Isolation and Characterization of the Inositol Cyclic Phosphate ...

THEJOURNAL OF BIOLOGICAL CHEMISTRY

Vol. 260, No. 25, Issue of November 5, pp. 1349613501 1985 Printed in i!S.A.

0 1985 by The American Society of Biological Chemists, Inc.

Isolation and Characterizationof the Inositol Cyclic Phosphate Products of Polyphosphoinositide Cleavageby Phospholipase C PHYSIOLOGICAL EFFECTS IN PERMEABILIZED PLATELETS AND LIMULUS PHOTORECEPTOR CELLS* (Received for publication, May 28, 1985)

David B. Wilson, Thomas M. Connolly, Teresa E. Bross, Philip W. MajerusS, William R.Sherman, Andrew N. Tyler, Leona J. Rubin, and JoelE. Brown From the Departments of Medicine, Biochemistry,Psychiatry, and Ophthulmology, Washington University School of Medicine, St. Louis,Missouri 63110

Cleavage of the polyphosphoinositides, catalyzed by phospholipase C purified from ram seminal vesicles, produces phosphorylated inositols containing cyclic phosphate esters (Wilson, D. B., Bross, T. E., Sherman, W. R., Berger, R. A., and Majerus, P. W. (1985)Proc. Natl. Acad. Sci. U. S. A. 82, 4013-4017). In the present study we describe the isolation and characterization of inositol l:%-cyclic4-bisphosphate and inositol l:2-cyclic 4,5-trisphosphate, the twocyclic phosphate products of phospholipase C catalyzed cleavage of phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate, respectively. We established the structures of these two cyclic compounds through '*O labeling of phosphate moieties, phosphomonoesterase digestion, and fast atom bombardment-mass spectrometry. We examined the physiological effects of these compounds in two systems: saponin-permeabilized platelets loaded with 4SCa2+and intact Limulus photoreceptors. Both inositol l:2-cyclic 4,5-trisphosphate and the noncyclic inositol 1,4,5-trisphosphate, but not inositol l:2-cyclic 4-bisphosphate, release 45Ca2+from permeabilized platelets ina concentrationdependent manner. Injection of inositol l:2-cyclic 4,5trisphosphate intoLimulus ventral photoreceptor cells conductance and a induces both a change in membrane transient increase in intracellular calcium ion concentration similar to those induced by light. We injected inositol 1,4,5-trisphosphate and inositol l:2-cyclic 4,5trisphosphateinto the same photoreceptor cell and found that the cyclic compound is approximately five times more potent than the noncyclic compound in stimulating a conductance change. We speculate/that inositol l:2-cyclic 4,5-trisphosphate may functiod as a / second messenger in stimulatedcells.

Stimulation of a variety of cells with the appropriate agonistsresults in the phospholipase C mediated cleavage of *This research was supported by National Institutes of Health Grants HLBI 14147 (Specialized Center for Research in Thrombosis), HL16634, and T32-HL 07088 (to P. W. M.); NS 05159 (to W. R. S.); EY 05166 and EY 05168 (to J. E. B.). The Mass Spectrometry Facility (RR00954) at Washington University is a Core Facility of the Diabetes Research and Training Center (AM 20579). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ To whom reprint requests should be addressed at: Division of Hematology-Oncology, Washington University School of Medicine, 660 South Euclid, St. Louis, MO 63110.

phosphatidylinositol (PtdInsl)andthe polyphosphoinositides, phosphatidylinositol 4-phosphate (PtdIns-4-P)and phosphatidylinositol 4,5-bisphosphate (PtdIns-4,5-Pz) (for reviews, see Refs. 1-4). Agonist-induced phosphoinositide metabolism generates a number of cellular messenger molecules. Phosphoinositide-derived diglyceride can serve to activate protein kinase C (5) or can be further metabolized by diglyceride and monoglyceride lipase activities to yield free arachidonic acid for icosanoid synthesis (6, 7). Another product of PtdIns-4,5-Pz hydrolysis by phospholipase C, inositol 1,4,5trisphosphate (IP3), is a Ca'+-mobilizing agent in a number of experimental systems (4; 8-16). Dawson et al. (17) have shown that cleavage of PtdIns by phospholipase C produces amixture of two water-soluble products: inositol 1-phosphate (IP1) and inositol l:2-cyclic phosphate (cIP,). We recently showed that all three phosphoinositides are hydrolyzed by the same phospholipase C (18).Therefore, by analogy to the PtdInsreaction, cyclic 1:2phosphate esters might be anticipated among the phospholipase C cleavage products of PtdIns-4-P and PtdIns-4,5-Pz. Recently, we employed "0 incorporation into the phospholipase Creaction products to demonstrate that cyclic phosphate esters are present among the water-soluble products of polyphosphoinositide cleavage by phospholipase C (19). We concluded that cleavage of PtdIns-4-P produces inositol 1,4bisphosphate (IPz),and the cyclic product inositol l:2-cyclic 4-phosphate (cIPz),while cleavage of PtdIns-4,5-Pz produces both inositol 1,4,5-trisphosphate (IP3) and the cyclic product inositol k2-cyclic 4,5-triphosphate (cIP3). These two cyclic inositol polyphosphates are candidatesfor second messengers in cells. In the present study we have isolated cIPz and cIP3 and examined the physiological effects of these compounds in two systems: saponin-permeabilized platelets and intactLimulus photoreceptor cells. EXPERIMENTAL PROCEDURES

