If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
* This work was supported by grants from the Wellcome Trust (to S. G. W., J. W., and B. V. L. P.) and from the Science and Engineering Research Council (Molecular Recognition Initiative; to B. V. L. P.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. § Lister Institute Research Professor.
Several natural and unnatural inositol phosphates and analogues were analyzed for their ability to inhibit the in vitro phosphatidylinositol 3-kinase (PI 3-kinase) activity immunoprecipitated from a leukemic T cell line by a p85 monoclonal antibody. A 3-position ring-modified analogue of D-myo-inositol 1,4,5-trisphosphate (Ins(1,4,5)P3), L-chiro-inositol 2,3,5-trisphosphate (L-chiro-Ins(2,3,5)P3) and its phosphorothioate analogue, L-chiro-inositol 2,3,5-trisphosphorothioate, as well as the analogue benzene 1,2,4-trisphosphate induced reversible inhibition of PI 3-kinase activity, which correlated with decreased Vmax but unchanged Kmvalues for PI 3-kinase. Other inositol phosphates, including D- and L-Ins(1,4,5)P3, D-myo-inositol 1,3,4,5-tetrakisphosphate, the enantiomers of myo-inositol 1,3,4-trisphosphate, DL-myo-inositol 1,4,6-trisphosphate (DL-Ins(1,4,6)P3), and DL-scyllo-inositol 1,2,4-trisphosphate (DL-scyllo-Ins(1,2,4)P3), did not inhibit PI 3-kinase activity under identical conditions. L-chiro-Ins(2,3,5)P3 closely resembles Ins(1,4,5)P3 and D-Ins(1,4,6)P3 except for a difference in the orientation of a single hydroxyl group at either the equivalent 3-OH or 2-OH position of Ins(1,4,5)P3, respectively. Similarly, L-chiro-Ins(2,3,5)P3 resembles D-scyllo-Ins(1,2,4)P3, but has a different orientation of both the equivalent 3-OH and 2-OH positions. Since Ins(1,4,5)P3, DL-Ins(1,4,6)P3, and DL-scyllo-Ins(1,2,4)P3 did not inhibit PI 3-kinase activity, this suggests that the orientation of the two hydroxyl groups at the 2- and 3-positions plays a pivotal role in the inhibitory action of inositol phosphate analogues on PI 3-kinase activity. Thus, inositol phosphate analogues inter alia are shown for the first time to inhibit PI 3-kinase and may be useful tools for determining the function of PI 3-kinase and its substrate binding specificities.
) that mediates the formation of D-3-phosphatidylinositol lipids by transferring the terminal phosphate of ATP to the D-3-position on inositol head groups of phosphatidylinositol (PtdIns), phosphatidylinositol 4-monophosphate (PtdIns(4)P), and phosphatidylinositol 4,5-bisphosphate to yield phosphatidylinositol 3-monophosphate (PtdIns(3)P), phosphatidylinositol 3,4-bisphosphate, and phosphatidylinositol 3,4,5-trisphosphate, respectively (
). Clarification of the precise function of such lipids is still hampered by a lack of agents to manipulate the activity of PI 3-kinase. Effective inhibitors of this enzyme may help to define the role of PI 3-kinase and its metabolic products in cells, but to date, only a few inhibitors have been described. Generally, these are derived from natural products, namely the fungal metabolite wortmannin (
), is the most effective PI 3-kinase inhibitor, being active at nanomolar concentrations. However, wortmannin at micromolar concentrations has been shown to inhibit other enzymes such as phospholipase D (
The PI 3-kinase phosphorylation target site on phosphatidylinositols is similar to that on myo-inositol 1,4,5-trisphosphate (Ins(1,4,5)P3) 3-kinase in that Ins(1,4,5)P3 3-kinase phosphorylates the phospholipase C metabolic product Ins(1,4,5)P3 (Fig. 1, compound 1), also at the D-3-position of the inositol ring (
) rather than to the existence of isoforms. We have adopted a novel approach in the search for PI 3-kinase inhibitors by analyzing the effect on PI 3-kinase activity of natural and unnatural inositol phosphates, which may compete with the D-3-phosphorylation site on the inositol head group of phosphatidylinositols for substrate recognition by PI 3-kinase. Hence, we reasoned that small molecule inositol phosphate analogues, which have been reported to inhibit the transfer of phosphate onto Ins(1,4,5)P3 catalyzed by Ins(1,4,5)P3 3-kinase (
), may additionally act as leads for compounds that inhibit the phosphorylation of phosphatidylinositols by PI 3-kinase.
