Wortmannin and its structural analogue demethoxyviridin inhibit stimulated phospholipase A2 activity in Swiss 3T3 cells. Wortmannin is not a specific inhibitor of phosphatidylinositol 3-kinase.

Wortmannin and its structural analogue demethoxyviridin (DMV) have been reported to be specific inhibitors of phosphatidylinositol 3-kinase activity. Here we report that these compounds are not as selective as assumed and demonstrate inhibition of bombesin-stimulated phospholipase A2 activity by both wortmannin and DMV with an IC40 (2 nM) which is slightly more potent than the inhibition of insulin-stimulated phosphatidylinositol 3,4,5-trisphosphate generation in these cells (⌼10 nM). While it has not been possible to fully block in vitro phospholipase A2 activity with wortmannin, inhibition cannot be a consequence of inhibition of PI 3-kinase activity since bombesin fails to generate 3-phosphorylated lipids in the intact cell. Therefore, while wortmannin is indeed a PI 3-kinase inhibitor, it is not as specific as previously reported, and experimental conclusions based solely on its use should be treated with caution.

Wortmannin and its structural analogue demethoxyviridin (DMV) have been reported to be specific inhibitors of phosphatidylinositol 3-kinase activity. Here we report that these compounds are not as selective as assumed and demonstrate inhibition of bombesin-stimulated phospholipase A 2 activity by both wortmannin and DMV with an IC 50 (2 nM) which is slightly more potent than the inhibition of insulin-stimulated phosphatidylinositol 3,4,5-trisphosphate generation in these cells (ϳ10 nM). While it has not been possible to fully block in vitro phospholipase A 2 activity with wortmannin, inhibition cannot be a consequence of inhibition of PI 3-kinase activity since bombesin fails to generate 3-phosphorylated lipids in the intact cell. Therefore, while wortmannin is indeed a PI 3-kinase inhibitor, it is not as specific as previously reported, and experimental conclusions based solely on its use should be treated with caution.
The fungal metabolite wortmannin and its structural analogue demethoxyviridin (DMV) 1 have been demonstrated to have an inhibitory effect upon phosphatidylinositol 3-kinase (PI 3-kinase) activity at nanomolar concentrations (1). Inhibition has been demonstrated upon both PI 3-kinase activity in anti-p85 immunoprecipitates and the stimulation of phosphatidylinositol 3,4,5-trisphosphate (PIP 3 ) generation in N-formylmethionylleucylphenylalanine (fMLP)-stimulated neutrophils (1). The binding of 17-[ 3 H]hydroxywortmannin to a number of neutrophil proteins has been demonstrated, and one of these has been identified as the 110-kDa subunit of PI 3-kinase (2).
On the basis of these findings, it has been proposed that wortmannin and DMV are specific inhibitors of PI 3-kinase, and, thus, addition of these compounds to cells will result in the specific inhibition of the PI 3-kinase pathway. Consequently, an increasing number of papers have described the use of wortmannin and, on the sole basis of such experiments, assigned a role for PI 3-kinase in a number of physiological responses (e.g. Refs. [3][4][5]. Wortmannin has also been reported to inhibit phospholipase D (PLD) (6), myosin light chain kinase, and plekstrin phosphorylation (see Ref. 1). Although these results could cast doubt upon the specificity of wortmannin, these effects have been shown to occur at concentrations greater than those reported to inhibit PI 3-kinase. However, the specificity of wortmannin for other lipid-metabolizing enzymes has not been examined. In this paper, we have examined the specificity of wortmannin and DMV and show that they inhibit stimulated PIP 2 -phospholipase C (PI-PLC), PLD, phospholipase A 2 (PLA 2 ) as well as PI 3-kinase in Swiss 3T3 cells and, in addition, in vitro PI 3-kinase and PLA 2 activities. We also demonstrate that both compounds are more potent inhibitors of PLA 2 than PI 3-kinase.

