HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Maier, U. Articles by Nürnberg, B. Search for Related Content PubMed PubMed Citation Articles by Maier, U. Articles by Nürnberg, B. Social Bookmarking                      What's this?

J Biol Chem, Vol. 274, Issue 41, 29311-29317, October 8, 1999

Roles of Non-catalytic Subunits in G-induced Activation of Class I Phosphoinositide 3-Kinase Isoforms  and ^*

Udo Maier, Aleksei Babich, and Bernd Nürnberg^§

From the Institut für Pharmakologie, Freie Universität Berlin, Thielallee 69-73, D-14195 Berlin (Dahlem), Germany

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

By using purified preparations we show that nanomolar concentrations of G significantly stimulated lipid kinase activity of phosphatidylinositol 3-kinase (PI3K)  and PI3K in the presence as well as in the absence of non-catalytic subunits such as p85 or p101. Concomitantly, G stimulated autophosphorylation of the catalytic subunit of PI3K (EC₅₀, 30 nM; stoichiometry 0.6 mol of P_i/mol of p110), which also occurred in the absence of p101. Surprisingly, we found that p101 affected the lipid substrate preference of PI3K in its G-stimulated state. With phosphatidylinositol as substrate, p110 but not p101/p110 was significantly stimulated by G to form PI-3-phosphate (EC₅₀, 20 nM). The opposite situation was found when PI-4,5-bisphosphate served as substrate. G efficiently and potently (EC₅₀, 5 nM) activated the p101/p110 heterodimer but negligibly stimulated the p110 monomer to form PI-3,4,5-trisphosphate. However, this weak stimulatory effect on p110 was overcome by excess concentrations of G (EC₅₀, 100 nM). This finding is in accordance with the in vivo situation, where activated PI3K catalyzes the formation of PI-3,4,5-trisphosphate but not PI-3-phosphate. We conclude that p101 is responsible for PI-4,5-bisphosphate substrate selectivity of PI3K by sensitizing p110 toward G in the presence of PI-4,5-P₂.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Phosphoinositides are integral constituents of eukaryotic lipid bilayers but also play a crucial role in transmembrane signaling (1, 2). An exponent that sets up one-third of all phosphoinositides in mammalian cells is phosphatidylinositol-4,5-bisphosphate (PI-4,5-P₂)¹ (3), which serves as a precursor for intracellular second messengers. On the one hand it is cleaved into inositol-1,4,5-P₃ and diacylglycerol by members of the phospholipase C family (4), which respond to receptor tyrosine kinases and G-protein-coupled receptors (GPCRs) (5, 6). On the other hand, the D-3 position of the inositol ring of PI-4,5-P₂ is sensitive to phosphorylation leading to phosphatidylinositol-3,4,5-trisphosphate (PI-3,4,5-P₃) (7). PI-3,4,5-P₃ is considered to act as a second messenger, since it is absent in quiescent cells but spikes instantly upon exposure to various stimuli (8, 9). Putative PI-3,4,5-P₃-dependent functions include regulation of cell proliferation, survival, cytoskeletal rearrangements, and vesicle trafficking (10-12). Hence, it is not surprising that PI-3,4,5-P₃ has also been implicated in pathophysiological processes leading to tumor growth and malignancy (13-15).

PI-3,4,5-P₃ is generated from PI-4,5-P₂ by members of a considerably large family of enzymes called class I phosphoinositide 3-kinases (PI3K) (16-20). They are heterodimers consisting of 110-120-kDa catalytic (p110, -, -, and -) and 50-100-kDa non-catalytic subunits (p85, -, p55, and p101), which are also capable of phosphorylating PI and PI-4-P in vitro, although they are assumed to exhibit a preference for PI-4,5-P₂ within the cell (21, 22). In contrast, class II and class III PI3Ks show a more restrictive lipid specificity (23).