We report here the first evidence that chemically synthesized analogues of Ins(1,4,5)P3 that act as inhibitors of Ins(1,4,5)P3 3-kinase and possess an inverted 3-hydroxyl group, namely L-chiro-inositol 2,3,5-trisphosphate (L-chiro-Ins(2,3,5)P3) (Fig. 1, compound 2) (
), a loosely structurally related analogue of Ins(1,4,5)P3, and demonstrate that it also inhibits PI 3-kinase.
Chemically synthesized Ins(1,4,5)P3 and D-myo-inositol 1,3,4,5-tetrakisphosphate (Ins(1,3,4,5)P4) were from the University of Rhode Island Chemistry Group. p85 monoclonal antibodies (mAbs) were gifts from Dr. Cantrell (Imperial Cancer Research Fund, London). ATP, wortmannin, and phosphatidylinositols (soybean PtdIns and bovine PtdIns(4)P) were purchased from Sigma (Poole, Dorset, United Kingdom). Wortmannin was dissolved in ethyl acetate to a concentration of 20 mM and stored in aliquots at -20°C in the dark. All other reagents were of analytical grade and were purchased from Sigma and Aldrich. TLC was performed on Silica Gel 60 F (Merck) with detection by UV light or phosphomolybdic acid. Flash column chromatography was performed on Silica Gel SORBSIL C60. Ion-exchange chromatography was performed using a Pharmacia Biotech medium pressure ion-exchange chromatograph, Q-Sepharose, and a gradient of triethylammonium bicarbonate. 1H NMR spectra (internal Me4Si reference) were recorded on a Jeol JMN-GX 270 or EX 400 spectrometer. 31P NMR spectra (external H3PO4 reference) were recorded on a Jeol EX 400 NMR spectrometer. Chemical shift(s) are reported as negative when downfield from H3PO4. 13C NMR spectra (internal CDCl3 reference) were recorded on a Jeol JMN-GX 270 spectrometer. Mass spectra were recorded at the Engineering and Physical Sciences Research Council Mass Spectrometry Service (Swansea, UK). Microanalysis was carried out by the Microanalysis Service at the University of Bath.
The leukemic T cell line Jurkat was grown in RPMI 1640 medium supplemented with 10% heat-inactivated fetal calf serum, 50 μg/ml streptomycin, and 50 units/ml penicillin at 37°C (
), were synthesized as described, and the structures of the most important of these are given in Fig. 1. References for further inactive analogues are given in the text. All compounds were purified by ion-exchange chromatography on Q-Sepharose Fast Flow (Pharmacia Biotech Inc.) using a gradient of triethylammonium bicarbonate, pH 8.0, quantified using the Briggs phosphate assay (
) and used as their triethylammonium salts. All compounds showed satisfactory 1H and 31P NMR and mass spectrometric data.
Synthesis of 1,2,4-Tris(diethylphospho)benzene
Benzene-1,2,4-triol (252 mg, 2 mmol) was suspended in dry dichloromethane (5 ml) and stirred under a blanket of nitrogen. Dry N,N-diisopropylethylamine (2.1 ml, 12 mmol) was added to the suspension, and the solution turned red in color. The solution was then cooled to -78°C, and diethyl chlorophosphite (1.57 ml, 9.0 mmol) was added dropwise, giving a pale yellow color, which indicated that the hydroxyl groups had been phosphitylated. The cooling bath was removed, and water (2 ml) was added to the solution, which was stirred for 30 min. t-Butyl hydroperoxide (1 ml, 7 mmol; 70% solution in water) was added dropwise, and the mixture was stirred for 15 min at room temperature. Analysis by TLC showed a new spot at Rf= 0.34. The solution was diluted with dichloromethane (100 ml) and washed with water (100 ml), 10% sodium metabisulfite (100 ml), 0.1 M hydrochloric acid (50 ml), saturated sodium hydrogen carbonate solution (100 ml), and water (100 ml). The organic layer was dried over magnesium sulfate and evaporated to give the crude product as an oil. Flash chromatography (9:1 ethyl acetate/ethanol) gave the pure title compound as a syrup (yield of 0.79 g, 76%). Found: C, 40.7; H, 6.42; C18H33O12P3 requires C, 40.45; H, 6.18. δH (CDCl3, 270 MHz), 1.34-1.40 (18H, m, BzOP(O)OCH2CH3), 4.17-4.32 (12H, m, BzOP(O)OCH2CH3), 7.03-7.39 (3H, m, H-3, H-5, H-6, Bz). δC (CDCl3, 68 MHz), 15.53-15.63 (2q, BzOP(O)OCH2CH3), 64.43, 64.53, 64.64 (3t, BzOP(O)OCH2CH3), 113.29-146.91 ring carbons. δP (CDCl3, 162 MHz), 6.86 (dtt, J = 7.83 Hz), -7.12 (dtt, J = 7.83 Hz), -7.28 (dtt, J = 7.83 Hz).