EXPERIMENTAL PROCEDURES
Materials-Radiochemicals were from Amersham International plc, except for the inositol phosphate standards which were from DuPont NEN, all tissue culture media and sera were from Life Technologies, Inc., thin layer chromatography plates and HPLC columns were from Whatman, and lipids were from Sigma, Avanti Polar Lipids, or Lipid Products, Nutley, Surrey, UK. Wortmannin was from Sigma, and DMV was a generous gift from Dr G. MacAully, Dept. of Chemistry, Glasgow University, UK. Anti-p85, clone U13 was from Serotec. SF9 cellexpressed human cytosolic PLA 2 (85-kDa enzyme) was a generous gift from Dr. C. Jackson, Fisons Pharmaceuticals, Loughborough, UK.
Culture and Labeling of Cells-Swiss 3T3 cells were cultured in Dulbecco's modified Eagle's medium containing 10% newborn calf serum at 37°C in a humidified atmosphere of 5% (v/v) CO 2 in air, all cells were quiesced in 2% (v/v) serum-containing medium for 24 h, with radiolabel as appropriate, prior to experiment. The cells were labeled in medium containing 2% (v/v) serum with  (7)(8)(9).
PI 3-Kinase Activity-Confluent quiesced cells were washed in Hanks' buffered saline containing 10 mM glucose and 0.1% bovine serum albumin, stimulated for the required period of time, washed with ice-cold phosphate-buffered saline, and then lysed in 1% (w/v) Nonidet P-40, 10% (v/v) glycerol, 20 mM Tris-HCl (pH 8), 137 mM NaCl, 1 mM MgCl 2 , 1 mM CaCl 2 , 0.5 mM Na 3 VO 4 , containing 10 g/ml leupeptin and 0.2 mM phenylmethanesulfonyl fluoride. The lysates were cleared by centrifugation, and their protein concentrations were determined by the BCA method (Sigma). A solution containing 100 g of protein was incubated with 15 l of anti-p85 PI 3-kinase subunit antibody (U13 hybridoma supernatant) for 2 h at 4°C, and 20 l of 50% (v/v) Protein G-Sepharose was then added for 2 h at 4°C. The immunoprecipitates were washed successively at 4°C as follows: 2 ϫ 1 ml of lysis buffer, 2 ϫ 1 ml of 0.5 M LiCl, 0.1 M Tris-HCl, pH 8.0, 1 ϫ 1 ml of 0.1 M NaCl, 1 * This work was supported by grants from the Wellcome Trust, the Medical Research Council UK and the Cancer Research Campaign. 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.
3-Phosphorylated Lipids-Cells labeled with [ 32 P]phosphate were washed twice with phosphate-free Dulbecco's modified Eagle's medium and then stimulated. Incubations were terminated by the addition of 4 ml of ice-cold methanol, 1 M HCl (3:1, v/v). The cells were scraped from the flask, and total lipids were extracted. They were then deacylated with methylamine and deglycerated using mild sodium periodate treatment (10 mM for 20 min in the dark) (11). The lipid-derived head groups were reconstituted in 1 ml of water, dried in vacuo, and reconstituted in an additional 1 ml of water. The phosphate-containing head groups were then separated by anion exchange HPLC on a 25-cm Partisphere 5 SAX column eluted with a linear gradient of ammonium dihydrogen phosphate (0 -0.5 M, pH 3.8 with phosphoric acid) in water at 1 ml/min over 110 min. Fractions were collected every 0.5 min, and 32 P radioactivity was determined by static liquid scintillation spectrometry. The retention times of radioactive peaks were compared to authentic external 3 H standards of Ins-1-P, Ins(1,4)P 2 , Ins(1,3,4)P 3 , Ins(1,4,5)P 3 , Ins(1,3,4,5)P 4 , and 32 P i . Standards, run every 5th injection, were detected by on-line liquid scintillation spectrometry (Radiomatic A500, Canberra Packard). The retention times of peaks and external standards throughout all the runs varied by less than 1 min.
Preparation of Demethoxyviridin (DMV) and Wortmannin-DMV and wortmannin were dissolved in dimethyl sulfoxide. For the phospholipase studies using whole cells, the inhibitors were diluted in buffer immediately prior to addition in the assay. A 5-min preincubation with the inhibitors was used prior to the addition of agonist. The level of dimethyl sulfoxide was below 0.1% (v/v) which had no observable effect in any assay.

