioL 358, 447-468.Hoffiiann, K. P., Pulvermuller, A., Buczylko, J., Van Hooser, P., and Palczewski, K.(1992) The role of arrestin and retinoids in the regeneration pathway of rhodopsin. I Blot.Chem. 267, 15701-15706.Hope, M. J., Bally, M. B., Webb, G., and Cullis, P. R. (1985) Production of largeunilamellar vesicles by a rapid extrusion procedure. Characterization of size distribution,trapped volume and ability to maintain a membrane potential. Biochim. Biophys. Acta 812,55-65.Hsu, S. -C. and Molday, R. S. (1990) Glyceraldehyde-3-phosphate dehydrogenase is amajor protein associated with the plasma membrane of retinal photoreceptor outersegments.J Blot. Chem. 265, 13308-13313.Hurley, J. B., Dizhoor, A. M., Ray, S., and Stryer, L. (1993) Recoverin’s role: conclusionwithdrawn. Science 260, 740.Hurley, J. B. and Stryer, L. (1982) Purification and characterization of the gammaregulatory subunit of the cyclic GMP phosphodiesterase from retinal rod outer segments.I Blot. C’hem. 257, 11094-11099.130lida, S. and Potter, J. D. (1986) Calcium binding to calmodulin. Cooperativity of thecalcium binding sites. I Biochem. 99, 1765-1772.Ingebritsen, T. S., Stewart, A. A., and Cohen, P. (1983) The protein phosphatasesinvolved in cellular regulation. 6. Measurement of type-i and type-2 protein phosphatasesin extracts of mammalian tissues: an assessment of their physiological roles. Eur. IBiochem. 132, 297-307.Ishiguro, S., Suzuki, Y., Tamai, M., and Mizuno K. (1991) Purification of retinoldehydrogenase from bovine retinal rod outer segments. I Biol. Chem. 266, 15520-15524.Jamieson, G. A. and Vanaman, T. C. (1979) Calcium-dependent affinity chromatographyof calmodulin on an immobilized phenothiazine. Biochem. Biophys. Res. Commun. 90,1048-1056.Jan, L. Y. and Jan, Y. N. (1990) A superfamily of ion channels. Nature 345, 672.Job, D., Fischer, E. H., and Margolis, R. L. (1981) Rapid disassembly of cold-stablemicrotubules by calmodulin. Proc. Nati. A cad. Sci. USA 78, 4679-4682.Johnson, L. V. and Blanks, J. C. (1984) Application of acrylamide as an embeddingmedium in studies of lectin and antibody binding in the vertebrate retina. Curr. Eye Res. 3,969-974.Kauer, J. C., Erickson-Viitanen, S., Wolfe Jr., H. R., and DeGrado, W. F. (1986) pBenzoyl-L-phenylalanine, a new photoreactive amino acid. Photolabeling of calmodulinwith a synthetic calmodulin binding peptide. I Biol. Chem. 261, 10695-10700.Kaupp, U. B. (1991) The cyclic nucleotide-gated channels of vertebrate photoreceptorsand olfactory epithelium. TINS 14, 150-157.Kaupp, U. B. and Koch, K. -W. (1992) Role of cGMP and Ca2 in vertebratephotoreceptor excitation and adaptation. Ann. Rev. Phys. 54, 153-175.Kaupp, U. B., Niidome, T., Tanabe, T., Terada, S., BOnigk, W., Stuhmer, W., Cook, N.J., Kangawa, K., Matsuo, H., Hirose, T., Miyata, T., and Numa, S. (1989) Primarystructure and functional expression from complementary DNA of the rod photoreceptorcyclic GMP-gated channel. Nature 342, 762-766.Kawamura, S. (1993) Rhodopsin phosphorylation as a mechanism of cyclic GMPphosphodiesterase regulation by S-modulin. Nature 362, 855-857.Kawamura, S. and Murakami, M. (1991) Calcium-dependent regulation of cyclic GMPphosphodiesterase by a protein from frog retinal rods. Nature 349, 420-423.131Keith, C. H., Ratan, R., Maxfielcl, F. R., Bajer, A., and Shelanski, M. L. (1985) Localcytoplasmic calcium gradients in living mitotic cells. Nature 316, 848-850.Keller, T. C. S., Jemiolo, D. K., Burgess, W. H., and Rebhun, L. I. (1982)Strongylocentrotus purpuratus spindle tubulin. II. Characterization