Synthesis of Benzene 1,2,4-Trisphosphate
1,2,4-Tris(diethylphospho)benzene (274 mg, 528 μmol) was dissolved in dry dichloromethane. Bromotrimethylsilane (0.836 ml, 6.3 mmol) was added dropwise to the solution, which was then stirred for 16 h. The solvents were evaporated, and the residue was stirred with water (1 ml). Final purification of the compound was by elution from a column of Q-Sepharose Fast Flow using triethylammonium bicarbonate buffer with a linear gradient of 0-1 M. The title compound (Fig. 1, compound 6) eluted between 0.2 and 0.5 M buffer and after evaporation was obtained as its glassy triethylammonium salt (yield of 456 μmol, 86%). δH (D2O, 400 MHz), 6.85 (1H, dd, J = 1.5 Hz, J = 8.85 Hz, H-5, Bz), 7.06 (1H, d, J = 1.5 Hz, H-3, Bz), 7.18 (1H, d, J = 8.85 Hz, H-6, Bz). δP (CDCl3, 162 MHz), -3.61 (s), -3.92 (s), -4.28 (s). Accurate mass spectrum requires the following: (M - H)- = 364.9228. Found: 364.9238.
Cell Lysis and Immunoprecipitation
Jurkat cells were washed, resuspended in RPMI 1640 medium containing 20 mM HEPES, aliquoted at 2 × 107 cells/ml, and incubated at 37°C for 5 min. Cells were pelleted, and the pellets were lysed in 1 ml of lysis buffer (1% Nonidet P-40, 100 mM NaCl, 20 mM Tris, pH 7.4, 10 mM iodoacetamide, 10 mM NaF, 1 mM phenylmethylsulfonyl fluoride, 1 μg/ml leupeptin, 1 μg/ml antipain, 1 μg/ml chymostatin, 1 μg/ml pepstatin A, 1 mM sodium orthovanadate). After centrifugation at 14,000 rpm, post-nuclear lysates were precleared with protein G-Sepharose to remove any nonspecific proteins that may have bound to protein G-Sepharose. This was followed by immunoprecipitation for 2 h at 4°C as described (
) using a lipid mixture of 100 μl of 0.1 mg/ml PtdIns and 0.1 mg/ml phosphatidylserine dispersed by sonication in 20 mM HEPES, pH 7.0, 1 mM EDTA. The reaction was initiated by the addition of a mixture of 20 μCi of [γ-32P]ATP (3000 Ci/mmol; DuPont NEN, Stevenage, UK) and 100 μM ATP to the immunoprecipitates suspended in 80 μl of lipid kinase buffer. The reaction was terminated after 15 min by the addition of 80 μl of 1 N HCl and 200 μl of chloroform/methanol (1:1, v/v). After centrifugation for 5 min at 14,000 rpm, the lower phase was removed, dried in vacuo, and redissolved in 50 μl of chloroform. Phospholipids were then separated by thin-layer chromatography in 2 M 1-propanol/acetic acid (65:35, v/v) developing solvents (
). [32P]PtdInsP was visualized by autoradiography, recovered from the TLC plate, and quantitated by scintillation counting (Canberra Packard). HPLC analysis of glycerophosphoryl derivatives of [32P]PtdInsP was performed as described (
). Alternatively, immunoprecipitates prepared as described above were washed three times in lysis buffer and twice in protein kinase assay buffer (100 mM NaCl, 25 mM HEPES, pH 7.4, 10 mM MgCl2, 5 mM MnCl2, 100 μM sodium orthovanadate) as described (
). Assays were initiated with 20 μl of protein kinase buffer containing 10 μM ATP and 10 μCi of [γ-32P]ATP. After 10 min at 37°C, the reaction was stopped by the addition of 1 ml of lysis buffer containing 20 mM EDTA. 32P-Labeled proteins were solubilized in SDS sample buffer prior to separation by SDS-polyacrylamide gel electrophoresis and were visualized by autoradiography at -70°C.