RESULTS AND DISCUSSION
In the Swiss 3T3 fibroblast cell line, bombesin stimulates the hydrolysis of phosphatidylinositol 4,5-bisphosphate by phospholipase C (PI-PLC) and phosphatidylcholine by phospholipase D (PLD) and phospholipase A 2 (PLA 2 ) (12). Bombesin has been reported to have no effect upon the activation of PI 3-kinase in Swiss 3T3 cells (10); therefore, insulin was used to stimulate PI 3-kinase activity.
DMV inhibited bombesin-stimulated PI-PLC activity, as determined by inositol phosphate accumulation, in Swiss 3T3 cells with an IC 50 value of approximately 30 nM (Table I). The wortmannin analogue also inhibited phosphatidylcholine-PLD activity, as measured by phosphatidylalcohol generation, with an IC 50 of approximately 50 nM, although maximum inhibition was never greater than 60% (Table I). Similar potency values were reported for fMLP-stimulated neutrophils (6) except that the sensitivity of the PLC response to DMV was greater in the 3T3 cells, and, in the neutrophil, the PLD response was completely inhibited. These differences may reflect different phospholipase isoforms in the different cell types.
DMV was an extremely potent inhibitor of bombesin-stimulated phosphatidylcholine-PLA 2 activity in Swiss 3T3 cells. Fig.  1 shows that inhibition was observable at concentrations as low as 0.1 nM, and that the IC 50 was approximately 2 nM. This inhibition of PLA 2 was also observed with wortmannin ( Fig. 1), demonstrating the potency of this class of compounds as PLA 2 inhibitors. Inhibition of PLA 2 was also observed in an in vitro assay. However, inhibition was only partial with approximately 30% inhibition being detected at 0.1 nM (Table II). This inhibition was dependent upon the inclusion of PIP 2 in the assay, which has been reported to enhance phospholipase A 2 activity (13). Incubation of a Swiss 3T3 cell lysate or an anti-PLA 2 immunoprecipitate with wortmannin for 10 min prior to addition of the lipid substrate failed to induce any PLA 2 inhibition in vitro.
Bombesin has been reported not to stimulate PI 3-kinase activity in Swiss 3T3 cells (10), and we have confirmed this observation both by assaying the activity of immunoprecipitated kinase (Fig. 2) and by determining changes in PIP 3 levels (see Fig. 3). Therefore, we examined the effects of wortmannin and DMV in insulin-stimulated cells. Fig. 2 shows that insulin and platelet-derived growth factor, but not bombesin, increase the PI 3-kinase activity in anti-p85 immunoprecipitates. In the in vitro PI 3-kinase assay, where the insulin-stimulated kinase was immunoprecipitated using an anti-p85 subunit antibody, DMV inhibited enzyme activity with an IC 50 value between 1 and 2 nM (Fig. 2). This is similar to the value of 3.4 nM previously reported (14) and compares with reported IC 50 values between 1 and 10 nM for PI 3-kinase  inhibition by wortmannin (14).

FIG. 1. Inhibition of bombesin-stimulated phospholipase A 2 in Swiss 3T3 cells by increasing concentrations of DMV (A) and wortmannin (B). Phospholipase
Inhibition of PI 3-kinase activity in anti-p85 immunoprecipitates where phosphatidylinositol rather than phosphatidylinositol 4,5-bisphosphate is presented as the substrate can only give an indication of the effects of a compound in the whole cell. The only definitive measure of effects upon in vivo PI 3-kinase activity is to examine changes in the cellular levels of the 3-phosphorylated lipids. Therefore, lipids were isolated from control and stimulated [ 32 P]P i -labeled cells, deacylated, and deglycerated, and the generated water-soluble phosphates were separated by ion exchange HPLC. Fig. 3 shows that insulin stimulates PIP 3 generation in Swiss 3T3 cells, the figure also shows that bombesin did not stimulate PIP 3 generation in agreement with Jackson et al. (10). Insulin-stimulated PIP 3 generation was inhibited by both DMV and wortmannin ( Fig. 3) with DMV being slightly more potent with an approximate IC 50 of 10 nM compared to a value greater than 10 nM for wortmannin (Fig. 4). The more potent effect of DMV compared to wortmannin upon 3-phosphorylated lipid generation is similar to that reported for effects upon PI 3-kinase activity in immunoprecipitates (14).