of its sensitivity toCa2 and the effects of calmodulin isolated from bovine brain and S. purpuratus eggs. JCell. Blot. 93, 797-803.Kennedy, M. B., Bennett, M. K., and Erondu, N. E. (1983b) Biochemical andimmunochemical evidence that the “major postsynaptic density protein” is a subunit of acalmodulin-dependent protein kinase. Proc. Nat. Acad. SeE. USA 80, 7357-7361.Kennedy, M. B., McGuinness, T., and Greengard, P. (1983a) A calcium/calmodulindependent protein kinase from manmialian brain that phosphorylates Synapsin I: partialpurification and characterization. I Neurosci. 3, 818-831.Kihira, M., Matsuzawa, K., Tokuno, H., and Tomita T. (1990) Effects of calmodulinantagonists on calcium-activated potassium channels in pregnant rat myometrium. Br. IPharmacol. 100, 353-359.Kilbride, P. and Ebrey, T. G. (1979) Light-initiated changes of cyclic guanosinemonophosphate levels in the frog retina measured with quick-freezing techniques. .1 Gen.Physiol. 74, 4 15-426.Kincaid, R. (1987) The use of melittin-Sepharose chromatography for gram-preparativepurification of calmodulin. Methods Enzymol. 139, 3-19.Klee, C. B. (1988) Interaction of calmodulin with Ca2 and target proteins. Mo!. Asp.Cell. Regul. 5, 35-56.Klee, C. B., Crouch, T. H., and Krinks, M. H. (1979) Calcineurin: a calcium- andcalmodulin-binding protein of the nervous system. Proc. Nati. Acad SeE. USA 76, 6270-6273.Klee, C. B. and Vanaman, T. C. (1982) Calmodulin. Adv. Prot. Chem. 35, 213-321.Kiockner, U. and Isenberg, G. (1987) Calmodulin antagonists depress calcium andpotassium currents in ventricular and vascular myocytes. Am. I Physiol. 253 (Heart Circ.Physiol. 22): H1601-H1611.Koch, K. -W. and Kaupp, U.B. (1985) Cyclic GMP directly regulates a cationconductance in membranes of bovine rods by a cooperative mechanism. I BioL Chem.260, 6788-6800.132Koch, K. -W. and Stryer, L. (1988) Highly cooperative feedback control of retinal rodguanylate cyclase by calcium ions. Nature 334, 64-66.Kohnken, R. E., Chafouleas, J. G., Eadie, D. M., Means, A. R., and McConnell, D. G.(1981) Calmodulin in bovine rod outer segments. J Biol. Chem. 256, 125 17-12522.Konig, B., Arendt, A., McDowell, J. H., Kahlert, M., Hargrave, P. A., and Hofhiann, K.P. (1989) Three cytoplasmic loops of rhodopsin interact with transducin. Proc. NathAcaci Sci. USA 86, 6878-6882.Korenbrot, J. I. and Miller, D. L. (1989) Cytoplasmic free calcium concentration in dark-adapted retinal rod outer segments. Vis. Res. 29, 939-948.Kretsinger, R., H., Moncrief N. D., Goodman, M., and Czelusniak, J. (1988) Homologyof calcium modulated proteins: their evolutionary and functional relationships. In TheCalcium Channel Structure, Function and Implication (Eds. Morad, M., Naylor, W. G.,Kazda, S., and Schramm, M.) p.16-34, Springer-Verlag, Berlin.Kretsinger, R. H. and Nockolds, C. E. (1973) Carp muscle calcium-binding protein. II.Structure determination and general description. .1 Biol. Chem. 248, 3313-3326.Kramer, R. H. and Siegelbaum, S. A. (1992) Intracellular Ca2 regulates the sensitivity ofcyclic nucleotide-gated channels in olfactory receptor neurons. Neuron 9, 897-906.Kuhn, H. (1980) Light- and GTP-regulated interaction of GTPase and other proteins withbovine photoreceptor membranes. Nature 283, 587-589.KUhn, H. and Dreyer, W. J. (1972) Light dependent phosphorylation of rhodopsin byATP. FEBSLett. 20, 1-6.Kuhn, H., Hall, S. W., and Wilden, U. (1984) Light-induced binding of a 48-kDa proteinto photoreceptor membrane is highly enhanced by phosphorylation of rhodopsin. FEBSLett. 