" a u-;.o'./» “If. uu.)l~u . . . wry»— ‘l-.) h. .1. m“ . {(-33 ’1‘ > P “1‘; rm 1‘) ., q: " t CL \ . A ~"" ‘K‘yiu‘v's‘ V ”V - . < 0‘- ‘1’“. 5“” ,fi 3’. V“,u’,ih,'1 ‘ .v v 'N I ‘V .E I: . l .1 1" e>2\.=.‘}=,;-,~ " 1 4“. ‘ x n, A x" .3- v, I. . . ‘ '1" 3,3 .9123); v~"‘~ v. a . 1,. 1'“ ~n u 3“ " ‘. . . r L I‘ .151? ‘V’??‘ .A V 1, I‘l' l‘ ' in»! '1 . 3, 2‘. .~ ‘1‘. «1.3,. ~ w _'\",':".-l .u‘ M (J'.l;v"‘t/ :. .‘u- H ‘- 4! u! i‘ (J .‘ V a . ‘ 1." .3 4.21:“: I ' ‘ ‘ - I . £7, ‘un ‘ l‘ ’ ‘ ' 'v: , . .g . .u u- a l I ‘;‘ ' .7; u a. .- 1. : l .,' , 3‘3"?) "r ‘ - 3- .. , -. "3.3 -.n . q. ("g-V . i ,1}.-. . ' .1. . . 4 3’3,“ ‘2'"? ,'.. u "r "_'|“.‘3" , ‘ {a ‘ ' . ‘4‘ _ l. STAT “HillHillllHUlHllHllHll 21le 3‘5” '0 93 00550 3457 hfiv #,- -.—. University This is to certify that the dissertation entitled STUDIES ON PHOST’HOINOSITIDE METABOLISM IN RETINAL ROD OUTER SEGMENTS presented by BARRY DEAN GEHM has been accepted towards fulfillment of the requirements for Ph. D. degree inBiochemistry Major professor BMW M8 U i: an Affirmative Action/Equal Opportunity Institution 0-12771 MSU LIBRARIES .——. \- RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. STUDIES ON PHOSPHOINOSITIDE METABOLISM IN RETINAL ROD OUTER SEGMENTS BY Barry Dean Gehm A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1988 54—54495? ABSTRACT STUDIES ON PHOSPHOINOSITIDE METABOLISM IN RETINAL ROD OUTER SEGMENTS By Barry Dean Gehm Phosphoinositides are an important class of phospholipids that help mediate signal transduction in many types of eukaryotic cells. They are believed to act as second messengers in the response of invertebrate photoreceptors to light. A similar role for them has been suggested in vertebrate phototransduction as well. The present work was undertaken to determine whether isolated mammalian photoreceptors (bovine rod outer segements, ROS) were capable of synthesizing and hydrolyzing phosphoinositides, and if so, how synthesis and hydrolysis might be regulated. Synthesis of phosphoinositides was demonstrated by the incorporation of radioactively labeled precursors. ROS incubated with [7-32P]ATP produced labeled phosphatidic acid (PA),phosphatidyl- inositol-é-phosphate (PIP) and phosphatidylinositol-4,5—bisphosphate (PIPZ). When Mn2+, CTP and inositol were added, labeled phosphatidylinositol (PI) was also produced. [3H]Inositol was incorporated into PI, PIP and PIP2, although prolonged incubation was required for detectable incorporation into PIP2. Incorporation of [3H]inositol was dependent on CTP, indicating that labeling proceeded via synthesis, not base exchange. Incubation with [a-32P]CTP produced labeled CDP-diacylglycerol, an intermediate in PI synthesis. 2+, Mn2+ and Incorporation of labeled precusors was stimulated by Mg spermine, but unaffected by light. Phosphoinositide hydrolysis was measured using exogenous 3H- labeled substrates. PI, PIP, and PIP2 were all hydrolyzed but most attention was devoted to PIP2. Hydrolysis occured via the phospholipase C (PLC) reaction, producing inositol trisphosphate and diacylglycerol. PLC was found in particulate and soluble ROS fractions, and displayed multiple forms on ion-exhange columns. PLC activity was Ca2+-dependent in the micromolar range. Crude enzyme preparations contained an endogenous inhibitor whose effects were Ca2+-dependently relieved by calmodulin antagonists. This inhibitor 2+ does nor appear to be calmodulin but may be a novel Ca -binding regulatory protein. No evidence for regulation of PLC activity by light or G-proteins was obtained. The effects on PLC activity of Mg2+, Mn2+, spermine, pH and detergents were also characterized. Copyright by BARRY DEAN GEHM 1988 To my parents, sine qua non. ACKNOWLEDGEMENTS It would require a volume as large again as this to properly thank everyone who has helped me along the tortuous road whose latest mile- stone is this dissertation; nevertheless I would be remiss if I did not acknowledge my special indebtedness to: the members of my committee, Drs. William Wells, Clarence Suelter, Paul Kindel and James Asher, and also Dr. John Wilson, for their support and advice; Dr. James Chafouleas and Florence D6 of the Université Laval, for assistance with calmodulin RIAs and immunoprecipitation; my helpers, Ken van Golen, Linda Doorenbos, and David Pepperl, for technical assistance and good cheer; my co-workers, Jeff Leipprandt, Donna Schultz, and Frank Wilkinson, for helpful discussions and caring fellowship; special friends who sustained me during rough times, especially Russell Kohnken, Ming Tien, Bill Higgins, Mike Toman, and Nikki Ballard; Dr. David McConnell, for support, guidance, and friendship, and for his extraordinary faith in and patience with me; and last, but also first and always, my parents, Noel and Leora Gehm, without whose unfailing love, care and support I could not have begun, much less completed, these labors. Vi TABLE OF CONTENTS LIST OF TABLES X LIST OF FIGURES xii ABBREVIATIONS xiv INTRODUCTION 1 Internal messengers 3 Phosphoinositides 5 Phosphoinositides and phototransduction 9 Choice of experimental system 10 Research goals 11 BIBLIOGRAPHY 12 CHAPTER I: SYNTHETIC PATHWAYS 14 LITERATURE REVIEW 15 MATERIALS AND METHODS 19 Materials 19 ROS isolation l9 Phospholipid labeling 19 Analysis of labeled lipids 20 Miscellaneous methods 21 vii RESULTS Synthesis of phosphoinositides in ROS Requirement for CTP Divalent metal ions and spermine Addition of unlabeled PIP Effects of washing ROS membranes Light and nucleotides Comparison of ROS and microsomes DISCUSSION BIBLIOGRAPHY CHAPTER II: ROS PHOSPHOLIPASE C LITERATURE REVIEW MATERIALS AND METHODS Materials Preparation of particulate and soluble ROS fractions Phospholipase C assays Miscellaneous methods RESULTS A. Demonstration of Phospholipase C in ROS. Hydrolysis of phosphoinositides labeled in situ Hydrolysis of exogenous [3H]PIP2 B. Characterization of ROS PLC Activities. Soluble and particulate forms of ROS phospholipase C Validation of assay Comparison of substrates Effects of divalent metal ions viii 22 22 32 32 39 39 39 48 52 54 56 57 61 61 61 61 62 63 63 63 65 65 65 69 69 69 Effect of spermine Effect of monovalent Effect of pH metal ions Chromatographic characterization C. Possible regulators of ROS PLC. 1. Nucleotides and GTP Binding Proteins 2. Light 3. Calmodulin Phenothiazines Haloperidol Calmidazolium Melittin Compound 48/80 Naphthalenesulfonamides CaM antagonists CaM antagonists CaM antagonists CaM antagonists Calcineurin CaM antagonists Addition of CaM DISCUSSION BIBLIOGRAPHY SUMMARY and Ca2+ and pH and BSA and detergent and PLC purification to PLC assays ix 73 73 73 83 83 83 96 101 101 104 104 104 108 113 120 126 126 131 131 131 133 135 144 150 CHAPTER Table Table Table Table Table Table Table H CHAPTER Table Table Table Table Table Table Table Table Table Table Table 9. . Effect of ionic strength and Ca LIST OF TABLES 32P labeling of ROS phospholipids. . Formation of labeled CDP-DG from [a-32P]CTP. . Effect of Ca2+ on phospholipid labeling. . Effect of washing on phospholipid labeling. . Absence of light effect on 32F labeling. . Guanine nucleotide effects on phospholipid labeling. . Comparison of phospholipid labeling in ROS and microsomes. II . No decrease in labeled phospholipids upon illumination. . Removal of phospholipase C from ROS by isotonic washing. . Comparison of substrates for ROS phospholipase C. 2+ on PLC activity. . Pertussis toxin does not affect PIP2 PLC activity. . Effect of nucleotides on particulate PIP2 PLC activity. . ROS PLC activity unaffected by light. . PLC activity in fractions from bleached and unbleached retinas. Effect of calmidazolium on PIP2 PLC activity. 10. Effect of W-7 on phosphoinositide hydrolysis. 11. Effect of pH on trifluoperazine stimulation of PLC. X 25 27 44 47 49 50 51 64 68 72 74 93 97 98 99 107 121 127 Table 12. Effect of added proteins on soluble PIP2 PLC activity. 128 Table 13. PLC activation by detergent and CaM antagonist not additive. 132 Table 14. Added CaM does not affect PLC activity. 134 Xi INTRODUCTION TABLE OF FIGURES Figure 1. A rod cell. Figure 2. CHAPTER I Figure 1. Figure Figure Figure Figure Figure Figure Figure Figure CHAPTER II 2. 3. b 0“ \l (13 k0 Phosphoinositide metabolic pathways. Autoradiogram of 32P-labeled ROS lipids. Time course of 32P-labeling of ROS phospholipids. Time course of [3H]inositol incorporation into PI. . Effect . Effect . Effect Effect . Effect Effect of CTP on [3H]inositol incorporation into PI. of Mg2+ on ROS phospholipid labeling. of spermine on ROS phospholipid labeling. 2+ of Mn on [3H]inositol incorporation. 2+ of Mn on 32F labeling of ROS lipids. of added PIP on 32P incorporation. Figure 1. Identification of labeled products of PIP2 hydrolysis. Figure 2. Effect of octylglucoside on quantity-activity relation- Figure 3. Effect of Ca ship of soluble PLC. 2+ on PIP2 PLC activity. Figure 4. Effect of Mg2+ on PIP2 PLC activity. xii 24 29 31 34 36 38 41 43 46 67 76 78 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure U1 \1 00 \O 10. . Effect of Mn2+ on PIP2 PLC activity. . Effect of spermine on PIP2 PLC activity. . Gel filtration of soluble PLC. Ion—exchange chromatography of soluble PLC. Effect of GMPPCP on PIP2 PLC activity. ADP-ribosylation of ROS fractions by pertussis toxin. Effect Effect Effect Effect Effect of GTPyS on particulate PIP2 PLC activity. of phenothiazines on soluble PIP2 PLC activity. of haloperidol on PIPZ PLC activity. of melittin on PIPZ PLC activity. of compound 48/80 on PIP2 PLC activity. Effects of naphthalenesulfonamides on soluble PIP2 PLC activity. Effects of naphthalenesulfonamides on particulate PIP2 PLC activity. Effect Effect of W-7 on activity-volume linearity. of W-7 on kinetics of soluble phospholipase C. Calcium-dependence of W-7 stimulation of PLC. Effect of W-7 and W-5 on inhibition by BSA. xiii 8O 82 85 87 89 92 95 103 106 110 112 115 117 119 123 125 130 ABBREVIATIONS ATP, ADP, AMP: adenosine triphosphate, diphosphate and monophosphate BAPTA: 1,2-bis(2-aminophenoxy)-ethane-N,N,N’,N’-tetraacetic acid BSA: bovine serum albumin CaM: calmodulin CDP-DG: cytidine diphosphate-diacylglycerol CTP: cytidine triphosphate ECSO: 50% effective concentration EGTA: ethyleneglycol-bis(fi-aminoethyl ether)—N,N,N’,N’-tetraacetic acid GDPflS: guanosine-S'-O-(2-thiodiphosphate) GMPPCP: guanylyl-(fi,7—methy1ene)-diphosphonate GMPPNP: guanylylimidodiphosphate GTP, GDP, GMP: guanosine triphosphate, diphosphate and monophosphate GTPyS: guanosine-S'-O-(3-thiotriphosphate) HEPES: N-2-hydroxyethylpiperazine—N’~2-ethanesu1fonic acid IBMX: 3-isobutyl-l-methylxanthine 1050: 50% inhibitory concentration IP: inositol-l-phosphate IP2: inositol-l,4-bisphosphate 1P3: inositol-1,4,5-trisphosphate Mr: relative molecular weight PI: phosphatidylinositol PIP: phosphatidylinositol-4-phosphate xiv PIPZ: phosphatidylinositol-4,5-bisphosphate PLC: phospholipase C ROS: retinal rod outer segments SDS: sodium dodecyl sulfate Tris: tris(hydroxyethyl)aminomethane W-S: N-(6-aminohexyl)-1-naphthalenesulfonamide W-7: N-(6-aminohexy1)-5-chloro-l-naphthalenesulfonamide W-12: N-(4-aminobutyl)-1-naphthalenesulfonamide W-13: N-(4-aminobutyl)-5-chloro-l-naphthalenesulfonamide. WROS: isotonically washed ROS INTRODUCTION Vision is mediated by specialized photoreceptor cells in the retina. Two morphologically distinct types of photoreceptors, rods and cones, are found in vertebrate retinas. Cones operate in bright light and comprise several subtypes with differing spectral sensitivities; they are responsible for color vision. Rods are more sensitive to light than cones and are responsible for dim-light vision. Because rods are larger and more numerous than cones in most retinas, and do not introduce the complicating factor of multiple spectral types, most work on the biochemistry of vision has been done with rods. As can be seen in Figure l, the rod cell is divided by a narrow constriction into an inner and outer segment. The inner segment contains the normal appurtenances of animal cells, such as mitochondria, endoplasmic reticulum, and the nucleus. A synaptic terminus connects the proximal end of the inner segment to other retinal neurons. (The morphology of photoreceptors varies considerably from species to species, and some workers do not consider the nucleus and synaptic terminus as parts of the inner segment.) The outer segment is a specialized structure in which the energy of light is absorbed, transduced and amplified to produce an electrical signal. Hundreds of flattened membranous vesicles (disks) form an orderly stack, surrounded by but not continuous with the cell’s plasma membrane. The disks contain the visual pigment, rhodopsin. Rhodopsin DISKS < OUTER " SEGMENT _ INNER SEGMENT \ifiifiiifié Figure l. A rod cell. This schematic drawing of a typical rod cell is oriented so that upwards represents the distal direction, i. e., towards the back of the eye. Light enters from below. consists of a protein, opsin, and a covalently attached lipid chromophore, ll-cis-retinal. Absorption of a photon causes retinal to isomerize to its all-trans form, causing a loss of color ("bleaching") and producing a series of conformational changes in the rhodopsin molecule. This ultimately results in electrical hyperpolarization of the rod cell. In the dark, a constant current of positive ions (primarily Na+, but some Ca2+ as well) exits the cell from the inner segment and re-enters through "sodium channels“ in the plasma membrane of the outer segment. Bleaching of rhodopsin indirectly causes the sodium channels to close, preventing influx of Na+ into the outer segment. The ATP-driven efflux of Na+ from the inner segment, without a balancing influx, drives the membrane potential from its resting value of -30 mV to as much as —70 mV. This hyperpolarization triggers firing of the synapse connecting the rod cell to retinal neurons. Internal messengers: How does the bleaching of rhodopsin, which is located in the disk membranes, bring about the closing of the sodium channels in the plasma membrane? One or