The data presented in Fig. 1 clearly define wortmannin and DMV as potent and complete inhibitors of stimulated phospholipase A 2 activity in Swiss 3T3 cells. However, in vitro inhibition was incomplete when the enzyme was activated by PIP 2 , diacylglycerol, and calcium (Table II). The contrast between the in vitro and in vivo inhibition of phospholipase A 2 suggests a complex mode of action for wortmannin and DMV. It is possible that they interact directly with the phospholipase and that the incomplete effect was due to a nonphysiological conformation taken up by the protein in the test tube. An alternative possibility is that inhibition may be due to an interaction with a protein which regulates phospholipase A 2 activity in the stimulated cell. Should this be the case, it is unlikely that the protein involved is PI 3-kinase since in Swiss 3T3 cells bombesin activates phospholipase A 2 without stimulating PI 3-kinase. Additionally, when stimulated phospholipase A 2 and PI 3-kinase activities are inhibited in vivo by wortmannin and DMV, both compounds are more potent against the release of arachidonate than the generation of 3-phosphorylated lipids.  The potency of DMV was greater than that of wortmannin for each of the enzymes examined. However, this difference is offset by the reduced stability of DMV when compared to wortmannin. Indeed, when DMV was incubated in a physiological buffer at 37°C for 10 min prior to addition to cells, it was rendered inactive (results not shown). In view of the instability of these compounds in aqueous solution, their experimental and therapeutic use will remain limited until the development of more stable analogues.
The results presented in this paper demonstrate that wortmannin and its structural analogue DMV are not specific or selective inhibitors of PI 3-kinase in intact cells. Inhibition of other enzymes by wortmannin, e.g. phospholipase D, myosin light chain kinase, has been reported previously; however, the concentrations required were much greater than those reported to inhibit PI 3-kinase. The inhibition of bombesin-stimulated phospholipase A 2 activity in Swiss 3T3 cells is apparent at subnanomolar concentrations and has an IC 50 (approximately 5 nM) which is similar to that previously reported for the inhibition of PI 3-kinase in vitro and lower than that determined for the inhibition of insulin-stimulated PIP 3 generation (Fig. 4). This value is also similar to that measured for the inhibition of PI 3-kinase activity in anti-p85 immunoprecipitates from Swiss 3T3 cells (Fig. 2). The proposed specificity of action of wortmannin upon PI 3-kinase has been supported by the demonstration of wortmannin binding to a 110-kDa protein in an SDS-polyacrylamide gel and to the 110-kDa subunit of PI 3-kinase (2). However, the 85-kDa phospholipase A 2 migrates at 110 kDa on SDS-polyacrylamide gel electrophoresis, highlighting the possibility that one of the bands shown to bind [ 3 H]wortmannin could be phospholipase A 2 . Accordingly, where a role for PI 3-kinase in a particular pathway or cellular function has been assigned on the basis purely of its inhibition by wortmannin, the conclusions may be incorrect and a role for other signaling pathways such as phospholipase A 2 must be considered. A clear example of this is in the regulation of neutrophil responses where a central role for PI 3-kinase has been proposed on the basis of wortmannin sensitivity (1). Other studies have suggested a role for the phospholipase A 2 product arachidonate in the activation of neutrophils (15), and, thus, the nonselective action of wortmannin would suggest a need for a reassessment. Many papers reporting the use of wortmannin have utilized concentrations of 100 nM and above; however, at these levels, as the results presented here demonstrate, there will be inhibition of PLC, PLD, PLA 2 , and PI 3-kinase. Despite these caveats, it is clear that in many cases where wortmannin has been used to identify a role for PI 3-kinase the conclusions are correct, but these studies have utilized further experimental evidence, e.g. dominant negative p85 subunit transfections (16), to confirm their findings. In other cases, it could be phospholipase A 2 that is the enzyme of importance.