176, 473-478.KUhn, H. and Hargrave, P. A. (1981) Light-induced binding of guanosine triphosphataseto bovine photoreceptor membranes: effect of limited proteolysis of the membranes.Biochemistry 20, 2410-2417.Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head ofbacteriophage T4. Nature 227, 680-685.Lamb, T. D., McNaughton, P. A., and Yau, K. -W. (1981) Spatial spread of activationand background desensitization in toad rod outer segments. J Physiol. 319, 463-496.133Lambrecht, H. -G. and Koch, K. -W. (1991) A 26 kDa calcium binding protein frombovine rod outer segments as modulator of photoreceptor guanylate cyclase. FJvIBO 10,793-798.Lamola, A. A., Yamane, T., and Zipp, A. (1974) Effects of detergents and high pressuresupon the metarhodopsin 1-metarhodopsin U equilibrium. Biochemistry 13, 73 8-745.Llinas, R., McGuinness, T. L., Leonard, C. S., Sugimori, M., Greengard, P. (1985)Intraterminal injection of Synapsin I or calcium/calmodulin-dependent protein kinase IIalters neurotransmitter release at the squid giant synapse. Proc. Nat!. A cad. Sci. USA 82,3035-3039.McDowell, L., Sanyal, G., and Prendergast, F. G. (1985) Probable role of amphiphilicityin the binding of mastoparan to calmodulin. Biochemistry 24, 2979-2984.McNaughton, P.A., Cervetto, L., and Nunn, B. 3. (1986) Measurement of the intracellularfree calcium concentration in salamander rods. Nature 322, 26 1-263.Matthews, H. R., Torre, V., and Lamb, T. C. (1985) Effects on the photoresponse ofcalcium buffers and cyclic GMP incorporated into the cytoplasm of retinal rods. Nature313, 582-585.Merat, D. L., Hu, Z. Y., Carter, T. E., and Cheung, W. Y. (1985) Bovine braincalmodulin-dependent protein phosphatase. Regulation of subunit A activity by calmodulinand subunit B. J Biol. Chem. 260, 11053-11059.Ivliller, 3. L., Fox, D. A., and Litman, B. J. (1986) Amplification of phosphodiesteraseactivation is greatly reduced by rhodopsin phosphorylation. Biochemistry 25, 4983-4988.Minowa, 0. and Yagi, K. (1984) Calcium binding to tryptic fragments of calmodulin. IBiochem. 96, 1175-1182.Molday, L. L., Cook, N. J., Kaupp, U. B., and Molday, R. S. (1990) The cGIvIP-gatedcation channel of bovine rod photoreceptor cells is associated with a 240-kDa proteinexhibiting immunochemical cross-reactivity with spectrin. I Biol. Chem. 265, 18690-18695.Molday, R. S. and Molday, L. L. (1987) Differences in the protein composition of bovineretinal rod outer segment disk and plasma membranes isolated by a ricin-gold-dextrandensity perturbation method. I Cell Bio!. 105, 2589-2601.Molday, R. S. and Molday, L. L. (1993) Isolation and characterization of rod outersegment disk and plasma membranes. Methods Neurosci. 15, 131-150.134Molday, R. S., Moiday, L. L., Dose, A., Clark-Lewis, I., Tiling, M., Cook, N. J., Eismann,E., and Kaupp, U. B. (1991) The cGMP-gated channel of the rod photoreceptor cellcharacterization and orientation of the amino terminus. .1 Biol. Chem. 266, 21917-21922.Morelli, A, Damonte, G., Panfoli, I., and Pepe, I. (1989) Proteins of rod outer segments oftoad retina: binding with calmodulin and with GTP. Biochem. Biophys. Res. Commun.163, 363-369.Murtaugh, T. J., Rowe, P. M., Vincent, P. L., Wright, L. S., and Siegel, F. L. (1983)Posttranslational modification of calmodulin. Methods Enzymol. 102, 158-170.Munjaal, R. P., Chandra, T., Woo, S. L. C., Dedman, J. R., Means, A. R. (1986) A clonedcalmodulin structural gene probe is complementary to DNA sequences from diversespecies. Proc. Nat!. Acad Sd. USA 78, 2330-2334.Nagao, S., Suzuki, Y., Watanabe, Y., and Nozawal, Y. (1979) Activation by calcium-binding protein of guanylate cyclase in Tetrahymena pyriformis. Biochem. Biophys. Res.Commun. 