Materials-Phospholipids, ATP, creatinine phosphate, and creatine phosphokinase were purchased from Sigma. Inositol l:2-cyclic phosphate (cyclohexylamine salt) was provided by Merck Sharp and The abbreviations used are: PtdIns, phosphatidylinositol; PtdIns4-P, phosphatidylinositol 4-phosphate; PtdIns-4,5-P2, phosphatidylinositol 4,5-bisphosphate; IP1, myo-inositol l-phosphate; IP2,myocIP1, inositol l,4-bisphosphate; IP3, myo-inositol1,4,5-trisphosphate; myo-inositol 1:2-cyclic phosphate; cIP2, myo-inositol 1:e-cyclic phosphate 4-monophosphate; cIP3, myo-inositol l:2-cyclic phosphate 4 5 bisphosphate; GroPIns-4,5-P2, glycerophosphorylinositol 4,5-bisphosphate; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; EGTA, ethylene glycol bis(8-aminoethy1ether)-N,N,N',N'-tetraacetic acid; FAB, fast atom bombardment; HPLC, high performance liquid chromatography.

13496

Isolation and Characterizationof the Inositol Cyclic Phosphate Products Dohme, Rahway, NJ. H;80 (>95% enriched) was obtained from Monsanto/Mound Laboratories, Miamisburg, OH. Aequorin was obtained from Dr. J. R. Blinks, Mayo Foundation, Rochester, MN. ["PI Orthophosphoric acid, my~-[~H]inositol, and %a'+ were obtained from New England Nuclear. [32P]Inositol1,4-bisphosphate and ["PI inositol 1,4,5-trisphosphate were prepared from labeled erythrocyte ghosts using the method of Downes et al. (20). A sample of ox brain IP3 for use in the Limulus photoreceptor studies was a gift of Dr. R. F. Irvine, ARC Institute, Babraham, Cambridge, United Kingdom. [32P]PtdIns-4-P and [32P]PtdIns-4,5-Pz were isolated from labeled human erythrocytes (21) by column chromatography on neomycinglass beads (22,23). [3H]PtdIns,[3H]PtdIns-4-P, and [3H]PtdIns-4,5Pz were isolated from [3H]inositol-labeledLM cells (24) by the same methods (22, 23). Phospholipase C enzymes were purified from ram seminal vesicles as described previously (18,24). Human plateletIP35-phosphomonoesterase was purified by the method of Connolly et al. (25). Small unilamellar phospholipid vesicles containing labeled phosphoinositide and phosphatidylethanolamine were prepared as described (26). Phospholipase C Reaction-A typical reaction mixture contained vesicles of phosphoinositide:phosphatidylethanolamine (1:0.4, mol/ mol; 300 p~ [32P]phosphoinositide(100-300 cpm/nmol by Cerenkov counting)), 10 mM Tris acetate, pH 5.3, 37 mM NaCl, 1 mM CaCl2, and 2 pg of phospholipase C in a total volume of 1 ml. The mixture was incubated 15 min a t 37 "C, and thereaction was then terminated by the addition of 5 ml of chloroform/methanol (l:l, v/v) and 1.5 ml of HzO. The upper phase of this reaction mixture was removed and dried in a shaker evaporatorafter theaddition of 10 pl of 1M NaHC03. The dried material was dissolved in 1 ml of water and the reaction products were separated by HPLC as described below. HPLC of Inositol Phosph~tes-~~P-labeled phosphorylated inositols or the tritiated compounds were resolved on a Whatman Partisil 10 SAX column (Cobert Associates, St. Louis, MO). The elution scheme consisted of isocratic elution with 50 mM ammonium formate, pH 6.25, for 10 min, followed by a linear gradient from 50 mM to 2.7 M ammonium formate, pH 6.25, over the next20 min, and thenisocratic elution a t 2.7 M ammonium formate, pH 6.25, for 20 min. The flow rate was 1 ml/min and 1-ml fractions were collected and assayed for radioactivity by Cerenkov counting. Under these conditions, the following retention times were obtained inositol, 4 min; cIPl, 8 min; IPl, 23 min; Pi, 25 min; cIP2, 26-27 min; IP2, 30 min; cIP3, 36 min; IP3, 38-39 min. Desalting-HPLC fractions containing cyclic or noncyclic inositol phosphates were desalted on a 1 X 90-cm Sephadex G-10 (Sigma) column equilibrated with 7% 1-propanol in water. The flow rate was 1.8 ml/min and 2-ml fractions were collected. Fractions containing inositol phosphates were pooled, lyophilized, and dissolved in either water or 100 mM Hepes, pH 7.4,200 mM KC1 at concentrations ranging from 0.2 to 10 mM. Fast Atom Bombardment (FAB)-Mass Spectrometry-FAB-mass spectra (27) were obtained using a VG ZAB 3F mass spectrometer equipped with a FAB source and operated in the double focusing mode. Instrument conditions were as follows: primary beam of Xenon 8 keV energy, 1 mA emission current. The ion source was operated a t 8 keV energy and instrument mass resolution was about 1000. Samples were dissolved in 2-3 p1 of water and applied to the FAB probe tip ontowhich 2-3 pl of glycerol had previously been deposited. Spectra were obtained by scanning at a rate of 100 s/decade from m/ z 800 and recorded on UV oscillographic paper. 