Quantitative changes in polyphosphoinositides 1,2-diacylglycerol and inositol 1,4,5-trisphosphate by platelet-derived growth factor and prostaglandin F2 alpha.

We developed a novel method to quantify trace amounts of phosphatidylinositol 4-phosphate (PIP) and phosphatidylinositol 4,5-bisphosphate (PIP2) using antibodies against PIP and PIP2. With this method, polyphosphoinositides can be measured in the range from 20 to 500 pmol. We applied the method to quantify changes in PIP and PIP2 levels in Balb/c/3T3 cells stimulated by platelet-derived growth factor (PDGF) and prostaglandin F2 alpha (PGF2 alpha), growth factors that stimulate the hydrolysis of PIP and PIP2. PIP2 content decreased rapidly to about 60% of control within 1 min while PIP content decreased gradually but significantly to 60% (PDGF) or 70% (PGF2 alpha) of control. Simultaneously we measured the mass levels of inositol 1,4,5-trisphosphate and 1,2-diacylglycerol (DG). Inositol 1,4,5-trisphosphate levels rapidly increased and reached a maximum at 30 s after PDGF or PGF2 alpha stimulation and then decreased to the control level within 2 min. On the other hand, DG formation showed biphasic changes. In the first phase, DG rapidly accumulated and reached a maximum at 30 s after PDGF or PGF2 alpha stimulation and then quickly decreased. In the second phase, DG accumulated gradually, but very markedly, 2 min after PDGF or PGF2 alpha stimulation. Considering the changes in PIP2, DG in the first phase seems to be derived mainly from PIP2 while most of the DG in the second phase derived from other lipids.

' The abbreviations used are: PIP2, phosphatidylinositol 4,5-bisphosphate; PA, phosphatidic acid; PI, phosphatidylinositol; PIP, phosphatidylinositol 4-phosphate; IPS, inositol 1,4,5-trisphosphate; DG, 1,2-diacylglycerol; PDGF, platelet-derived growth factor; PGF2,, prostaglandin F2m; ELISA, enzyme-linked immunostaining assay. only metabolical changes have been measured due to the inability to detect low mass amounts of polyphosphoinositides, IPS, and DG. These studies used the radiolabeling techniques, which provide useful qualitative information, but show only metabolical changes rather than quantitative changes in these materials (3, 13,14). Recently, new methods to measure trace amounts of IP3 and DG have been developed (15, 16). Determination of polyphosphoinositide levels in tissue culture also requires a new method capable of detecting extremely low levels of these lipids.
Recently, we developed a monoclonal antibody against PIP2 and demonstrated that the antibody abolishes the mitogenic effect of PDGF and bombesin, suggesting the essential role of PIPz in the action of these mitogens (17,18). Now we have also developed a monoclonal antibody against phosphatidylinositol 4-phosphate (PIP) in addition to the anti-PIP, antibody. Using these antibodies, we established a sensitive assay method capable of measuring mass amounts of PIP2 and PIP. Next, we have used these assays to examine the kinetic changes in PIP2 and PIP when Balb/c/3T3 cells are stimulated by PDGF and PGF2,. In addition, we have measured the quantitative changes in IP3 and DG following PDGF or PGF, stimulation.
Cell Culture"Balb/c/3T3 cells were maintained at 37 "C in Dulbecco's modified Eagle's medium supplemented with 10% calf serum.
The cells were plated at 6.4 X lo5 cells in 9-em dishes. Two days after plating, subconfluent cells were serum-starved by washing 2 times with Dulbecco's modified Eagle's medium containing 5 pg/ml transferrin and 100 gg/ml bovine serum albumin and then incubated for 24 h at 37 "C in the same medium.
Production of Anti-PIP Antibody-Anti-PIP antibody was made by a method similar to that described before (18). Briefly, BALB/c mice were immunized with liposomes containing dimyristoyl phosphatidylcholine, cholesterol, PIP (0.5 mg), and lipid A (molar ratio, 1:1.5:1:0.08) every 2 weeks for 4 months. The immune spleen cells were fused with P3-X63-Ag8 cells using polyethylene glycol. Two weeks after cell hybridization, hybridoma supernatants were evaluated by ELISA. The hybridoma cells secreting anti-PIP antibody were doubly cloned by limiting dilution. The cells were inoculated into pristane-primed mice, and the resulting ascites fluid was used to measure PIP content. The antibody, named KD2, was found to be IgM by subclass analysis. The specificity of the antibody was evaluated by ELISA and TLC immunostaining as described before (18).
Mass Analysis of PIP and PZP2-Cells treated with growth factors were incubated at 30 "C for the indicated time, and then the incubations were terminated by aspirating the medium, washing with phosphate-buffered saline, and immediately adding 1.5 ml of ice-cold methanol containing 1 N HCl. The cells were scraped, and the dishes were further washed with 1.5 ml of chloroform/methanol (1:2, v/v). After adding 1.5 ml of chloroform, lipids were then extracted and washed with 1 N HCI as described before (20). The lipids were spotted on a TLC polygram (Macherey-Nagel) and then developed with chloroform/methanol/aqueous ammonia/water (9065:12:8). The TLC plate was soaked overnight in phosphate-buffered saline containing 3% bovine serum albumin, 1% polyvinylpyrrolidone 40. The plate was then treated for 2 h at room temperature with ascites of anti-PIP and anti-PIP2 antibodies diluted 100-fold with PBS. Treated plates were washed 4-5 times with 50 mM Tris/HC1 buffer (pH 7.4) containing 0.15 M NaCl and 0.05% Tween 20, treated with peroxidaseconjugated anti-mouse immunoglobulins, and stained with a Konica staining kit (Konica, Tokyo). The contents of PIP and PIP2 were measured by an image processer system (ADS Co., Nara, Japan).
Mass Analysis of ZP3-Cultures were incubated with growth factors for the indicated times at 30 "C, and incubations were terminated by aspirating the medium, washing with phosphate-buffered saline, and immediately adding 1 ml of trichloroacetic acid and then 1 ml of distilled water. Both fractions were collected in a tube. The extracts were centrifuged at 5000 X g for 10 min, and the resulting supernatant was washed 4 times with diethyl ether saturated with distilled water to remove trichloroacetic acid. The pH of the supernatant was adjusted to pH 7.4 with 10% NaHC03, and aliquots of the supernatant were used for IP3 analysis using an IP3 assay kit (Amersham Corp.).
Mass Analysis of DG-Cells were incubated with growth factors at 30 "C for the indicated times. The reactions were terminated by aspirating the culture medium, washing with PBS, and adding 1.5 ml of ice-cold methanol. The cells were scraped off, and the dishes were washed with 1.5 ml of chloroform/methanol (1:2, v/v). Lipids were extracted with 1.5 ml of chloroform and washed with 1 M KC1. An aliquot of the extracts (corresponding to 5 X lo5 cells) was used for DG analysis according to the method of Preiss et al. (16) using a DG assay kit.

