Isolation and Characterization of the Inositol Cyclic Phosphate ...

THEJOURNALOF BIOLOGICAL CHEMISTRY 0 1986 by The American Society of Biological Chemists,

Vol. 261, No. 1, Issue of January 5,!pP: 122-126, 1986 nnted m U.S.A.

Inc.

Isolation and Characterization of the Inositol Cyclic Phosphate Products of Phosphoinositide Cleavageby Phospholipase C METABOLISM IN CELL-FREEEXTRACTS* (Received for publication, July 11, 1985)

Thomas M. Connolly, David B. Wilson, Teresa E.Bross, and PhilipW. Majerus From the Division of Hematology-Oncology, Departments of Internal Medicine and Biological Chemistry, Washington University School of Medicine, St. Louis, Missouri 63110

The phosphoinositides are metabolized by phospholipase C inresponse to hormone or agonist stimulation in many cell types to produce diglyceride and watersoluble inositol phosphates. We have recently shown that the phospholipase C reactionproducts include cyclic phosphate esters of inositol. One of these, inositol 1, 2-cyclic 4,5-trisphosphate, is active in promoting Ca2+mobilization in platelets andin inducing changes in conductance in Limulus photoreceptors similar to those produced by light (Wilson, D. B., Connolly, T. M., Bross, T. E., Majerus, P. W., Sherman, W. R., Tyler, A., Rubin, L. J., and Brown, J. E. (1985) J. Biol. Chem. 260, 13496-13501. In the current study, we have examined the metabolism of the inositol phosphates. We find that both cyclic and non-cyclic inositol trisphosqhates are metabolized by inositol 1,4,5-trisphosphate 5-phosphomonoestera,to inositol 1,2cyclic bisphosphate and inositol l,4-bisphosphate, respectively. However, the apparent K , of the enzyme for thecyclic substrate is approximately 10-fold higher than for the non-cyclic substrate. These inositol bisphosphates are more slowly degraded to inositol 1,2cyclic phosphate and inositol 1-phosphate, respectively. Inositol 1,2-cyclic phosphate is thenhydrolyzed to inositol 1-phosphate, which in turn is degraded to inositol and inorganic phosphate by inositol l-phosphate phosphatase. The human platelet inositol 1,2cyclic phosphate hydrolase enzyme and a similar rat kidney hydrolase do not utilize the cyclic polyphosphate esters of inositol as substrates. These results suggest that theinositol cyclic phosphates and thenoncyclic ir?ositol phosphates are metabolized separately by phosphatases to cyclic and non-cyclic inositol monophosphates. The cylic monophosphate is then converted to inositol 1-phosphate by a cyclic hydrolase. We suggest that the enzymes that metabolize the inositol phosphates may serve to regulate cellular responses to thesecompounds.

Hormones and other agonists which promote Ca2+mobilization also effect the increased metabolism of the phosphoi-

* This research was supported by Grants HLBI 14147 (Specialized Center for Research in Thrombosis),HLBI 16634, and Training Grant T32 HLBI 07088 from the National Institutes of Health, and by National Institutes of Health Research Medical Scientist Service Award GM-07200 from the National Institute of General Medical Sciences. This is paper I1 of a series, Ref. 13 is number I. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solelyto indicate this fact.