90, 261.Nagao, S., Yamazaki, A., and Bitensky, M. W. (1987) Calmodulin and calmodulin bindingproteins in amphibian rod outer segments. Bochemist,y 26, 1659-1665.Nagle, B. and Burnside, B. (1984) Calmodulin-binding proteins in teleost retina, rod innerand outer segments, and rod cytoskeletons. Eur. I Cell Blot. 33, 248-257.Nakatani, K. and Yau, K. -W. (1988a) Calcium and magnesium fluxes across the plasmamembrane of the toad rod outer segment. I Physiol. 395, 695-729.Nakatani, K. and Yau, K. -W. (1988b) Calcium and light adaptation in retinal rods andcones. Nature 334, 69-71.Nicol, G. D., Schnetkamp, P. P. M., Saimi, Y., Cragoe Jr., E. J., and Bownds, M. D.(1987) A derivative of amiloride blocks both the light- and cyclic GMP-regulatedconductances in rod photoreceptors. I Gen. Physiol. 90, 651-670.Nicol, G. D. (1993) The calcium channel antagonist, pimozide, blocks the cGMPactivated current in rod photoreceptors. I Pharmacol. Exp. Ther. 265, 626-632.Nikonov, S. S., Filatov, G. N., and Fesenko, E. E. (1993) On the activation ofphosphodiesterase by a 26 kDa protein in vitro. I. Purification and characterization. FEBSLetters 316, 34-3 6.Nishida, E. and Sakai, H. (1980) A new protein factor that modulates assembly ofmicrotubules in vitro. I. Purificaion and characterization. I Biochem. 88, 1577-1586.135Nojima, H. and Sokabe, H. (1987) Structure of a gene for rat calmodulin. I Mol. Biol.193, 439-445.Ogawa, Y. and Tanokura, M. (1984) Calcium binding to calmodulin: effects of ionicstrength, Mg2 pH and temperature. I Biochem. 95, 19-28.O’Neil, K. T. and DeGrado, W. F. (1990) How calmodulin binds its targets: sequenceindependent recognition of amphiphilic cL-helices. TIBS 15, 59-64.O’Neil, K. T., Erickson-Viitanen, S., and DeGrado, W. F. (1989) Photolabelling ofcalmodulin with basic, amphiphilic alpha-helical peptides containing pbenzoylphenylalanine. I Biol. Chem. 261, 14571-14578.Palczewski, K. and Benovic, J. L. (1991) G-protein-coupled receptor kinases. TIBS 16,387-391.Palczewski, K., McDowell, J. H., Jakes, S., Ingebritsen, T. S., and Hargrave, P. A. (1989)Regulation of rhodopsin dephosphorylation by arrestin. I Biol. Chem. 264, 15770-15773.Pallen, C. J., Valentine, K. A., Wang, J. H., and Hollenberg, M. D. (1985) Calcineurinmediated dephosphorylation of the human placental membrane receptor for epidermalgrowth factor urogastrone. Biochemistry 24, 4727-4730.Penn, R. D. and Hagins, W. A. (1972) Kinetics of the photocurrent of retinal rods.Biophys. 1 12, 1073-1094.Penniston, J. T. (1982) Plasma membrane Ca2-pumping ATPase. Ann. NY. Acaci Sci.402, 296-303.Piazza, G. A. and Wallace, R. W. (1985) Calmodulin accelerates the rate ofpolymerization of human platelet actin and alters the structural characteristics of actinfilaments. Proc. Nati. Acaci Sci. USA 82, 1683-1687.Pittler, S. 3., Lee, A. K., Altherr, M. R., Howard, T. A., Seldin, M. F., Hurwitz, R. L.,Wasmuth, 3. J., and Baehr, W. (1992) Primary structure and chromosomal localization ofhuman and mouse rod photoreceptor cGMP-gated cation channel. I Biol. Chem. 267,6257-6262.Poenie, M., Alderton, J., Tsien, R. Y., and Steinhardt, R. A. (1985) Changes of freecalcium levels with stages of the cell division cycle. Nature 315, 147-149.Pugh Jr., E. N. and Lamb, T. D. (1990) Cyclic GMP and calcium: The internal messengersof excitation and adaptation in vertebrate photoreceptors. Vision Res. 30, 1923-1948.136Putkey, J. A., Ts’ui, K. F., Tanaka, T., Lagace, L., Stein, J. P., Lai, E. C., and Means, A.R. (1983) Chicken calmodulin genes. A species comparison of eDNA sequences andisolation of a genomic clone. J Biol. Chem. 258, 11864-11870.Ratto, G. M., Payne, R., Owen, W. G., and Tsien, R. Y. (1988) The concentration ofcytosolic free calcium in vertebrate rod outer segments measured with Fura-2. .1Neurosci. 