more water-soluble messengers must carry information across the cytoplasm. The identity of the messenger molecule(s) has been one of the principal problems in Visual biochemistry for more than a decade. The two major candidates that have been proposed for the role of internal messenger are cyclic GMP and Ca2+ (Pugh and Cobbs, 1986; Lamb, 1986). cGMP: It has been known for more than a decade that rod outer segments (ROS) of frogs and cattle contain a light-activated cylic GMP phosphodiesterase (PDE) (Miki et al., 1973; Manthorpe and McConnell, 1975). Activation is mediated by a CTP—binding protein that Fung er a1. (1981) named "transducin" but which earlier vision investigators (Godchaux & Zimmerman, 1979) called ROS G-protein. More recently, it has been demonstrated that cGMP can increase the Na+ conductance of isolated patches of ROS plasma membrane, apparently by opening the sodium channels (Fesenko et al., 1985). cGMP is now the leading candidate for internal messenger in vertebrate rods. In the "cGMP hypothesis", cGMP holds the sodium channels open in the dark. Photobleaching of rhodopsin activates PDE via ROS G-protein. Hydrolysis of cGMP by the activated PDE then causes the sodium channels to close. Ca2+: Large amounts of calcium are stored inside the disks. The 2+ is released "calcium hypothesis“ proposed that, on illumination, Ca from the disks into the cytoplasm and blocks the sodium channels (Hagins, 1972). This idea gained support from numerous experiments showing Ca2+—regulation of photoreceptor electrical behavior (summarized in Pugh and Cobbs, 1986). Alterations in external Ca2+ concentration, which are believed to bring about concomitant changes in internal concentration via Na+/Ca2+ exchange and/or entry of Ca2+ through the "sodium" channels, alter membrane conductance. Specifically, high [Ca2+] decreases dark current and increases membrane potential, and low [Ca2+] has opposite effects. These effects are enhanced by calcium ionophores. Microinjections of Ca2+ or EGTA into toad rods cause hyperpolarization and depolarization respectively (Brown et al., 1977). However, introduction of chelators into the ROS cytoplasm does not prevent the photoresponse (Matthews et al.,l985), and recent studies of Ca2+ fluxes in rods (reviewed in Pugh and Cobbs, 1986, and Lamb, 5 1986) indicate that cytoplasmic Ca2+ concentration may not increase in response to light as previously thought -- in fact, it may decrease. Additionally, experiments with isolated patches of outer segment membranes show no regulation of conductance by physiological levels of Ca2+, in contrast to cGMP (Fesenko et al., 1985). Hence, the calcium hypothesis in its original form has been largely abandoned as a model for vertebrate phototransduction. Nevertheless, Ca2+ is still believed to play a role in the function of the rod cell, possibly in adjusting sensitivity at different light levels or terminating the photoresponse. Treatment of frog photoreceptors with EGTA and the calcium ionophore A23187 decreases sensitivity to low light levels (Nicol et al., 1987). 2+ chelator BAPTA into salamander rods extends the Infusion of the Ca duration of the photoresponse (Matthews et a1., 1985) and slows light-adaptation (Torre et 31., 1986). These findings suggest that Ca2+ may act as a modulator, rather than a transmitter, of the photoresponse. Phosphoinositides: The phosphoinositides are a family of phospholipids found in membranes of eukaryotic cells. They contain an inositol moiety in their polar head groups, and comprise phosphatidylinositol (PI), phosphatidylinositol-4-phosphate (PIP) and phosphatidylinositol-4,5-bisphosphate (PIPZ). Evidence for the involvement of phosphoinositides in signal transduction was first uncovered twenty-five years ago by Hokin and Hokin (1953) in blowfly salivary gland. As the presence of this pathway has been revealed in a host of different cell types in recent years, interest in it has increased almost explosively, and has focused on the hydrolysis of PIP2 by phospholipase C (PLC) (for reviews, see Berridge, 1984, Berridge and Irvine, 1984; Nishizuka, 1984, Joseph, 1984; Hokin, 1985; Majerus et al., 1985, Majerus et al., 1886, Berridge, 1987). This hydrolysis produces diacylglycerol, an activator of protein kinase C, and inositol-1,4,5-trisphosphate (IP3), which mobilizes Ca2+ from intracellular stores. A portion of the inositol trisphosphate may be produced in the form of a 1,2-cyclic isomer (Wilson et al., 1985); it appears to be similar to the non—cyclic isomer in its effects but is metabolized more slowly (Majerus et al., 1986). Figure 2 presents a simplified schematic of phosphoinositide metabolism. PI is synthesized in a two-step, CTP-requiring process from phosphatidic acid (PA) and inositol. Some of it is phosphorylated to produce PIP and PIPZ. PI, PIP and PIP2 are substrates for one or more PLCs, producing diacylglycerol and the respective inositol phosphates. A series of phosphatases degrade the inositol phosphates back to inositol, which can then re-enter the synthetic pathway. (1P phosphatase is inhibited by Li+.) For the sake of simplicity, some reactions have been omitted from Figure 2. The 1,2-cylic isomers of the inositol phosphates are not shown, as they are beyond the scope of the present work. Similarly, the phosphorylation of IP3 to IP4, 1P5 and 1P6, which has been reported in some types of cells (Batty et al., 1985; Vallejo, M. et al., 1987) has not been demonstrated in photoreceptors and is omitted from Figure 2. Phosphatases which dephosphorylate PIP and PIP2 are not shown; they have not been studied in photoreceptors previously and are touched on only briefly in this dissertation. Although a Figure 2. Phosphoinositide metabolic pathways. This diagram presents a simplified schematic of phosphoinositide synthesis and breakdown. Abbreviations: DG, diacylglycerol; PA; phosphatidic acid; CDP-DG, GDP-diacylglycerol (also called phosphatidyl-CMP); PI, phosphatidlyinositol; PIP, phosphatidylinositol-4-phosphate; PIP2, phosphatidylinositol-4,5-bisphosphate; IP3, inositol-1,4,5-trisphosphate; IP2, inositol-l,4-bisphosphate; IP, inositol—l-phosphate; Ins, inositol. The hydrolysis of PI, PIP, and PIP2 may be carried out by the same or different PLCs. PIP PI CDP-DG PA DG J phospholipase C OOCR ‘ COCR' A I phospholipase C 0 0H 0H OOCR CMP OOCR' PI synthase \ ~k .6 CMP~®E OOCR , OOCR’ Pplji{ CDPzdiacylglycerol CTP [:00CR ‘ OOCR' DG kinase ADP ATP LA.) I O OCR OCR' cytidylyltransferase L/ IP3 0H (Ni P'i (iIP3 phosphatase 0H 9 0 0H (Mi P' 1P2 phosphatase 1 <___L_ OH IP 0H (Ni _‘ 0.1 mM and is not shown on the graph. [3H]PI, kcpm 41 ICDQJ h/Ir]ZZ-+- C]Cj(j€3(j: “A Figure 7 42 Figure 8. Effect of Mn2i on 32F labeling of ROS lipids. ROS (2.6 nmol rhodopsin/assay) were incubated in minimal 32P labeling medium as described in Materials and Methods, 2+ 321) except that the Mn concentration was varied as shown. incorporation into PA (0), PIP (A) and PIP2 (u) is shown as percent of control (no added Mn2+). 43 32P incorporated, % of control O H I 1 I I I I —oo —4 +3 —2 log Mn2+ added, M Figure 8 44 Table 3. Effect of Ca2+ on phospholipid labeling. 32F incorporated, cpm [Ca2+],mM PA PI PIP PIP2 0 39721 1060 1702 627 :4094 :89 :50 :169 1 41396 1018 937 274 :4500 :272 :137 :10 2 43119 880 905 226 :3011 :18 :71 :108 ROS (1.3 nmol rhodopsin/assay) were incubated in complete 32P labeling medium as described in Materials and Methods, with the addition of the indicated concentration of CaClZ. Labeled lipids were separated by TLC and quantitated by liquid scintillation counting. Results are shown as means i standard errors for duplicate assays. 45 Figure 9. Effect of added PIP on 32F incorporation. ROS (1 nmol rhodpsin/assay) were incubated in minimal 32P labeling mixture as described in Materials and Methods, with the addition of the indicated concentration of unlabeled PIP (sodium salt, purchased from Sigma and added as a lmM aqueous suspension). Results are expressed as % of no-added-PIP control. 32P incorporated, % of control 46 200 150 50- — - H P|P2 . “ H PIP ' ' H PA 7 O ' I ' I ' I ' I 0.0 0.1 0.2 0.3 0.4 [PIP] added, mM Figure 9 47 Table 4. Effect of washing on phospholipid labeling. Product Relative Labeling (ROS = 100) ROS 1501 1502 Hypol Hyp02 [32P]PA 100 33.6 46.5 35.9 36.7 $14.3 $3.8 $2.6 $0.3 $0.3 [3H]PI 100 78.0 68.7 69.9 65.8 $3.5 $4.6 $4.1 $1.1 $5.5 [32P]PIP 100 99.0 110.3 112.5 126.2 $10.7 $9.7 $8.0 $2.0 $2.8 [32P]PIP2 100 88.9 91.8 88.9 74.6 $8.4 $15.9 $4.6 $1.7 $16.2 ROS were washed twice isotonically and twice hypotonically, as described in Methods. Samples of the pellets were retained at each step. Aliquots of the original ROS, the lst isotonic pellet (1501), the 2nd isotonic pellet (1502), the lst hypotonic pellet (Hypol) and the 2nd hypotonic pellet (Hyp02), each containing 2 nmol of rhodopsin, were incubated in minimal 32F labeling medium (to label PA, PIP, and PIPZ) and in 3H labeling medium (to label PI). Labeling of each lipid in the washed membranes is expressed as a percentage of the labeling seen in ROS. Values shown are means ($ s.e.) of duplicate assays. 48 Various calcium concentrations were tried. No effects of light were found on label incorporation from either [32P]ATP or [3H]inositol. Table 5 shows the results of a representative experiment. No effects of cyclic nucleotides on 32F or 3H labeling were observed, nor did the cyclic nucleotide phosphodiesterase inhibitor IBMX, reported to decrease phosphoinositide labeling in octopus photoreceptors (Yoshioka et al., 1983), have any effect. GTP (1mM) and its non-hydrolyzable analog, guanylylimidodiphos- phate, produced a modest decrease in labeling of PA, PIP and PIP2 in minimal 32F labeling medium (Table 6). This is probably attributable to competition with ATP for the active sites of the kinases. Comparison pf ROS g_g microsomes: Although our ROS preparations are estimated to contain <1% microsomal contamination as determined by glucose 6-phosphatase (McConnell, 1965), NADPH-cytochrome C reductase and electron microscopy (McConnell et al., 1969), the possible contribution of microsomes to phospholipid labeling was examined. Retinal microsomes were assayed for incorporation of labeled precursors into phospholipids. As shown in Table 7, the patterns of labeling in ROS and microsomes are quite different. PI synthesis is about lO-fold higher (on a pmole/min/mg protein basis) in microsomes than in outer segments, but microsomes cannot account for all the PI synthesis exhibited by ROS, as this would require ca. 10% microsomal contamination in our ROS preparations. 49 Table 5. Absence of light effect on 32F labeling. 3 2? 1110015130171de ‘ cpm PA PI PIP 2122 Dark 20677 276 1361 320 $956 $5 $189 $63 Light 19726 292 1340 311 $175 $44 $22 $12 ROS (1.2 nmol rhodopsin/assay) phospholipids were labeled with 32F as described in Materials and Methods, except that the incubation was done under dim red light. ROS were either maintained in darkness or exposed to normal laboratory illumination before labeling. Results are expressed as means $ standard errors for duplicate assays. 50 Table 6. Guanine nucleotide effects on phospholipid labeling. 32P incorporated, cpm Nucleotide PA PIP PIP2 None 46385 2930 144 $883 $124 $7 GTP 32483 2272 105 $1506 $117 $4 GMPPNP 40208 2371 116 $2180 $169 $18 ROS (2.1 nmol rhodopsin) were incubated in minimal 32P labeling medium, as described in Materials and Methods, with the addition of the indicated nucleotides (1mM). Results are expressed as means $ standard errors for duplicate determinations. Table 7. Comparison of phospholipid labeling in ROS and microsomes. 51 Rate of synthesis, Product pmol/min/mg protein _ROS_ Microsomes [32P]PA 20 2.6 [32P]PIP 0.95 3.0 [32P]PIP2 0.22 0.44 [3H]PI 0.24 2.5 ROS and retinal microsomes were incubated for 30 min with [32P]ATP or 45 min with [3H]inositol. Labeled products were isolated and quantitated as described in Materials and Methods. duplicate determinations. Ratio ROS/Microsomes 0.32 Values shown are averages of 52 DISCUSSION These experiments establish that ROS contain a complete pathway for synthesis of phosphoinositides. Phosphoinositdes and PA incorporate radioactive phosphate from ATP much more rapidly than do the "major" phospholipids. Turnover of these lipids is metabolically expensive: synthesis of PIP2 from inositol and diacylglycerol requires hydrolysis of four high-energy phosphate bonds. Based on the experiments presented here, it is logical to presume that phosphoinositides have some functional role in ROS. The disproportionately high level of PA synthesis is somewhat puzzling. It may be that high levels of diacylglycerol accumulate in ROS membranes after slaughter, due to lack of perfusion (Matthys et al., 1984). If so, it could be expected that PA would be rapidly synthesized when ATP became available. 32P-labeling experiments were less than ideal for measuring PI synthesis, due to the proximity, on thin—layer chromatograms, of the P1 to the tail of the intensely labelled PA spot. Drying the plates in vacuo between the first- and second-dimension developments decreased but did not eliminate tailing. Use of [3H]inositol provided a much better way to measure PI synthesis. However, [3H]inositol can also be incorporated into PI by a non-synthetic base-exchange reaction. Base exchange can be distinguished from de novo synthesis by the latter's requirement for CTP (Eisenberg and Hasegawa, 1981). Omission of CTP from 3H-labeling experiments greatly reduced labeling of PI, which indicates that the bulk of [3H]inositol incorporation represents PI synthesis. 