sign in sign up
<<
Home , Inositol

Proc. Natl. Acad. Sci. USA Vol. 84, pp. 1206-1209, March 1987 Biochemistry Synthesis of 1,2-(cyclic)-4,5-trisphosphate (31P NMR spectroscopy/inositol cyclic trisphosphate/ phosphates/) RICHARD J. AUCHUS, SUSAN L. KAISER, AND PHILIP W. MAJERUS* Division of Hematology-Oncology, Departments of Internal Medicine and Biological Chemistry, Washington University School of Medicine, St. Louis, MO 63110 Contributed by Philip W. Majerus, November 7, 1986

ABSTRACT We have developed a method for synthesis of This goal has been achieved, and we describe here the inositol 1,2-(cyclic)-4,5-trisphosphate from inositol 1,4,5- synthesis and purification of inositol 1,2-(cyclic)-4,5-tris- trisphosphate using a water-soluble carbodiimide. We obtained phosphate. The cyclic product was converted to inositol 1-1.5 Amol of the inositol cyclic trisphosphate starting with 5 1,4,5-trisphosphate by acidification and to inositol 1,2-(cy- jAmol of inositol 1,4,5-trisphosphate. The cyclized product was clic)-4-bisphosphate with a specific 5-phosphomonoesterase. isolated by HPLC on Partisil SAX. The identity of the cyclic 31P NMR spectroscopy is consistent with the conclusion that product was verified by its to inositol 1,4,5- the structure of the compound is inositol 1,2-(cyclic)-4,5- trisphosphate in acid and by its conversion to 1,2-(cyclic)-4- trisphosphate. bisphosphate by a specific 5-phosphomonoesterase from plate- lets. We also identified the product by 31P NMR spectroscopy, which showed a peak at 17.2 ppm, characteristic of a five- MATERIALS AND METHODS membered cyclic phosphodiester ring, and peaks at 4.1 ppm Materials. The following chemicals and reagents were and 0.8 ppm, indicative of phosphomonoesters. This relatively obtained from the indicated suppliers: type I Folch fraction simple method for producing inositol 1,2-(cyclic)-4,5-trisphos- I from bovine brain, sodium m-periodate, 2-(N-morpho- phate will facilitate studies of the physiology of this compound lino)ethanesulfonic acid, and triethylamine, Sigma (before in signal transduction. use, the triethylamine was treated by passing it through a column of activity grade I alumina); , ammo- The breakdown of phosphatidylinositol phosphates by phos- nium formate, chloride, and Scinti Verse I, Fisher; pholipase C generates a variety of messenger molecules that methylamine gas in a pressurized canister, Fluka; [2- include several different inositol phosphates. The inositol 3H]phosphatidylinositol 4,5-bisphosphate, New England Nu- phosphate products of phospholipase C hydrolysis of phos- clear; 1,1-dimethylhydrazine and ethyl formate, Aldrich, phatidylinositol, phosphatidylinositol 4-phosphate, and Milwaukee, WI; Celite, Manville, Denver, CO; Dowex AG phosphatidylinositol 4,5-bisphosphate are mixtures of the 50W-X8 (100-200 mesh) hydrogen form, Dowex AG 1-X8 corresponding inositol phosphates and 1,2-(cyclic)phos- (200-400 mesh) formate form, Dowex AG 1-X4 (100-200 phates (1, 2). Recent studies using intact cells stimulated by mesh) chloride form, and Chelex 100 (200-400 mesh) sodium agonists also indicate that inositol 1,2-(cyclic)phosphate form, Bio-Rad; 1-ethyl-3-(3-dimethylaminopropyl)carbodi- (3-6) and inositol 