8, 3240-3246.Reid, D. M., Friedel, U., Molday, R. S., and Cook, N. 3. (1990) Identification of thesodium-calcium exchanger as the major ricin-binding glycoprotein of bovine rod outersegments and its localization to the plasma membrane. Biochemistry 29, 160 1-1607.Roberts, D. M., Rowe, P. M., Siegel, F. L., Lukas, T. 3., and Watterson, D. M. (1986)Trimethyllysine and protein function. Effect of methylation and mutagenesis of lysine 115of calmodulin on NAD kinase activation. .1 Biol. Chem. 261, 1491-1494.Roof D. 3. and Heuser, J. E. (1982) Surfaces of rod photoreceptor disk membranes:integral membrane. J Cell Biol. 95, 487-500.Rowe, P. M., Wright, L. S., and Siegel, F. L. (1986) Calmodulin N-methyl transferase.Partial purification and characterization components. I Biol. Chem. 261, 7060-7069.Saimi, Y. and Ling, K. -Y. (1990) Calmodulin activation of calcium-dependent sodiumchannels in excised membrane patches of Paramecium. Science 249, 144 1-1444.Schnetkamp, P. P. M. (1990) Cation selectivity of and cation binding to the cGMPdependent channel in bovine rod outer segment membranes. I Gen. PhysioL 96, 5 17-534.Schoenlein, R. W., Peteanu, L. A., Mathies, R. A., and Shank, C. V. (1991) The first stepin vision: femtosecond isomerization of rhodopsin. Science 254, 412-415.Schworer, C. M., el-Maghrabi, M. R., Pilkis, S. 3., and Soderling, T. R. (1985)Phosphorylation of L-type pyruvate kinase by a calmodulin-dependent protein kinase. IBioL C’hem. 260, 13018-13022.Schworer, C. M., McClure, R. W., and Soderling, T. R. (1985) Calmodulin-dependentprotein kinases purified from rat brain and rabbit liver. Arch. Biochem. Biophys. 242, 137-145.Seckler, B. and Rando, R. R. (1989) Schiff-base deprotonation is mandatory for light-dependent rhodopsin phosphorylation. Biochem. 1 264, 489-493.Shapley, R. and Enroth-Cugell, C. (1984) Visual adaptation and retinal gain controls. InProgress in Retinal Research 3 (Eds. Osborne, N.and Chader, G.), p.263-346.137Shenolikar, S., Cohen, P. T. W., Cohen, P., Nairn, A. C., and Perry, S. V. (1979) The roleof calmodulin in the structure and regulation of phosphorylase kinase from rabbit skeletalmuscle. Eur. I Biochem. 100, 329-337.Sherry, J. M. F., Gorecka, A., Aksoy, M. 0., Dabrowska, R., and Hartshorne, D. J.(1978) Roles of calcium and phosphorylation in the regulation of the activity of gizzardmyosin. Biochemistry 17, 4411-4418.Shichi, H. (1983) in Biochemistry of Vision. Academic Press, New York.Simmerman, H. K. B., Collins, 3. H., Theibert, J. L., Wegener, A. D., and Jones, L. R.(1986) Sequence analysis of phospholamban. Identification of phosphorylation sites andtwo major structural domains. I Biol. Chem. 261, 13333-13341.Sitaramayya, A., Wright, L. S., and Siegel, F. L. (1980) Enzymatic methylation ofcalmodulin in rat brain cytosol. .1 BioL Chem. 255, 8894-8900.Smith, V. L., Doyle, K. E., Maune, J. F., Munjaal, R. P., and Beckingham, K. (1987)Structure and sequences of the Drosophila melanogaster calmodulin gene. I Mo!. Biol.196, 471-485.Smith, J. S., Rousseau, E., and Meissner, G. (1989) Calmodulin modulation of singlesarcoplasmic reticulum Ca2+rele se channels from cardiac skeletal muscle. Circ. Res. 64,352-359.Sobue, K., Muramoto, Y., Fujita, M., and Kakiuchi, S. (1981) Purification of acalmodulin-binding protein from chicken gizzard that interacts with F-actin. Proc. Nat!.Acad Sd. USA 78, 5652-5655.Steinberg, R. H., Fisher, S. K., and Anderson, D. H. (1980) Disc morphogenesis invertebrate photoreceptors. I Comp. Neurol. 