Activation of class I PI3Ks is observed in response to a wide array of cellular ligands including hormones, neurotransmitters, growth factors, and cytokines. Although an ever increasing number of cellular responses are elicited by these lipid kinases, a remarkable degree of specificity is maintained within pleiotropic PI3K-dependent signaling pathways, allowing the assignment of intracellular effects to extracellular stimuli. This surprising specificity is certainly due to a structural heterogeneity of class I PI3K isoforms in concert with different expression patterns as well as spatial and/or temporal compartmentation. Based on their tight association with non-catalytic binding proteins, catalytic subunits of class I PI3Ks are subdivided into class I_A p85- or class I_B p101-associated heterodimers. Until recently this grouping went strictly parallel with the isoform-specific regulation by different signaling pathways. Whereas class I_A-isoforms (p110, -, and -) were assumed to be sensitive to tyrosine kinases, the only class I_B member p110 is activated by G-proteins. This structure-function relationship was challenged by reports detailing a synergistic activation of PI3K and other so far unidentified isoforms by either regulator (24-26). In line with their sensitivity toward tyrosine kinases, the catalytic p110 subunits of PI3K, -, and - are associated with p85 adaptors, which are indispensable for activation (27). These adaptors harbor various regions such as SH2-, SH3-, or proline-rich domains, which contribute to further selectivity within tyrosine kinase-dependent signaling pathways (18, 28-30). In contrast, the G-protein-regulated p110 does not bind to p85 adaptors but instead associates with a non-catalytic p101 subunit (31, 32). Surprisingly, p101 shows no homology with any known protein. Nevertheless at first it was assumed to serve as an adaptor for G liberated by GPCRs. However, we and others (26, 31, 33) have found significant stimulation of the purified catalytic subunit of PI3K by G in the absence of p101. Thus, the function of this p101 non-catalytic subunit of PI3K has yet to be identified.

Versatility in PI3K-dependent signaling appears to be accomplished by a second enzymatic function, i.e. protein-serine kinase activity, which is inherent to the class I PI3Ks (34). Accordingly, first evidence emerged that cellular signaling bifurcates at the level of the PI3K (35). Whereas lipid kinase activity of PI3K stimulated protein kinase B, its protein kinase activity signaled to the extracellular signal-regulated kinase/mitogen-activated protein kinase pathway. In this context recent reports are of interest which suggested that the non-catalytic p101 subunit was crucial for supporting p110-induced protein kinase B and c-Jun amino-terminal kinase activation but had only little effect on extracellular signal-regulated kinase/mitogen-activated protein kinase activation (36, 37). Therefore, p101 may function as a regulatory subunit of PI3K that differently modulates protein and lipid kinase activity of PI3K.

The intention of the present study was to examine the sensitivity of PI3Ks toward G. We therefore purified recombinant proteins including all four known class I PI3K isoforms. Monomeric and heterodimeric enzyme preparations were analyzed for the role of non-catalytic subunits as adaptors for G-induced stimulation of PI3Ks. Furthermore, this experimental approach was successfully applied to assign a novel functional role to the p101 subunit of PI3K.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Recombinant Proteins-- Construction of recombinant baculoviruses for expression of PI3K subunits was described previously (31, 33, 38). For protein expression, cells were incubated at a multiplicity of infection (m.o.i.) of 1 virus per cell. Subunits of heterodimeric PI3Ks were coexpressed at equal m.o.i. numbers in Sf9 cells and used for functional studies. After 48-60 h of infection cells were pelleted by centrifugation (1,000 × g) and washed with phosphate-buffered saline twice. For purification of GST fusion proteins cells were resuspended in ice-cold buffer A containing 150 mM NaCl, 1 mM NaF, 1 mM EDTA, 50 mM Tris/HCl, pH 8.0, 10 mM dithiothreitol, 10 µg/ml each of L-1-tosylamido-2-phenylethyl chloromethyl ketone, benzamidine, leupeptin, and 0.2 mM Pefabloc® SC (Roche Molecular Biochemicals). They were disrupted by N₂ cavitation (30 min at 4 °C, 25 bar) or by forcing the cell suspension through a 22-gauge needle (5 times) and subsequently through a 26-gauge needle (10 times). Nuclei and debris were discarded. The cytosolic fraction was separated from the particulate by centrifugation at 100,000 × g for 50 min. Cytosol was incubated 3-4 h with glutathione-Sepharose 4B beads (Amersham Pharmacia Biotech) prewashed with buffer A. The Sepharose-bound GST fusion proteins were stored at 20 °C in buffer B containing 50% glycerol, 1 mM EDTA, 40 mM Tris/HCl, pH 8.0, 1 mM dithiothreitol, and 1.57 mg/ml benzamidine. For enzymatic assays GST fusion proteins were freshly eluted with buffer C consisting of buffer A with 10 mM glutathione for 1 h at 4 °C. For purification of hexa-His-tagged PI3K cells were disrupted using buffer D (20 mM HEPES, 150 mM NaCl, 10 mM -mercaptoethanol) containing 10 µg/ml each of L-1-tosylamido-2-phenylethyl chloromethyl ketone, benzamidine, leupeptin, and 0.2 mM Pefabloc^® SC (Roche Molecular Biochemicals) using the same procedure as described above. The cytosolic fraction was incubated 1-2 h with Ni²⁺-NTA-agarose (Qiagen, Hilden, Germany) prewashed with buffer D containing 20 mM imidazole. After extensive washing with buffer D, proteins were eluted with buffer D containing 150 mM imidazole. Purified proteins were quantified by Coomassie Blue staining following SDS-PAGE with bovine serum albumin as the standard. For coexpression experiments with recombinant subunits of PI3K, i.e. p101, p110, and mutants thereof, equal m.o.i. numbers for all recombinant baculoviruses were used. Cell lysates were incubated with glutathione-Sepharose 4B, and eluted GST fusion proteins were analyzed for binding of p101. For coexpression experiments with recombinant G₁₂, cell lysates were obtained by forcing the Sf9 cell suspension through needles as described above and subsequent incubation (30 min) with buffer A supplemented with 0,5% Lubrol PX. After incubation with glutathione-Sepharose 4B eluted proteins were analyzed for bound G.