1,2-(cyclic)-4,5-trisphosphate are found in imide hydrochloride, Pierce; activity grade I alumina, Uni- vivo (7). Inositol 1,4,5-trisphosphate has been shown to versal Scientific, Atlanta, GA; and absolute , U.S. mobilize ions from intracellular stores in many Industrial Chemicals, Tuscola, IL. types (8). Inositol 1,2-(cyclic)-4,5-trisphosphate also mobi- Inositol 1,4,5-[32P]trisphosphate was prepared according lizes calcium ions in permeabilized platelets (9) and 3T3 cells to the procedure of Downes et al. (12). D,L-Inositol 1,2- (10) with a potency similar to its noncyclic counterpart. (cyclic)phosphate was prepared as described (4). Inositol cyclic and noncyclic trisphosphates induce an in- Preparation of Inositol 1,4,5-Trisphosphate. Inositol 1,4,5- crease in membrane conductance when injected into Limulus trisphosphate was prepared by a method derived from one ventral photoreceptors, although the cyclic trisphosphate is suggested to us by R. F. Irvine (see ref. 10) as detailed below. five times more potent than the noncyclic compound in this from 1 g of Folch fraction I were extracted tissue (9). The relative proportion of cyclic vs. noncyclic twice with 14.5 ml of chloroform//0.1 M HCl, inositol phosphates formed in vivo is currently unknown. It 8:4:2.5. The two lower phases were combined and dried on is also uncertain whether these two compounds have similar a shaker evaporator without heating, and the extracted or different functions in various systems. residue was dissolved in 10 ml of chloroform Efforts to study the physiological effects and metabolism of and stored at -40C. inositol 1,2-(cyclic)-4,5-trisphosphate have been stalled by The phospholipids were deacylated by a modification of the low yields achieved by enzymatic production of the the procedure described by Clarke and Dawson (13). A tracer compound from phosphatidylinositol 4,5-bisphosphate. amount (3.5 x 106 cpm) of [2-3H]phosphatidylinositol 4,5- Structural characterization of inositol 1,2-(cyclic)-4,5- bisphosphate was added to the phospholipid extract, which trisphosphate has been difficult because ofthe scarcity of the was dried under a flow of nitrogen and redissolved in a material and the relative chemical instability ofinositol cyclic methylamine reagent made by bubbling monomethylamine phosphates (11). gas through a solution (40 ml) of methanol/water/1-, To obtain adequate amounts of inositol 1,2-(cyclic)-4,5- 20:15:5, until the volume increased to 65 ml (13). The trisphosphate for structure verification and for further phys- round-bottom flask with ground glass stopper was tightly iological studies, we attempted to chemically convert inositol closed, heated at 530C for 1 hr, and then cooled on ice. Cold 1,4,5-trisphosphate to inositol 1,2-(cyclic)-4,5-trisphosphate. 1-propanol (32.5 ml) was added and the solvent mixture was

The publication costs of this article were defrayed in part by page charge *To whom reprint requests should be addressed at: Division of payment. This article must therefore be hereby marked "advertisement" Hematology-Oncology, Washington University School of Medi- in accordance with 18 U.S.C. §1734 solely to indicate this fact. cine, 660 South Euclid, St. Louis, MO 63110.