RESULTS
Specificity of Anti-PIP and Anti-PIPz Antibodies and Application of the Antibodies to the Measurement of PIP and PIP2-Antibody KD2 bound to PIP very specifically. On ELISA, KD2 had little cross-reactivity with PIPz, PI, or other lipids (Fig. 1A). Moreover, TLC immunostaining ( Fig. 1B) showed that the antibody reacted only with PIP, even when rat brain total phospholipids (15 pg of phosphorus/spot) were used as antigens. The anti-PIP, antibody has already been demonstrated to react very specifically with PIP, (18). We tried to measure PIP and PIP, content by a TLC immunostaining assay method using the anti-PIP and anti-PIPp antibodies (Fig. 2). PIP and PIP, contents could be measured in the range from 20 to 500 pmol. To ascertain the reliability of this assay, a determined amount (20,50, 100, or 200 pmol of PIP and PIP,) was added to cells (1 x lo6), and then PIP and PIPz were extracted and quantitated. PIP and PIP2 levels were shifted in proportion to the amounts added to the original cells. In this system, recovery of PIP and PIPz was found to be more than 95%. This assay system enables us to measure trace amounts of PIP, and PIP.
Time Course of PIP and PIP2 Decreases by PDGF and PGFz,"Serum-starved quiescent cultures of Balb/c/3T3 cells contain PIP (130 pmol/106 cells) and PIP, (110 pmol/106 cells). Stimulation of these cells by 2 units/ml PDGF or 3.3 ~L M PGF2, caused a rapid decrease in PIP and PIPz as shown in Fig. 3. Especially, the decrease in PIP, was very marked, with levels decreasing very rapidly to about 60% of the prestimulation level within 1 min after stimulation and remaining almost the same for up to 20 min (PGF,,) or with a continuing slight decrease (PDGF). PIP levels decreased grad-  (0). Other lipids did not react with the antibody. B, TLC plates loaded with rat brain total lipid (15 pg of phosphorus) were developed with chloroform/methanol/aqueous ammonia/water (9065:12:8, v/v) in the first dimension and n-butyl alcohol/acetic acid/water (9015:15, v/v) in the second dimension, and immunostained with antibody KD2. To determine the position of PIP, [3H]PIP (2000 dpm/pmol) was added to total lipid extract. After immunostaining, the spot was scraped and counted for radioactivity. ually over 20 min to 60% of the original level following PDGF stimulation while in PGF2,-stimulated cells, the decrease in PIP reached a maximum 7 min after stimulation, and then the level gradually increased up to 20 min. The results suggest two possibilities. The first is that PIP, breakdown continues for a long time, while the resynthesis of PIP, is enhanced simultaneously. The second possibility is that PIP, breakdown is only transiently enhanced but that compensatory resynthesis is very slow. To examine these possibilities, the effect of PDGF or PGFp, on the synthesis of PIP and PIP, was studied. As shown in Fig. 4, PDGF and PGF,, caused enhanced incorporation of [32P]Pi into PA and PI. However, 32P incorporation into PIP and PIPz was not significantly increased by PDGF and only very slightly increased by PGF,,. These results suggest that the compensatory resynthesis of PIP and PIP, is not so active in Balb/c/3T3 cells. The doseresponse curve for PDGF and PGF2, on 32P incorporation was also examined. As shown in Fig. 5, even high doses of PDGF and PGFz, did not enhance the 32P incorporation into PIP and PIPz, although the incorporation into PI and PA was dramatically increased at doses of PDGF greater than 0.5 unit/ml or of PGF2, greater than 0.033 p~. On the other hand, mass levels of PIP2 and PIP were markedly decreased at doses higher than 0.5 unit/ml PDGF or 0.033 PM PGFz, (Fig. 6).