Preparation of G Subunits and Phosphotyrosyl Peptides-- For isolation of bovine brain G, we employed standard techniques with modifications. Bovine brain G-proteins were purified to apparent homogeneity in the presence of aluminum fluoride. Isolation and final purification of G was achieved using a Mono Q (Amersham Pharmacia Biotech) fast protein liquid chromatography column (39). G complexes were identified by their immunoreactivity to specific antisera and quantified by the method of Lowry and by Coomassie Blue staining of G following SDS-PAGE with bovine serum albumin as the standard. Contamination by pertussis toxin-sensitive (PT) G subunits was excluded by analysis of autoradiographic signals after PT-mediated [³²P]ADP ribosylation with a BAS 1500 Fuji-Imager (Raytest, Straubenhardt, Germany). Purified G were highly concentrated and contained Lubrol PX (0.1%) and CHAPS (11 mM) as detergents and were stored at 70 °C until use. The doubly tyrosine-phosphorylated peptide used in this study, CGGY(P)MDMSKDESVDY(P)VPMLDM, was based on that of the human platelet-derived growth factor receptor (40) and supplied by Schering AG, Berlin. A non-phosphorylated peptide was used as a control and had no effect.

Gel Electrophoresis, Immunoblotting, and Antibodies-- Characterization of the monoclonal antibody against p110, antisera against p101, and G subunits (AS 398) were detailed elsewhere (33, 41). The polyclonal anti-GST antibody was purchased from Santa Cruz Biotechnology (Heidelberg, Germany). For detection of GST fusion proteins, p110, p101, or G, proteins were fractionated by SDS-PAGE and transferred to nitrocellulose or polyvinylidene difluoride membranes (Millipore, Eschborn, Germany). Visualization of specific antisera was performed using the ECL chemiluminescence system (Amersham Pharmacia Biotech) or the CDP-Star chemiluminescence reagent (Tropix, Bedford, MA) according to the manufacturers' instructions.

Lipid Kinase Assay-- Lipid kinase activity was determined basically as detailed previously (33, 41). In brief, the assays were conducted in a final volume of 50 µl containing 0.1% bovine serum albumin, 1 mM EGTA, 120 mM NaCl, 40 mM HEPES, pH 7.4, 1 mM dithiothreitol, 1 mM -glycerophosphate (buffer E) as described with some modifications. Lipid vesicles (30 µl containing 320 µM phosphatidylethanolamine, 300 µM phosphatidylserine, 140 µM phosphatidylcholine, 30 µM sphingomyelin supplemented with either 300 µM PI or 40 µM PI-4,5-P₂ in buffer E) were mixed with stimuli as indicated and incubated on ice for 10 min. For measuring the effects of G on PI3K activity, it was ensured that stimuli-containing vehicles did not suppress enzymatic activity. In addition, assay samples containing different amounts of G were adjusted to identical detergent concentrations, i.e. 0.0004% for Lubrol PX and 0.044% for CHAPS. Mg²⁺ at 10 and 7 mM for PI and PI-4,5-P₂ as substrates, respectively, was added to lipids before sonification. For inhibition assays, kinase preparations were preincubated with wortmannin or 17-OH wortmannin (Schering AG) at 37 °C. Thereafter, the enzyme fraction (1-10 ng) was added, and the mixture was incubated for a further 10 min at 4 °C in a final volume of 40 µl. The assay was then started by adding 40 µM ATP (1 µCi of [-³²P]ATP, NEN Life Science Products) in 10 µl of the above assay buffer (30 °C). After 15 min the reaction was stopped with ice-cold 150 µl of 1 N HCl and placing tubes on ice. The lipids were extracted by vortexing samples with 500 µl of chloroform/methanol (1:1). After centrifugation the organic phase was washed twice with 200 µl of 1 N HCl. Subsequently, 40-80 µl of the organic phase were resolved on potassium oxalate-pretreated TLC plates (Whatman) with 35 ml of 2 N acetic acid and 65 ml of n-propyl alcohol as the mobile phase. Dried TLC plates were exposed to Fuji imaging plates, and autoradiographic signals were quantitated with a BAS 1500 Fuji-Imager.