Identification of in Vitro Lipid Kinase Activity Immunoprecipitated by p85 mAb
The p85 immunoprecipitates were assayed for lipid kinase activity, and the production of [32P]PtdIns was shown to be linear for at least 30 min under the conditions described under “Experimental Procedures” (Fig. 2A). The lipid kinase activity present in p85 mAb immunoprecipitates (Fig. 2B, lane 2) was also inhibited by the reported PI 3-kinase inhibitor wortmannin (Fig. 2A, lane 3). The lipid kinase activity immunoprecipitated by p85 (Fig. 2B, lane 2) was further confirmed as PI 3-kinase by HPLC analysis of glycerophosphorylinositol derivatives of the [32P]PtdInsP formed (Fig. 2C) and by comparison with known standards. In contrast, HPLC analysis of glycerophosphorylinositol derivatives formed following incubation of total cell lysates with PtdIns under identical conditions (Fig. 2B, lane 1) revealed predominant formation of [32P]PtdIns(4)P (Fig. 2D). The apparent Kmfor PtdIns is 15.7 ± 5.7 μM (n = 7), which was determined by varying the concentration of PtdIns (keeping the molar ratio of phosphatidylinositol to phosphatidylserine at 1).
Effect of L-chiro-Ins(2,3,5)P3andL-chiro-Ins(2,3,5)PS3on PI 3-Kinase Activity
L-chiro-Ins(2,3,5)P3 (Fig. 1, compound 2) and L-chiro-Ins(2,3,5)PS3 (compound 3)induced concentration-dependent inhibition of the PI 3-kinase activity present in p85 mAb immunoprecipitates with IC50values of 5 ± 3.5 and 20 ± 5 μM, respectively (calculated using a substrate concentration of 100 μMPtdIns) (Table I). The inhibition byL-chiro-Ins(2,3,5)P3(Fig. 3) andL-chiro-Ins(2,3,5)PS3(data not shown) was reversible (Fig. 3), with apparentKivalues of 39 ± 14 and 76 ± 34 μM, respectively (Table I). L-chiro-Ins(2,3,5)P3(Fig. 4and) and Table IIL-chiro-Ins(2,3,5)PS3 (Table II) had no effect on theKmfor immunoprecipitated PI 3-kinase, whileVmaxwas markedly decreased, indicating noncompetitive inhibition of PI 3-kinase by these inositol phosphate analogues. The percentage inhibition of PI 3-kinase activity induced by these compounds was similar after preincubation times ranging from 1 to 30 min (data not shown).
Table ICharacteristics of the inhibition of Ins(1,4,5)P35-phosphatase, Ins(1,4,5)P33-kinase, and PI 3-kinase by inositol phosphate analogues and Bz(1,2,4)P3
Kivalues for Ins(1,4,5)P3 and Ins(1,4,5)P3 3-kinase are representative and may vary according to source of enzyme and individual investigators.
Table IIKmand Vmaxvalues for immunoprecipitated PI 3-kinase
The presented representative data were determined by double-reciprocal plots. Kmand Vmax values vary ~2-4-fold from preparation to preparation. Data for each analogue are derived from separate preparations. Mean Kmvalues (± S.E.) for immunoprecipitated PI 3-kinase with respect to PtdIns are given in parentheses and are derived from seven separate preparations.
Effect of Other Inositol Phosphate Analogues on PI 3-Kinase Activity
The PI 3-kinase activity present in p85 immunoprecipitates was not affected by D-Ins(1,4,5)P3, L-Ins(1,4,5)P3, or the major Ins(1,4,5)P3 metabolites Ins(1,3,4,5)P4 (Table I) and D- and L-myo-inositol 1,3,4-trisphosphate. A number of other synthetic analogues, including DL-Ins(1,4,6)P3, L-chiro-inositol 1,4,6-trisphosphorothioate (L-chiro-Ins(1,4,6)PS3), L-myo-inositol 1,4,5-trisphosphorothioate (L-Ins(1,4,5)PS3), and myo-inositol 1,3,5-trisphosphorothioate (Ins(1,3,5)PS3) (Table I) at concentrations of 100 μM, also did not inhibit PI 3-kinase activity (see “Discussion”).