190, 501-5 18.Stern, 3. H., Kaupp, U. B., and MacLeish, P. R. (1986) Control of the light-regulatedcurrent in rod photoreceptors by cyclic GMP, calcium, and l-cis-diltiazem. Proc. Nat!.Acad Sd. USA 83, 1163-1167.Stewart, A. A., Ingebritsen, T. S., and Cohen, P. (1983) The protein phosphatasesinvolved in cellular regulation. 5. Purification and properties of a Ca2/calmodulin-dependent protein phosphatase (2B) from rabbit skeletal muscle. Eur. .1 Biochem. 132,289-295.Stryer, L. (1986) Cyclic GMP cascade of vision. Ann. Rev. Neurosci. 9, 87-119.138Sugimoto, Y., Yatsunami, K., Tsujimoto, M., and Khorana, H. G. (1991) The amino acidsequence of a glutamic acid rich protein from bovine retina as deduced from the cDNAsequence. Proc. Natl. Acaci Sci. USA 88, 3116-3119.Swamy, M. S. and Abraham, E. C. (1992) Glycation of lens IvllP26 affects thepermeability in reconstituted liposomes. Biochem. Biophys. Res. Commun. 186, 632-63 8.Takio, K., Wade, R. D., Smith, S. B., Krebs, E. G., Walsh, K. A., and Titani, K. (1984)Guanosine cyclic 3’-5’-phosphate dependent protein kinase, a chimeric proteinhomologous with two separate protein families. Biochemistry 23, 4207-42 18.Tallant, E. A. and Cheung, W. Y. (1983) Calmodulin-dependent protein phosphatase: adevelopmental study. Biochemistry 22, 3630-363 5.Tanimura, A., Matsumoto, Y., and Tojyo, Y. (1991) Mastoparan increases membranepermeability in rat parotid cells independently of action on G-proteins. Biochem. Biophys.Res. Commun. 177, 8023-8028.Teo, T. S. and Wang, 3. H. (1973) Mechanism of activation of a cyclic adenosine 3’:5’-monophosphate phosphodiesterase from bovine heart by calcium ions. Identification of theprotein activator as a Ca2 binding protein. I Biol. Chem. 248, 5950-5955.Thomas, D. D. and Stryer, L. (1982) Transverse location of the retinal chromophore ofrhodopsin in rod outer segment disc membranes. I Mol. Biol. 154, 145-157.Torre, V., Matthews, H. R., and Lamb, T. D. (1986) Role of calcium in regulating thecyclic GMP cascade of phototransduction in retinal rods. Proc. Nati. AcacL Sci. USA 83,7109-7113.Tschudi, C., Young, A. S., Ruben, L., Patton, C. L., and Richards, F. F. (1985)Calmodulin genes in trypanosomes are tandemly repeated and produce multiple mRNAswith a common 5’ leader sequence. Proc. Nati. Acad. Sci. USA 82, 3 998-4002.Vanaman, T. C., Sharief F. M., and Watterson, D. M. (1977) Structural homologybetween brain modulator protein and muscle TnCs. In Calcium binding proteins andcalcium function (Eds. Siegel, F. L., Carafoli, E., Kretsinger, R. H., MacLennan, D. H.,and Wasserman, R. H.) p.107-116, Elsevier/North-Holland.Vuong, T. M. and Chabre, M. (1990) Subsecond deactivation of transducin byendogenous GTP hydrolysis. Nature 346, 71-74.Vulliet, P. R., Woodgett, 3. R., and Cohen, P. (1984) Phosphorylation of tyrosinehydroxylase by calmodulin-dependent multiprotein kinase. I Biol. Chem. 259, 13680-13683.139Watterson, D. M., Sharief F., and Vanaman, T. C. (1980) The complete amino acidsequence of theCa2+dependent modulator protein (calmodulin) of bovine brain. J Biol.Chem. 259, 13680-13683.Wawrzynczak, E. J. and Perham, R. N. (1984) Isolation and nucleotide sequence of acDNA encoding human calmodulin. Biochem. Internati. 