Protein Kinase Assay-- The assays were performed as described for lipid kinase activity with some modifications. The assay volume was 25 µl (2-3 µCi of [-³²P]ATP per tube) and usually contained 7 mM Mg²⁺. Lipid vesicles were devoid of PI3-kinase lipid substrates such as PI or PI-4,5-P₂. The reaction was stopped with 25 µl of ice-cold 2× sample buffer according to Laemmli containing 10 mM ATP. Following separation on SDS-PAGE, proteins were transferred to nitrocellulose membranes. Dried membranes were exposed to Fuji imaging plates, and autoradiographic signals were quantitated with a BAS 1500 Fuji-Imager.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

G Sensitivity of Class I PI3Ks-- Based on the observation that the structural subdivision of class I PI3Ks does not correlate with regulation by different signaling pathways, all four members were examined for their G sensitivity. Therefore we isolated heterodimeric enzymes following expression in Sf9 cells (Fig. 1A). As previously reported for p110 (42), coexpression of p110, -, and - catalytic subunits together with p85 increased the amount of protein in Sf9 cytosolic fractions. In contrast, p101 did not enhance expression of p110 but preserved the catalytic subunit from rapid degradation during storage. In initial experiments we found no difference in the enzymatic activities of PI3K preparations, regardless whether the GST tag was fused to p110 or p101. For convenience we used p101-GST/p110 heterodimeric preparations throughout the study.

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Assignment of class I PI3Ks to receptor tyrosine kinase and GPCR-coupled pathways. A, heterodimeric PI3K isoforms were purified from Sf9 cells as detailed under "Experimental Procedures." Proteins were subjected to SDS-PAGE and analyzed by Coomassie staining. Apparent molecular masses of marker proteins are indicated. B, purified recombinant heterodimeric PI3K, -, -, and - were examined for their sensitivity toward G (150 nM; open bars), a tyrosine-phosphorylated peptide derived from the platelet-derived growth factor receptor (1 µM; gray bars), and synergistic activation by either stimulus (black bars). All experiments were carried out as detailed under "Experimental Procedures." Formation of PI-3,4,5-P3 out of PI-4,5-P2 is shown as fold stimulation of basal activity. Stimulations of PI3K and - by G alone ranged between 20- and 40-fold in all experiments, whereas synergistic activation of PI3K following incubation with both stimuli were consistently 100-200-fold. Basal activities of PI3K, -, and - ranged between 0.023 and 0.067 mol/min/mol enzyme. For PI3K a significantly higher basal activity was found, i.e. 0.154 ± 0.024 mol/min/mol enzyme. Shown is one typical experiment out of five. C, non-catalytic subunits of PI3K and PI3K are not required for stimulation by G. Monomeric (p110 and p110) and heterodimeric preparations (p85/p110 and p101/p110) of PI3K and - isoforms were assayed for their sensitivity toward G (150 nM) and tyrosine-phosphorylated peptides (1 µM). Formation of PI-3,4,5-P3 was measured upon incubation with the stimuli indicated as detailed under "Experimental Procedures." Maximal fold stimulations of responsive preparations were similar and ranged between 20 and 40-fold. Shown is one representative experiment out of three.

The isolated heterodimers were assayed for their sensitivity toward purified bovine brain G and phosphotyrosyl peptides resembling an intracellular p85-binding region of the platelet-derived growth factor receptor (Fig. 1B). As expected, PI3K was solely activated by G, whereas PI3K only responded to phosphotyrosyl peptides. However, coincubation of PI3K with both stimuli led to a remarkable synergistic activation of lipid kinase activity in accordance with previous reports (24). Moreover, unlike these published results, PI3K was also significantly activated by G alone, suggesting that PI3K represents an effector of G (see Fig. 1B). Interestingly, PI3K like PI3K was only responsive to phosphotyrosyl peptides but not to G, although p110 exhibits the highest degree of overall amino acid sequence identity to p110. It should be mentioned that all stimulatory effects were observed in a concentration-dependent manner (data not shown).