Downloaded by guest on October 2, 2021 1206 Biochemistry: Auchus et al. Proc. Natl. Acad. Sci. USA 84 (1987) 1207 removed under reduced pressure. The residue was redis- were separated by HPLC on Partisil SAX as described solved in 20 ml of water and 24 ml of 1-butanol/light above. petroleum ether (bp, 40-600C)/ethyl formate, 20:4:1. After NMR Spectroscopy. 31p {1H}-NMR spectra were recorded extraction, the aqueous phase containing deacylated phos- at 121 MHz on a Varian XL 300 spectrometer equipped with pholipids was washed with 15 ml of the same mixture and a 5-mm variable temperature probe using a 10-gsec pulse stored at -40C. width, 0.8-sec acquisition time (no delay), and continuous Degradation of deacylated phospholipids (640 ,xmol) was WALTZ-16 decoupling. Chemical shifts are reported in ppm done essentially as described by Brown and co-workers (14, relative to 85% H3PO4 external reference at 0.0 ppm. Samples 15). Sodium m-periodate (205 mg, 960 tmol) was allowed to were dissolved in 2H20 for room temperature spectra and in react with the deacylated phospholipids for 10 min in the dark 40% methanol/60% 2H20 for the -20'C spectrum. at room temperature. The reaction was stopped by the addition of 1% ethylene glycol (1.78 ml, 320 gmol). After an RESULTS additional 15 min at room temperature, 1% 1,1-dimethylhy- drazine (4.86 ml, 640 4mol; pH 4.5) was added, and the We attempted to cyclize inositol 1,4,5-trisphosphate using reaction was allowed to proceed for 4 hr in the dark at room dicyclohexylcarbodiimide with various amounts of tri-n- temperature. The reaction mixture was treated with 20 ml of butylamine, tert-butanol, dimethylformamide, and pyridine. slurried Dowex AG 5OW-X8 (100-200 mesh) hydrogen form These experiments either gave no reaction or multiple unde- and then filtered through a 1-cm bed of Celite. The eluate was sired products. Poor solubility of inositol 1,4,5-trisphosphate combined with two 5- to 10-ml water washes and lyophilized. in the mixed solvent systems needed to dissolve dicyclohex- The sample dissolved in water was loaded on a 2-ml column ylcarbodiimide appears to be the major problem with this of Dowex AG 1-X8 (200-400 mesh) formate form. The choice of conditions. column was eluted as described by Downes et al. (16). The In initial experiments, we found that the reaction ofinositol samples were desalted by treatment with Dowex AG 5OW-X8 1,4,5-trisphosphate with the water-soluble carbodiimide, 1- (100-200 mesh) hydrogen form. The ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochlo- was lyophilized and phosphate was assayed as described by ride, rapidly generated several products, none of which Ames and Dubin (17). Following this procedure, 17-20 ,mol cochromatographed on HPLC with inositol 1,2-(cyclic)-4,5- of inositol 1,4,5-trisphosphate was obtained from 1 g of brain trisphosphate. We therefore added triethylamine to suppress extract. The yield based on recovery of 3H was -70%. intermolecular reactions and to prevent further reaction of Synthesis of Inositol 1,2-(Cyclic)-4,5-Trisphosphate. A so- the cyclic phosphate (21). We found that the reaction would lution of 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hy- not proceed in triethylamine until the pH was reduced to 8-9. drochloride (40 mg, 104 ,mol) and triethylamine (400 ,u1, 0.29 When the pH was further lowered to 5-6, multiple undesired g, 2.87 mmol) in 50 ml of water was adjusted to a pH of 8.2 products were formed. Thus, we carried out the reaction at with HCl. Inositol 1,4,5-trisphosphate in 0.5 ml of H20 (5 pH 8.2 at room temperature. Under these conditions, inositol kkmol, 700,000 cpm) was added to the above solution, and the 1,4,5-trisphosphate was