Synthesis of Benzene 1,2,4-Trisphosphate and Effect on PI 3-Kinase Activity
Benzene 1,2,4-trisphosphate (Fig. 1, compound 6) was synthesized by a markedly improved method to that previously reported (
), involving isolation and full spectral and analytical characterization of an intermediate 1,2,4-tris(diethylphospho)benzene, followed by deprotection, rigorous purification of the anionic product, and full spectral characterization. This very loosely related structural analogue of Ins(1,4,5)P3 was found to inhibit PI 3-kinase activity in p85 mAb immunoprecipitates with an IC50 value of 25 ± 2.5 μM (n = 5) (Table I) when using 100 μM PtdIns as a substrate. This inhibition of PI 3-kinase by Bz(1,2,4)P3 was reversible (Fig. 5). Bz(1,2,4)P3 (Fig. 6Table II) had no effect on the Kmfor immunoprecipitated PI 3-kinase, while Vmax was markedly decreased, indicating noncompetitive inhibition of PI 3-kinase with an apparent Kiof 93.6 ± 15 μM.
Effect of L-chiro-Ins(2,3,5)P3, L-chiro-Ins(2,3,5)PS3, and Bz(1,2,4)P3on PI 4-Kinase Activity
The specificity of the inhibition of PI 3-kinase by L-chiro-Ins(2,3,5)P3, L-chiro-Ins(2,3,5)PS3, and Bz(1,2,4)P3 was determined by analyzing the effects of these compounds on PtdIns 4-kinase. L-chiro-Ins(2,3,5)P3, L-chiro-Ins(2,3,5)PS3, and Bz(1,2,4)P3 at concentrations of 100 μM had no effect on the predominantly PtdIns 4-kinase activity present in T cell lysates (data not shown). The lack of effect of these compounds with respect to the inhibition of kinase activity in lysates may be due to instability of these analogues in cell lysates. Previous work has shown that PtdIns 4-kinase binds to protein A (
). Hence, protein A-Sepharose beads that had been previously incubated with post-nuclear lysates at 4°C contained PtdIns 4-kinase activity as determined by HPLC analysis of glycerophosphorylinositol derivatives formed during in vitro lipid kinase experiments (data not shown). L-chiro-Ins(2,3,5)P3 and Bz(1,2,4)P3 at concentrations of 100 μM had no effect on the PtdIns 4-kinase activity associated with protein A (Fig. 7).
Effect of L-chiro-Ins(2,3,5)P3, L-chiro-Ins(2,3,5)PS3, and Bz(1,2,4)P3on in Vitro Protein Kinase Activity
L-chiro-Ins(2,3,5)P3 (Fig. 8), L-chiro-Ins(2,3,5)PS3, and Bz(1,2,4)P3 (data not shown) had no effect on the activity of a coprecipitated protein-serine kinase that is known to phosphorylate the p85 subunit (
) that have enabled the preparation of natural and unnatural inositol phosphates and their structurally modified analogues. This approach has yielded inositol phosphate analogues that have been shown to act as inhibitors of enzymes controlling the metabolism of Ins(1,4,5)P3, particularly inositol-phosphate 5-phosphatase and Ins(1,4,5)P3 3-kinase (
). We have also demonstrated the inhibition of leukemic T cell-derived PI 3-kinase by two synthetic inositol phosphate analogues, L-chiro-Ins(2,3,5)P3 (Fig. 1, compound 2) and L-chiro-Ins(2,3,5)PS3 (compound 3), and by a loosely related Ins(1,4,5)P3 structural analogue, Bz(1,2,4)P3 (compound 6). These compounds did not affect the lipid kinase activity present in total cell lysates, which was predominantly PtdIns 4-kinase, or a serine kinase activity (