Non-catalytic Subunits of PI3K and PI3K Are Not Required for G Activation-- Since PI3K and - were sensitive to G but complexed with different non-catalytic subunits, we next examined the influence of p85 or p101 on G-induced activation of p110 and p110 lipid kinase activity, respectively (Fig. 1C). In initial experiments we have confirmed that p110 copurified with p85 but not with p101, whereas p110 showed the opposite association pattern to the non-catalytic subunits (data not shown and Ref. 31). In addition, we ascertained that the type of tag fused to p110, i.e. GST or polyhistidine, did not affect the enzymatic activity of the catalytic subunit. p110 as well as p110 were significantly stimulated by G irrespective of non-catalytic subunits (see Fig. 1C, upper panel). In contrast, sensitivity to phosphotyrosyl peptides was seen only when p110 was complexed with p85 but not when p110 was associated with p101 (see Fig. 1C, lower panel). Accordingly, in the absence of p85, costimulation of p110 by G together with phosphotyrosyl peptides showed no synergistic effect (data not shown). These findings provide unequivocal evidence that structural elements necessary for G-induced stimulation of PI3K isoforms are located on the catalytic subunits of PI3K and PI3K. Hence, p101 does not function as a p110-specific adaptor that is necessary for activation by Gs. This obviously contrasts with the role of p85 for tyrosine kinase-induced activation of PI3Ks.

As another putative parallel between class I_A and I_B PI3Ks, the p101-binding site on p110 was hypothesized to lie within a region analogous to the p85-binding site on class I_A members, i.e. the amino terminus (16, 43). To examine this assumption we studied the binding of p101 to a GST-fused full-length or an amino-terminally truncated construct of p110 (1-97) by copurification from Sf9 cells (Fig. 2A). The specificity of this approach was proven by expression of p101 with or without GST as a control. p101 was bound equally well by both p110 constructs suggesting that the non-catalytic subunit associates with regions of p110 outside the amino terminus. Nevertheless, as an indication that p101 may be still involved in G-induced activation of p110, we confirmed that p101 bound much stronger to G than p110 (Fig. 2B and Ref. 32).

View larger version (23K): [in this window] [in a new window]   Fig. 2.   A, binding of p101 to the p110 catalytic subunit of PI3K. The p101 subunit was coexpressed with GST fusion proteins of either full-length p110 or an amino-terminally truncated mutant thereof (1-97). As controls p101 was expressed alone or with GST. Expression of similar amounts of proteins were confirmed by immunoblotting cell lysates using specific antisera against GST and p101 (upper and center). Subsequently GST-fused proteins were purified on glutathione-Sepharose beads. Eluted fractions were subjected to SDS-PAGE and analyzed for bound p101 (bottom) by immunoblotting. B, binding of G12 to PI3K subunits. Recombinant G12 was coexpressed with constructs encoding p110-GST, p101-GST, or GST alone and purified as described under "Experimental Procedures." Lysates from Sf9 cells were analyzed for similar levels of expression of either GST constructs (top) or G (center) by immunoblotting using anti-GST and anti-G antisera. Following purification on glutathione-Sepharose, bound proteins were separated by SDS-PAGE and analyzed by immunoblotting with a G-specific antiserum AS 398 (bottom). Apparent molecular masses of marker proteins are indicated.

p101 Determines PI-4,5-P₂ Preference of G-stimulated p110-- Previous studies by two different laboratories have resulted in an apparent discrepancy concerning the function of p101 in G-induced activation of p110 (31-33, 44). However, some experimental conditions varied which may have contributed to the observed differences. To analyze the function of p101 in more detail, we measured the basal and G-stimulated activity of p110 in the presence and absence of p101 using PI or PI-4,5-P₂ as substrates. Under basal conditions p110 exhibited an approximately 2-fold higher specific enzymatic activity for PI than for PI-4,5-P₂. Furthermore, basal activity of p110 was hardly affected by p101 (Fig. 3A). Our observation was somewhat surprising since p101 was claimed to inhibit non-stimulated lipid kinase activity of p110 by up to 76% (32). This would be in parallel to the effect of p85 on p110 (29) which we confirmed (data not shown) and extended to p110 (basal catalytic activity of p110: 1.3 ± 0.6 mol/min/mol enzyme versus p85/p110: 0.05 ± 0.02 mol/min/mol enzyme). In contrast to the basal enzymatic situation of PI3K, the G-stimulated p101/p110 heterodimer was found to be a PI-4,5-P₂-selective enzyme, whereas G-induced PI-phosphorylation was observed only in the absence of p101 (see Fig. 3A). This not only explains the apparent discrepancy of results found in previous studies but also implies that p101 affects the substrate preference of the G-stimulated p110.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Substrate selectivity of PI3K. A, purified recombinant monomeric ("" p101) and heterodimeric PI3K ("+" p101) were stimulated with 30 nM G using PI or PI-4,5-P2 as substrate. 32P-Labeled lipid products were isolated and quantified as described under "Experimental Procedures." Shown are an autoradiography of one typical experiment and mean values (± S.D.) of three independent experiments. B and C, additionally concentration response curves of G-stimulating monomeric p110 () or heterodimeric p101/p110 () using PI (B) and PI-4,5-P2 (C) as substrate is shown. Shown are mean values (±S.D.).