converted to inositol 1,2-(cyclic)- pH was readjusted to 8.2. The reaction mixture was stirred at 4,5-trisphosphate in -25% yield with only two additional room temperature for 3 hr, quenched with 4 M ammonium products (Fig. 1). Incubation for longer times increased the formate (800 ,ul, 3.2 mmol), and lyophilized. The sample was early eluting peaks without increasing the yield of the desired redissolved, desalted as described by Rittenhouse and Sas- product. No reaction occurred at 0°C. son (18) using a 1-ml column of Dowex AG 1-X4 (100-200 mesh) chloride form, and lyophilized again. The HPLC fractions containing the putative cyclic product HPLC of Inositol Phosphates. 32p and 3H-labeled inositol (40-43) were pooled, desalted, and examined by 31P NMR phosphates were resolved on a Whatman Partisil 10 SAX spectroscopy. At ambient temperature in 100% 2H20, the 31p column (Cobert, St. Louis, MO) as described (19) except that {'H}-NMR spectrum consisted of a broad (64 Hz) peak at 17.2 a small 0.5 x 8 cm Chelex column was added in front of the ppm and a second broad feature at 1 ppm. Cooling the sample precolumn. The elution scheme consisted of isocratic elution to 0°C gave a spectrum with a sharper down-field signal (23 with 50 mM ammonium formate (pH 6.25) for 10 min; this was Hz) and two broad but partially resolved signals at 1.8 and 0.2 followed by a linear gradient from 50 mM to 2.7 M ammonium ppm. At -20°C in 40% methanol/60% 2H20, the spectrum formate (pH 6.25) over the next 20 min, which was followed by 30 min of isocratic elution with 2.7 M ammonium formate 1ns(>P)P, Ins-1,4,5-P, (pH 6.25). A rate of 1 ml/min was used, and 1-ml fractions 350 were collected and assayed for radioactivity in a Beckman 300 liquid scintillation counter either by measuring Cerenkov radiation to determine 32p content or with Scinti Verse I .2250 scintillation fluid to determine the 3H content. Acid Hydrolysis of Inositol 1,2-(Cyclic)-4,5-Trisphosphate. '~200 A sample (2000 cpm, 14 nmol) was mixed with an equal volume of 2 M HCl. After 2 min at room temperature, the 2150 sample was frozen in a dry ice/acetone bath and lyophilized to remove acid. The reaction mixture was taken up in water, z100 and the products were separated by HPLC. Treatment with 5-Phosphomonoesterase. 50- The 5-phosphomonoesterase assay was performed as de- scribed by Connolly et al. (20). Briefly, [3H]inositol 1,2- 10 20 30 40 50 60 (cyclic)-4,5-trisphosphate (170 uM), inositol 1,4,5-[32P]tris- Time, min 20-40 and various amounts of phosphate (2000 cpm, liM), FIG. 1. Partisil 10 SAX HPLC of products of a reaction mixture 5-phosphomonoesterase were incubated in 50 til at of inositol 1,4,5-trisphosphate, 1-ethyl-3-(3-dimethylaminopropyl)- 370C with 3 mM MgCl2/50 mM 2-(N-morpholino)ethanesul- carbodiimide hydrochloride, and triethylamine in water. The elution fonic acid, pH 6.5, for 60 min to 24 hr in various experiments. positions of standards of inositol 1,2-(cyclic)-4,5-trisphosphate The reactions were stopped by freezing, and the products [Ins(>P)P2] and inositol 1,4,5-trisphosphate (Ins-1,4,5-P3) are noted. Downloaded by guest on October 2, 2021 1208 Biochemistry: Auchus et al. Proc. Natl. Acad. Sci. USA 84 (1987) exhibited a sharp down-field signal at 17.2 ppm and two Ins(>P)P, Ins-1,4,5-Pli completely resolved signals at 4.1 and 0.8 ppm (Fig. 2). For I I comparison, the 31P {1H}-NMR spectra of inositol 1,2- (cyclic)phosphate and inositol 1,4,5-trisphosphate are shown in Fig. 2. We further established the identity ofthe cyclic product by 0 acid hydrolysis and enzymatic degradation. The addition of hydrochloric acid to inositol 1,2-(cyclic)-4,5-trisphosphate 0. hydrolyzed it to inositol 1,4,5-trisphosphate (Fig. 