9, 177-185.Welsh, M. J., Dedman, J. R., Brinkley, B. R., and Means, A. R. (1978) Calcium-dependent regulator protein: localization in the mitotic spindle of eucaryotic cells. Proc.Nati. Acad Sd. USA 75, 1867-1871.Wensel, T. 0. and Stryer, L. (1990) Activation mechanism of retinal rod cyclic GMPphosphodiesterase from bovine rod outer segments. Biochemisty 29, 2155-2161.Wilden, U. and KUhn, H. (1982) Light-dependent phosphorylation of rhodopsin: numberof phosphorylation sites. Biochemistry 21, 3014-3022.Wilden, U., Wust, E., Weyand, I., and Kuhn, H. (1986) Rapid affinity purification ofretinal arrestin (48 kDa protein) via its light-dependent binding to phosphorylatedrhodopsin. FEBS Lell. 207, 292-295.Williams, D. S. (1991) Actin filaments and photoreceptor membrane turnover. Bioessays13, 171-178.Wolilfart, P., Haase, W., Molday, R. S., and Cook, N. J. (1987) Antibodies againstsynthetic peptides used to determine the topology and site of glycosylation of the cGMPgated channel from bovine rod photoreceptors. .1 Blot. Chem. 267, 644-648.Wohifart, P., Muller, H., and Cook, N. J. (1989) Lectin binding and enzymaticdeglycosylation of the cGMP-gated channel from bovine rod photoreceptors. J Blot.Chem. 264, 20934-20939.Wohifart, P., Vienhues, R., and Cook, N. J. (1990) Spectrophotometric determination ofphotoreceptor cGMP-gated channel Mg2-fluxes using dichiorophosphonazo HI.Biochim. Biophys. Acta 1022, 283-290.Wolniak, S. M., Hepler, P. K., and Jackson, W. T. (1980) Detection of the membrane-calcium distribution during mitosis in Haemanthus endosperm with chlorotetracycline. .1Cell Blot. 87, 23-32.Woodrufi M. L. and Bownds, M. D. (1979) Amplitude, kinetics and reversibility of alight-induced decrease in guanosine Y:5’-cyelic monophosphate in frog photoreceptormembranes. I Gen. Physiot. 73, 629-653.140Yamamoto, K. and Shichi, H. (1983) Rhodopsin phosphorylation occurs at metarhodopsinII level. Biophys. Struct. Mech. 9, 259-267.Yamauchi, T. and Fujisawa, H. (1983) Purification and characterization of the braincalmodulin-dependent protein kinase (kinase II), which is involved in the activation oftryptophan 5-monooxygenase. Eur. .1 Biochem. 132, 15-21.Yamauchi, T., Nakata, H., and Fujisawa, H. (1981) A new activator protein that activatestryptophan 5-monooxygenase and tyrosine 3-monooxygenase in the presence of Ca2,calmodulin-dependent protein kinase. Purification and characterization. .1 Biol. Chem.256, 5404-5409.Yamazaki, A, Sen, I., Bitensky, M. W., Casnellie, J. E., and Greengard, P. (1980) CyclicGMP-specific, high affinity, noncatalytic binding sites on light-activated phophodiesterase.J Biol. Chem. 255, 11619-11624.Yau, K. -W. and Baylor, D. A. (1989) Cyclic GMP-activated conductance of retinalphotoreceptor cells. Annu. Rev. Neurosci. 12, 289-328.Yau, K. -W. and Nakatani, K. (1985) Light-suppressible, cyclic GMP-sensitiveconductance in the plasma membrane of a truncated rod outer segment. Nature 317, 252-255.Yellen, G., Jurman, M. E., Abramson, T., and MacKinnon, R. (1991) Mutations affectinginternal TEA blockage identifj the probable pore-forming region of a K channels.Science 251, 939-942.Yoshida, M., Minowa, 0., and Yagi, K. (1983) Divalent cation binding to wheat germcalmodulin. .1 Biochem. 94, 1925-1933.Yoshikami, S. and Hagins, W. A. (1970) Ionic basis of dark current and photocurrent ofretinal rods. Biophys. 1 10, 60a.Zimmerman, A. L. and Baylor, D. A. (1986) Cyclic GMP-sensitive conductance of retinalrods consists of aqueous pores. Nature 321, 70-72.

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0088090/manifest

Comment

Related Items