These experiments were carried out using half-effective G concentrations (33, 41). We also examined whether this p101-regulated substrate preference of p110 could be overcome by excess G (Fig. 3, B and C). Employing PI as the substrate and monomeric p110, we confirmed our previous findings that purified bovine brain G significantly activated PI-3-P formation with half-maximal concentrations of about 20 nM (see Fig. 3B). Maximum effects were seen 100 nM G. Using heterodimeric p101/p110 resulted in similar G EC₅₀ values, although a significantly lower efficiency of stimulation was evident. However, the opposite situation was seen when phosphorylation of PI-4,5-P₂ was stimulated. G-induced formation of PI-3,4,5-P₃ was catalyzed by monomeric as well as heterodimeric enzymic preparations with similar efficacy but G differed remarkably in its potency (see Fig. 3C). The concentrations necessary for a similar extent of stimulation of monomeric p110 as compared with the heterodimeric enzyme were more than an order of magnitude higher, i.e. 100 and 500 nM G, to elicit half-maximal and maximal stimulation of p110, respectively. In contrast, for the p101/p110 heterodimer the mean EC₅₀ and the maximum effect were shifted to approximately 5 and 20 nM, respectively. This finding is in accordance with the in vivo-situation, where PI3K is assumed to occur as a heterodimer and G enhances formation of PI-3,4,5-P₃ but not of PI-3-P. Therefore, the non-catalytic subunit p101 functions in two ways to establish PI3K as a PI-4,5-P₂-selective enzyme upon stimulation by G, i.e. p101 suppresses G-induced PI-phosphorylation and concurrently facilitates the G-induced activation in the presence of PI-4,5-P₂.

Wortmannin Sensitivity of PI3K Isoforms-- By having found that the regulatory p101 subunit affected stimulation of p110 in a lipid substrate-dependent manner, we examined the influence of p101 on wortmannin-induced inhibition of p110 lipid kinase activity. The potency by which wortmannin inhibited PI3K was similar, irrespectively whether PI or PI-4,5-P₂ was used (IC₅₀, approximately 4 nM, Fig. 4). However, monomeric p110 was inhibited at slightly lower concentrations (0.75-1.75 nM) than heterodimeric p101/p110, suggesting that p101 may stabilize p110 or sterically hinders wortmannin from attacking p110. Considering the different IC₅₀ values of wortmannin found in this (4 nM) and previous studies (approximately 2-40 nM) (26, 40, 43), one has to recall that wortmannin acts in an irreversible manner by covalently binding to a lysine residue within the catalytic core (43, 45). Since wortmannin-induced inactivation of PI3Ks does not reach equilibrium during incubation, variations of the published IC₅₀ values are caused by the experimental conditions applied such as incubation time or temperature.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Sensitivity of monomeric and heterodimeric PI3K to wortmannin. Basal lipid kinase activity of p110 (open symbols) and p101/p110 (filled symbols) was inhibited by indicated concentrations of wortmannin using PI (circles) and PI-4,5-P2 (triangles) as substrate. Shown is one representative experiment out of three.