3), in 20. agreement with earlier results on the characterization of this I) compound (1, 9). It is possible that the 10% ofthe sample that a, appears acid-resistant reflects a contaminant in our product, or it may represent inositol 2,4,5-trisphosphate, which is formed in small amounts by acid hydrolysis of the cyclic compound. Incubation of [3H]inositol 1,2-(cyclic)-4,5-trisphosphate 30 with the platelet 5-phosphomonoesterase as described in Time, min Materials and Methods resulted in conversion to [3H]inositol FIG. 3. HPLC of inositol 1,2-(cyclic)-4,5-trisphosphate (0-0) 1,2-(cyclic)-4-bisphosphate. Under conditions in which 50% and inositol 1,2-(cyclic)-4,5-trisphosphate after acidification of an inositol 1,4,5-[32P]trisphosphate internal standard was (o--- o). See Fig. 1 for abbreviations. converted to inositol 1,4-bisphosphate, we found 22% con- version of [3H]inositol 1,2-(cyclic)-4,5-trisphosphate to inosi- ofinositol 1,4,5-[32P]trisphosphate was completely converted tol 1,2-(cyclic)-4-bisphosphate. After exhaustive digestion, to inositol 1,4-bisphosphate. This result is similar to the we found 78% conversion to [3H]inositol 1,2-(cyclic)-4- conversion achieved by using enzymatically prepared bisphosphate under conditions in which an internal standard inositol 1,2-(cyclic)-4,5-trisphosphate, where 85% was con- verted to the cyclic bisphosphate product after exhaustive digestion (22). A DISCUSSION The cyclization reaction mixture using 1-ethyl-3-(3-dimeth- ylaminopropyl)carbodiimide hydrochloride contains three products (Fig. 1). When the reaction was allowed to proceed for longer periods, the two side products increased at the expense of inositol 1,2-(cyclic)-4,5-trisphosphate, implying that they may be oligomers. We monitored the reaction by HPLC and found that the best yields and the fewest products were obtained at about 25% conversion to inositol 1,2- B (cyclic)-4,5-trisphosphate. Unreacted inositol 1,4,5-trisphos- phate is recoverable and can undergo reaction again. To acquire a 31p {1H}-NMR spectrum that had three resolved signals, we cooled the sample to -20TC. Since the sample could not be run in 2H20 at this temperature, methanol was added as a cryosolvent. Apparently, some conformational flexibility creates line broadening through chemical exchange above 0C. This motion is partially frozen out at lower temperatures. Comparison of the 31p {'H}-NMR PA I. MU.&AMW.AMhII Li L.& lkL kIL/AMIfI1hh&PdL.hbW.YUF.LkbAL I L lh A. . . i-JIA iA dfiJ. IflAilIL AM, Ri'1.F.NffWYUW'IL ik- 9 I 11,AKILIkAl spectrum of the synthetic material with spectra of inositol jf",rrjwlIvlyw lyl T IT' -171-T 1,4,5-trisphosphate and inositol 1,2-(cyclic)phosphate (Fig. 2) shows the presence of a five-membered cyclic phospho- C diester ring and two phosphomonoesters. A 31P NMR spectrum of inositol 1,4,5-trisphosphate has been reported by Lindon et al. (23). The spectrum in that study was run at alkaline pH, which may explain the differences in chemical shifts from our findings. The char- acteristic down-field position of the five-membered cyclic phosphate has been observed (24) in the 2', 3' series and has been attributed to a change in phosphorous hybridization by the 70 reduction ofthe O-P--O bond angle. Connolly and co-workers (19, 22) found that inositol 1,4,5-trisphosphate is a better substrate than inositol 1,2-(cy- 25 20 clic)-4,5-trisphosphate for 5-phosphomonoesterase. When ppm inositol 1,4,5-trisphosphate and inositol 1,2-(cyclic)-4,5-tris- phosphate are incubated together with the 5-phosphomono- FIG. 2. 