Time Course of IPS and DG Formation by PDGF or PGFZa-
Both PDGF and PGFza caused a rapid accumulation of IPS as shown in Fig. 7. IP3 levels reached a maximum at 30 s after stimulation and then decreased very rapidly to control level. At 30 s after stimulation, the amounts of accumulated IP3 were about 13 pmol/106 cells in PDGF-stimulated cells and 16 pmol/106 cells in PGF2,-stimulated cells. On the other hand, these growth factors induced a biphasic accumulation of DG (Fig. 8). The early phase showed a transient peak of DG at 30 s with a rapid decline between 30 and 60 s. In the second phase, DG accumulated gradually up to 7 min (PGFZ,) or up to 20 min (PDGF). The increase in DG 30 s after PDGF stimulation was approximately 40 pmol/106 cells; at 20 min after stimulation the increase was 140 pmol/106 cells. DG levels 30 s after PGF2, stimulation were increased by about 40 pmol/106 cells and maximally increased by 120 pmol/106 cells after 7 min.

DISCUSSION
The novel methods we developed enabled us to measure trace amounts of polyphosphoinositides. These novel methods also produce different results from the [3H]inositol-labeling methods so far used (3,13,14). Several attempts to quantify polyphosphoinositides have been made using steady-state labeling techniques. However, the labeling efficiency of polyphosphoinositide with [3H]myoinositol is very low, perhaps due to the metabolically inactive pools (21). Therefore, it has been difficult to determine the precise levels of polyphosphoinositides. We have measured the mass levels of signal-producing lipids and two second messengers, IPS and DG, since Quiescent cells (2 X lo6) were treated without (0)  also reported the biphasic accumulation of DG following a-thrombin stimulation in Chinese hamster embryo fibroblast, suggesting that DG is produced through a pathway independent of PIP2 hydrolysis in the second phase. Considering the changes in PIP2 and PIP levels, DG formation in the first phase, which is finished within 1 min, is caused by PIPz hydrolysis, since the decrease in PIPz is almost equivalent to the increase in DG in both PDGFit is thought that the absolute content is the most important gradually up to 7 min, but then the levels begin to recover. factor in deciding the strength of the signals. Our novel However, compensatory resynthesis of polyphosphoinositides methods show that the decrease in PIP and PIPz content by is not so active in Balb/c/3T3 cells (Fig. 4). Most of the DG PDGF or PGFz, stimulation is more marked than that pre-formed in the second phase, therefore, may be derived from viously found by steady-state labeling methods (3,12,13). other lipids. But in cells stimulated by PDGF, there is the PIP2 content decreased by almost 40% within 1 min following possibility that a part of the DG still arises from PIP,. The difference might be due to the different mechanisms of action between PDGF and PGF, (7). We previously demonstrated that the mitogenic effect of PDGF or bombesin in NIH/3T3 cells is abolished by anti-PIP2 antibody (17). As well, in Balb/ c/3T3 cells the antibody inhibits PDGF-induced cell proliferation (data not shown) showing that PIPz plays a key role in PDGF-induced mitogenesis. However, there have been several reports suggesting that PIPz hydrolysis is not essential for cell growth (22-24). Although it is not clear whether the discrepancy depends only on the type of cells examined, our novel method for quantifying PIP, and PIP may afford a key to explaining the discrepancy. IPS accumulation is very transient even in the presence of LiC1, suggesting rapid metabolism of IP3, while PIP, hydrolysis continues (especially in the case of PDGF stimulation) after IP3 levels return to basal levels (Fig. 3). Therefore, it is difficult to know whether PIPz hydrolysis occurs solely through the measurement of IP3 formation. Also in ras-transformed cells, PIP, content is significantly low compared with that in normal cells, although IP3 levels do not differ.' T o measure PIPz levels may be one of the best ways to solve the problem. Besides the role as signal-producing lipids, polyphosphoinositides have been implicated in cell motility through modulation of profilactin and gelsolin which regulate actin polymerization (25, 26). Therefore, the development of assay methods for polyphosphoinositides may also supply a powerful tool to study the involvement of these lipids in cell motility.