45Ca2+Release from Permeabilized Platelets-The 45Ca2+release assay was developed by Dr. Lawrence Brass, University of Pennsylvania. Human platelets were washed as described elsewhere (28) and resuspended just prior to use a t a final concentration of 1.4 X IO9 cells/ml in 137 mM NaC1,2.7 mM KCl, 1mM MgC12, 3.3mM NaH2P04, 5.6 mM glucose, 20 mM Hepes, pH 7.4. A 0.8-ml portion of the platelet suspension was added to 2.4 ml of a solution containing 160 mM KCl, 5.3 mM MgC1, 3.3 mM ATP, 26.7 mM Hepes, pH 7.1,0.67 mM EGTA, 6.7 mM creatine phosphate, and 13 units/ml creatinephosphokinase. The platelets were permeabilized by the addition of 56 pl of a 1 mg/ ml saponin solution (final saponin concentration 5 pg/108 platelets) followed by incubation a t room temperature for 2 min. The permeabilized platelets were loaded with 45CaZ+ by the addition of 3.2 p1 of a 0.15 M solution of%aC12 (80 pCi/pmol) followed by incubation at room temperature for 20 min. The platelet suspension was then diluted with 9.6 ml of 120 mM KCl, 4 mM MgCl,, 0.5 mM EGTA, and 20 mM Hepes, pH 7.1. Ca'+ release was determined by adding 0.5 ml of the diluted suspension to tubes containing varying amounts of 30

13497

p~ stock solutions of I&, IP2, cIP3, or cIP, followed by incubation for 3 min at room temperature. The reactions were stopped by dilution with 4 ml of 120 mM KC1,4 mM MgC12,0.5 mM EGTA, 20 mM Hepes, pH 7.1, and immediate filtration on 0.45-pm HAWP Millipore filters previously soaked in 1%bovine serum albumin. The filters were washed twice with 4-ml portions of 120 mM KCl, 4 mM MgCl2, 0.5 mM EGTA, 20mM Hepes, pH 7.1, and then counted in 10 ml of Scintiverse I scintillation mixture to determine the &Ca" remaining in the cells. The difference in 45Ca2+retained in control platelets versus those treated with 1 ~ L MA23187 was considered to be 100% release. The &Ca released by inositol phosphates is expressed as a percentage of the *Ca'+ released by A23187. Intracellular Injection into Limulus Photoreceptors-Ventral rudimentary eyes (29) were dissected from Limulus polyphemus and prepared for electrophysiology (30). Single photoreceptors were impaled with double-barreled (0 glass) micropipettes that were used both to measure membrane voltage and to pressure inject intracellularly solutions outof both barrels. By this technique, the effects of injection of two soutions could be made at thesame locus in a cell; thus spatial differences in effects of injection (31,32) did not confound the findings. The relative sizes of injections out of both barrels were compared optically (at 940 nm) by the technique of Corson and Fein (33). The volumes of individual injections were estimated to vary from about 10"l to less than 10"' liters; the volume of a typical Limulus ventral photoreceptor is about 5 X 10-l' liters. Values are given for the concentrations of substances in the electrode-filling solutions. In some experiments the cell was impaled with a second micropipette filled with 2 M KC1; this pipette was used to pass current in order to voltage-clamp a cell (34,35). Changes in calcium ion concentration within Limulus photoreceptors were monitored by aequorin (36). The aequorin luminescence was detected by a photomultiplier tube (R105UH, Hamamatsu) operated in the photon-counting mode at dry ice temperature. Other Methods-The concentration of inositol phosphate solutions was determined by phosphate analysis (37). In some experiments cIP3 and IP3were isolated from the same phospholipase C reaction mixture and used for injection into photoreceptors thereby allowing for estimation of the concentrations of the two compounds by measurement of radioactivity. Protein was measured by a Coomassie Blue protein assay (Bio-Rad).

RESULTS

In a previous study (19) we demonstrated two water-soluble products of phospholipase C cleavage of PtdIns-4-P, cIP2, and IP2. The cyclic and noncyclic products were not fully resolved by anion exchange HPLC at pH 4.3. Because cyclic inositol phosphates are unstable at low pH, we developed an HPLC method that operates at pH 6.25. This method resolves the cyclic and noncyclic inositol phosphate products of phospholipase C cleavage of all three polyphoinositides. A chromatogram of the water-soluble products of phospholipase C-mediated cleavage of phospholipid vesicles containing [32P]PtdIns4-P and [32P]PtdIns-4,5-Pzis shown in Fig. 1A. Four radioactive components are resolved on this chromatogram. Two of the components co-migrate with ["P]IP2 and [32P]IP3 isolated from labeled erythrocyte ghosts (Fig. 1C). The other two represent [32P]cIPzand [32P]cIP3based on criteria described below. A chromatogram similar to that inFig. 1A was obtained when [3H]inositol-labeled PtdIns-4-P and PtdIns4,5-P2 weresubstituted for the 32P-labeledlipids in the phospholipase C reaction; this result suggests that all four peaks contain inositol. When the individual products were desalted and re-chromatographed, they eluted assingle peaks with the same retention times. In thepresence of acid, inositol cyclic phosphate esters are rapidly hydrolyzed to phosphomonoesters. Treatment of the p,roducts of the phospholipase C reaction with acid prior to HPLC converted the [32P]cIPzand [32P]cIP3 products into their noncyclic counterparts (Fig. 1B).When acid hydrolysis of cyclic phosphate esters is carried out inHz180,the resultant phosphomonoesters contain a single atom of "0. The HPLC fractions corresponding to [32P]cIPz, [32P]IPz, [32P]cIP3,and

13498

Isolation and Characterization of the Inositol Cyclic Phosphate Products TABLEI "0 incorporation into the phosphate groups of cyclic and noncyclic inositol phosphates HPLC fractions corresponding to [32P]cIP2, [32P]IP2, [32P]cIP3, and [32P]IP3 were desalted, resuspended in 50 mM Hepes, pH 7.5,200 mM KCl, and tested for their ability to incorporate "0 into phosphate groups in response to treatment with acidified H2"O. Samples of Pi and chemically synthesized cIPl were also treated in this fashion. The samples (approximately 3 nmol each) were exposed to acidified HZ1'O, lyophilized, and treated with alkaline phosphatase as described (19). The resultant inorganic phosphate was converted to the tertbutyldimethylsilyl derivative(32) for gas chromatography/mass spectroscopy using OV-17 resin(19). "0 enrichment of phosphate groups was measured using selected ion monitoring where the ratio of the m/z 385 peak to the m/z 383 peak was compared. The values shown represent the average of duplicate determinations. 0.1554

Compound

mlz 385:383

Pi CIPl cIP2 IPZ C1P.q IPS

0.4522 0.2507 0.1562 0.2200 0.1585

Samples of [32P]cIP3and IP, were further characterized by FAB-mass spectrometry, a technique useful for the determination of the molecular weights of non-volatile compounds. The negative ion mass spectra for IP3 andcIP, are shown in Fig. 2. The positive ion spectra were of inferior quality and are not presented. The sample of IP, analyzed was prepared from erythrocyte ghosts and was the free acid. The mass spectrum of IPS contained a peak at m/z 419, which corresponds to the mass of a compound whose molecular weight is FRACTION NUMBER 420 minus a proton. There was also a small peak at m/z 441 FIG. 1. Partisill0 SAX HPLC of the water-soluble products which corresponds to themonosodium salt of that compound of ["P]PtdIns-4-P and [s2P]PtdIns-4,5-Pacleavage by phos- minus a proton. These molecular weights agree with the actual pholipase C. Small unilamellar vesicles containing[32P]PtdIns-4,5- molecular weightof IPS and itssalt. The sample of [32P]cIP3 Pz (100 cpm/nmol), [32P]PtdIns-4-P(150 cpm/nmol), and phospha- analyzed was prepared by HPLC chromatographyfollowed by tidylethanolamine in a molar ratio of (1:0.25:0.4) were prepared and gel filtration on Sephadex G-10 to remove excess salt. The treated with phospholipase C as described under "Experimental Pronegative ion mass spectrum of cIP, contained a peak at 401, cedures." The water-soluble products of the reaction were divided into two portions. One portion was subjected to HPLC (panel A ) . which corresponds to the mass of a compound whose molecThe otherportionwastreatedwith 1 N HC1 for 1min,frozen, ular weightis 402 minus a proton. Peaks at 423 and 445 lyophilized, resuspended in HzO, and then subjected to HPLC (panel represent mono- and disodium salts of this compound. The B). Panel C shows the chromatogram for a mixture of erythrocyte cIP3 spectrum has a higher background,indicative of a lower [32P]IP2(eluting at fraction 30) and[32P]IP3(eluting at fraction38). sample concentration. The molecularweights obtained by The ordinate is cpm of Cerenkov radiation. FAB-mass spectrometry agree with the actual molecular [32P]IP3were desalted and tested for their ability to incorpo- weights of cIP3 and its salts. The m/z 419 peak found in the rate "0 into phosphate groups in acidified H2"0. The "0 IP, spectrum was not present in the cIP3; this finding indicontent of the phosphomonoesters was measured after alka- cates that thecIP3 sample was not contaminated with IP3. Samples of [32P]IP3and [32P]cIP3(derived from the phosline phosphatase treatment and conversion of the resultant pholipase C cleavage of the same preparation of [32P]PtdInsinorganic phosphate to a volatile tert-butyldimethylsilyl derivative for gas chromatography/mass spectroscopy (38). "0 4,fi-p~)were treated with IP3-5-phosphomonoesterasepurified enrichment was determined by selected ion monitoring of the from human platelets, and theproducts were characterized as m/z 385 and 383 peaks of the Tris-tert-butyldimethylsilyl/ described in a separate report (39). Treatment of [32P]IP3 phosphate spectrum (19). In the absence of enrichment, the with the 5-phosphomonoesteraseproduced compounds that ratio of the m/z 385 peak to them/z 383 peakis about 0.1555 co-migrated with [32P]IP2and ,'Pi. Treatment of [32P]cIP3 (19). Ratios greater than 0.1555 are a measure of "0 enrich- produced compoundsthat co-migratedwith [32P]cIP2and 32Pi. ment. The m/z 385:383 ratios obtained for [32P]IP2,[32P]cP12, These results confirm that theHPLC peaks identified as cIP2 [32P]IP3,and [32P]cIP3,are shown in Table I along with the andcIP3 differ by a phosphate moietyin the 5-position. ratios for unenriched Pi and synthetic cIP1. The [32P]cIP2and Moreover, these findings suggest that cIP3 may be metabo[32P]cIP3samples showed incorporation of "0 into phosphate, lized in uiuo in a fashion analogous to IP3, which appears to while IP2 and IP3 samples did not. The m/z 385:383 ratio was be degraded by a 5-phosphomonoesterase (9, 20, 25, 40-42). found to decrease in theorder cIPl >cIP2 > cIP3. This reflects The metabolism of cIP3 by cell extractsis detailed in a dilution of the "0-enriched 1-positionphosphate by the unen- sep&ate report (39). In seven experiments, cIPz comprised an average of 43% riched 4- and 5-position phosphates that arealso released by alkaline phosphatase treatment. These data confirm the ex- (range 28-58%) of the total water-soluble products of phosistence of an acid labile cyclic phosphate ester in theproducts pholipase C-mediated cleavage of labeled PtdIns-4-P, based on chromatograms of the reaction mixtures. In nine experiidentified as cIP2 andcIP3.

Isolation and Characterization of the Inositol Cyclic Phosphate Products

[5G-H]-

["HIm/Z 401

FIG. 2. Negative ion FAB-maw spectrum of [3zP]cIP3 (9 nmol) (right)and IP3 (20 nmol) (left)dissolved in glycerol. [5G-H]- pentamer

13499

[MtNa-ZHIm/z 423

\

of glycerol. Markings under upper trace indicatemassrange 400-450. Lower trace, XI; upper trace, X10.

-

I

L

Intracellular injection of IP, into intact Limulus ventral photoreceptors has been shown to induce electrical responses generated across the membrane mimicking those induced by IPS light (31, 32). Also, injection of IP, has been shown to induce cIP,--an increase in cytoplasmic Ca2+ as assessed by aequorin 'P* x luminescence (16). As shown in Fig. 4 A , injection of cIP3 also CIpZ 0 induces an inward current similar to that induced by light. 50 Brief pressure injection of cIP3 elicited a depolarization that resembles the response of cells to illumination (Fig. 4, upper tracings) and an increase in intracellularCa2+as measured by aequorin luminescence (Fig. 4, middle tracing). The change in membrane voltage induced by microinjection of solutions containing equal concentrations of cIP, and IP3 is shown in Fig. 4C. The depolarizations in response to cIP3 were consistently larger than those induced by IP3injections of approximately the same volume. Fig. 4 0 shows the tracing for a voltage clamp experiment in which a 0.2 mM solution of cIP3 was compared to a 1.0 mM solution of IP3 using a doublebarrel micropipette. Nearly equal inward currents were induced by equal size injections of these two solutions, suggestFIG. 3. Release of ",aa+ from saponin-permeabilized plateing that cIP3 isabout five times as potent as IP3in inducing lets inresponse to IP3 and cIP3. The IP3 used in these experiments was from erythrocyte ghosts. The &Ca2+release assays were per- a response inthese cells. This difference in potency was formed as described under "Experimental Procedures." Data is consistently observed in several experiments in which various graphed as a percentage of the &Ca2+release elicited by 1p~ calcium amounts of IP, and cIP3were injected. Intracellular injection ionophore A23187. Each point represents the mean & S.D. of 7-14 of 1 mM c1P2 or 1mM cIPl into Limulus photoreceptor cells determinations using plateletsfrom multiple donors, The permeabil- induced no consistent changes in membrane voltage or current ized platelets accumulated an average of 3600 cpm of &(=a2+ prior to stimulation. In the absence of ATP the permeabilized cells accumu- (data notshown). MOBILIZATION OF CALCIUM IN PLATELETS

-

4.0"

released lated only about 300 cpm over this same interval. The 45Ca2+ by A23187 averaged 1700 cpm.

DISCUSSION

In this study we have isolated cIPz and cIP,, the cyclic ments, cIP, comprised an average of 36% (range 19-55%) of phosphate productsof polyphosphoinositide cleavage by phospholipase C, and confirmed the structure of these compounds the water-soluble products of PtdIns-4,5-P2 cleavage. Physiological Actionsof Inositol Cyclic Phosphates-We also through "0 labeling, phosphomonoesterase digestion, and examined the physiological effects of CIPZand cIP3 in two FAB-mass spectrometry. Additionally, we have shown that systems. A number of studies have demonstrated that addition one of these compounds, cIP3, elicits physiological responses of IP3 to permeabilized cells results in mobilization of Caz+ in two systems. In a previous report (19) we used "0 labeling of phosphate from intracellular stores (for review, see Ref. 4). Consistent with observations in other permeabilized cell systems, sa- to show that phosphorylated inositols containing cyclic phosphate esters were present among the water-soluble products ponin-treated platelets accumulate 45Ca2+into intracellular storage pools in an ATP-dependent fashion. As shown in Fig. of phospholipase C mediated cleavage of all three polyphos3, addition of IP, to these 45Ca2+-loadedcells resulted in a phoinositides. Based on "0 labeling we concluded that the concentration-dependent release of 45Caz+from intracellular ratio of cyclic to noncyclic product in the phospholipase C storage pools as shown in Fig. 3. At low concentrations (0.3- reaction differed for the three substrates. About 70% of the 0.6 PM) IP, mobilized Ca2+more efficiently than cIP3. Higher product of PtdIns cleavage was cyclic, whereas with PtdInsconcentrations of the two compounds induced release of sim- 4-P and PtdIns-4,5-P2 thesepercentages were 35% and 15%, ilar amounts of Caz+.The Caz+release induced by cIP3 does respectively. In the present study we find using HPLC that not require conversion to IP, prior to Ca" release. When 43% of the PtdIns-4-P cleavage product is cyclic and that [32P]cIP3(400 PM) was incubated with saponin-treated plate- 36% of the product of the PtdIns-4,5-Pz cleavage product is lets, there was no conversion to [32P]IP3.Addition of IP2 or cyclic. We are uncertain why the "0 method used previously cIPz at levels up to 4 PM did not cause 45Ca2+release from failed to detect the higher cIP3 levels that wenow obtain permeabilized platelets (Fig. 3). consistently by direct isolation of the compound.

Isolation and Characterization of the Inositol Cyclic Phosphate Products

13500

40

SM

L

n

n

I

I

I

I

4

a

4

L

0

&

Q

r

FIG. 4. Electrophysiological effects of intracellular injections of cIPs and IPS.Samples of IPS andcIP3 were dissolved in 50 mM Hepes, pH 7.5, 200 mM KCl. The bottom tracing of each set is a stimulus monitor. L denotes illumination, i denotes voltage-clamped current, V indicates membrane voltage, SM denotes stimulus monitor, LUM indicates aequorin luminescence monitor, filled arrows indicate cIP3 injections, and empty urrows indicate IP3 injections. A , membrane current recorded by a voltage clamp technique in response to aninjection of a 0.2 mM solution of cIP3. B, change in membrane voltage and aequorin luminescence elicited by intracellular injection of 0.2 mM cIP3. C, change in membrane voltage induced by injection of cIP, and IP3 using a doublebarreled electrode containing 0.4 mM cIP3 inone barrel and 0.4 mM IPSin theother. D, membrane current recorded by voltage clamp technique. The injections were made out of a double-barreled pipette containing 0.2 mM cIP3 in one barrel and 1.0 mM IP, in the other.

We find that cIP3causes release of Ca2+from permeabilized platelets, but that this compound is less effective than IP,. cIPz does not cause Ca2' release from these cells. These results are consistent with a study by Irvine et al. (12) who investigated the ability of various phosphorylated inositols to elicit Ca2+release in permeabilized cells. These workers found the order of potency of the phosphorylated inositols studied to be Ins-1,4,5-P3 > Ins-2,4,5-P3 2 GroPIns-4,5-Pz > Ins-4,5-P2 >> Ins-l,4-P~.They concluded that phosphate groups at the 4and 5-positions of inositol are required for Ca2' release, and that inaddition there isa requirement for a phosphate moiety on the opposite end of the molecule, with a preference for the 1-position. cIP, has a pair of phosphate groups on the 4- and 5-positions of inositol and another phosphate &esterified to the 1- and 2-positions, so it is not surprisingthat thismolecule elicits Ca2+release in permeabilized cells. Previous studies have shown that microinjection of IP, in Limulus photoreceptor cells induces changes in membrane conductance similar to those produced by light (31, 32). Ins4,5-P2 and GroPIns-4,5-P2 also elicit changes in membrane conductance inthese cells. Ins-1,4-P2 induces weaker responses than IPS;IP, and phytic acid do not elicit responses. In this study we have shown that intracellular injection of cIP, also induces changes in membrane conductance similar to those induced by light. Based on direct comparisons in the same cell, cIP3 appearsto be about five times as potent asIP3 in eliciting these responses. The finding that cIP3is less potent thanIPSin releasing Ca2+from permeabilized platelets may not be inconsistent with cIP3 being more potentin eliciting changes in membrane conductance in the Limulus system. In Limulus photoreceptor cells, light-induced changes in membrane conductance do not require an increase in

cytoplasmic Ca" (43). However, increases in cytoplasmic Ca2+ function in adaptation to light (36,43,44). cIP, may prove to be less potent than IP, in inducing increases in cytoplasmic Ca" in Limulus photoreceptor cells. Therefore, we speculate that IPS and cIP, may be distinct signal transducing molecules that have different functions. Several unresolved questions remain: 1)what levels of inositol cyclic phosphates are produced in stimulated cells? Previous studies measuring levels of IPSwould have overlooked cIP3because of its acid lability. 2) Are there factors which alter the proportion of IPS uersus cIP3 produced and could this in turn yield varied cellular responses? 3) What are the functions of IP2and cIP2 in signal transduction? In a separate paper (39) we present evidence that cIP3 is degraded in cell extracts by 5-phosphomonoesterase cleavage, the same mechanism by which IP, is degraded in cells. Acknowledgments-We thank Monita Peacock for technical advice, and Dr. Michael Grayson and theMcDonnell Douglas Corp. for permitting us to use the FAB-mass spectrometer. We also thank Dr. Lawrence Brass for providing us with his unpublished protqcol for 45Ca2+release from platelets. REFERENCES 1. Majerus, P. W., Wilson, D. B., Connolly, T. M., Bross, T. E., and Neufeld, E.J. (1985) Trends Biochem. Sci. 10,168-171 2. Berridge, M. J. (1984) Biochem. J. 220,345-360 3. Majerus, P. W . , Neufeld, E. J., and Wilson, D. B. (1984) Cell 3 7 , 701-703 4. Berridge, M. L.,and Irvine, R. F. (1984) Nature 312,315-321 5. Nishizuka, Y. (1984) Nature 308,693-697 6. Bell, R. L., Kennerly, D.A., Stanford, N., and Majerus, P. W . (1979) Proc. Natl. Acud. Sci. U. S. A. 76, 3238-3241

Isolation and Characterization

of the Inositol Cyclic Phosphate Products

7. Prescott, S. M., and Majerus, P. W. (1983) J. Biol. Chem. 258, 764-769 8. Streb, H., Imine, R. F., Berridge, M. J., and Schultz, I. (1983) Nature 306,67-69 9. Joseph, S. K.. Thomas, A. P., Williams, R. J., Imine, R. F., and Williamson; J. R. (1984) J.-Biol. Chem. 259,3077-3081 10. Burgess, G. M., Godfrey, P. P., McKinney, J. S., Benidge, M. J., Imine, R. F., and Putney, J. W., Jr. (1984) Nature 309,63-66 11. Berridge, M. J., Heslop, J. P., Imine, R. F., and Brown, K. D. (1984) Biochem. J. 222,195-201 12. Imine. R. F.. Brown. K. D., and Berridge, - . M. J. (1984) Biochem. J. 222,269-272 13. Prentki, M., Biden, T. J., Janjic, D., Imine, R. F., Berridge, M. J.. and Wollheim, C. B. (1984) Nature 309,562-564 14. Dawson, A. P., and Irvine, R. F . (1984) Biochem. Biophys. Res. Commun. 120,858-864 15. Suematsu, E., Hirata, M., Hashimoto, T., and Kuriyama, H. (1984) Biochem. Biophys. Res. Commun. 120,481-485 16. Brown, J. E., and Ruhin, L. J. (1984) Biochem. Biophys. Res. Commun. 125, 1137-1142 17. Dawson, R. M.C., Freinkel, N., Jungalwala, F. B., and Clarke, N. (1971) Biochem. J. 122,605-607 18. Wilson, D. B., Bross, T. E., Hofmann, S. L., and Majerus, P. W. (1984) J. Biol. Chem. 259,11718-11724 19. Wilson, D. B., Bross, T. E., Sherman, W. R., Berger, R. A., and Majerus, P. W. (1985) Proc. Natl. Acud. Sci. U. S. A. 82, 4013-4017 20. Downes, C. P., Mussat, M. C., andMichel1, R. H. (1982) Biochem. J. 203,169-177 21. Downes, C. P., and Michell, R. H. (1981) Biochem. J. 198, 133140 22. Schacht, J. (1978) J. Lipid Res. 19, 1063-1067 23. Palmer, F. B. (1981) J. Lipid Res. 22,1296-1300 24. Hofmann, S. L., and Majerus, P. W. (1982) J. Biol. Chem. 257, 6461-6469 25. Connolly, T. M., Bross, T. E., and Majerus, P. W. (1985) J. BWl. Chem. 260,7868-7874 '

13501

26. Hofmann, S. L., and Majerus, P.W. (1982) J. Bid. Chem. 257, 14359-14364 27. Barber, M., Bordoli, R. S., Elliot, G. J., Sedgwick, R. D., and Tyler, A. N. (1982) Chem.Anal. 54, 645A 28. Baenziger, N. L., and Majerus, P. W. (1974) Methods Enzymol. 31,149-155 29. Demoll, R. (1914) 2001.Juhrb. Abt. Anut. Ontog. Tiere 38, 443464 30. Millecchia, R., and Mauro, A. (1969) J. Gen. Physiol. 54, 310330 31. Brown, J. E., Rubin, L. J., Ghalayini, A. J., Tamer, A. P., Imine, R. F., Berridge, M. J., andAnderson, R. E. (1984) Nature 311, 160-163 32. Fein, A., Payne, R., Carson, D. W., Berridge, M. J., and Imine, R. F. (1984) Nature 311, 157-160 33. Corson, D. W., and Fein, A. (1983) Biophys. J. 44, 299-304 34. Brown, J. E., Harary, H. H., and Waggoner, A. (1979) J. Physiol. 296,357-372 35. Millecchia, R., and Mauro, A. (1969) J. Gen. Physiol. 54, 331351 36. Brown, J. E., and Blinks, J. R. (1974) J. Gen. Physiol. 64, 643665 37. Ames, B. W., and Dubin, D. T. (1960) J. Biol. Chem. 235, 769775 38. Mawhinney, T. P. (1983) J. Chromatogr. 257,37-44 39. Connolly, T. M., Wilson, D. B., Bross, T. E., and Majerus, P. W. (1985) J. Biol. Chern. 260, in press 40. Berridge, M. J., Dawson, R. M. C., Downes, C. P., Heslop, J. P., and Imine, ,R. F. (1983) Biochem. J. 212,473-482 41. Roach, P. D., and Palmer, F. B. St. C. (1981) Biochim. Biophys. Acta 661,323-333 42. Storey, D. J., Shears, S. B., Kirk, C. J., and Michell, R. H. (1984) Nature 312,372-376 43. Lisman, J. E., and Brown, J. E. (1975) J. Gen. Physiol. 66,489506 44. Lisman, J. E., and Brown, J. E. (1972) J. Gen. Physiol. 59, 701719