31p {1H}-NMR spectra of inositol 1,4,5-trisphosphate (4 esterase, the inositol 1,4,5-trisphosphate is hydrolyzed be- Amol) in 100%0 2H20, ambient temperature, 500 scans (A), inositol be- 1,2-(cyclic)phosphate (4 /imol) in 50%o 2H20, ambient temperature, fore the inositol 1,2-(cyclic)-4,5-trisphosphate hydrolysis 500 scans (B), and inositol 1,2-(cyclic)-4,5-trisphosphate (1 Amol) in gins. Similar results were obtained by using inositol 1,2- 40% methanol/60% 2H20, -200C, 524 scans (C). All spectra were (cyclic)-4,5-trisphosphate prepared enzymatically compared processed with 5-Hz exponential line broadening. to that synthesized in this study, implying that the two Downloaded by guest on October 2, 2021 Biochemistry: Auchus et al. Proc. Natl. Acad. Sci. USA 84 (1987) 1209

products are the same. The synthetic material possesses 8. Berridge, M. J. & Irvine, R. F. (1984) Nature (London) 312, chemical, enzymatic, and spectral characteristics consistent 315-321. with the proposed structure of inositol 1,2-(cyclic)-4,5- 9. Wilson, D. B., Connolly, T. M., Bross, T. E., Majerus, P. W., trisphosphate. Now that micromole quantities of inositol Sherman, W. R., Tyler, A. N., Rubin, L. J. & Brown, J. E. (1985) J. Biol. Chem. 260, 13496-13501. 1,2-(cyclic)-4,5-trisphosphate are available, many more 10. Irvine, R. F., Letcher, A. J., Lander, D. J. & Berridge, M. J. physiological studies can be made. (1986) Biochem. J. 240, 301-304. 11. Pizer, F. L. & Ballou, C. E. (1959) J. Am. Chem. Soc. 81, We thank Drs. Tom Connolly and Doug Covey for their helpful 915-921. suggestions concerning this project. Assistance was provided by the 12. Downes, C. P., Mussast, M. C. & Mitchell, R. H. (1982) Washington University High Resolution NMR Service Facility Biochem. J. 203, 169-177. supported in part by National Institutes of Health Grant 1 S10 13. Clarke, N. G. & Dawson, R. M. C. (1981) Biochem. J. 195, RR00204 and a gift from the Monsanto Company. This research was 301-306. supported by Grants HLBI 14147 (Specialized Center for Research 14. Brown, D. M. & Stewart, J. C. (1966) Biochim. Biophys. Acta in Thrombosis) and HL 16634 from the National Institutes of Health 125, 412-421. and in part by National Institutes of Health Research Service Award 15. Brown, D. M., Hall, G. E. & Letters, R. (1959) J. Chem. Soc., GM 07805 from the National Institute of General Medical Sciences 3547-3552. (R.J.A.). 16. Downes, C. P., Hawkins, P. T. & Irvine, R. F. (1986) Bio- chem. J. 238, 501-506. 17. Ames, B. W. & Dubin, D. T. (1960) J. Biol. Chem. 235, 1. Wilson, D. B., Bross, T. E., Sherman, W. R., Berger, R. A. & 769-775. Majerus, P. W. (1985) Proc. Natl. Acad. Sci. USA 82, 18. Rittenhouse, S. E. & Sasson, J. P. (1985) J. Biol. Chem. 260, 4013-4017. 8657-8660. 2. Dawson, R. M., Freinkel, N., Jungalwala, F. B. & Clark, N. 19. Connolly, T. M., Wilson, D. B., Bross, T. E. & Majerus, (1971) Biochem. J. 122, 605-607. P. W. (1986) J. Biol. Chem. 261, 122-126. 3. Dixon, J. F. & Hokin, L. E. (1985) J. Biol. Chem. 260, 20. Connolly, T. M., Bross, T. E. & Majerus, P. W. (1985) J. Biol. 16068-16071. Chem. 260, 7868-7874. 4. Shayman, J. A., Auchus, R. J. & Morrison, A. R. (1986) 21. Smith, M., Moffatt, J. G. & Khorana, H. G. (1958) J. Am. Biochim. Biophys. Acta 888, 171-175. Chem. Soc. 80, 6204-6212. 5. Binder, H., Weber, P. C. & Siess, W. (1985) Anal. Biochem. 22. Connolly, T. M., Bansal, V. S., Bross, T. E., Irvine, R. F. & 148, 220-227. Majerus, P. W. (1987) J. Biol. Chem., in press. 6. Koch, M. A. & Diringer, H. (1974) Biochem. Biophys. Res. 23. Lindon, J. C., Baker, D. J., Farrant, R. D. & Williams, J. M. Commun. 58, 361-367. (1986) Biochem. J. 233, 275-277. 7. Ishii, H., Connolly, T. M., Bross, T. E. & Majerus, P. W. 24. Cozzone, P. J. & Jardetzky, 0. (1976) Biochemistry 15, 4853- (1986) Proc. Natl. Acad. Sci. USA 83, 6397-6401. 4859. Downloaded by guest on October 2, 2021