). Bz(1,2,4)P3, which has the inositol ring replaced by a flat benzene ring but possesses three phosphate groups in a similar spatial arrangement to that of Ins(1,4,5)P3, inhibits Ins(1,4,5)P3 5-phosphatase and Ins(1,4,5)P3 3-kinase less effectively than L-chiro-Ins(2,3,5)P3 or L-chiro-Ins(2,3,5)PS3 and does not mobilize intracellular calcium (
), all of which were found not to inhibit PI 3-kinase at a concentration of 100 μM (Table I and data not shown). Hence, the inhibition of PI 3-kinase action appears not to be a general feature of analogues that inhibit 5-phosphatase. In addition, several other naturally occurring and synthetic inositol polyphosphate analogues were examined for their ability to inhibit PI 3-kinase activity. These compounds included DL-3-O-methyl-myo-inositol 1,4,5-trisphosphate (
) resembles Ins(1,4,5)P3 (compound 1) in all respects except one: the equatorial 3-hydroxyl group of Ins(1,4,5)P3 is replaced by an axial hydroxyl group at the L-chiro-Ins(2,3,5)P3 1-position. In addition, the analogue D-Ins(1,4,6)P3 (Fig. 1, compound 4) (
), in its expected receptor binding conformation relative to Ins(1,4,5)P3, also resembles L-chiro-Ins(2,3,5)P3 except that the axial 6-hydroxyl group of the latter, which is equivalently located at the 2-position of Ins(1,4,5)P3, is replaced by an equatorial 3-hydroxyl group in D-Ins(1,4,6)P3. These subtle modifications reduce the potency of both compounds with respect to calcium mobilization in comparison with Ins(1,4,5)P3 since both L-chiro-Ins(2,3,5)P3 and DL-Ins(1,4,6)P3 are less potent than Ins(1,4,5)P3(
). scyllo-Ins(1,2,4)P3 (Fig. 1, compound 5) is identical in structure to Ins(1,4,5)P3 except that it possesses an equatorial 5-hydroxyl group and is equivalent to the 2-position in Ins(1,4,5)P3, which has an axial hydroxyl group. scyllo-Ins(1,2,4)P3 is only marginally less potent than Ins(1,4,5)P3 with respect to calcium mobilization (
) and was inactive as a PI 3-kinase inhibitor. The ability of inositol phosphate analogues to inhibit PI 3-kinase thus appears to be very sensitive to orientations of hydroxyl groups of the inositol ring since Ins(1,4,5)P3, Ins(1,4,6)P3, and scyllo-Ins(1,2,4)P3, which all resemble L-chiro-Ins(2,3,5)P3 except for orientation of the equivalent 2- and 3-position hydroxyl groups (Ins(1,4,5)P3 numbering), do not inhibit PI 3-kinase. Moreover, alternative modification at the 3-position of Ins(1,4,5)P3 only, as in 3-O-Me-Ins(1,4,5)P3, produced an inactive compound with respect to PI 3-kinase inhibition.
Thus, there appears to be an absolute requirement for axial hydroxyl groups at both 2- and 3-hydroxyl positions (D-Ins(1,4,5)P3 numbering) for inositol phosphate analogues to exhibit the ability to inhibit PI 3-kinase activity. It is not surprising that a structural perturbation at the equivalent myo-inositol D-3-position, the site of phosphorylation by PI 3-kinase, results in inhibitor generation. Interestingly, the orientation of hydroxyl groups of the inositol ring has also been reported to determine recognition of Ins(1,4,5)P3 by its receptor and the metabolic enzymes Ins(1,4,5)P3 5-phosphatase and Ins(1,4,5)P3 3-kinase (
). Hence, it appears from the data presented in this study that modification of hydroxyl group orientation at the crucial 3-position can also confer PI 3-kinase inhibitory properties upon certain inositol phosphate analogues.