53 PI synthesis was strongly dependent on Mn2+. Routine phospholipid labeling reactions were carried out in the presence of 1 mM MnC12. Later work (Chapter 11) showed that this concentration completely inhibits ROS PIP2 phospholipase C. This simplifies interpretation of some labeling experiments by lessening the possibility that decreased labeling attributed to diminished synthesis is actually due to increased hydrolysis of labeled phosphoinositides. However, it raises the possibility that light-dependent decreases in labeling were not observed because of inhibition of the putative light-activated PLC. Since their substrates are lipids and presumably confined to membranes, it might well be expected that the enzymes involved in phosphoinositide synthesis would also be tightly membrane-bound. Generally, this was found to be so. Washing of ROS produced a modest reduction in P1 labeling and a more substantial reduction in PA labeling. These decreases in label incorporation may be due to removal of the relevant enzymes (diglyceride kinase for PA, phosphatidate cytidylyltransferase and/or PI synthase for PI) or loss of polar lipid precursors. After the phospholipid labeling experiments presented here had been completed, Giusto and Ilincheta de Boschero (1986) reported labeling of PA, PIP and PIP2 in bovine ROS and isolated disks incubated with [7-32P]ATP. Label incorporation into PA was ca. 2 orders of magnitude greater than into PIP. No experiments examining the effects of metal ions, spermine, or light were reported. 54 BIBLIOGRAPHY Anderson, R. E., and Hollyfield, J. G. (1981) Biochim. Biophys. Acta 665, 619-622. Anderson, R. E., and Hollyfield, J. G. (1984) J. Cell Biol. 99, 686-691 Anderson, R. E., and Kelleher, P. A. (1981) Exp. Eye Res. 32, 729-736 Anderson, R. E., Maude, M. B., and Kelleher, P. A. (1980) Biochim. Biophys. Acta 620, 236-246 Bazan, H. E. P., and Bazan, N. G. (1975) Life Sciences 17, 1671—1678 Bazan, H. E. P., and Bazan, N. G. (1976) J. Neurochem. 27, 1051-1057 Bazan, H. E. P., and Bazan, N. G. (1977) Adv. Exp. Med. Biol. 83, 489-495 Bazan, H. E. P., Careaga, M. M., and Bazan, N. G. (1981) Biochim. Biophys. Acta 666, 63-71 Bazan, N. G., Ilincheta de Boschero, M. G., and Giusto, N. M. (1977) Adv. Exp. Med. Biol. 83, 377-388 Bonner W. M. and Stedman J. D. (1978) Anal. Biochem. 89, 247-256. Eichberg, J., Zetusky, W. J., Bell, M. E., and Cavanagh, E. (1981) J. Neurochem. 36, 1868-1871 Eisenberg, F., Jr., and Hasegawa, R. (1981) Trends Biochem. Sci. 6, 9-10 Fisher, S. K., and Agranoff, B. W. (1987) J. Neurochem. 48, 999-1017 Fisher, S. K., and Agranoff, B. W. (1986) in Receptor Biochemistry and Methodology: Receptors and Phosphoinositides (Putney, J. W., Jr., ed.) pp. 219-243, Alan R. Liss, New York Gehm, B. D., and McConnell, D. G. (1985) Fed. Proc. 44, 1227 Giusto, N. M., and Bazan, N. G. (1979) Exp. Eye Res. 29, 155-168 Giusto, N. M., and Ilincheta de Boschero, M. G. (1986) Biochim. Biophys. Acta 877, 440-446 Gumber, S. C. and Lowenstein, J. M. (1986) Biochem. J. 235, 617-619 Hayashi, F. and Amakawa, T. (1985) Biochem. Biophys. Res. Comm. 128, 954-959 Kohnken, R. E., Eadie, D. M., Revzin, A., and McConnell, D. G. (1981a) J. Biol. Chem. 256, 12502—12509 55 Kohnken, R. E., Chafouleas, J. G., Eadie, D. M., Means, A. R. and McConnell, D. G. (1981b) J. Biol. Chem. 256, 12517-12522 Laskey R. A. and Mills A. D. (1977) FEBS Letters 82, 314-316. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275 Matthys, E., Patel, Y., Kreisberg, J., Stewart, J. H., and Venkatachalam, M. (1984) Kidney Int. 26, 153-161 McConnell, D. G. (1965) J. Cell Biol. 27, 459-473 McConnell, D. G., Dangler, C. A., Eadie, D. M., and Litman, B. J. (1981) J. Biol. Chem. 256, 4913-4918 McConnell, D. G., Ozga, G. W., and Solze, D. A. (1969) Biochim. Biophys. Acta 184, 11-28 Mercurio, A. M., and Holtzman, E. (1982) J. Neurocyt. 11, 295-322 Nielsen, N. G., Fleischer, 8., and McConnell, D. G. (1970) Biochim. Biophys. Acta 211, 10-19 Schacht, J. (1981) Meth. Enzymol. 72, 626-631 Schmidt, S. Y. (1983a) J. Cell Biol. 97, 832-837 Schmidt, S. Y. (1983b) J. Biol. Chem. 258, 6863-6868 Seyfred, M. A., Kohnken, R. E., Collins C. A., and McConnell, D. G. (1984) Biochem. Biophys. Res. Comm. 123, 121-127 Urban, P. F., Dreyfus, H., Neskovic, N., and Mandel, P. (1973) J. Neurochem. 20, 325-335 Vandenberg, C. A., and Montal, M. (1984) Biochemistry 23, 2347-2352 Yandrasitz, J. R., Berry, G., and Segal, S. (1985) in Inositol and Phosphoinositides: Metabolism and Regulation (Bleasdale, J. E., Eichberg, J. and Hauser, G., eds) pp. 601-620, Humana Press, Clifton, N. J. Yoshioka, T., Inoue, H., Takagi, M., Hayashi, F., and Amakawa, T. (1983) Biochim. Biophys. Acta 755, 50-55 CHAPTER II ROS PHOSPHOLIPASE C 56 57 LITERATURE REVIEW Published research on phospholipase C in photoreceptors has focused almost exclusively on its regulation by light and possible involvement in phototransduction. As noted above, much of the evidence for this involvement has been developed in invertebrates. Light-induced decreases in labeled PIPZ in squid (Vandenberg and Mental, 1984) and octopus (Yoshioka et al., 1983a) photoreceptors were discussed in the review of labeled precursor incorporation into retinal phospholipids (Chapter 1), but are also consistent with increased PLC activity. Most reports of light-stimulated phosphoinositide hydrolysis employ substrates that are labeled in situ by incorporation of radioactive precursors. Brown et a1. (1984) reported that light increased [3H]IP3 and decreased 3H- and 32P-labeled PIP2 in Limulus (horseshoe crab) ventral eyes that had been preincubated with labeled precursors. Light flashes produced transitory increases in [3H]IP3 in squid retinas prelabeled with [3H]inositol (Szuts et al., 1986). In both cases essentially intact retinas were used; cells distal to photoreceptors (i. e. neurons) may account for some or all of the phosphoinositide turnover. Devary et a1. (1987) labeled fly eyes with [3H]inositol, homogenized them, and prepared a crude membrane fraction which, even after frozen storage, displayed light-stimulated release of tritiated IP3 and IP2. This light-stimulation was enhanced by GTPyS and inhibited by GDPBS, indicating regulation by a G-protein. The use of a homogenized preparation instead of intact retinas would seem to reduce the possibility that light-stimulated PIP2 hydrolysis is due simply 58 to neurotransmitter-stimulated neurons, but considerable non-photoreceptor material must be present in this preparation. Yoshioka et al. (1983b), working with blind Drosophila mutants (norpA) whose photoreceptors lack normal electrophysiological response to light, reported decreased incorporation of [32F] into PA and increased labeling of PIP and PIP2 compared to wild types. This was originally interpreted as an abnormality of phospholipid synthesis, but can also be interpreted as evidence of decreased PLC activity: less phosphoinositide hydrolysis resulting in decreased availability of DG to be phosphorylated to PA. The latter interpretation was confirmed by later work from the same laboratory (Inoue et al., 1985) showing that homogenates of the eyes of wild-type Drosophila, but not norpA mutants, diplayed PIP2 PLC activity against exogenous [32P]PIP2. No attempt to measure the effect of light on PLC activity was reported. Recently, Baer and Saibil (1988) reported that light stimulated hydrolysis of exogenous [3H]PIP2 by isolated squid photoreceptor outer segments in the presence of GTP and Ca2+. Hydrolysis was measured by formation of water-soluble labeled products, primarily 1P3. The experimental system used by these experimenters (exogenous substrate and isolated outer segments) is quite similar to that used in the present work, although there are some methodological differences as well as the difference in organisms. Another type of experiment that supports a role for PLC in phototransduction involves injection of various substances into photoreceptors, whose electrical responses are measured with microelectrodes. The comparatively large size of Limulus photoreceptors makes them popular for such experiments. Injections of 59 1P3 mimicked the effect of light (Fein et al., 1984; Brown et al., 1984). 1,2-cyclic IP3 was also effective (Wilson et al.,l985). Although response to light was blocked by injection of GDPBS, response to IP3 was not (Fein, 1986). IP3 injection increased intracellular Ca2+ concentration (Brown and Rubin, 1984) apparently via release from stores within the photoreceptor (Payne and Fein, 1987). Injections of calcium salts also produced a light—like response (Payne et al., 1986), as did injections of phospholipase C (Rubin et al., 1986). These results all strongly support a model of G-protein mediated light-activation of PIP2 PLC producing IP3 which 2+ 2+ releases Ca , which then (in the invertebrate version of the Ca hypothesis) directly or indirectly increases Na+ conductance and depolarizes the cell. The picture in vertebrate photoreceptors is less clear-cut, but a few reports have appeared suggesting a role for PIPZ hydrolysis, at least in amphibians. Ghalayini and Anderson (1984), using prelabeled frog retinas, found that ROS isolated from retinas exposed to a flash of light contained less [3H]PIP2 than those isolated from retinas kept in darkness. Hayashi and Amakawa (1985) obtained similar results using frog ROS labeled and illuminated after isolation. Using prelabeled toad retinas flash-frozen at timed intervals after light stimulation, Brown et al. (1987) found that [3H]IP3 in ROS rose 50-80% within 250 msec but returned to the unstimulated level by 1 sec. Levels of the breakdown products of 1P3 were not reported. In experiments analogous to those performed in Limulus, Waloga and Anderson (1985) found that microinjection of 1P3 into outer 60 segments of salamander rods caused hyperpolarization and decreased the rods' response to dim light. Conversely, light decreased the response to 1P3 injection. No analysis of the response kinetics was presented, but inspection of receptor potential recordings suggest that the response elicited by IP3 is considerably slower than the photoresponse. The experiments that follow were undertaken to determine the presence of PIP2 PLC in mammalian rod outer segments and to characterize its properties and regulation. The possibility of regulation by light or G protein was of significant but not exclusive interest. 61 MATERIALS AND METHODS Materials: Tritiated phospholipids were purchased from New England Nuclear. Calmodulin antagonists, nucleotides and unlabeled phosphoinositides were purchased from Sigma. Non-hydrolyzable GTP analogs were purchased from Boehringer Mannheim. BAPTA (1,2-bis(2- aminophenoxy)-ethane-N,N,N’,N’-tetraacetic acid) was purchased from Fluka. Preparation pf particulate gpg soluble RLO fractions: ROS were isolated as described in Chapter 1. Isolated unbleached ROS were washed 4 times at 100,000 x g in 10 - 20 volumes of 10 mM Tris-HCl, pH 7.8, 150 mM KCl, 1 mM dithiothreitol, to yield the particulate fraction (washed ROS = WROS). The soluble fraction (isotonic supernate) consisted of the supernate of the first washing; alternatively, supernate from washing at 15,000 x g was clarified by filtration through Durapore type HA 0.45 pm filter membranes (Millipore). No differences were observed between soluble fractions prepared by these two methods. Phospholipase O assays: [3H]PIP2 (2.5 pCi/pmole) was prepared from [inositol-2,3-3H]PIP2 and PIP2 sodium salt, and used as a 1 mM aqueous solution. [3H]PIP (5.0 uCi/pmol) and [3H]PI (9.1 pCi/pmol) were prepared analogously, but required sonication to produce 1 mM dispersions. Unless otherwise indicated, assays were performed at 30° C. under normal room illumination in 200 mM potassium HEPES, pH 7.5, 50 mM KCl, 30 pM CaC12 and 10 pM tritiated phosphoinositide plus indicated additions. Total assay volume was 500 pl. After 30 min, reactions were terminated by addition of 500 pl 10% trichloracetic acid 62 followed by 100 pl 5% bovine serum albumin (Inoue et al., 1985). After centrifugation and neutralization, total acid-soluble 3H in the supernate was determined by liquid scintillometry, with subtraction of no-enzyme blanks. Miscellaneous methods: Protein concentrations were determined by the method of Lowry (1951) or by the Read and Northcote (1981) modification of the Bradford (1976) method. ROS and washed ROS were treated with pertussis toxin (List Biological Laboratories, Campbell CA) by the method of Van Dop et al. (1984) and with cholera toxin (Sigma) by the method of Abood et a1. (1982). Isotonic supenatant was treated with pertussis toxin by the method of Manning et a1. (1984). 63 RESULTS A. Demonstration of Phospholipase C in ROS. Hydrolysis p: phosphoinositides labeled i_ situ: Breakdown of phospholipids can be measured by disappearance of labeled lipids or formation of water-soluble products. The work of Hayashi and Amakawa (1985) illustrates the first method. They incubated isolated frog ROS with [7-32P]ATP and PIP, producing labeled PA, PIP and PIPZ. A five-second flash of light immediately before termination of reaction and extraction of lipids resulted in a 20% decrease in labeled PIP2. Table 1 shows the results of a similar experiment with bovine ROS. No effects of light were found. Initial experiments examining formation of water-soluble products from labeled phosphoinositides employed ROS pre-incubated with [3H]inositol. ([32P]ATP was not used because of the large number of labeled products possible in addition to inositol phosphates.) After several hours incubation, ROS were pelleted by centrifugation and resuspended in fresh buffer containing no label. Various concentrations of metal ions and other effectors were used in the resuspension buffer, and the resuspended membranes were incubated with or without illumination. Water-soluble products were analyzed by ion-exchange chromatography (Downes and Michell, 1981). The amount of labeled 1P3 recovered from such experiments was very small (<0.001% of original 3H). As noted in Chapter I, [3H]inositol incorporation into PIP2 proceeds quite slowly, and it was concluded that the amounts of [3H]PIP2 produced by this method 64 Table 1. No decrease in labeled phospholipids upon illumination. PA PIP RLRZ Dark 84560 3728 801 $3524 $170 $33 Light 89536 3801 812 $3474 $64 $11 ROS (1 nmol rhodopsin/assay) were incubated under dim red light in 100 , pl minimal 32P labeling medium with the following modifications: [7—32P]ATP concentration was 20 pH, with a specific activity of 5 mCi/pmol; unlabeled PIP was added to a final concentration of 100 pM. After 5 min, reactions were terminated by addition of 50 pl 1N HCl. Some tubes were exposed to a 5 sec flash of white light from an incandescent bulb immediately before termination. Labeled phospho- lipids were separated and quantified as described in Chapter 1. Results are shown as means $ standard errors for 3 (light) or 4 (dark) replicates. 65 were inadequate. It was therefore decided to use exogenous [3H]PIP2 as a substrate for measuring phospholipase C activity. Hydrolysis pi exogenous 3H PIPg: ROS incubated with [3H]PIP2 released 3H that remained in solution when protein and unhydrolyzed substrate were precipitated with trichloracetic acid. Analysis of the acid-soluble label by ion-exchange chromatography (Figure 1) indicated it was primarily [3H]IP3 with smaller amounts of [3H]IP2, [3H]IP and [3H]inositol. (This technique will not resolve isomers.) Since ion-exchange analysis of the products of every assay would be extremely time-consuming and limit the number of experiments that could be performed, PLC activity was routinely measured by determining total acid soluble 3H. B. Characterization of ROS PLC Activities. Soluble app particulate fppmp pf ROS phospholipase O: PLC activity was found both in soluble and particulate fractions derived from ROS. The soluble activity was released from ROS during isotonic washing, while particulate activity remained associated with the membranes even after several washings (Table 2). However, repeated washing failed to eliminate PLC activity in the supernates, suggesting that washed ROS membranes (WROS) continued to release small amounts of soluble PLC. This may represent leakage of cryptic PLC from compartments in the ROS, or an equilibrium between soluble and membrane-bound forms of PLC. As a result of this slow release of soluble activity, assays of WROS do not represent purely membrane-bound PLC. Unless otherwise noted, assays of soluble PLC were performed using the lst isotonic supernate from a washing such as that shown in Table 2. Particulate material was removed by 100,000 x g ultracentrifugation or 0.45 pm filtration. 66 Figure ;. Identification pf labeled products _f 2122 hydrolysis. ROS (320 pg protein) were incubated with 10 uM [3H]PIP2 (5 pCi/pmol) in 50 mM Tris-HEPES, pH 7.5, 150 mM KCl and 50 pM CaC12 for 30 minutes at 30°in a final volume of 500 pl. After precipitation of unhydrolyzed PIP2 with trichloracetic acid and BSA, the supernate was neutralized, diluted to 10 m1, and applied to a l-ml Dowex-l (formate) minicolumn. The column was then washed with 10 ml H20. Inositol phosphates were eluted with a series of formic acid-ammonium formate solutions as described by Downes and Michell (1981). l-ml fractions were collected and counted by liquid scintillometry. For the flowthrough (F) and water wash (W), l-ml aliquots were counted and the results (less background) multiplied by 10. (Ins = inositol.) 67 300- 200- E o. 0. I co 100- IFS IP IP2 IP3 24 28 32 4 8 12 16 20 FRACTION h 0.1 M HCOOH .1. 0.1 M HCOOH _h 0.1 M HCOOH _, I‘o.2 M HCOONH4T 0.4 M HCOONH4' 1.0 M HCOONH,’I Figure l 68 Table 2. Removal of phospholipase C from ROS by isotonic washing. Protein Volume [3H]PIP2 Fraction Concentration Assayed hydrolyzed mg/ml pl ____gpm___ ROS 13.3 50 4663 $ 52 lst isotonic supernate 0.13 100 2423 $ 33 2nd isotonic supernate 0.15 100 843 $ 26 3rd isotonic supernate 0.03 100 493 $ 30 4th isotonic supernate 0.02 100 323 $ 49 Final pellet 6.6 50 1537 $ 76 ROS were washed 4 times with 20 volumes of isotonic washing solution. The final pellet (WROS) was resuspended in one volume of the same solution. The pellet, supernates, and ROS were assayed for PIP2 phospholipase C as described in Materials and Methods. Results are presented as means $ std. errors for duplicate assays. 69 The apparent specific activities of ROS and washed pellet are much lower than those of the supernates, due in part to the large amount of protein (rhodopsin) present in the membranes. Additionally, dilution of label by endogenous PIP2 may produce a lower effective specific radioactivity for substrate in assays of ROS and particulate fraction. Comparison of activity in soluble and membranous preparations must therefore be made with caution. Validation pi ppppy: As would be expected, due to the presence of unlabelled substrate in the membranes, the particulate fraction did not show a linear relationsip between volume of enzyme and [3H]PIP2 hydrolysis. Activity in the soluble fraction was also non-linear with respect to volume of enzyme, suggesting the presence of an endogenous inhibitor. Detergent (0.3% octylglucoside) relieved this inhibition, producing several-fold increases in activity when large amounts of crude enzyme were used (Figure 2). The nature of the endogenous inhibitor is discussed in more detail below. With respect to time, assays of unwashed ROS and isotonic supernate were approximately linear for 30 minutes, but activity in WROS began to decrease after ~15 minutes. Comparison pf substrates: ROS hydrolyzed all three phosphoinositides (Table 3), as did the soluble and particulate fractions (Table 10, columns 1 and 3). In separate experiments using [3H]phosphatidylcholine as a substrate, no hydrolysis could be detected in ROS or isotonic supernate. Effects of divalent metal ions: PLC in both isotonic supernate and 2+ WROS displayed an absolute requirement for Ca Treatment with EGTA abolished activity, which was restored by CaClZ. Both fractions 70 Figure 2. Effect pi octylglucoside pp Quantity-activity relationship pf soluble PLC. Varying amounts of isotonic supernate were assayed for PIP2 PLC activity as described in Materials and Methods, with the addition of 0.3% octylglucoside to some tubes. Results are presented as means $ std. errors for duplicate assays. [3H]PIP2 hydrolyzed, pmoIe/min 71 75 I I I I I I I' j l I I l l I H + 00 H -— OG 50- — 25— — O I l I I I I I I I I l I l r 0 3 6 9 Isotonic supernatant, Figure 2 12 15 pg protein 72 Table 3. Comparison of substrates for ROS phospholipase C. Specific Activity Substrate Hydrolyzed Substrate pCi/umol dpm [3H]PI 10 31987 $ 2444 [3H]PIP 5 33897 $ 169 [3H]PIP2 5 41610 $ 1537 ROS (0.9 mg protein/assay) were incubated with tritiated phosphoinositides as described in Materials and Methods. Results are shown as means $ standard errors for duplicate assays. [3H]PIP2 specific activity in this experiment was twice that used routinely. Dilution of substrate radioactivity by endogenous phosphoinositides in ROS was not considered in calculating specific activities. 73 2+ contained sufficient endogenous Ca to display substantial activity 2+ in the absence of added Ca or chelators, but were usually stimulated by additional Ca2+. This stimulation was potentiated by isotonic K+ or Na+ (Table 4). High (2 300 pM) Ca2+ concentrations were inhibitory, however. Half-maximal activity was obtained from isotonic supernate at ~0.1 pM free Ca2+, from WROS between 1 and 10 pM (Figure 3). Mg2+ at concentrations above 0.1 mM strongly inhibited the soluble PLC. In contrast, PLC activity in WROS was stimulated by 1 mM Mg2+ but inhibited at higher concentrations (Figure 4). Mn2+ at concentrations above 30 pM strongly inhibited both soluble and particulate activity (Figure 5). The possibility that inhibition by these ions was an artifact caused by precipitation of metal-IP3 salts was examined by adding an inhibitory concentration of each ion at the end of an assay. No decrease in water-soluble 3H was observed. Effect pf spermine: Spermine at low concentrations (0.1 mM) stimulated the particulate activity but had no effect on the soluble. Higher (millimolar) concentrations inhibited both soluble and particulate activity (Figure 6). Effects pf monovalent metal ions: Low ionic strengths produced submaximal activity, as seen in Table 4, but no specific effects of K+, Na+ or Li+ were observed; replacement of one of these ions by another at constant ionic strength had no significant effect. Effect pi pR: Both the soluble and particulate activities were maximal at pH 6.5. Activity at pH 7.5 was 40% - 50% less than at 6.5; nevertheless, assays were routinely performed at the higher pH to 74 2+ Table 4. Effect of ionic strength and Ca on PLC activity. Fraction Added Salt |3H|PIP2 hydrolyzed, dpm No added CaC12 30 pM CaC12 Particulate none 929 $ 13 933 $ 120 140 mM KCl 1346 $ 109 2344 $ 151 140 mM NaCl 1457 $ 13 2223 $ 25 Soluble none 1099 $ 21 643 $ 47 140 mM KCl 1184 $ 8 1807 $ 47 140 mM NaCl 982 $ 25 1584 $ 156 WROS and isotonic supernate (250 and 32 pg protein/assay, respectively) were incubated with [3H]PIP2 (10 pM, 2.5 pCi/pmol) for 15 minutes in 50mM Tris-HEPES pH 7.5 plus the indicated additions. Enzyme aliquots contributed an additional 7.5 mM KCl to each assay. Results are presented as means $ std. errors for duplicate assays. 75 Figure S. Effect pf gag: p_ PI P C activity. Particulate (n, 250 pg protein/assay, 10 min incubation) and soluble (o, 2.8 pg protein/assay, 30 min incubation) fractions were incubated with [3H]PIP2 as described in Materials and Methods, except that Ca2+ concentration was adjusted to the indicated values with a Ca-BAPTA buffer. All assays contained 100 pM total BAPTA; free Ca2+ concentrations were calculated using a value of 107 nM for the Ca-BAPTA dissociation constant (Tsien, 1980). Results are presented as means $ std. errors for duplicate assays. [3H]PIP2 hydrolyzed, dpm 76 2500-. Ill Particulhte' I r . ' H Soluble I ' i ‘ -I 2000' I L I 1 15003 L .10001 3 500- J 4 d O I ' I j I ' F ' I r -8 -7 -6 -5 —4 -:5 log [002+] M Figure 3 77 Figure fl. Effect pg flggi 0 IP P C activity. Soluble (o, 30 min assays) and particulate (I, 10 min assays) fractions were assayed for PIP2 phospholipase C as described in Materials and Methods, with the addition of the indicated concentrations of MgC12. Activity at each concentration is 2+ expressed as percent of no-Mg control, mean $ std. error for duplicate assays. 78 B _‘Illl‘ IIIIrIIIIIIrrlrIrrTTIUr *5 2001 H Particulate . c O 4 H Soluble . 0 q H— s j 0 1504 - 8\° ' . p“ . 0 1 g, 1004 _ (D j ‘ L d 'U j ., > .5: - . (\I 50: .. Q.- l d E‘ ”H O llt‘illrfrITIIlllIITrTUIrTITrr 0 1 2 3 M92+ added, mM Figure 4 79 Figure S. Effect pf Mn2i 0 PI P C activity. Soluble (o) and particulate (u) fractions were assayed for PIP2 phospholipase C as described in Materials and Methods, with the addition of the indicated concentrations of MnClZ. Activity at each concentration is expressed as mean $ std. error for duplicate assays. [3H]PIP2 hydrolyzed, dpm 80 48 0| 0 O O O O O ljl+lllll (A O O O L 20001 l 000 I + 1 . H Soluble H Particulate j Mn2+ added, M Figure 5 U «I .1 81 Figure p. Effect p: spermine p_ R;_ _LO acitivity. Particulate (u, 10 min assays) and soluble (o, 30 min assays) fractions were assayed for PIP2 phospholipase C as described in Materials and Methods, with the addition of the indicated concentrations of spermine tetrahydrochloride. Activity at each concentration is expressed as percent of no-spermine control, mean $ standard error for duplicate determinations. 82 E 300‘II{1"leIrUUUTIIIUUU.]TIrfl E H PartIculate o : H Soluble 0 250-: - . O L\° 200-: _ci’ 4 i3 50' 2 1 " O . L '2. 100 .C N e 50 0. F1 :I: r’) L—-J q .l a 1L1 l LLJJII O UUIU UUUU 'UII FITS TUrT I I I I I O 0 IV 0.0 05 1.0 ' 15 2.0 2.5 [Spermine], mM Figure 6 83 maintain consistency with experiments done before the pH optimum was determined. Chromatographic characterization: Gel filtration of isotonic supernate produced a broad activity peak, suggesting poorly resolved multiple forms of PLC (Figure 7). The position of the peak's center corresponded to a Mr of ~160,000, however, the peak is wide enough to include proteins of twice or half that Mr' This may represent a multimer/monomer equilibrium. Ion-exchange chromatography of isotonic supernate produces two peaks of PLC activity, typically eluting from DE-52 at 0.22 and 0.35 M NaCl (Figure 8). C. Possible Regulators of ROS PLC. l.Nuc1eotide§ and 932 Binding Proteins. GTP had no effect on either the soluble or particulate form of the PLC. Since ROS contain GTPase activity (Godchaux and Zimmerman, 1979), non-hydrolyzable GTP analogs were also used. In the presence of Mg2+, high (> 100 pM) concentrations of guanylyl-(fl,7-methylene)- diphosphonate (GMPPCP) inhibited both the soluble and particulate activities (Figure 9). At similar concentrations, guanosine-5'-O- (2-thiodiphosphate) (GDPfiS) was found to be inhibitory and guanylyl- imidodiphosphate (GMPPNP) moderately stimulatory, but these effects were variable, possibly due to seasonal differences in the ROS. The concentrations required for these effects were considerably higher than those usually associated with G-protein-mediated effects. Inclusion of detergent (0.3% octylglucoside) in assays eliminated the effects of GTP analogs. 84 Figure 1. gel filtration of soluble 2L9. Isotonic supernate (500 mg protein) in a volume of 1.5 ml was applied to a 100 x 1 cm G-150 column and eluted with 10 mM NaHEPES, pH 7.5, 250 mM NaCl, 30 pM CaClz. 4-ml fractions were collected and assayed for PLC activity by adding 5 pl lmM [3H]PIP2 (2.5 pCi/pmol) to 495 pl fraction aliquots. After 30 min, reactions were terminated and acid soluble 3H determined. Solid line = activity; dashed line = A28O' [3H]PIP2 hydrolyzed, pmol/min 85 1.0 20.0 IIIIT—rlIIIIIIIIIfTTTIrTITIUII I I I I I I l : A I 5.0-4 g 1 -: .I I J D- -' |I 1P----.‘- 0.0 IIIIIIIIIIIIIITfiTIIIIIIIIIfiT 1 10 20 3o Fraction Number Figure 7 (---) 093v 86 Figure 8. log-exchange chromatography of soluble ELQ. Isotonic supernate (ca. 20 mg protein) was applied to a lO-ml DE-52 column and eluted with a linear 0 - 1.0 M NaCl gradient (dotted line) in 10 mM Tris-HCl, pH 7.8, 1 mM DTT. S-ml fractions were collected and assayed for PIP2 phospholipase C activity (solid line) and A280 (dashed line). [3H]PIP2 hydrolyzed, pmoI/min . 25.0 87 N F3 o l 141 l I IIIIUITIIIIUIIIIllll'lIIIIIIIIIIIUUTTTO1.0 O ‘T 1 r .5 U1 .------------—.' 1 n _k‘ ' 'r'I'IIl'I'IIrjIIIIIIIIIUI—I1IIIIIUUUIrrO 10 . 20 30 4o Fraction Number Figure 8 (___) OBZV 9 w ‘IODN 88 Figure 2. Effect of GMPPCP on PIP2 PLC activity. Particulate (u, 3.4 mg protein/assay) and soluble (o, 0.1 mg protein/ assay) fractions were assayed as described in Materials and Methods, with the addition of 3 mM MgC12 and the indicated concentrations of GMPPCP. Activities are shown as means i std. errors for triplicate (particulate) or duplicate (soluble) assays. 1 500 89 H Particulate 40—4» Skfluhfle E o. .0 . J .0 o 1000- N £Z~ ‘ o L. .0 > L '1 N ' .. EE: ENDED 0.. l""'1 1 I to L——J C) C) UTTI'TTTI rill r'rr‘I I""l‘ 100 200 300 400 500 [GMPPCPLLIM Figure 9 9O Pertussis toxin inhibits G-protein-mediated stimulation of phospholipase C in some cells (Cockcroft, 1987). Pertussis toxin ADP-ribosylated G-protein (Figure 10) but did not affect PLC activity (Table 5). The effects of GMPPCP, GDPBS and GMPPNP were not affected by pertussis toxin treatment. ROS G-protein can also be ADP-ribosyl- ated by cholera toxin; this toxin too had no effect on PLC activity. Guanosine-S'-0—(3-thiotriphosphate) (CTPyS) stimulates G-protein- regulated PLCs in many cell types (Banno et al., 1986; Huque and Bruch, 1986; Hepler and Harden, 1986; Uhing et al., 1985). In at least some cases this activation results from decreasing the PLC's Ca2+ requirement (Smith et al., 1986; Deckmyn et al., 1986). GTPyS at concentrations from 0.1 to 100 pM had no effect on particulate PIP2 PLC activity in the presence of added Ca2+ (Figure 11). A modest inhibition by 100 pM GTPyS observed in the absence of added Ca2+ is probably due to Chelation of endogenous Ca2+ by the nucleotide. GTPyS in concentrations from 0.1 to 100 pM was entirely without effect on the soluble PLC in the presence or absence of added Ca2+. GTPyS was also without effect at low Ca2+ concentrations maintained by Ca-EGTA buffers, in the presence of added transducin, in whole (unwashed) ROS and under various conditions of illumination. Fluoride ion activates G-protein-regulated PLCs (Martin et al., 1986). The activation is enhanced by aluminum ion, presumably via formation of AlFA', which appears to be the true activator (Cock- croft and Taylor, 1987). ROS G-protein is similarly affected (Kanaho et al., 1985; Bigay et al., 1985). ROS PLC activity was unaffected by NaF at concentrations up to 5 mM. A13+ was not used because it inhibited ROS PLC at concentrations (~10 pM) used in the above papers. 91 Figure IQ. _QE-ribosylation 9: Egg fractions by pertussis toxin. ROS, particulate fraction, and soluble fraction were incubated for 30 min with (lanes 1, 3, and 5) or without (lanes 2, 4, and 6) pertussis toxin using the conditions described in the caption of Table 7, except that unlabelled NAD was replaced with 5 uCi (soluble fraction) or 10 Ci (ROS and particulate fraction) of [a—32P]NAD (7 Ci/mmol). Incubation was terminated by addition of SDS (soluble fraction) or trichloracetic acid (ROS and particulate fraction). Labeled proteins were analyzed by SDS-polyacrylamide gel electrophoresis. An autoradiogram of the dried gel is shown. The autoradiogram was deliberately overexposed in lanes 3 and 5 to secure adequate exposure of lane 1. Lanes 1 and 2: Soluble fraction (6 pg protein/lane). Lanes 3 and 4: particulate fraction (6 nmol rhodopsin/lane). Lanes 5 and 6: ROS (8 nmol rhodopsin/lane). Lane 7: purified ROS G protein (marker indicates position of a subunit). 92 Figure 10 93 Table 5. Pertussis toxin does not affect PIP2 PLC activity. [3HIPIP2 hydrolyzed. dpm Treatment Particulate Soluble Control 5199 i 330 12823 i 167 Pertussis toxin 5097 i 413 12955 i 1220 Particulate fraction was treated with pertussis toxin by the method of van Dop et al. (1984). Briefly, WROS (25 nmol rhodopsin) were incubated at 30° C under dim red light in 500 p1 of 8 mM sodium 3-(E—morpholino)propanesulfonic acid, pH 7.4, 35 mM NaCl, 50 mM KCl, 2mM MgC12, 2 mM DTT, and 0.2 mM NAD, with or without 12.5 pg of pertussis toxin. After 30 min, 75—p1 aliquots were assayed for phospholipase C activity as described in Materials and Methods, with the addition of 2.4 mM MgC12. The [3H]PIP2 used in these assays had a specific activity of 10 pCi/pmol. Results are shown as means i std. errors of triplicate assays. Soluble fraction was treated with pertussis toxin by a modification of the method of Manning et a1. (1984). Aliquots of isotonic supernate containing 60 pg Lowry protein were incubated at 30 °C in 100 ul of 100 mM Tris-HCl, pH 7.8, 2.5 mM MgC12, 10 mM thymidine, 2 mM DTT, 0.2 mM ATP, 0.1 mM GTP, and 0.1 mM NAD, with or without 1.2 pg of pertussis toxin. After 3 hr, the total contents of each tube were assayed for PIP2 PLC activity as described in Materials and Methods. Results are shown as means i std. errors of duplicate assays. Pertussis toxin by itself had no PLC activity. 94 Figure 11. Effect of GTPIS g_ particulate Pl P C activity. WROS (7 pmol rhodopsin/assay) were assayed for PLC activity as described in Materials and Methods, with the following modifications: GTPyS was added in the indicated concentrations, and CaC12 was omitted from some assays (filled circles) and present in others at a concentration of 10 pM (open circles). 95 5000 L l DJ «P O O O O O O 1 l FD O O O L P l A [3H]P|P2 hydrolyzed, cpm 3 8 9—0 + C02+ II CF“. " (30:24- ” I ' I ' 1 10 [GTP78],pM Figure 11 100 96 Ghalayini and Anderson (1987) report inhibition of PIP2 PLC in bovine ROS by 2.5 mM ATP in the presence of 5 mM MgC12 and 100 pM CaClz. As shown in Table 6, this inhibition also occurs in washed ROS but is not specific for ATP. The regulatory significance of such inhibition is unclear. 2. Light. The possibility that light might regulate ROS PIP2 PLC was examined under a variety of conditions, including the use of washed and unwashed ROS; exposure to light in advance of, at the start of, and during the assay; presence of GTP and non-hydrolyzable analogs; and 2+ 2+ and Mg . Although occasional varying concentration of added Ca modest increases in activity apparently caused by bleaching have been observed, these have not been consistently reproducible. Table 7 illustrates unresponsiveness of PLC activity to illumination; in contrast to PLC, cyclic nucleotide phosphodiesterase activity in isolated ROS is stimulated an order of magnitude or more by light in the presence of Mg2+ and GTP (Kohnken et al., 1981a). It is possible that the ROS PLC loses the capacity to respond to light during purification. The PLC activity of total retinal homogenates was compared in dark and light; no difference was found. Intact retinas cannot be assayed for PLC using exogenous substrate, but ROS and other retinal fractions isolated from retinas that were exposed to light were compared to those isolated from unbleached retinas (Table 8). The differences detected were minor at most; considering that the bleached and unbleached materials were necessarily handled separately during processing, small disparities do not provide significant support for light effects. 97 Table 6. Effect of nucleotides on particulate PIP2 PLC activity. Added Nucleotide |3H PIP2 hydrolyzed, dpm None 1033 i 62 (4) ATP 286 i 13 (3) ADP 323 i 48 (3) AMP 394 i 55 (3) GTP 818 i 17 (3) GDP 814 i 40 (2) GMP 603 i 42 (2) WROS (0.3 mg protein/assay) were assayed for PLC activity as described in Materials and Methods, except that each assay contained 100 pM CaC12, 5 mM MgCl2 and 2.5 mM of the indicated nucleotide. Assay duration was 15 min. Results are shown as means i std. errors for n replicates (n = number in parentheses). 98 Table 7. ROS PLC activity unaffected by light. Nucleotide None GTP 100 pM GMPPCP 100 pM l3HlPIP2 hydrolyzed. cpm Dark 4503 4184 3027 i i i ROS (6.6 pmol rhodopsin/assay) were Light 39 4761 i 167 364 4720 i 141 131 3110 i 23 assayed for PIP2 PLC activity as described in Materials and Methods, except that 100 pM MgClz, or 200 pM Mg012 and the indicated nucleotide, were added to the reaction mixture. Assays were performed under dim red light ("dark") or normal laboratory illumination ("light"). Results are shown as means i std. errors of duplicate assays. 99 Table 8. PLC activity in fractions from bleached & unbleached retinas. ROS were isolated as described in Chapter I. 100 retinas were exposed to normal laboratory illumination ("bleached") after excision but before shaking. These were processed in parallel with 200 retinas that were exposed to dim red light only. Fractions derived from the bleached retinas are labeled "(1)", those from the unbleached, "(d)". Volumes of the unbleached fractions were twice those of the corresponding bleached fractions, to maintain equivalent volume per retina. The supernants from the 1100 x g centrifugations (which pellets crude ROS) were centrifuged at 100000 x g for 1 hr to produce a pellet of microsomes. "Microsomal super" is the supernate from this ultracentrifugation. Aliquots of bleached and unbleached ROS were washed as described in Materials and Methods to produce WROS and isotonic supernates. Fractions produced from the two sets of retinas were assayed for PIP2 PLC activity as described in Materials and Methods. ROS and WROS were assayed under dim red light. 0.3% octylglucoside (0G) was added to some assays to release cryptic or inhibited enzyme. (W-7 was used instead of 0G in assays of isotonic supernates). Results are shown as dpm of [3H]PIP2 hydrolyzed in 15 min, means i standard errors for duplicate assays. 100 Table 8. PLC Activity in Fractions from Bleached & Unbleached Retinas Fraction ROS (d) ROS (1) Microsomes (d) Microsomes (l) Microsom. super(d) Microsom. super(l) WROS (d) WROS (l) Isotonic super (d) Isotonic super (1) Lowry Volume Protein Assayed mggml pl 14 5 0.5 13 5 0.5 8.4 5 0.5 8.4 5 0.5 4.9 5 0.5 4.6 5 0.5 8.6 5 9.4 4.5 0.02 50 0.02 50 2624 743 2214 840 3580 3457 3156 3746 2502 1624 2950 1749 1153 1353 699 842 a 100 pM W-7 was used instead of 0.3% 0G [3H]PIP2 hydrolyzed, dpm 06 i 139 i 20 i 195 i 86 i 354 i 131 i 277 i 33 i 201 i 171 i 36 i 35 i 53 i 21 i 17 i 16 + DC 5003 781 5266 944 8778 5722 _ 8697 _ 6986 12279 1592 11546 1607 728 933 1714 2337 in these assays. i1385 i 23 i 723 i 17 i 322 i1152 i 828 1+ 52 i 444 i 12 60a |+ 101 3. Calmodulin. Sensitivity of ROS PLCs to Ca2+ suggested possible regulation by calmodulin (CaM). Although its target enzymes have not been identified, CaM is present in ROS (Kohnken et al., 1981c). Because of the presence of endogenous CaM, initial experiments probing the possibility of CaM-regulation did not use added CaM but instead examined the effects of CaM antagonists. These are compounds that antagonize the effects of CaM by interfering with its binding to target proteins. Chemically they are quite diverse. The effects of a number of CaM antagonists from several different chemical categories are described below. Unexpectedly, CaM 3‘ antagonists stimulated PIP2 PLC. Phenothiazines: The phenothiazines, a family of antipsychotic tranquilizers, were the first category of CaM antagonists to be discovered. Chlorpromazine (2-chloro-10-(3-dimethylaminopropyl)- phenothiazine) and trifluoperazine (10-[3-(4-methy1piperazin-l-yl)- propyl]-2-(trifluoromethyl)-10H-phenothiazine) inhibit CaM-stimulated cyclic nucleotide phosphodiesterase with ICSOS of 10 and 42 pM respectively (Weiss and Levin, 1978). Figure 12 shows the concentration-dependent stimulation of soluble PLC by the phenothiazines. The apparent ECSOS were somewhat higher than those for phosphodiesterase inhibition. Gietzen (1986) points out that reported ICSOS for CaM antagonists are highly variable, and depend greatly on the concentration of CaM and any other hydrophobic materials (such as proteins or lipids) present. Disparities of an order of magnitude or more for ICSOS of these and other CaM antagonists can be found in the literature. Excessive weight should therefore not be given to differences between experimental and literature values for effective concentrations; neither should CaM 102 Figure _2. Effect of phenothiazines Qfl soluble PIP; PLC activity. Isotonic supernate (30 pg protein/assay) was assayed for PIP2 PLC activity as described in Materials and Methods, but with the addition of the indicated concentration of trifluoperazine (A) or Chlorpromazine (6). Results are presented as % of no—drug control, mean i std. error of duplicate determinations. 103 505 . k—i Tfifluoperafine ‘ H Chlorpromazine [3H]PIP2 hydrolyzed, % of control ’TFIIIIIIII O I l I I I 1 I I l I I I I I 0 50 100 I l l 150 [Phenothiazine],plvl Figure 12 . I 200 104 involvement be inferred simply from similarities. Concentration curves for these drugs were not performed on WROS, but in separate experiments, WROS PLC was activated 31% by 200 pM Chlorpromazine (p < .05) and 104% by 100 pM trifluoperazine (p < .001). Haloperidol: Although unlike the phenothiazines structurally, haloperidol (4-[4—(4-chlorphenyl)-4-hydroxy-l—piperidinyl]—l—(4-fluoro- phenyl)-1-butanone), like them, is used clinically as an antipsychotic tranquilizer and in Vitro as a CaM antagonist, with an ICSO of 60 pM for CaM-stimulated phosphodiesterase (Weiss and Levin, 1978). Haloperidol at concentrations in the 10'4 M range stimulates PIP2 PLC in both isotonic supernate and WROS (Figure 13), the effect being greater on the supernate. Calmidazolium: Calmidazolium (Compound R24571, l-[bis(4—chloro- phenyl)methyl]-3-[2,4-dichloro—fi-(2,4-dichlorobenzyl-oxy)phenethyl]- imidazolium chloride) is reportedly a highly specific CaM antagonist, with an ICSO for inhibition of erythrocyte CaM-dependent Ca2+- ATPase of ~0.4 pM (Gietzen et al., 1981). Table 9 shows the effects of 1 pM and 10 pM calmidazolium on PLC activity in isotonic supernate and WROS. Melittin: Melittin is a 26-amino-acid polypeptide found in bee venom and is a potent CaM antagonist. Its reported Ki for CaM-dependent protein kinase is 80 nM (Katoh et al., 1982). Melittin is not highly specific for CaM, however, and also stimulates phospholipase A2 (Mollay et al., 1976), inhibits protein kinase C (Katoh et al., 1982) and interacts directly with phospholipid membranes (Dawson et al., 1978). 105 Figure 3. Effect 9f haloperidol g_ PIP2 PLC activity. WROS (u, 7 pmol rhodopsin/assay) and isotonic supernate (o, 30 pg protein/assay) were assayed for PLC activity as described in Materials and Methods, with the addition of the indicated concentration of haloperidol. Results are presented as % of no-drug control, means i standard errors for duplicate assays. [3H]PIP2 hydrolyzed, % of control 106 :355() A ' ' ' ' l ' ' ' ' l ' ' ' ' I ‘ ‘ ' ' L soof 250{ 200{ 150{ 100: 50{ j . H Soluble 1 1 H Particulate - O I I I I I l I l I I I I I I I I I I I I o 50 100 150 200 [Haloperidol], pM Figure 13 107 Table 9. Effect of calmidazolium on PIP2 PLC activity. lCalmidazoliuml, EM l3HlPIP2 hydrolyzed. dpm Soluble Particulate 0 9348 i 342 2155 i 85 1 12400 i 73 2293 i 120 10 25007 11734 3234 i 67 Isotonic supernate and WROS were assayed for PIP2 PLC activity as described in Materials and Methods, with the addition of the indicated concentration of calmidazolium. Ethanol (solvent for calmidazolium) was present in all assays at a concentration of 1% v/v. Results are shown as means i std. errors of duplicate (soluble) or triplicate (particulate) assays. 108 Figure 14 shows the effect of melittin on PLC activity in WROS and isotonic supernate. The soluble activity was activated by concentrations comparable to those that inhibit CaM-activated enzymes, but inhibited at higher concentrations. WROS required higher concentrations for activation. Compound 48188: Compound 48/80 is actually a mixture of compounds produced by the condensation reaction of N-methyl-p-methoxyphen- ethylamine and formaldehyde (Baltzy et al., 1949). It is reportedly a highly specific CaM antagonist with an ICSO of 0.85 pg/ml for CaM-dependent Ca2+ -ATPase (Gietzen et al., 1983). Compound 48/80 is more hydrophilic than other CaM antagonists and therefore less likely to exert non-specific hydrophobic effects on enzymes or substrates (Gietzen and Bader, 1985). The response of PLC to Compound 48/80 was biphasic: stimulation at concentrations up to 2.5 pg/ml, inhibition at higher concentrations. This pattern was seen in both WROS and isotonic supernate (Figure 15). Bronner et al. (1987) report that Compound 48/80 inhibits PI PLC from human platelets. ICSOS were 2 pg/ml for soluble PLC and 5 pg/ml for the particulate fraction. Phospholipase A2 gave a biphasic response, similar to those shown for PIP2 PLC in Figure 15. They interpret their results as non-CaM mediated action of Compound 48/80 on the phospholipases or their substrates, but present no experimental evidence against CaM-mediation. Inhibition of PI hydrolysis is not necessarily inconsistent with stimulation of PIP2 hydrolysis: the CaM-antagonist W-7 produced both effects in ROS isotonic supernate (see below). 109 Figure 14. Effect of melittin on £18; 2L8 activity. WROS (u) and isotonic supernate (0) were assayed for PLC activity as described in Materials and Methods, with the addition of the indicated concentration of melittin. Results are presented as % of no-drug control, means i std. errors for duplicate (particulate) or triplicate (soluble). [3H]PlP2 hydrolyzed, % of control 110 250_ 200% 150—: 100: 509 : H Soluble ' H Particulate 0 I 0.0 0.2 I I I I I 1 I ' I 0.4 0.6 0.8 1.0 [Melittin], pM Figure 14 111 Figure 18. Effect pf compound 48/80 on IP P C activity. WROS (n) and isotonic supernate (0) were assayed for PLC activity as described in Materials and Methods, with the addition of the indicated concentration of compound 48/80. Results are presented as % of no-drug control, means i std. errors for duplicate assays. [3H]PIP2 hydrolyzed, % of control 0 112 'I'I'I'l'l‘l'l'l‘l'l ' H Soluble ' H Particulate 'l'l'l'l'l‘l'f‘l'l'l' 012345678910 [Cmpd 48/80], pg/ml Figure 15 113 Naphthalenesulfonamides: Hidaka and Tanaka (1983) have prepared a series of N-aminoalkyl-naphthalenesulfonamides that are widely used as probes of CaM regulation. Among the most useful are N-(6-amino- hexyl)-l-naphthalenesulfonamide (W—S), N-(6-aminohexyl)-5-chloro-l- naphthalenesulfonamide (W-7), N-(4-aminobutyl)-1-naphthalenesulfon- amide (W-12), and N-(4-aminobutyl)-5-chloro-l-naphthalenesulfonamide (W-l3). W-7 and W-13 are much more potent CaM antagonists than their non-chlorinated analogs, W—5 and W-12, but are similar to them in hydrophobicity. Comparing W-l3 with W-12 and W-7 with W-5 has been proposed as a means of distinguishing CaM-mediated from non-CaM-mediated effects of CaM antagonists (Hidaka et al., 1981; Chafouleas et al., 1982). Figures 16 and 17 show concentration curves for the naphthalensulfonamides applied to isotonic supernate and WROS. W-13 and W-7 stimulated PLC activity much more effectively than W-l2 and W-5. Since W-7 is one of the most widely used and well-characterized CaM antagonists, it was selected for further experiments to examine the effects of these drugs in more detail. W-7 eliminated the tendency of soluble PLC activity to "shoulder off" with increasing enzyme concentration, making activity approximately linear with volume of enzyme (Figure 18). This is consistent with antagonism of an endogenous inhibitor. Figure 18 illustrates another property of CaM-antagonist PLC stimulation: the fold activation obtained was dependent on the quantity of enzyme used in the assay. This was true not only of W-7 but also of the other CaM antagonists examined. 114 Figure 18. Effects pf naphthalenesulfongmides pp soluble PIPz PLC activity. Isotonic supernate (32 pg protein/assay) was assayed for PIP2 PLC activity as described in Materials and Methods, with the addition of the indicated concentrations of W-series compounds. A: W-l3 (A) and W-12 (I). B: W-7 (A) and W-5 (I). Results are presented as means i std. errors for duplicate assays. [3H]PlP2 hydrolyzed, pmol/min 115 A—i W—13 I—I W—12 OIIIIIIIII’TIWIII TWI T ' I 50 100 150 200 Concentration, plvi Figure 16 116 Figure 11. Effects pf naphthalenesulfonamides pp particulate PIP2 PLC activity. WROS (0.26 mg protein/assay) was assayed for PIP2 PLC activity as described in Materials and Methods, with the addition of the indicated concentrations of W-series compounds. A: W-l3 (A) and W-l2 (I). B: W—7 (A) and W-5 (I). Results are presented as means i std. errors for duplicate assays. 117 : H W—13 . H W—12 O I I I l I I I l I I I I I I I I I‘rI‘I 0 50 100 150 200 : B [3H]P|P2 hydrolyzed, dpm 1000- _ A—A W—7 - H W—5 O I l I I I I I I I I I I I l I I I I I I 0 50 100 150 200 Concentration, ,u.l\/1 Figure 17 118 Figure _8. Effect pf fl-l pp activity-volume linearity. The indicated volumes of soluble fraction (0.13 mg protein/ml) were assayed for PIP2 PLC activity in the presence (0) or absence (0) of 100 pM W-7. Results are presented as means i std. errors of duplicate assays. on O 119 \l C.’ 0’) O l [3H]PIP2 hydrolyzed, pmol/min lllL1111'llllllllllll-Lllllllllllllll IllIIUITIlerYIIIIIIIIIII 50 100 150 200 250 Volume of enzyme, pL Figure 18 120 W-7 affected hydrolysis of all three phosphoinositides (Table 10), but in different ways. It appears to alter the specifity of the PLCs in favor of the polyphosphoinositides at the expense of PI. PI hydrolysis was inhibited in the soluble fraction and only slightly activated in the particulate fraction, in contrast to PIP and PIP2. Similar results were obtained with trifluoperazine. Inhibition of PI PLC by CaM antagonists has been reported previously (Wightman et al., 1981; Craven and DeRubertis, 1983; Benedikter et al., 1985; Schwertz et al., 1987; Bronner et al., 1987) but these studies did not attempt to determine the role of CaM, if any, in this inhibition. CaM increases the Vmax of adenylate cyclase and cyclic nucleotide phosphodiesterase without altering their Kms, although conflicting results have also been reported (Cheung et al., 1978). W-7 increased the Vmax of the soluble PLC without altering the apparent Km of approximately 6 pM for PIP2 (Figure 19). The linearity of the double-reciprocal plots is somewhat surprising considering the presence of multiple PLCs as revealed by ion-exchange chromatography. The different PLCs may be similar kinetically, or different forms of a single species. CaM antagonists and 882i: Stimulation by W-7 was dependent on Ca2+ concentration, with half maximal stimulation occuring between 0.1 and 1 pM (Figure 20). This figure also shows that the effects of W-7 and Ca2+ are dependent on the quantity of crude enzyme assayed. W-7 is most effective in the presence of concentrated enzyme, as noted above. In the absence of W-7, calcium's effect is largest in the presence of dilute enzyme. The latter observation can be interpreted 2+ as indicating the presence of a Ca -dependent inhibitor which 121 Table 10. Effect of W-7 on phosphoinositide hydrolysis. Specific Acid-soluble label released. dpm Activity Particulate Soluble Substrate pCi/pmol — W—7 + W-7 - W-7 + W-7 [3H]PI 9.1 4235 i 75 4808 i 53 1022 i 62 348 i 40 [3H]PIP 5.0 6102 i 388 11322 i 390 4652 i 475 5548 i 65 [3H]PIP2 2.5 5348 i 277 7868 i 902 1558 i 118 2898 i 30 WROS and isotonic supernate (400 and 36 pg protein/assay, respectively) were assayed for phospholipase C activity as described Materials and Methods, with or without the addition of 100 pM W-7. Results are shown as means i standard errors for triplicate (particulate) or duplicate (soluble) assays. Values for specific activities of substrates do not reflect dilution of the label by endogenous lipids in the enzyme preparations, which could be significant for the particulate fraction. Similar results were obtained using trifluoperazine instead of W-7. in 122 Figure _2. Effect pf W;Z pp kinetics pf soluble phospholipase 8. Soluble phospholipase C activity was assayed as described in Materials and Methods, but with varying concentrations of [3H]PIP2 and in the presence (0) or absence (0) of 100 pM W-7. Data are presented in a double-reciprocal plot; each point represents the mean of two assays. Data were analyzed by the method of Wilkinson (1961) and numerical results are shown below: - W-7 + W-7 Km 5.8 i 0.9 6.5 i 0.7 V 11.8 i 0.6 28.2 i 1.1 123 1/V, pmol—1 min [Trlill‘ll II IIIII —o.2 —o.1 o"'o31"'o32'"bfzi' 034 0.5 1 /[PIP2], pM-i Figure 19 124 Figure _8. Calcium dependence of W-7 stimulation of PLC. Isotonic supernate was assayed for PIP2 PLC activity in the presence (triangles) and absence (circles) of 100 pM W-7. Ca2+ concentrations were set with a Ca-BAPTA buffer as described in Figure 3. Assays were performed on 15 pg (filled symbols) or 1.5 pg (open symbols) of isotonic supernate protein. Inset: replot (with expanded y axis) of 1.5 pg results, showing 2+ activation by Ca even in absence of W-7. 125 . . . . _ . . .J — . . u . _ . . . . — . . . . O. 15 _ 0. 16 :_E\u_oEQ flog—0.6»; Nan—naming 0. I7 4 7. _ W W + _ A o O A 0 I8 . a a . _ . J a a A q 4 . . — . . . . A a . a . _ 5 O 5 O 5 O 2 O 7 5 2 1 1 c_E\o_oEa .Uombotgg NQEHIE log [002+], M Figure 20 126 dissociates from the enzyme on dilution; this is also consistent with the response of the enzyme to W-7. This point is discussed in more detail below. 88M antagonists 888 p8: The Ca2+-dependent, high affinity binding of trifluoperazine to CaM is pH-dependent, and decreases sharply above pH 7.5. As shown in Table 11, the effect of trifluoperazine on soluble PLC activity increases in the interval from pH 7.5 to 8.5: PLC activity was less inhibited by increasing pH in the presence of trifluoperazine. Similar effects were observed with W—7, mellitin and compound 48/80. QaM antagonists £29 888: Addition of bovine serum albumin to PLC assays in concentrations above 10 pg/ml resulted in decreased activity, which was initially interpreted as a non-specific protein effect. However, as shown in Table 12, this inhibition appeared to be specific to BSA. Inhibition by BSA was antagonized by W-7 but not by W-5 (Figure 21). Albumin binds hydrophobic compounds in serum, and BSA preparations contain bound lipids and other non-polar substances. The inhibitory effect of BSA may be due to such contaminants, but a commercial preparation of BSA that had been specially processed to decrease such contamination (Sigma A-7030) also inhibited PLC. It is possible that substrate or the PLC bind to hydrophobic sites on BSA molecules. On SDS-polyacylamide gels, isotonic supernate shows a minor (<1%) component with an Mr equal to BSA's. Even if this is BSA, its quantity is too small to account for the observed endogenous inhibiton, which typically appears at total isotonic supernate protein concentrations of 5 pg/ml in assay or higher. 127 Table 11. Effect of pH on trifluoperazine stimulation of PLC. l3HlPIP2 hydrolyzed. dpm _2fl_ - TFP + TFP 7.5 4324 i 466 5928 i 332 8.0 3606 i 105 5419 t 32 8.5 2369 i 112 4207 i 300 Isotonic supernate (5 pg protein/assay) was assayed for PIP2 PLC activity as described in Materials and Methods, in the presence or absence of 50 pM trifluoperazine (TFP). The pH of the HEPES buffer was adjusted to the indicated values with KOH. Results are expressed as means i std. errors for duplicate determinations. 128 Table 12. Effect of added proteins on soluble PIP2 PLC activity. Protein None BSA Ovalbumin Sheep IgG Parvalbumin Gelatin Calmodulin [SHIPIPZ hydrolyzed. dpm 2566 498 2257 2386 2594 2505 2570 Isotonic supernate (3.3 pg protein/assay) i 94 i 169 i 98 i 233 i 319 i 185 i 94 was assayed for PLC activity as described in Materials and Methods, with the addition of 25 pg of the indicated proteins. Gelatin (bacteriological grade) was purchased from Difco, calmodulin from Calbiochem, the others from Sigma. Results are shown as means i standard errors for triplicate assays. 129 Figure 81. Effect of W-7 and W-S on inhibition of soluble PLC 8y BSA. Soluble PIP2 phospholipase C activity was assayed in the presence of the indicated concentration of BSA and either 100 pM W-7 (I), 100 pM W-5 (A), or no drug (0). Protein concentration in assays due to enzyme solution was 15 pg/ml. Activities are shown as percent of no-BSA control for each drug treatment. Each point represents the mean i std. error of duplicate assays. P|P2 PLC activity, % of control 130 100 60— Control 404 204 4 O I I T I I I I I l I I l I I I 7 7 T T _I O 50 100 150 200 [BSA] added, lug/ml Figure 21 131 QaM antagonists £29 detergent: 100 pM trifluoperazine and 0.3% octylglucoside (10 mM) produced equal stimulation of soluble PLC (Table 13). Together they had no greater effect than either one alone, suggesting that both prevent the same inhibitory interaction. Calcineurin: Calcineurin is a CaM-dependent protein phosphatase that binds CaM extremely tightly in the presence of Ca2+. Because of its tight binding it has been used as a CaM antagonist: by binding CaM to itself, calcineurin prevents it from binding to other proteins (Klee et al., 1983). Calcineurin added in up to 1000-fold excess over CaM (estimated from radioimmunoassays described below) had no effect on soluble or particulate PIP2 PLC activity (data not shown), although a low level of PLC activity in a commercial (Sigma) calci— neurin preparation was initially mistaken for activation of the enzyme. Qéfl antagonists apg E_8 purification: Preliminary attempts to render soluble PLC insensitive to CaM antagonists by purifying it away from CaM have been unsuccessful. Ion exchange and gel filtration chromatography yielded peaks of PLC activity that could still be stimulated by trifluoperazine. Isotonic supernate was applied to an affinity column of W-7-agarose, to which CaM binds in a Ca2+-dependent manner (Endo et al., 1981). It was hoped that CaM would bind to the column and CaM-depleted enzyme would flow through; instead, the PLC activity bound tightly to the column in a non-Ca2+ ~dependent fashion. EGTA did not elute it, but about half the activity could be eluted with octylglucoside. The eluted enzyme could still be stimulated by W-7. The failure of these chromatographic techniques to produce a CaM-antagonist-insensitive preparation would seem to argue for 132 Table 13. PLC activation by detergent and CaM antagonist not additive. [3H]PIP2 Addition hydrolyzed, dpm None 9065 i 489 Trifluoperazine (100 pM) 18276 i 463 Octylglucoside (10 mM) 17301 i 1448 Trifluoperazine + Octylglucoside 17162 i 537 Isotonic supernate (45 pg protein/assay) was assayed for PIP2 PLC activity as described in Materials and Methods, with the indicated additions. Results are shown as means i std. errors for duplicate determinations. 133 CaM-independent stimulation of PLC, but radioimmunoassays performed in the laboratory of Dr. James Chafouleas (at the Centre Hospitalier de l'Université Laval in Ste. Foy, Quebec) indicated that the PLC preparations produced by these techniques were not depleted of CaM. These RIAs were not part of the original experimental designs, and in some cases no unfractionated starting material was available for analysis, but it appears that some of the PLC fractions were enriched in CaM during partial purification, suggesting co-purification. CaM content of isotonic supernate is ~4 ng CaM/ pg protein as measured by RIA (in the experiments described here and in Kohnken et al., 1981c). Addition pf QQM pp 2L8 assays: If CaM is the endogenous inhibitor, excess CaM should inhibit PLC activity in isotonic supernate that is sufficiently dilute. Sufficiently dilute enzyme would display PLC activity that is proportional to volume assayed and unresponsive to CaM antagonists (Figure 18, lower left corner). Table 14 shows the results of such an experiment. No effects of added CaM were found. 134 Table 14. Added CaM does not affect PLC activity. [3H1PIP2 hydrolyzed. dpm CaM added, pg - W-7 + W-7 O 2875 i 310 2525 i 56 0.2 2741 i 95 2477 i 98 2.0 2498 i 150 2701 i 60 Isotonic supernate was incubated overnight at 4° C in the presence of 10 pM CaC12 and varying amounts of added CaM (from bovine brain, purchased from Calbiochem). Aliquots containing 1.2 pg of isotonic supernate protein and the indicated quantity of CaM were assayed for PIP2 PLC activity as described in Materials and Methods. Results are shown as means i std. errors for triplicate determinations. 135 DISCUSSION Demonstration of phosphoinositide hydrolysis complements demonstration of phosphoinositide synthesis. Analysis of the labeled products of PIP2 hydrolysis shows 1P3 as the major product. The smaller amounts of 1P2, IP and inositol detected presumably result from the action of phosphatases on 1P3, although hydrolysis of [3H]PIP and [3H]PI (produced from [3H]PIP2 by phosphatases) may also contribute. Experiments using [3H]PIP and [3H]PI as substrates indicate that they can be hydrolyzed by soluble and membrane-bound ROS PLCs, although less rapidly than PIP2. Their hydrolysis may be due to multiple enzymes or broad substrate specificity of PIP2 PLC. Although the particulate PLC can be ascribed to the ROS with confidence, the attribution of the soluble activity is less certain. It can be found throughout the sucrose gradients used for isolation of ROS (see Methods), and in the supernate of the crude ROS pellet. ROS prepared by our method have osmotically leaky plasma membranes [R. E. Kohnken, unpublished results, using the method of Yoshikami et al. (1974)] and it is possible that a cytosolic enzyme could escape during preparation. Retinal cells other than rods may also be damaged during shaking, and these could release cytoplasmic enzymes as well. Taking into account the dilution at each step, the activity in the isotonic supernate cannot be accounted for by carry-over from the crude supernate, indicating that soluble PLC continues to leak out of ROS. At least a portion of the soluble PLC activity can therefore by ascribed to the ROS. ‘ .._.._..,,,, 136 In terms of the amount of label released by hydrolysis during assay, the total soluble activity would seem to be much greater than the total particulate activity. However, quantitative comparison of the soluble and particulate activities is difficult, due to the presence of endogenous PIP2 in the membranes, which may decrease the specific radioactivity of the substrate. There is also the problem of accessibility of the particulate enzyme to an exogenous substrate. [3H]PIP2 added to WROS in the presence of Ca2+ chelators (to prevent hydrolysis) was quickly inserted into membranes as judged by centrifugation (data not shown), but it has been reported that some cells maintain discrete pools of phosphoinositides -- one pool available to agonist-stimulated PLC and the other not (Monaco and Woods, 1983). The mechanism of such segregation is unclear, but it further complicates interpretation of results obtained with ROS and WROS. Another question arises concerning the soluble PLC activity: how does it function in vivo? The sensitivity to Mg2+ and spermine suggest that it would be inhibited under cytoplasmic conditions. Majerus et a1. (1986) propose that cytosolic PLCs may associate reversibly with cell membranes and that this may be a part of their regulation. The soluble PLC activity may represent a pool of inactive enzyme that can be activated by association with membranes (plasma or diskal) under the right conditions. If so, the loss of large amounts of this enzyme during preparation could account for failure to observe regulation by light. Some cations appear to have opposite effects on the anabolic and catabolic pathways of phosphoinositide metabolism. Mg2+, Mn2+ and 137 spermine stimulated precursor incorporation and inhibited hydrolysis, Ca2+ at micromolar concentrations has no especially by soluble PLC. effect on phosphoinositide synthesis but is required for hydrolysis. Al3+ inhibited PLC; its effects on synthesis were not examined. Eichberg et a1. (1981) have suggested that polyamines and metal ions form complexes with the negatively charged head groups of phosphoinositides. The observed effects of Mg2+, Mn2+ , Ca2+, Al3+ and spermine may be due in whole or in part to formation of such complexes, and not to the effects of the ions on the enzymes. Hormonal regulation of PIP2 PLC via G-proteins occurs in a wide variety of cells (reviewed by Cockcroft, 1987), and provides an obvious paradigm for possible regulation by light in ROS. The G-protein that mediates light-activation of cGMP phosphodiesterase constitutes a major part of the non-rhodopsin protein in ROS. The non-hydrolyzable GTP analog GTPyS is highly effective in stimulating G-protein regulated PIP2 PLC in several cell types (Uhing et al., 1985; Deckmyn et al., 1986; Smith et al., 1986; Hepler and Harden, 1986), and cGMP PDE in ROS (Yamanaka, et al., 1986). The absence of any effect by GTPyS on ROS PLC, either soluble or particulate, argues against regulation by G-protein. Inhibition by GMPPCP, and the inconstant effects of GDPfis and GMPPNP, occur at concentrations considerably higher than those usually associated with G-protein activation. Pertussis toxin inactivates phospholipase-regulating G proteins in many, but not all, cell types (Cockscroft, 1987); both pertussis and cholera toxin inactivate transducin, Neither affected ROS PLC's activity or sensitivity to GMPPCP. In sum, these results provide little support for regulation of ROS PLC by a G-protein. 138 Nor do they provide any persuasive evidence of regulation by light. It is difficult to compare these results directly with those obtained by some other workers claiming light-stimulation of PLC, due to differences in assay techniques. Many workers, instead of using an exogenous labeled substrate, label retinal lipids in situ, e. g. by intraocular injection of [3H]inositol or 32Pi. Phospholipase activation is then inferred from a decrease in labeled PIP2 in response to light. This technique is impractical with live cattle. Early in this project, analogous experiments were done with isolated ROS; these were not promising and led to the adoption of the exogenous-substrate assay. The recent successful use of exogenous [3H]PIP2 to detect light-stimulated PLC in isolated squid photoreceptor segments (Baer and Saibil, 1988) suggests that this choice of methodology should not have prevented detection of light—regulation. If light does regulate ROS PLC, the regulatory system must be more labile than that of the cGMP PDE, since we have failed to detect any light-activation of PLC under conditions that result in a several-fold activation of PDE. It is also possible that light-regulation of ROS PIP2 PLC is indirect -- that changes in activity result from, rather than cause, the changes in ionic concentrations that attend the rod photoresponse. Isolated ROS are of course separated from the ion pumps of the inner segment; furthermore, ROS isolated in sucrose have leaky plasma membranes. Secondary light-regulation of PLC by ionic changes could hardly be preserved in such a preparation. 139 The stimulatory effects of CaM antagonists are especially provocative, suggesting as they do inhibition by CaM. This would be doubly novel, as PLC has not previously been shown to be CaM-regulated, and those enzymes whose regulation by CaM is well established are all activated by it, not inhibited. However, as Hartshorne (1985) has observed, "there is no a priori reason whey calmodulin interaction must activate enzymatic activity, and it is conceivable that calmodulin-dependent inhibition may occur." The following observations are consistent with CaM inhibition of PIP2 PLC: 1. CaM is present in ROS, both in the soluble and particulate fractions, as determined by radioimmunoassay (Kohnken et al., 1981c). 2. The apparent specific activity of crude soluble PLC decreases with increasing amounts of enzyme, suggesting the presence of an endogenous inhibitor. 3. All anti-CaM drugs tested, including two that are reportedly highly specific, stimulated PLC activity. 4. The relative potencies of W—7 compared to W-5, and W-l3 compared to W-12, for stimulating PLC are consistent with their potencies as CaM antagonists. 5. Stimulation of PLC activity by W-7, a representative CaM antagonist, was dependent on the quantity of enzyme; at high enzyme concentrations W-7 appeared to release the enzyme from the inhibition noted in 2. 6. The effect of W-7 was Ca2+ dependent, increasing sharply between 0.1 and 1 pM Ca2+. 140 Although these observations seem persuasive, pharmacolgical probes of CaM regulation are seldom unambigous. No CaM antagonist is perfectly specific. Most of them are fairly hydrophobic and can interact with lipids and hydrophobic proteins. Phenothiazines and haloperidol inhibit protein kinase C (Schatzman et al., 1281), as do naphthalenesulfonamides (Tanaka et al., 1982) and melittin (Katoh et al., 1982), which also activates phospholipase A2 (Mollay et al., 1976). Other results argue against CaM regulation: 1. The concentrations of the CaM antagonists required for full effectiveness are somewhat higher than those reported for CaM-activated enzymes; W-7 and W-l3, for example, show no sign of saturation in isotonic supernate at concentrations up to 100 and 200 pM, respectively. Their respective ICSOS for CaM-stimulated cylic nucleotide phosphodiesterase are 26 and 66 pM, respectively (Hidaka and Tanaka, 1983). Gietzen (1986), warns against attaching great significance to effective concentrations, however. 2. Calcineurin, which binds CaM with high affinity, did not affect PLC activity. It is arguable that the affinity of PLC for CaM may also be very high. 3. The effect of pH on trifluoperazine stimulation of PIP2 PLC does not match that reported for its binding to CaM. 4. Most significantly, addition of CaM to PLC assays does not produce inhibition. Initially this point was not considered compelling, due to the presence of endogenous CaM and the possibility that PLC was saturated with CaM under assay conditions. As the relationship between CaM antagonist stimulation and enzyme concentration became clearer, it became apparent that the enzyme is not 141 saturated with inhibitor when small quantities are assayed. The addition of CaM under such conditions should inhibit PLC activity if CaM is the endogenous inhibitor; it does not. Nevertheless, the behavior of the CaM antagonists are difficult to reconcile with non—specific hydrophobic interactions with PLC or its substrate. A possible solution to this dilemma appears in the form of non-CaM regulatory proteins. Although CaM is the best characterized Ca2+-dependent regulatory proteins, others are also known, some structurally similar to CaM (e. g. troponin C) and some quite dissimilar. Many of these proteins interact with CaM antagonists in a Ca2+-dependent manner; this property has been exploited in their isolation via affinity chromatography. For instance, W-7 affinity resin has been used to purify S-lOO (Endo et al., 1981) a CaM-like protein originally isolated from nervous tissue. Moore and Dedman (1982) isolated several proteins that bind Ca2+ —dependently to phenothiazine and W-7 affinity columns; they later coined the name "calcimedins" to describe these. Unlike S-lOO and troponin C, these proteins differ considerably from CaM in size. Phenothiazine resins also bind S-lOO and troponin C (Marshak et al., 1981). Calmidazolium and compound 48/80 apparently have not been employed as affinity-resin substituents and their effects on these non-CaM proteins is unknown. It should be borne in mind that these antagonists are described as "highly specific" on the basis of the relative concentrations required to inhibit a CaM-dependent reaction and a similar CaM-independent reaction,e. g. the CaM-stimulated and basal activities of phosphodiesterase. Selectivity for CaM vs. related proteins is not implied. 142 Except for troponin C, the physiological functions of most of these proteins are unknown. The endogenous inhibitor of PIP2 PLC may be such a protein. Definitive testing of this hypothesis will entail purification of the PLC and inhibitor, followed by reconstitution. Laying aside the question of identity, there is also the question 2+ 2 of why a Ca -activated enzyme should have a Ca +-dependent inhibitor. Two possible functions (neither mutually exclusive nor exhaustive) that have suggested themselves are negative feedback and substrate switching. Negative feedback: The Ca2+-dependence of PIP2 PLC has been a topic of some controversy. Some laboratories find that PIP2 hydrolysis is not affected by variations in [Ca2+] in the physiological range, others report that it is (Berridge, 1984). These differnces may result from differences in tissues and/or assay techniques. In Vitro, at least, the ROS-derived PLC responds to 2+ 2+ physiogical Ca concentrations. As a Ca -stimu1ated enzyme one of whose products elevates cytoplasmic Ca2+, it would seem to possess a potential for positive feedback, allowing it to function as a “Ca2+ amplifier." The release of many Ca2+ ions could be triggered by a few. Such an amplifier would require "damping" -- some form of negative feedback to prevent runaway hydrolysis. A Ca2+-dependent inhibitor could provide such feedback. Substrate switching: The CaM antagonists promote hydrolysis of PIP2 and inhibit hydrolysis of PI. If these effects are produced by antagonizing a Ca2+ -dependent regulatory protein, that protein would presumably have the opposite effect: switching the substrate preference from PIP2 to P1. Just such a switch has been proposed in 143 agonist-stimulated phosphoinositide hydrolysis to account for production of diacylglycerol in excess of 1P3 (Majerus et al., 1986). PI hydrolysis appears to follow PIP2 hydrolysis. Diacylglycerol activates protein kinase C, which can produce comparatively long—term changes in cell function by protein phosphorylation. In ROS, protein kinase C phosphorylates rhodopsin, inhibiting its interaction with ROS G-protein (Kelleher and Johnson, 1986). This suggest a means of controlling light-sensitivity via protein kinase C. It is also possible that the inhibition is exerted not by a separated protein but by the PLC itself. Formation of less-active or inactive multimers from active monomers could also account for the results presented here if the multimerization were prevented by CaM antagonists. Purification of the PLC, which is being undertaken by another worker in our laboratory, should help decide this issue. Regardless of the identity of the endogenous inhibitor, the effects of CaM antagonists on PIP2 PLC are significant to anyone using these compounds as probes of CaM function, especially in whole cells or crude fractions. Additionally, the stimulation of PIP2 hydrolysis may play a role in the clinical effects of the anti-psychotic tranquilizers. Another worker in our laboratory (Frank Wilkinson) has shown that PIP2 PLC from bovine brain is also stimulated by CaM antagonists. 144 BIBLIOGRAPHY Abood, M. E., Hurley, J. B., Pappone, M.-C., Bourne, H. R. and Stryer, L. (1982) J. Biol. Chem. 257, 10540-10543 Baltzy, R., Buck, J. S., DeBeer, E. J., and Webb, F. J. (1949) J. Am. Chem. Soc. 71, 1301-1305 Baer, K. M., and Saibil, H. R. (1988) J. Biol. Chem. 263, 17-20 Banno, Y., Nakashima, S., Tohmatsu, T., Nozawa, Y. and Lapetina, E. G. (1986) Biochem. Biophys. Res. Comm. 140, 728-734 Benedikter, H., Knopki, G., and Renz, P. (1985) Z. Naturforsch. 40, 68-72 Bigay, J., Deterre, P, Pfister, C. and Chabre, M. (1985) FEBS Letters 191, 181-185 Bradford, M. M. (1976) Analyt. Bioch. 72, 248-254 Bronner, C., Wiggins, C., Monte, C., Marki, F., Capron, A., Landry, Y., and Franson, R. C. (1987) Biochim. Biophys. Acta 920, 301-305 Brown, J. E., Rubin, J. L., Ghalayini, A. J., Tarver, A. P., Irvine, R. F., Berridge, M. J., and Anderson, R. E. (1984) Nature 311, 160-163 Brown, J. E., Blazynski, C., and Cohen, A. I. (1987) Biochem. Biophys. Res. Comm. 146, 1392-1396 Brown, J. E., and Rubin, L. J., (1984) Biochem. Biophys. Res. Comm. 125, 1137-1142 Chafouleas, J. G., Bolton, W. E., Hidaka, H., Boyd, A. E., and Means, A. R., (1982) Cell 28, 41-50 Cheung, W. Y., Lynch, T. J., and Wallace, R. W. (1978) Adv. Cyc. Nucl. Res. 9, 223—251 Cockcroft, S. (1987) Trends Bioch. Sci. 12, 75-78 Cockcroft, S., and Taylor, J. A. (1987) Biochem. J. 241, 409-414 Craven, P. A., and DeRubertis, F. R. (1983) J. Biol. Chem. 258, 4814- 4823 Dawson, C. R., Drake, A. F., Helliwell, J., and Hider, C. (1978) Biochim. Biophys. Acta 510, 75-86 Deckmyn, H., Tu, S.-M., and Majerus, P. W., (1986) J. Biol. Chem. 261, 16553-16558 145 Devary, 0., Heichal, 0., Blumenfeld, A., Cassel, D., Suss, E., Barash, S., Rubinstein, C. T., Minke, B., and Selinger, Z. (1987) Proc. Natl. Acad. Sci. USA 84, 6939-6943 Downes, C. P., and Michell, R. H. (1981) Biochem. J. 198, 133—140 Eichberg, J., Zetusky, W. J., Bell, M. E., and Cavanagh, E. (1981) J. Neurochem. 36, 1868-1871 Endo, T., Tanaka, T., Isobe, T., Kasai, H., Okuyama, T., and Hidaka, H. (1981) J. Biol. Chem. 256, 12485-12489 Fein, A., Payne, R., Corson, D. W., Berridge, M. J. and Irvine, R. F. (1984) Nature 311, 157-160 Fein, A. (1986) Science 232, 1543-1545 Ghalayini, A. and Anderson, R. E. (1984) Biochem. Biophys. Res. Comm. 124, 503-506 Ghalayini, A. and Anderson, R. E. (1987) Invest. Opthalmol. Vis. Sci. 28 (ARVO abstracts supplement), 96 Gietzen, K. (1986) in Intracellular Calcium Regulation (Bader, H., Gietzen, K., Rosenthal, J., Ruedel, R., and Wolf, H. U., eds) pp. 405-423, Manchester University Press, Manchester Gietzen, K., Wfithrich, A., and Bader, H. (1981) Biochem. Biophys. Res. Comm. 101, 418-425 Gietzen, K., and Bader, H. (1985) in Calmodulin Antagonists and Cellular Physiology (Hidaka, H., and Hartshorne, D. J., eds.) pp. 347-362, Academic Press, Orlando Gietzen, K., Sanchez-Delgado, E., and Bader, H. (1983) IRCS Med. Sci. Libr. Compend. 11, 12-13 Godchaux, W., and Zimmerman, W. F. (1979) J. Biol. Chem. 254, 7874-7884 Hartshorne, D. J. (1985) in Calmodulin Antagonists and Cellular Physiology (Hidaka, H., and Hartshorne, D. J., eds.) pp. 3—12, Academic Press, Orlando Hayashi, F. and Amakawa, T. (1985) Biochem. Biophys. Res. Comm. 128, 954-959 Hepler, J. R., and Harden, T. K. (1986) Biochem. J. 239, 141-146 Hidaka, H., and Tanaka, T. (1983) Meth. Enzymol. 102, 185—194 Hidaka, H., Yasuharu, S., Tanaka, T., Endo, T., Ohno, S., Fujii, Y. and Nagata, T. (1981) Proc. Natl. Acad. Sci. USA 78, 4354-4357 146 Huque, T., and Bruch, R. C. (1986) Biochem. Biophys. Res. Comm. 137, 36-42 Inoue, H., Yoshioka, T., and Hotta, Y. (1985) Biochem. Biophys. Res. Comm. 132, 513-519 Kanaho, Y., Moss, J., and Vaughan, M. (1985) J. Biol. Chem. 260, 11493-11497 Katoh, N., Raynor, R. L., Wise, B. C., Schatzman, R. C., Turner, R. S., Helfman, D. M., Fain, J. N., and Kuo, J.-F. (1982) Biochem. J. 202, 217-224 Kelleher, D. J., and Johnson, G. L. (1986) J. Biol. Chem. 261, 4749—4757 Klee, C. B., Krinks, M. H., Manalan, A. S., Cohen, P., and Stewart, A. A. (1983) Math. Enzymol. 102, 227-224 Kohnken, R. E., Chafouleas, J. G., Eadie, D. M., Means, A. R. and McConnell, D. G. (1981c) J. Biol. Chem. 256, 12517-12522 Levin, R. M., and Weiss, B. (1978) Biochim. Biophys. Acta 540, 197-204 Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275 Majerus, P. W., Connolly, T. M., Deckmyn, H., Ross, T. S., Bross, T. E., Ishii, H., Bansal, V. S. and Wilson, D. B. (1986) Science 234, 1519-1526 Manning, D. R., Fraser, B., Kahn, R. A., and Gilman, A. G. (1984) J. Biol. Chem. 259, 749-756 Manthorpe, M., and McConnell, D. G. (1975) Biochim. Biophys. Acta 403, 438-445 Marshak, D. R., Watterson, D. M., and Van Eldik, L. J. (1981) Proc. Natl. Acad. Sci. USA 78, 6793-6797 Martin, T. F. J., Bajjalieh, S. M., Lucas, D. 0., and Kowalchyk, J. A. (1986) J. Biol. Chem. 261, 10041-10049 Mollay, C., Kreil, G., and Berger, H. (1976) Biochim. Biophys. Acta 426, 317-324 Monaco, M. E., and Woods, D. (1983) J. Biol. Chem. 258, 12705-12708 Moore, P. B., and Dedman, J. R. (1982) J. Biol. Chem. 257, 9963-9967 Nelson, G. A., Andrews, M. L., and Karnovsky, M. J. (1983) J. Cell Biol. 96, 730-735 147 Payne, R., and Fein, A. (1987) J. Cell Biol. 104, 933-937 Payne, R. Corson, D. W., and Fein, A. (1986) Biophys. J. 49, 31a Read, S. M., and Northcote, D. H. (1981) Analyt. Biochem. 116, 53-64 Rubin, L. J., Brown, J. E., and Poznansky, M. (1986) Biophys. J. 49, 30a Schatzman, R. C., Wise, B. C., and Kuo, J. F. (1981) Biochem. Biophys. Res. Comm. 98, 669-676 Schwertz, D. W., Halverson, J. B., Palmer, J. W., and Feinberg, H. (1987) Arch. Biochem. Biophys. 253, 388-398 Smith, C. D., Cox, C. C., and Snyderman, R. (1986) Science 232, 97-100 Stern, J. H., Knutsson, H., and MacLeish, P. R. (1987) Science 236, 1674-1678 Szuts, E. 2., Wood, S. F., Reid, M. S., and Fein, A. (1986) Biochem. J. 240, 929—932 Tanaka, T., Ohmura, T., Yamakado, T., and Hidaka, H. (1982) Mol. Pharmacol. 22, 408-412 Tsien, R. Y. (1080) Biochemistry 19, 2396-2404 Uhing, R. J., Jiang, H., Prpic, V. and Exton, J. (1985) FEBS Letters 188, 317-320 Vandenberg, C. A., and Montal, M. (1984) Biochemistry 23, 2347-2352 van Dop, C., Yamanaka, G., Steindberg, F., Sekura, R. D., Manclark, C. R., Stryer, L. R. and Bonner, H. R. (1984) J. Biol. Chem. 259, 23-29 Waloga, G. and Anderson, R. E. (1985) Biochem. Biophys. Res. Comm. 126, 59-62 Weiss, B., and Levin, R. M. (1978) Adv. Cyc. Nucl. Res. 9, 285-303 Wightman, P. D., Dahlgren, M. E., Hall, J. C., Davies, P., and Bonney, R. J. (1981) Biochem. J. 197, 523-526 Wilkinson, G. N. (1961) Biochem. J. 80, 324-332 Wilson, D. B., Connolly, T. M., Bross, T. E., Majerus, P. W., Sherman, W. R., Tyler, A. N., Rubin, L. J., and Brown, J. E. (1985) J. Biol. Chem. 260, 13496-13501 Yamanaka, G., Eckstein, F., and Stryer, L. (1986) Biochemistry 25, 6149-6153 148 Yoshikami, S., Robinson, W. E., and Hagins, W. A., (1974) Science 185, 1176-1179 Yoshioka, T., Inoue, H., Takagi, M., Hayashi, F., and Amakawa, T. (1983a) Biochim. Biophys. Acta 755, 50-55 Yoshioka T., Inoue, H., and Hotta, Y. (1983b) Biochem. Biophys. Res. Comm. 111, 567-573 SUMMARY 149 150 The experiments presented here demonstrate the presence of a complete pathway for synthesis and hydrolysis of phosphoinositides in isolated vertebrate photoreceptor outer segments, which had not previously been demonstrated. Radioactively labeled precursors were incorporated into phosphoinositides much more rapidly than into the major phospholipids. Phosphoinositides were also hydrolyzed preferentially. Phosphoinositide synthesis and hydrolysis were also powerfully influenced by divalent metal ions and spermine, a polyamine. Phosphoinositides are believed to play a central role in invertebrate phototransduction, and the presence of this pathway in bovine ROS suggests that they have a role in vertebrate photoreceptor function as well. However, regulation of phosphoinositide turnover in ROS appears to be unlike that in invertebrate photoreceptors, suggesting that the function of phosphoinositides in the two systems may be quite different. In invertebrates, hydrolysis of PIP2 is stimulated by light via a G-protein which activates PLC. No regulation by light of PIPZ PLC was observed in ROS, nor was any clear indication of regulation by G—protein obtained in experiments using GTP, non-hydrolyzable CTP analogs, fluoride ion, added ROS G-protein, or pertussis toxin. In contrast, evidence for a previously unexplored regulatory mechanism was obtained from experiments examining the effects of calmodulin antagonists on PLC activity. ROS were found to contain an endogenous inhibitor of PIP2 PLC, the effects of which were opposed in a Ca2+-dependent fashion by calmodulin antagonists. This inhibitor appeared not to be calmodulin per se, but may be a member of a recently-described class of Ca2+ -binding proteins. 151 Investigation of the function and regulation of PIP2 PLC in ROS will be complicated by the presence of multiple forms of the enzyme. PLC activity was found in soluble and particulate fractions derived from ROS, and both fractions contained multiple forms as revealed by ion exchange chromatography. Resolution and purification of the different forms of ROS PLC, along with the endogenous inhibitor, would seem to be the next logical step in the elucidation of the role of phosphoinositides in vertebrate vision. MICHIGAN STRTE UNIV. LI mlllilllilfl