G Stimulates Protein Kinase Activity of PI3K-- Evidence is accumulating that GPCRs may signal also through protein kinase activity of PI3K, which anticipates that G-proteins regulate this enzymatic activity (35). However, a G-induced activation of PI3K protein kinase activity was not demonstrated. In order to prove if G is an appropriate stimulus for protein kinase activity, we isolated PI3K from cytosolic preparations following expression of p110 alone or together with p101 in Sf9 cells (Fig. 5A). Subsequently these protein preparations were incubated with [-³²P]ATP in the presence of divalent cations, i.e. Mg²⁺ or Mn²⁺, and subjected to SDS-PAGE. Overlay analysis of autoradiographies and immunoblots revealed that the catalytic p110 subunit but not p101 incorporated significant amounts of ³²P-labeled phosphate (Fig. 5, A and B). Initially, coincubation of PI3K with G did not stimulate autophosphorylation of p110 (not shown). Alternatively, the presence of lipid vesicles readily established conditions to study G-induced stimulation of protein kinase activity (Fig. 5B). G (250 nM) significantly stimulated autophosphorylation of purified p110 independent of p101 coexpression. Whereas protein kinase activity of p110, -, and - was reported to be predominantly Mn²⁺-dependent (38, 46-48), basal autophosphorylation of p110 has been shown in the presence of Mg²⁺ and Mn²⁺ (43). Therefore, we analyzed the influence of these two divalent cations on G-induced autophosphorylation of PI3K (Fig. 5C). Stimulated PI3K protein kinase activity was assessed by recording the intensity of p110 autophosphorylation which, unlike autophosphorylation of p110, did not affect enzymatic activity of the catalytic subunit (43, 49). The G-induced autophosphorylation was more pronounced in the presence of Mg²⁺ (see Fig. 5C). When G was coincubated with the heterodimer, stoichiometry of autophosphorylation was estimated to be at least 0.6 mol of phosphate per mol of p110.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Autophosphorylation of the catalytic subunit of PI3K is stimulated by G. A, recombinant p110-GST and p101-GST/p110 were expressed in Sf9 cells and affinity purified from cytosol. Proteins were separated by SDS-PAGE and Coomassie-stained or immunoblotted using antisera as indicated. Note that the wild-type catalytic subunit p110 is migrating at a lower apparent molecular weight as does the p101-GST fusion protein (lower panel, right lanes). B, purified recombinant monomeric and heterodimeric PI3K were assayed for incorporation of 32P phosphate in the absence and presence of G complexes isolated from bovine brain membranes as detailed under "Experimental Procedures." Proteins were separated on SDS-PAGE, and subsequent to Western blotting phosphorylation of proteins was quantified using a phosphorimaging system. In the absence of G, amounts of phosphate incorporation into the catalytic subunit were approximately 36 and 24 mmol of phosphate/mol of p110 for monomeric and heterodimeric PI3K, respectively. C, metal dependence of autophosphorylation of PI3K. Purified recombinant p101-GST/p110 PI3K heterodimer was assayed for protein phosphorylation using Mg2+ (circles) and Mn2+ (triangles) in the absence (open symbols) and presence (filled symbols) of G (250 nM). Shown are the results of one typical experiment out of two.

p101 Modulates of G-stimulated p110 Autophosphorylation-- In order to examine a regulatory role of p101, we studied the effect of G on p110 protein kinase activity in the presence and absence of p101 (Fig. 6). p101 enhanced the efficiency of G to stimulate autophosphorylation of p110 almost 5-fold. However, unlike the effect of p101 on G-induced PI-4,5-P₂-phosphorylation, it did not affect the potency by which G stimulated the protein kinase activity of p110 (EC₅₀, 15-30 nM). Interestingly, these G EC₅₀ values were in the same range as those found for stimulation of lipid phosphorylation (see Fig. 3, A and B), assuming that lipid substrate and autophosphorylation may occur simultaneously.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   G stimulates autophosphorylation of PI3K. Monomeric () and heterodimeric () recombinant PI3K purified from Sf9 cells were assayed for autophosphorylation in response to increasing concentrations of G. Incorporation of 32P into p110 catalytic subunit is illustrated as fold stimulation of basal activities. Shown are mean values (±S.D.) of four independent experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The present study was designed to examine class I PI3Ks as targets of G signaling. Here we demonstrate that two different PI3Ks are effectors of G-proteins by direct interaction of G with the catalytic subunits of PI3K and PI3K. Moreover, we collected compelling evidence that the non-catalytic subunit of PI3K, p101, is not a G adaptor necessary to convey kinase stimulation. Instead, as a major result of these studies, we postulate that p101 determines lipid substrate preference of PI3K by sensitizing p110 for G in the presence of PI-4,5-P₂.

Testing purified recombinant heterodimeric class I PI3Ks representing all known catalytic subunits revealed that two isoforms, PI3K as well as PI3K, responded to nanomolar concentrations of G, whereas three isoforms, PI3K, -, and -, were significantly stimulated by phosphotyrosyl peptides. The extent of lipid kinase activation by maximally effective concentrations of these stimuli were within the same range for all four isoforms tested. The stimulation of PI3K by G was independent of concurrent growth factor-induced signaling. This is in line with an increasing number of reports describing G-protein-induced signaling pathways linked to p85-associated PI3Ks (24, 50-54). Accordingly, we add PI3K to the rapidly growing list of G-protein-regulated cellular effectors. In addition, by using purified components, we confirmed previous findings in a cell-free system on the synergistic activation of PI3K by G and phosphotyrosyl peptides (24). The data suggest that tyrosine kinase- and G-protein-dependent signaling converges at the level of PI3K in two different ways. On the one hand, each pathway signals independently through PI3K. In this scenario PI3K represents a branch point allowing G-dependent signaling to switch into p85-connected pathways. On the other hand, PI3K also functions as a coincidence detector integrating and amplifying concurrent signals. Thus, by forwarding signals of different amplitude, PI3K may contribute to specificity of PI3K-dependent signaling.