The observation that the inhibition of PI 3-kinase by L-chiro-Ins(2,3,5)P3, L-chiro-Ins(2,3,5)PS3, and Bz(1,2,4)P3 is reversible is not surprising since the inhibitory action of L-chiro-Ins(2,3,5)P3 on Ins(1,4,5)P3 3-kinase is of a reversible competitive nature (
). Indeed, these compounds were used in this study on the basis that they may compete with the lipid substrate for the active site on the PI 3-kinase. The data shown here suggest that the inhibition of PI 3-kinase induced by L-chiro-Ins(2,3,5)P3 and its phosphorothioate derivative as well as Bz(1,2,4)P3 is noncompetitive and may involve mechanisms other than competition with the lipid substrate at the active site of the enzyme. There are several explanations for these observations. First, these inositol phosphate analogues may modulate enzyme activity by interacting with PI 3-kinase at a site other than the active site. This may be an important regulatory site present on either the p85 or p110 subunit and a potential target site for future inhibitor development. The possible existence of such a regulatory site and its endogenous ligand (and even whether such a ligand is an inositol phosphate) remains to be established since the most obvious candidates for binding to this site are Ins(1,4,5)P3 and Ins(1,3,4,5)P4, which have been shown to have no effect on PI 3-kinase activity. Second, L-chiro-Ins(2,3,5)P3 and Bz(1,2,4)P3 may differ sufficiently in structure from the numerous inactive inositol phosphates we have studied in that they are recognized by PI 3-kinase and are able to compete with the substrate lipid for access to the active site. However, these inositol phosphate analogues lack the glycerol backbone and acyl groups of the natural substrate lipids that may mediate or facilitate optimum enzyme-substrate interaction. In these circumstances, it is possible to envisage the PI 3-kinase inhibitors binding to both enzyme and enzyme-lipid substrate complexes and thus fulfilling the criteria for noncompetitive inhibition (
). Third, the apparent Kivalues for inhibition of PI 3-kinase by L-chiro-Ins(2,3,5)P3, L-chiro-Ins(2,3,5)PS3, and in particular Bz(1,2,4)P3 are substantially higher (Table I) than the IC50 values. This discrepancy may imply mixed competitive and noncompetitive inhibition, which can be envisaged given the proposed interactions described above. The data Table II in show that the calculated Kmvalue for PI 3-kinase has an error of ~30%, while the data Table I in show that the calculated Kivalues have errors of ~15-30%. The apparent failure to detect an increased Kmfor PI 3-kinase in the presence of the inhibitory inositol phosphate analogues that would be expected for mixed competitive and noncompetitive inhibition (
) may be obscured by the relatively large error margins associated with these experiments. The apparent Kivalues for inhibition of PI 3-kinase by L-chiro-Ins(2,3,5)P3, L-chiro-Ins(2,3,5)PS3, and in particular Bz(1,2,4)P3 are substantially higher (Table I) than the reported Kivalues for inhibition of Ins(1,4,5)P3 3-kinase. It should be noted from, however, that the Kifor Ins(1,4,5)P3 3-kinase is much lower than the Kmfor PI 3-kinase. While it is not certain precisely how these inhibitors interact with PI 3-kinase, one possibility is that their inhibitory action may be due to nonselective inhibition of the transfer of phosphate from ATP. However, this is unlikely since the analogues had no effect on the serine phosphorylation of p85 mediated by a tightly associated serine kinase (
) and have been suggested to be important lead compounds in the development of specific small molecule inhibitors of these enzymes. At present, no compound has been identified as a PI 3-kinase inhibitor that is not also an Ins(1,4,5)P3 5-phosphatase or 3-kinase inhibitor. Bz(1,2,4)P3 represents an especially important lead since chemical modification of this compound is much easier than are the often extensive synthetic routes for preparation of inositol phosphate analogues. We have developed a much improved and high yielding synthetic route to Bz(1,2,4)P3 than that previously reported by other workers (
), who neither purified nor fully characterized their intermediate and final products.
This study suggests that such simplified structures may also represent lead compounds in the development of PI 3-kinase inhibitors. Clearly, a future strategy to exploit these observations could include synthesis of related analogues possessing part or all of the missing diacylglycerol element. This should enhance recognition of the analogues by PI 3-kinase and potentially enhance potency of the inhibitory effect on PI 3-kinase activity. In the development of second generation analogues based on these lead compounds, it should also be an important consideration to avoid dual specificity, which may compromise their use as specific inhibitors of PI 3-kinase or Ins(1,4,5)P3 3-kinase. Furthermore, the analogues described in this study are highly polar, and these agents therefore do not permeate the cell membrane. Thus, while they can be considered useful agents in cell-free or permeabilized cell preparations, this does not apply when considering D-3-phosphatidylinositol lipid generation in intact cells, where it will be necessary to make analogues with further refinements and developments to generate effective apolar membrane-permeable inhibitors. Nevertheless, the identification of inositol phosphate analogues that act as PI 3-kinase inhibitors and that presumably compete with the phosphatidylinositols for substrate recognition by PI 3-kinase provides novel key pharmacological tools for intervention at a crucial component of this postulated signaling pathway.
We thank A. Riley and Dr. D. Lampe for provision of inositol polyphosphate analogues and Dr. R. Eisenthal for helpful discussions.