nositides (1)phosphatidylinositol, PtdIns’; phosphatidylinositol 4-phosphate) PtdIns-4-P; and phosphatidylinositol 4,5bisphosphate, PtdIns-4,5-P,. This accelerated phosphoinositide metabolism is catalyzed by a phospholipase C enzyme. The reaction products of PtdIns, PtdIns-4-P, andPtdIns-4,5Pz breakdown are diglyceride and the water-soluble inositol phosphates: IPl,IP2,and IP3, respectively. IP3has been suggested to be a “second messenger’’ (2). Studies of both permeabilized cells and cellular microsomal fractions have shown that this inositol phosphate can promote Ca2+mobilization from non-mitochondrial cellular stores (Ref. 3, reviewed in Ref. 4). Further evidence of a role for IPSin cellular responsiveness was demonstrated in studies in which injection of IP, into intact cells of Limulus ventral photoreceptors (5, 6), sea urchin eggs (7), Xenopus oocytes (8),and salamander rods (9) in each case produced a physiological response similar to thatpromoted by the naturalstimulus. Dawson et al. (10) observed that the products of phospholipase C-mediated breakdown of PtdIns included both IP1 and inositol 1,2-cyclic phosphate, the latter containing a cyclic phosphate ester between the 1 and 2 carbons of the inositol ring. Wilson et al. (11)have shown that a single phospholipase C from ram seminal vesicles cleaves all three phosphoinositides to yield both the non-cyclic IP2and IP3 and the corresponding inositol cyclic phosphates, cIPz and cIP3 (12). We have recently isolated the inositol cyclic phosphates from the reaction products of phospholipase C-mediated cleavage of the polyphosphoinositides (13). We have shown that cIP3 is active in in uitro assays; it mobilizes Ca2+ from saponinpermeabilized human platelets and it evokes a response similar to that produced by light when injected into Limulus photoreceptors (13). These results suggest that cIP, is likely to be involved in cellular responses in uiuo. If IP3 and cIP, are mediators of cellular function, their metabolism could be a signal terminating step. Mechanisms for metabolism of non-cyclic inositol phosphates have been demonstrated in several tissues. The conversion of IP3 toIPz has been shown in crude homogenates of the salivary blow fly (14)) erythrocytes (15, 16), and rat liver (17-19), which also metabolizes IPz to a lesser extent (17-19). We have recently isolated a soluble enzyme from human platelets that specifically removes the 5-phosphate from IP3 to form IPz, an l The abbreviations used are:PtdIns, phosphatidylinositol; PtdIns4-P, phosphatidylinositol 4-phosphate; PtdIns-4,5-Pz, phosphatidylinositol 4,5-bisphosphate; IP1, myo-inositol l-phosphate; IPp,myoinostiol 1,4-bisphosphate; Ips, myo-inositol 1,4,5-trisphosphat.q cIP1, myo-inositol 1,2-cyclic phosphate; cIPz, myo-inositol 1,2-cyclicphosphate-4-bisphosphate; clP3, myo-inositol l,2-cyclic phosphate-4,5trisphosphate;HPLC, high performance liquid chromatography; MES, 2-(N-morpholino)ethanesulfonicacid.

122

Metabolism of Inositol Cyclic Phosphates

123

inositol phosphates as described previously (13). We incubated [32P]cIP3with the soluble fraction from human platelets and subsequently fractionated the reaction mixture by HPLC. We found three separate radioactive compounds,correspond.ing to unmetabolized cIP3 and reaction products, [32P]cIPz and [32P]Pi(Fig. 1B), demonstrating that cIP3 was metabolized by a 5-phosphomonoesterase.When ["P]IP3 was incubated with the same soluble platelet fraction, phosphatase activity was also observed in agreement with our previous studies on the platelet IPS 5-phosphomonoesterase (20). HPLC of this reaction mixture showed three compounds, corresponding to the substrate, IP3, and the products of a 5phosphomonoesterasereaction, IP, and Pi (Fig. lA). In these experiments on inositol trisphosphate metabolism we found EXPERIMENTAL PROCEDURES no evidence for a cyclic hydrolase which uses cIP3 as substrate. Materials-myo-2-[3H]Inosit~l,[3zP]phosphoric acid, and 13H] Thus, there was no conversion of cIP3 to IP3 (Fig. 1B) and phosphatidylinositol 4-phosphate were from New England Nuclear. there was also no IP, formed, which would result from the The Partisil SAX HPLC column, silicic acid, and Partisil SAX resins were from Whatman. Inositol 1,2-cyclic phosphate was provided by action of the 5-phosphomonoesterase on any IP3 that was Merck Sharp and Dohme, Rahway, NJ. All other materials are as formed. When the platelet-soluble fraction (enzyme) concenlisted or from Sigma or Fisher. tration was increased 10-fold and the incubation time was Preparation of Radiolabeled inositol Phosphates and Inositol Phos- lengthened 4-fold, noinositol 1,2-cyclic hydrolaseactivity that pholipids-32P- and tritium-labeledphosphoinositidesand thevarious acts on cIP3 was detected (data not shown). No inositol 1,2cyclic and non-cyclic inositol phosphates, prepared from reaction mixtures of phospholipase C with 32P-or tritium-labeled phospho- cyclic hydrolase activity acting on cIP3 was detected using inositides, were prepared as described previously (13). The isolated undialyzed solublefraction or in incubations using a platelet that platelets inositol phosphates were desalted by lyophilization or Sephadex G- membrane fraction. Thus, we found no evidence 10 chromatography and assayed for phosphorus (23) as previously convert cIP3 to IP3, rather it appears that cIP3 is metabolized described (13). We now use a 4 X 250-mm precolumn of silicic acid to cIPZ and Pi. (mounted before the injector loop) and a 4 X 50-mm Partisil SAX We next examined whether or not CIP,, the cyclic inositol guard column which are changed after every 6 runs to prevent of PtdIns-4-P, degradation of the column by the high concentrations of ammonium product of phospholipase C-mediated cleavage is metabolized by a platelet phosphomonoesterase or by an formate used to elute the inositol phosphates. Human Platelet Sonicate-Fresh human platelets were obtained inositol 1,2-cyclic phosphate hydrolase. We incubated [3H] from normal donors as previously described (24). The platelets were cIPz with the platelet-soluble fraction and again fractionated washed (24) and suspended at 5 X lo9 platelets/ml in 50 mM Tris, pH 7.8, 10 mM 2-mercaptoethanol, and 200 mM NaCl, and in the the reaction mixture by HPLC. No evidence for an inositol 1,2-cyclic hydrolasewas detected since no [3H]IP2was formed same buffer without 2-mercaptoethanol and at 8 X IOgplatelets/ml for the IPZ and cIPzmetabolism studies. The platelets were sonicated (Fig. 1D). The products of this reaction contained unmeta3 X 15 s on ice at 100 W with 20-s cooling intervals (Biosonik bolized substrate, [3H]cIPz, anda phosphomonoesterase resonicator, Bronville Scientific), centrifuged for 20 min a t 10,000 X g, action product, [3H]~IP1. There was also a small amount of and the supernatant and particulatefractions were separated. A free inositol found indicating some inositol 1-phosphatephosportion of each fraction was dialyzed against 2 X 1000 ml of 20 mM Tris, pH 7.8, 150 mM NaCl, 3 m M MgCl,, and 10 m M Z-mercaptoeth- phatase activity. Similar findings were also observed using anol. Both the dialyzed and undialyzed fractions were tested for [3H]IP2incubated with a soluble platelet fraction (Fig. 1C). apparently inert compound (20). IP, breakdown to inositol and Pi has been demonstrated in a number of different tissues. The IP, phosphatase involved is inhibited by lithium ions (21). Finally, Dawson and Clarke (22) demonstrated that crude extracts from many tissues can convert inositol 1,2cyclic phosphate to IP1. Thus, a potential pathway for the degradation of the inositol cyclic polyphosphates also exists. We now describe the metabolism of the inositol cyclic and non-cyclic phosphates by human platelets and ratkidney. We find that only cIPl is metabolized to its corresponding noncyclic inositol phosphate. The otherinositol cyclic phosphates as well as thenon-cyclic inositol phosphates are metabolized by phosphomonoesterases.

inositol 1,2-cyclic hydrolase and inositol phosphate phosphomonoesterase activities. The human platelet Ips 5-phosphomonoesterase was isolated from human platelets as previously described (20). Rat Kidney Inositol 1,2-Cyclic Phosphate Hydrolase-Extracts of rat kidney homogenate were prepared as described by Dawson and Clarke (22). Briefly, 1.7 g of rat kidney were homogenized in 15 ml of HzO and centrifuged for 20 min at 10,000 X g. The supernatant was dialyzed overnight against HzO. Precipitatedproteins were sedimented and the resulting supernatant used as a crude inositol 1,2cyclic hydrolase. Inositol 1,2-CyclicHydrolase and inositol Phosphate Phosphomonoesterase Reactions-Crude enzyme (1.6-120 pg of protein) was incubated a t 37 "C with 50 mM Tris-HC1, pH 7.8,5mM MgClz and the various radiolabeled cyclic or non-cyclic inositol phosphates (20-1200 p M ) in 25 pl. The assays were terminated by dilution with ice cold Hz0 to 200 pl followed by filtration on a CentriconTMmicroconcentrator a t 4 "C for 2 h (Amicon Corp.) to remove protein. The filtered sample was applied to a Partisil SAX HPLC column and fractions were monitored for radioactivity as previously described (13). For some of the phosphatase assays with the purified IP3 5-phosphomonoesterase, the release of [32P]Piwas measured following the isolation of the molybdate-phosphate complex (16,20,25). Protein was measured by the Bio-Rad assay kit with y-globulin as a standard. RESULTS

Metabolism of cZP3,ZP3, c Z P ~ZPz, cZP1,and ZP,by Human Platelet Extracts-We used HPLC to separate the various

. . .

.

.

.

.

.

FRACTION NUMBER

FIG. 1. Partisil SAX HPLCchromatography of reaction mixture of the human platelet-soluble fractionwith inositol phosphates. The radioactively labeled inositol phosphates were in25-pl final volume as described cubated with crude platelet extract ain under "Experimental Procedures." The substrate concentration, incubation times, and protein content, respectively, were the following: 270 p M [32P]IP3for 15 min, 13 pg (A);220 p M [32P]cIP3for 15 min, 13pg ( B ) ;120 pM, ['H]IPZ for 30 min, 24 pg (c);160p M [3H]cIPzfor 50 p~ [3H]IPl for 15 min, 24 pg (E);and 20 p~ 60 min, 95 pg (D); [3H]~IPI for 60 min, 11 pg ( F ) . The elution positions for standard inositol phosphates are shown in A and B and are applicable for all panels.


124

The phosphatase-mediated breakdown of IP2 was confirmed in studies in which [32P]IP2was the substrate and [32P]Pi released was measured. The amount of IP2 hydrolyzed by the platelet supernatantswas less than thatfor IP3, 0.9 uersus 20 nmol/min/mg protein, respectively. Similarly, the amount of cIP2hydrolyzed bythe platelet supernatantswas less than for cIP3, 0.4 uersm 3.4 nmol/min/mg, respectively. The dialyzed soluble fraction and dialyzed and undialyzed membranes showed no inositol 1,2-cyclic hydrolase activity which used cIP2 as a substrate. Indeed, dialysis markedly diminished the IP2phosphatase activity in the soluble fraction. When [3H]~IPl was incubated with the dialyzed plateletsoluble fraction, its breakdown by both a hydrolase and a phosphatase was detected. HPLC chromatography of the reaction mixture showed three compounds: cIP,, the substrate; IPl, theproduct of hydrolase activity; and inositol, the product of phosphomonoesterase hydrolysis of IP1 (Fig. 1F). The activity of the platelet inositol 1,2-cyclic hydrolase enzyme was 1.91 nM cIPl hydrolyzed per min/mg of protein. Use of less enzyme or shorter incubation time showed only cIP, and IP1 in the reaction mixtures suggesting that the hydrolase converts cIPl to produce IP1, as described by Dawson and Clarke (22), which is the substrate for the IPl phosphatase. Incubation of [,H]IP1 with the nondialyzed plateletsoluble fraction resulted in very little hydrolysis to inositol while incubation of [,H]IP1 with the dialyzed soluble fraction resulted in its breakdown to inositol by a phosphatase as shown in Fig. 1E. When cIPl was incubated with the plateletsoluble fraction dialyzed against Tris buffer, pH 7.8, more inositol was formed than when cIPl was incubated with the

nondialyzed soluble platelet fraction (data notshown). These results suggest that dialysis of the platelet-soluble fraction yields a more active inositol 1-phosphate phosphomonoesterase. Incubation of either IP, or cIPl with the platelet membrane fraction resulted in no apparent metabolism of either substrate (data not shown). Metabolism of CIP3 by Piutelet IP3 5-Phosphornonoesterme-The metabolism of cIP3 to cIP2, as shown in Fig. 1B, could be an important regulatory step in platelets whereby the active cIP, is inactivated by its conversion to cIP2. We next incubated the isolated platelet IP3 5-phosphomonoesterase, that removes the 5-phosphate from IP,, with cIP,. HPLC chromatographs of the reaction mixtures of the purified 5phosphomonoesterase with IP, and cIP3 are shown in Fig. 2, A and B, respectively. As expected, IP3 was metabolized to IP2 and Pi. The compound eluting in fraction 33 is an unknown component that was present in the IP3substrate. cIP3 was metabolized by this phosphatase to cIP2 andPi. Previous examination of substrates hydrolyzed by the 5-phosphomonoesterase showed that only IP3 is hydrolyzed (20). We next examined the rate of cIP, hydrolysis by the 5-phosphomonoesterase in comparison to that for IP3. The dependence of hydrolysis rate on IP3 or cIP3 concentration is shown in Fig. 3, A and B, respectively. The K,of the enzyme for IPS was 75 pM and for cIP3was approximately 1000 pM. A more thorough evaluation of the enzyme activity towards cIP3 was not possible due to the limited availability of radiolabeled substrate [32P]cIP3.These results, however, suggest that the cyclic inositol trisphosphate is hydrolyzed by the platelet 5phosphomonoesterase at a much slower rate than thatfor the corresponding non-cyclic compound. Metabolism of cIPl, cIP2, and cIP3 by Rat Kidney Cyclic Hydrolase-We were surprised that only cIPl was a substrate for the platelet inositol 1,2-cyclic hydrolase. Thus, we examined the ability of rat kidney cIP, hydrolase (22) to metabolize the inositol cyclic phosphates. Fig. 4A shows theHPLC chromatogram of the reaction mixture of therat kidney hydrolase incubated with [32P]cIP3.Although more enzyme

n

N

n

1100

400

2

2.0

-

v)

200

n

100 20

30

40

FRACTION NUMBER

FIG. 2. Partisil SAX HPLC of the products of [saP]IPs and by IPS 5-phosphornonesterase. 400 PM Ia2P] [32P]cIPs cleavage IP3and 500 PM [3*P]cIP3at 90 cpm/nmol were incubated20 min with 140 ng of purifiedIP, 5-phosphomonoesterasein 50 mM MES buffer, pH 6.5,3 mM MgC12 in a 50-p1 final volume. The reaction wasstopped by dilution to 500 pl with water and subjected to HPLC. Panel A is the IP3 reaction mixture, panel B is the cIP3 reaction mixture. The position of elution of standard compoundsis indicated.

60

120 180 240 303 360 420 480

[SI,PM

FIG. 3. Inositoltrisphosphate 5-phosphomonoester~activity as a function of substrate concentration. Assays were performed as described in the legend to Fig. 2, except the reaction was stopped by perchloricacid treatment followed by phosphate extraction as previously described (20). [3zP]IP3was the substrate in A and [3zP]cIP3was the substrate in B.


125

pholipase C converts PtdIns toboth IP, andinositol 1,2-cyclic phosphate. Wilson et al. (12) later showed that phospholipase metabolism of the polyphosphoinositides also resulted in the formation of the inositol cyclic polyphosphates, cIP, and cIP3. Recently we demonstrated (13) that these inositol cyclic phosphates can be isolated by HPLC chromatography, yielding sufficient quantities (albeit limited) to examine their physiological properties. cIP3 injection into Limulw photoreceptors evoked a physiological response similar to that produced by light (13). It is more potent than its counterpart non-cyclic inositol phosphate, IP,, in promoting this physiological response. cIP, is inactive as is IP2, suggesting the necessity of the trisphosphate structure for biological activity. cIP3 aswell as IP3 also stimulates 45Ca mobilization from saponin-permeabilized human platelets. Thus, the formation and physiological activity of these two water-soluble products of PtdIns-4,5-P2 breakdown can be demonstrated i n uitro. It is likely these compounds mediate cellular functions in vivo. If inositol 1,2-cyclic trisphosphate is a cellular second messenger, its enzymatic breakdown to inositol 1,2-cyclic 4-bisphosphate or to non-cyclic ionsoitol trisphosphate could be important regulatory steps. Also, the further breakdown of FRACTION NUMBER the inositol phosphates to yield free inositol is necessary to FIG. 4. Partisil SAX HPLC chromatography of the reaction provide a supply of free inositol to resynthesize phosphoinomixture of crude rat kidney hydrolase with the cyclicinositol phosphates. Assays wereperformed as described under "Experimen- sitides. A summary of our findings on the metabolism of cIP3 tal Procedures." The substrate concentration and incubation times and the other inositol phosphates is shown in Fig. 5. We find were 180 p M [32P]cIP3for 60 min ( A ) , 1200 p M [32P]cIP2for 60 min that cIP, is metabolized to cIP, which is further metabolized ( B ) ,and 620 p~ [3H]~IP1 for 15 min ( C ) . The crude hydrolase used to cIP, by soluble enzymes from human platelets. While cIPl was 16 pg of protein in A and B, and 1.6 pg of protein in C. The is broken down by a platelet, inositol 1,2-cyclic hydrolase, to retention times for standard inositol phosphates are shown in A and IPI, cIP3, and cIP2 are not metabolized by this enzyme. We are applicable for B and C. also used a preparation of inositol 1,2-cyclic hydrolase from was used than required to breakdown cIPl, no hydrolase rat kidney to demonstrate that the enzyme metabolized cIPl but not cIPz or cIP3. The non-cyclic inositol phosphates are activity towards [32P]cIP3was observed as no ["P]1P3was formed. Under these conditions, a small amount of 5-phos- metabolized by a series of soluble phosphomonoesterases to phomonoesterase activity was demonstrated as [32P]cIPzwas sequentially remove phosphates to convert IP3 to IP2 toIP,. formed. No [32P]cIPzwas observed in standard [32P]cIP3or The inositol cyclic phosphates aremetabolized independently in the reaction mixture when less hydrolase was used (data of the non-cyclic inositol phosphates. When IP, is formed not shown). When we incubated [32P]cIP2with the hydrolase, from cIP,, the pathways of metabolism of the cyclic and nonagain no [32P]IP2was formed, suggesting that cIP2 isalso not cyclic inositol phosphates converge. The fact that a specific enzyme, IP3 5-phosphomonoestera substrate for the kidney inositol 1,2-cyclic hydrolase. As ase (20), exists that can metabolize cIP, to cIP, supports the shown in the HPLCchromatograph of Fig. 4C, incubation of [3H]cIPlwith the hydrolase formed two products, L3H]IP1and hypothesis that cIP3 is formed in uiuo. In previous studies we [3H]inositol. The enzyme activity we observed, 370 nmol/ found that this enzyme could only metabolize IP3 (20). Bemin/mg, was comparable to thatfound by Dawsonand Clarke cause the amount of [32P]cIP3available was limited, we could (22). When we incubated cIPl under conditions comparable not thoroughly examine the properties of its metabolism by to those used to examine hydrolase activity on cIP2 and cIP3 the purified IPB 5-phosphomonoesterase. Thus, we do not (10-fold more enyzme and 4-fold longer incubation), the sub- know if its metabolism is enhanced by other assay conditions. strate was completely converted to free inositol. Thus, we An increase in assay pH to 7.8, as well as a decrease to 5.5, conclude that theinositol 1,2-cyclic hydrolase does not utilize did not greatly change the rate of cIP3 hydrolysis as compared cIPz or cIP3, as we could have detected reaction products if to thatfor IPS. The current results are not in conflict with the rapidly their rate of breakdown was as little as0.25% that of cIP,. growing evidence that IP3 is involved in cellular activation (4). A rapid decrease in PtdIns-4,5-P2 inresponse to hormone DISCUSSION stimulation has been documented in human platelets by sevThe recent discovery that IP,may be involved in thesignal eral investigators (26-31), as well as the appearance of IP, transduction of those hormones which are associated with Ca2+mobilization (4), hasfocused attention on the breakdown METABOLISM OF INOSITOLPHOSPHATES of the polyphosphoinositides and on the generation of their metabolites, the water-soluble inositol phosphates. While the PI=PI 4 P e p 1 4,5 P2 hormone specific changes in the phospholipid compounds fPhoy0&75eC/ have been extensively studied, much less is known about the cIP,-cIP2-cIP3 formation and subsequent metabolism of the inositol phosphates. We have examined the metabolism of the inositol /Uydrohse/ t + /5p'tose/+ phosphates formed by the phospholipase C-mediated breakIIP,-IP*-IP, down of the threephosphoinositides, PtdIns, PdtIns-4-P, and PdtIns-4,5-P2. FIG. 5. Proposed pathways of phosphoinositide and inositol Dawson et al. (10) firstreported that PtdIns-specific phos- phosphate metabolism. See text for discussion.

1

t

'

I

126


N. (1971)Biochem. J. 122, 605-607 (30,31).In these studies the inositol cyclic phosphates would not have been detected as the plateletswere extracted under 11. Wilson, D. B., Bross, T. E., Hofmann, S. L., and Majerus, P. W. (1984)J. Bwl. Chem. 269,11718-11724 acidic conditions, whichconverts the cycliccompoundsto 12. Wilson, D. B., Bross, T. E., Sherman, W. R., Berger, R. A., and their corresponding non-cyclic forms (12, 13).Also, although Majerus, P. W. (1985)Proc. Natl. Acad. Sci. U.S. A. 82,4013the mass ofPtdIns-4,5-P2is much less than that of PtdIns-44017 P and PtdIns in human platelets(ll),interestingly platelets 13. Wilson, D. B., Connolly, T. M., Bross, T. E., Majerus, P. W., Sherman, W. R., Tyler, A., Rubin, L. J., and Brown, J. E. contain 40- to60-fold morephosphataseactivityforthe (1985)J. Biol. Chem. 260, 13496-13501 inositol trisphosphates as compared to phosphatases for the 14. Berridge, M. J., Dawson, R. M. C., Downes, C. P., Heslop, J. B., inositol diphosphates and the inositol monophosphates. Thus, and Irvine, R. F. (1983)Bwchem. J. 212,473-482 the regulation of cellular IP3 and cIP3 levels may be critical 15. Roach, P. D., and Palmer, F. B. St. C. (1981)Biochim. Biophys. Acta 661,323-333 to normal cell function and awaits further examination. 16. Downes. C. P.. Mussat. M. C.. and Michell. R. H. (1982)Biochem. Finally, the existence of a pathway for the metabolism of J. 203, 169-177 ' the cyclic inositol phosphates in platelets supports the hy17. Sevfred. M.A.. Farrell. L. E.. and Wells. W. W. (1984)J. Bwl. pothesis that they areformed in vivo. Furtherstudieswill 269,13204-13208 ' determine the proportion of cyclic inositol phosphates formed 18. Storey, D. J., Shears, S.B., Kirk, C. J., and Michell, R. H. (1984) Nuture 312,374-376 in uiuo. 19. Joseph, S. K., and Williams, R. J. (1985)FEBS Lett. 180, 150-

&?rn:

154 20. Connolly, T. M., Bross, T. E., and Majerus, P. W. (1985)J. Bwl. Chem. 260,7868-7874 21. Hallcher, L. M., and Sherman, W. R. (1980)J. Biol. Chem. 256, REFERENCES 10896-10901 22. Dawson, R. M. C., and Clarke, N.(1972)Biochem. J. 127, 1131. Michell, R. H.,Kirk, C. J., Jones, L. M., Downes, C., and Creba, 118 J. A. (1981)Phibs. Trans. R. SOC.hnd. Bwl. Sci. 296,123- 23. Ames, B. W., and Dubin, D. T. (1960)J. Bwl. Chem. 236, 769137 775 2. Berridge, M. J. (1983)Bwchem. J. 212,849-858 24. Baenziger, N. L., and Majerus, P. W. (1974)Methods Enzyml. 3. Streb, H.,Irvine, R.F., Berridge, M. J., and Schulz, I. (1983) 31,149-155 Nature 306,67-69 25. Martin, J. B., and Doty, D. M. (1949)Anal. Chem. 21,965-967 4. Berridge, M. J., and Irvine, R. F. (1984)Nature 312, 315-321 26. Billah, M. M., and Lapetina, E. G . (1982)J. Bwl. Chem. 257, 5. Brown, J. E., Rubin, L. J., Chalayini, A. J., Tarver, A. P., Irvine, 12705-12708 R. F., Berridge, M. J., and Anderson, R. E. (1984)Nature 31 1, 27. Broekman, M. J. (1984)Bwchem. Bwphys. Res. Commun. 120, 160-163 226-231 6. Fein, A., Payne, R. Carson, D. W., Berridge, M. J., and Irvine, R. 28. Wilson, D. B., Neufeld, E. J., and Majerus, P. W. (1985)J. Biol. F. (1984)Nature 311, 157-160 Chem. 260,1046-1051 7. Whitaker, M., and Irvine, R. F. (1984)Nature 312,636-639 29. Agranoff, B.W., Murthy, P., and Sequin, E. B. (1983)J. Biol. 8. Oron, Y.,Dascal, N., Nadler, E., and Lupu, M. (1985)Nature Chem. 268,2076-2078 30. Watson, S. P., McConnell, R. T., and Lapetina, E. G . (1984)J. 313, 141-143 9. Waloga, G., and Anderson, R. E. (1985)Biochem. Biophys. Res. Bid. Chem. 259,13199-13203 31. Rittenhouse, S.,and Sasson, J. (1985)J.BioZ. Chem. 260,8657Commun. 126,59-62 8660 10. Dawson, R. M. C., Freinkel, N., Jungalwala, F. B., and Clarke,

Acknowledgment-We thank Cecil Buchanan for technical assistance.