Our data provide evidence that for stimulation of enzymatic activity G has to target the catalytic subunits of PI3K and -. In other words, this would exclude the necessity of adaptors that in turn are indispensable for coupling of PI3Ks to tyrosine kinase-linked pathways. However, our studies indicate a functional role for the non-catalytic p101 subunit of PI3K, i.e. determining a preference for PI-4,5-P₂ as the substrate. Interestingly, p101 affects only the G-stimulated but not basal lipid kinase activity of p110. This is reasonable from a physiological point of view, because 3-phosphorylated products are absent in quiescent cells. All the data argue that the stimulus G favors a complex consisting of three components, i.e. the catalytic subunit p110, the regulatory module p101, and the physiological substrate PI-4,5-P₂. The structural basis for this particular behavior remains to be elucidated. However, direct interaction of p110 as well as p101 with G has already been proven. In this context the PH domain of p110 may be of interest since previous studies have shown that this structural element is capable of mutual binding of PI-4,5-P₂ and G (55, 56). Furthermore, for purified -adrenergic receptor kinase, it was demonstrated that a cooperative binding of G and PI-4,5-P₂ leads to membrane association and enzyme stimulation. Since a similar situation is conceivable for PI3K (26), we are currently focusing on the function of the p110 PH domain as an element involved in the lipid substrate preference of PI3K.

We also showed for the first time that G stimulates autophosphorylation of PI3K. In contrast to class I_A PI3Ks, protein kinase activity of p110 does not lead to inhibition of its lipid kinase activity. Considering the fact that G-induced activation of lipid kinase activity and autophosphorylation of p110 went parallel and required the same conditions such as Mg²⁺ concentrations and the presence of lipids, we speculate that autophosphorylation contributes to regulation of lipid kinase activity. A phosphorylated p110 may exhibit altered affinities to interacting proteins and substrates including p101, G, and phospholipids. In fact, we obtained preliminary evidence that the site of phosphorylation lies within the PH domain of p110. In this context it is reminiscent that phosphorylation of PH domains such as the PH domain has been shown to enhance mutual binding of G and PI-4,5-P₂ (56)_. Furthermore, in our study p101 enhanced G-induced autophosphorylation. Combining the effects of p101 on G-induced lipid and protein kinase activities, a speculative picture is imaginable in which p101 enhances p110 autophosphorylation, thereby supporting mutual binding of G and PI-4,5-P₂ to p110, which in turn contributes to the observed substrate selectivity. Experiments are under way to prove this tempting hypothesis.

	ACKNOWLEDGEMENTS

We are grateful to Prof. Michael Waterfield and Dr. Bart Vanhaesebroeck for providing p110-His and p110-GST encoding viruses as well as Drs. Osamu Hazeki and Savvas Christoforidis for a p110-His construct and virus. We also thank Dr. Andreas Steinmeyer for providing phosphotyrosyl peptides and wortmannin derivatives. Valuable discussions with Drs. Marcus Thelen, Bern, Switzerland, and Dr. Doris Koesling, Berlin, Germany, are appreciated. We are indebted to Günter Schultz for critical reading of the manuscript and for support.

	FOOTNOTES

^* This work was supported by Deutsche Forschungsgemeinschaft and Fonds der Chemischen Industrie.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Recipient of the Stiftung Stipendien Fonds des Verbandes der Chemischen Industrie.

^§ To whom correspondence should be addressed. Tel.: 49-30-8445- 1830; Fax: 49-30-8445-1818; E-mail: bnue@zedat.fu-berlin.de.

	ABBREVIATIONS

The abbreviations used are: PI-4,5-P₂, phosphatidylinositol-4,5-bisphosphate; PI-3,4,5-P₃, phosphatidylinositol-3,4,5-trisphosphate, G, -subunits from bovine brain; GPCR, G-protein-coupled receptor; GST, glutathione S-transferase; His, hexahistidine tag; m.o.i., multiplicity of infection; PH, pleckstrin homology; PI3K, phosphoinositide 3-kinase; PI, phosphatidylinositol (locants of other phosphates on inositol ring shown in parentheses); PT, pertussis toxin, an exotoxin from Bordetella pertussis; PtdEtn, PtdCho, choline, -serine; p85, regulatory subunit of class I_A PI3Ks; p101, subunit associated with p110; p110, catalytic subunit of PI3Ks; SH, src homology; PAGE, polyacrylamide gel electrophoresis; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES