Selective recognition of phosphatidylinositol 3,4,5-trisphosphate by a synthetic peptide.

The present study takes a novel approach to explore the mode of action of phosphoinositide 3-kinase lipid products by identifying a synthetic peptide W-NG(28-43) (WAAKIQASFRGHMARKK) that displays discriminative affinity with phosphatidylinositol 3,4,5-trisphosphate (PtdIns(3,4,5)P3). This PtdIns(3,4,5)P3-binding peptide was discovered by a gel filtration-based binding assay and exhibits a high degree of stereochemical selectivity in phosphoinositide recognition. It forms a 1:1 complex with PtdIns(3,4,5)P3 with Kd of 2 microM, but binds phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2) and phosphatidylinositol 3,4-bisphosphate (PtdIns(3,4)P2) with substantially lower affinity (5- and 40-fold, respectively) despite the largely shared structural motifs with PtdIns(3,4,5)P3. Other phospholipids examined, including phosphatidylserine, phosphatidylcholine, phosphatidylinositol, and phosphatidylethanolamine, show low or negligible affinity with the peptide. Several lines of evidence indicate that this phosphoinositide-peptide interaction is not due to nonspecific electrostatic interactions or phospholipid aggregation, and requires a cooperative action among the hydrophobic and basic residues to exert the selective recognition. CD data suggest that the peptide acquires an ordered structure upon binding to PtdIns(3,4,5)P3. Further, we demonstrate that PtdIns(3,4,5)P3 enhances the phosphorylation rate of this binding peptide by protein kinase C (PKC)-alpha in a dose-dependent manner. In view of the findings that this stimulatory effect is not noted with other PKC peptide substrates lacking affinity with PtdIns(3,4,5)P3 and that PKC-alpha is not susceptible to PtdIns(3,4,5)P3 activation, the activity enhancement is thought to result from the substrate-concentrating effect of the D-3 phosphoinositide, i.e. the presence of PtdIns(3,4,5)P3 allows the peptide to bind to the same vesicles/micelles to which PKC is bound. Moreover, it is noteworthy that neurogranin, the full-length protein of W-NG(28-43) and a relevant PKC substrate in the forebrain, binds PtdIns(3,4,5)P3 with high affinity. Taken together, it is plausible that, in addition to PKC activation, PtdIns(3,4,5)P3 provides an alternative mechanism to regulate PKC activity in vivo by recruiting and concentrating its target proteins at the interface to facilitate the subsequent PKC phosphorylation.


EXPERIMENTAL PROCEDURES
,-bisphosphate (PtdIns-(3,4)P 2 ) were prepared as previously reported (21). Both synthetic phosphoinositides were characterized by 1 H and 31 P NMR and fast atom bombardment mass spectrometry, in which no appreciable impurity was detected. PtdIns(4,5)P 2 , phosphatidylcholine (PtdCho), phosphatidylethanolamine (PtdEA), phosphatidylinositol (PtdIns), and phosphatidylserine (PtdSer) were products from Sigma. 1,2-Dioctanoyl-snglycerol was obtained from Calbiochem. Inositol phosphates used in this study were synthesized according to the reported procedures (22). Colorimetric PKC assay and lissamine rhodamine-labeled PKC and PKA substrates (including glycogen synthase peptide, myelin basic protein peptide 4 -14 , PKC pseudosubstrate, epidermal growth factor receptor peptide, neurogranin peptide 28 -43 ), epsilon peptide, and Kemptide) were obtained from Pierce. PKC-␣ was kindly provided by Drs. Alex Toker and Kiyotaka Nishikawa. Other peptides used in this investigation were either purchased from commercial sources or synthesized by an Applied Biosystems Peptide Synthesizer in the Macromolecular Structure Analysis Facility at University of Kentucky. All custom-synthesized peptides were characterized by fast atom bombardment mass spectrometry in the Mass Spectrometry Facility at the University of Kentucky. Micelles containing individual phospholipids were prepared by sonicating in distilled water in a model 1210 Bransonic ultrasonic cleaner for 5 min.
Identification of a PtdIns (3,4,5)P 3 -binding Peptide by Gel Filtration Analysis-The screening method is based on the principle that binding of a peptide to PtdIns(3,4,5)P 3 -containing micelles will change its elution profile on gel filtration chromatography. This technique has been applied to the study of protein-phospholipid interactions (23,24). Individual peptides (125 M) were incubated with micelles consisting of pure PtdIns(3,4,5)P 3 (theoretical concentration, 125 M), in a final volume of 100 l, for 30 min at room temperature. The mixture was chromatographed on a Sephacryl S-200 column (1 ϫ 10 cm) equilibrated with 10 mM Tris/HCl, pH 7.5, containing 75 mM KCl. The column was eluted with the same buffer at 0.5 ml/min, and fractions of 0.45 ml were collected. Protein assays were performed by a Coomassie Blue dyebinding method or by UV absorbance at 230 nm. The micelle-bound peptide was eluted in the void volume, and was well separated from the free peptide. The amount of micelle-bound peptide was calculated as the difference between the total amount applied to the column and the amount of the free peptide. The dissociation constant (K d ) was estimated according to Equation 1. Circular Dichroism Spectroscopy-CD spectra were recorded with a JASCO J720 spectropolarimeter at room temperature in a 20 mm path length cell. The solution contained 25 M W-NG 28 -43 and 50 M individual phosphoinositides in 25 mM Tris/HCl and 100 mM KCl, pH 7.5. The following settings were used: wavelength range, 200 -250 nm; bandwidth, 1 nm; step resolution, 0.5 nm; scan speed, 10 millidegree/ min. Each spectrum represented an average of 10 scans with base-line subtraction. 31 P NMR Spectroscopy-Decoupled 31 P NMR spectra were recorded at 25°C in 10-mm tubes on a Varian VRX400 spectrometer at 161.9 MHz. Aliquots of a W-NG 28 -43 solution were introduced into a 1-ml solution containing 0.8 mM PtdIns(3,4,5)P 3 and 10% (v/v) deuterium oxide. The spectrum width was 20 kHz with 10-s pulse width and 0.8-s acquisition time. For each spectrum, 5,000 acquisitions were obtained over a 4,000-s period. External H 3 PO 4 and deuterium oxide were employed as a chemical shift reference and a locking signal, respectively.
Protein Kinase C Assay-Effect of PtdIns(3,4,5)P 3 on PKC-mediated phosphorylation of NG 28 -43 and other PKC substrates was examined using a colorimetric method developed by Pierce (25). The reaction mixture (25 l) contained 30 mM Tris/HCl, pH 7.4, 50 mM NaCl, 2 mM ATP, 10 mM MgCl 2 , 0.1 mM CaCl 2 , 0.002% Triton X-100, PtdSer (1 mg/ml), 1,2-dioctanoyl-sn-glycerol (20 g/ml), PKC-␣ (0.1 unit), 300 M dye-labeled PKC substrate, and varying amounts of PtdIns(3,4,5)P 3 . After incubating at 30°C for 20 min, 20 l of the mixture were applied to a separation unit containing an affinity membrane that would retain the phosphorylated peptide. The membrane was washed under reduced pressure with 750 l of a phosphopeptide-binding buffer consisting of 0.1 M sodium citrate, pH 5.0, 0.5 M NaCl, and 0.02% sodium azide to remove the unreacted peptide. The phosphopeptide was then eluted by washing the membrane with 600 l of 15% formic acid. Quantitation of the phosphorylated product was accomplished by measuring its absorbance at 570 nm in reference to a standard curve constructed from known amounts of the phosphopeptide generated from the reaction.
Production of Recombinant Neurogranin-The bacterial expression vector for rat neurogranin pDGRC3 (a kind gift from Dr. Dan D. Gerendasy) was transformed into Escherichia coli BL21 (DE3) (pLysS), and recombinant neurogranin was produced and purified according to a procedure described by Gerendasy et al. (26). In brief, The bacterial culture was grown in LB broth containing ampicillin (100 g/ml) and chloramphenicol (30 g/ml) to A 600 nm ϭ 0.4, isopropyl-␤-D-thiogalactoside was added to a final concentration of 0.4 mM, and the cells were grown for another 6 h. The cells were collected by centrifugation and were suspended in 10 ml of cold lysis buffer consisting of 50 mM Tris/HCl, pH 7.5, 100 mM NaCl, 2 mM EDTA, 1 mM EGTA, and 50 mM dithiothreitol. The suspension was frozen and thawed, and sonicated. After removing cell debris by centrifugation, the crude homogenate was treated with perchloric acid (final concentration, 2.5% v/v), followed by trichloroacetic acid (final concentration, 15% w/v). The perchloric acidsoluble, trichloroacetic-insoluble material was dissolved in 50 mM Tris/ HCl, pH 7.5, containing 200 mM NaCl, 2 mM EDTA, 1 mM EGTA, and 50 mM dithiothreitol (washing buffer), and applied to a column containing 15 ml of calmodulin-Sepharose 4B. The column was washed with 5 column volumes of washing buffer and the adsorbed protein was eluted with 50 ml of elution buffer consisting of 50 mM Tris/HCl, pH 7.5, 200 mM NaCl, 7 mM CaCl 2 , and 50 mM dithiothreitol. Fractions were collected throughout, and the pooled solution was subjected to the aforementioned acid treatments. The resulting pellet was washed twice with ethanol-ether (1:1) and dissolved in 10 mM Tris/HCl, 75 mM KCl, and 1 mM dithiothreitol. The homogeneity of the purified protein was indicated by a single band on SDS-PAGE with silver staining, and the identity was confirmed by N-terminal sequencing. The sequence of the first 15 amino acids at the N terminus was MDCCTESACSKPDDD, which was identical to that reported in the literature (26). Concentrations of neurogranin were determined by the BCA method with bovine serum albumin as a standard. (3,4,5)P 3 by a Synthetic Peptide-In this study, a gel filtration-based binding assay was employed to search for PtdIns(3,4,5)P 3 -binding peptides. Individual peptides were incubated with micelles consisting of pure PtdIns(3,4,5)P 3 , and the mixture was applied onto a short path Sephacryl S-200 column. In principle, binding to PtdIns(3,4,5)P 3 would be indicated by the co-elution of the tested peptide with the micelles in the void volume. This expe-dient assay obviated the use of radioligands and laborious separation procedures to detect ligand binding. Numerous unrelated peptides with lengths ranging from 10-to 20-amino acid residues were subjected to this analysis. Of more than 100 peptides evaluated, only one peptide W-NG 28 -43 displayed tight binding to PtdIns(3,4,5)P 3 -containing micelles. Fig. 1 shows the elution profile of the peptide in the presence of PtdIns(3,4,5)P 3 at an 1:1 molar ratio (solid line). As shown, the peptide existed exclusively in the micelle-bound form, indicating the high affinity with PtdIns(3,4,5)P 3 .

Interfacial Recognition of PtdIns
W-NG 28 -43 was a synthetic peptide with the amino acid sequence of WAAKIQASFRGHMARKK (single-letter amino acid code). It is worthy to note that after deleting the Nterminal Trp, this PtdIns(3,4,5)P 3 -binding peptide would be identical to NG 28 -43 that constituted the PKC phosphorylation and calmodulin-binding domain of neurogranin (27). W-NG 28 -43 was originally designed in our laboratory for the kinetic study of PKC-mediated phosphorylation of the parent peptide by fluorescence spectrophotometry.
This gel filtration analysis was repeated at different PtdIns(3,4,5)P 3 /W-NG 28 -43 ratios, and the amount of remaining free peptide was plotted against the theoretical concentration of PtdIns(3,4,5)P 3 (Fig. 1, inset). Accordingly, the apparent molecular stoichiometry was calculated from the dose-dependence curve to be 1:1, indicating that each W-NG 28 -43 was associated with 1 PtdIns(3,4,5)P 3 at saturation. In addition, from the data of individual experiments, the apparent dissociation constant (K d ) was estimated to be 2 M.
Further evidence that W-NG 28 -43 exhibited high affinity with PtdIns(3,4,5)P 3 was provided by the fluorescence titration experiment ( Fig. 2A). As this peptide contained a tryptophan residue, its interaction with the phospholipid could be assessed by monitoring the fluorescence with excitation wavelength set to 292 nm.
As shown in Fig. 2A, PtdIns(3,4,5)P 3 quenched the fluorescence of W-NG 28 -43 in a dose-dependent and saturable manner accompanied by a rapid shift of the fluorescence maximum from 350 nm to 328 nm. Saturation of the fluorescence titration was attained at 1 molar equivalent of PtdIns(3,4,5)P 3 , which is consistent with that noted in the gel filtration study. Furthermore, preincubation of W-NG 28 -43 with Ins(1,3,4,5)P 4 or Ins(1,4,5)P 3 , even at a molar ratio of 1:50, did not perturb the fluorescence titration with PtdIns(3,4,5)P 3 , suggesting that the affinity between W-NG 28 -43 and inositol phosphates was negligible.
Enhancement of PKC-mediated Phosphorylation of W-NG 28 -43 by PtdIns (3,4,5)P 3 -Another important issue that warranted investigation was the mode of complex formation between W-NG 28 -43 and PtdIns(3,4,5)P 3 . It has been reported that polycationic polypeptides such as histone III-S and neurogranin peptide analog 29 -47 formed aggregates with mixed micelles containing PtdSer and diacylglycerol in a time-dependent manner (28). Concomitant with the formation of these aggregates there was a progressive loss of substrate phosphorylation by Ca 2ϩ -dependent PKC. Conceivably, embedment of the basic peptide into the micelles due to aggregation blocked its access to PKC, thereby preventing the subsequent phosphorylation. To distinguish between specific peptide-phospholipid binding and phospholipid aggregation, fully activated PKC ␣ was used to examine the phosphorylation of NG 28 -43 in the presence of varying amounts of PtdIns(3,4,5)P 3 . A number of PKC substrates that lacked affinity with PtdIns(3,4,5)P 3 were also tested as control, which included myelin basic protein peptide 4 -14 , glycogen synthase peptide, PKC pseudosubstrate, epidermal growth factor receptor peptide, and ⑀ peptide. As mentioned, all these polycationic peptides failed to bind micellar PtdIns(3,4,5)P 3 according to the gel filtration assay. Fig. 3 shows the dose dependence of PKC ␣ activity on PtdIns(3,4,5)P 3 with myelin basic protein peptide 4 -14 (open bars) and NG 28 -43 (shaded bars) as substrates. The initial rate of PKC phosphorylation of NG 28 -43 showed a significant increase with elevated levels of PtdIns(3,4,5)P 3 , up to 3.3-fold at a lipid/peptide ratio of 0.6, while no appreciable modulatory effect was noted with myelin basic protein peptide 4 -14 or other peptide substrates (data not shown) at comparable concentrations. It is worth mentioning that the enhancement of PKC activity was independent of the preparation of the phospholipid vesicle/micelle. PtdIns(3,4,5)P 3 could be externally added to PtdSer/1,2-dioctanoyl-sn-glycerol vesicles in a micelle form or subjected to sonication with PtdSer and 1,2-dioctanoyl-sn-glycerol to form mixed vesicles. The results obtained with these two preparations were virtually identical.
Clearly, the Ser residue of NG 28 -43 was accessible to PKC after binding to the phospholipid, indicating that the interaction between these two molecules was highly specific. Differential Recognition of Phospholipids by W-NG 28 -43 -To examine the binding specificity of W-NG 28 -43 , various phospholipids were subjected to the aforementioned analyses, including PtdIns(4,5)P 2 , PtdIns(3,4)P 2 , PtdIns, PtdSer, PtdCho, and PtdEA. The respective molecular stoichiometry and K d values were estimated by the gel filtration assay and summarized in Table I. Among these phospholipids, PtdIns(4,5)P 2 cross-reacted with W-NG 28 -43 , however, with an affinity 5-fold lower than PtdIns(3,4,5)P 3 . This cross-reactivity was presumably due to their largely shared structural motifs. In contrast, the affinity with PtdIns(3,4)P 2 was 40-fold weaker than that of PtdIns(3,4,5)P 3 . This structure-activity correlation suggests the importance of the 5-phosphate in the peptide binding, which was supported by the 31 P NMR examination of the interaction between W-NG 28 -43 and PtdIns(3,4,5)P 3 (Fig. 4).
The 31 P NMR signals for the phosphates at positions 1, 3, 4, and 5 of PtdIns(3,4,5)P 3 were noted at 2.92, 3.08, 3,88, and 3.49 ppm, respectively (spectrum a). Due to the size of the peptide, changes in the chemical shifts upon binding were small. Among these phosphate functions, the 5-phosphate exhibited the highest sensitivity to the peptide-induced change in chemical shift, which underscored the involvement of the 5-phosphate in the ionic interactions. For instance, when 0.75 molar equivalents of W-NG 28 -43 were added, a downfield shift of 0.12 ppm was noted for the phosphate, vis à vis 0.02 ppm (upfield), 0.09 ppm (downfield), and 0.02 ppm (upfield) for 1-, 3-, and 4-phosphates, respectively. No such changes were noted when an equivalent amount of PtdSer was added to the peptide.
Further assessment of the interfacial interactions between W-NG 28 -43 and various phosphoinositides was conducted by  fluorescence and CD spectroscopy. Fig. 2, A-C, compares the fluorescence spectral changes of W-NG 28 -43 by PtdIns(3,4,5)P 3 , PtdIns(4,5)P 2 , and PtdIns(3,4)P 2 . No significant change in the fluorescence spectrum was noted with other phospholipids examined such as PtdCho, PtdIns, and PtdEA at comparable concentrations. Modifications of the spectrum by phosphoinositides appear to be structurally dependent. In comparison, the molar equivalencies of PtdIns(3,4,5)P 3 , PtdIns(4,5)P 2 , and PtdIns(3,4)P 2 to attain saturation in the fluorescence titration were 1, 1.6, and 6, respectively, which were in accord with that determined by gel filtration. Although PtdIns(3,4,5)P 3 binding caused a significant change in the emission maximum, especially between the molar ratios of 0.46:1 and 0.65:1, no appreciable blue shift was noted for PtdIns(4,5)P 2 (Fig. 2B) and PtdIns(3,4)P 2 (Fig. 2C). Moreover, the extent of fluorescence quenching by PtdIns(4,5)P 2 was greater than the other two counterparts. These results suggest that the modes of binding with W-NG 28 -43 among these three phosphoinositides differed.
The same conclusion could be drawn by the results from CD spectroscopy. CD spectra were recorded in the far UV range between 200 and 250 nm (Fig. 5) for W-NG 28 -43 alone (curve a) or in the presence of PtdIns(3,4)P 2 (curve b), PtdIns(4,5)P 2 (curve c), or PtdIns(3,4,5)P 3 (curve d). It is worth mentioning that these phosphoinositides alone did not give any appreciable spectra.
As shown, the CD spectrum of the free peptide is characterized by a weak, negative band centered at 204 nm, which might be attributed to a random coil conformation (29). The spectra underwent marked changes when phosphoinositides were added. This result indicates that W-NG 28 -43 exhibited different structural behaviors when moving from a free state to a micelle-bound state at the interface and that the ligand-induced conformational change was dependent upon the phosphoinositide structure. However, lack of significant absorption at 222 nm in these spectra indicates that the peptide did not acquire ␣-helical structures after interfacial binding. When inositol phosphates or other phospholipids that lacked affinity with the peptide (e.g. PtdCho and PtdEA) were added, no change in the CD spectrum was observed.
Neurogranin Binding to PtdIns (3,4,5)P 3 -In view of the fact that W-NG 28 -43 represented the PKC recognition/calmodulinbinding domain of neurogranin, the interaction between PtdIns(3,4,5)P 3 and the full-length protein was assessed by the gel filtration assay. Fig. 6 illustrates the representative elution profiles of recombinant neurogranin (21.3 M) alone (A) and with increasing amounts of PtdIns(3,4,5)P 3 . By using data determined at six different lipid/protein ratios, the K d value and binding stoichiometry were estimated to be 2.2 Ϯ 0.5 M and 10, respectively.
Conceivably, this peptide model provides a useful tool to study the biomolecular recognition of phosphoinositides and to delineate PtdIns(3,4,5)P 3 -binding motifs in putative targets such as the nonconventional PKC isozymes and the SH2 domains on the p85 subunit. Structurally, this 17-mer peptide can be divided into two discrete regions composed largely of apolar and basic residues, respectively. It is noteworthy that the sequence of the C-terminal polybasic segment (i.e. RGH-MARKK) bears resemblance to the consensus sequences for PtdIns(4,5)P 2 -binding motifs deduced by Yin and co-workers (24), (K/R)XXXKX(K/R)(K/R) and (K/R)XXXXKX(K/R)(K/R), which are present in phospholipase C isozymes (24,32) and various actin-regulating proteins such as profilin, gelsolin, cofilin, and villin (Table II). More recently, Fukami et al. (33) reported another consensus sequence by comparing the putative PtdIns(4,5)P 2 -binding site of ␣-actinin (Table II) with the homologous sequences on spectrin, and the pleckstrin homology (PH) domain of various proteins such as phospholipase C ␦1 , pleckstrin, Grb 7, Ras-GAP, and racK␤. Again, the sequence RXXXXXXX(H/R/K)XX(X)W(K/R) is analogous to that of W-NG 28 -43 with regard to the spacing of the basic residues.
Most of these putative binding peptides listed in Table II were able to block binding of the parent proteins to PtdIns-(4,5)P 2 and inhibit the consequent physiological responses. For example, the two synthetic peptides from gelsolin, gelsolin- , and gelsolin-(150 -169), prevented PtdIns(4,5)P 2 from inhibiting actin filament severing by gelsolin (24,31). Also, the putative PtdIns(4,5)P 2 -binding peptide from ␣-actinin inhibited phospholipase C activity by competing for substrate binding (33). Nevertheless, its phospholipase C␤ 2 counterpart showed a stimulatory effect on the enzyme activity (32). Furthermore, binding studies indicated that synthetic peptide from the binding motifs of actin-binding proteins, such as gelsolin-(150 -169), bound PtdIns(4,5)P 2 with an affinity comparable to that of the native proteins (24).
However, no information is available concerning the differential affinity of these PtdIns(4,5)P 2 -binding motifs toward different phosphoinositides. Our recent study showed that PtdIns(3,4,5)P 3 was able to bind actin-regulating proteins such as profilin and gelsolin with an affinity greater than or at least in the same order of magnitude as PtdIns(4,5)P 2 (19). Taken together with the fact that these two phosphoinositides are closely related in their structures, it is plausible that the binding motif for PtdIns(3,4,5)P 3 and that for PtdIns(4,5)P 2 have similar structural features, which may account for the crossreactivity between the two phosphoinositides. Although principles governing the ligand selectivity remain unclear, it is plausible that the number and composition of amino acid residues located between those basic residues affects the phosphoinositide preference.
To assess the role of the polybasic domain of W-NG 28 -42 in interacting with the acidic phospholipid, a truncated peptide, WFRGHMARKK, was prepared by deleting a hydrophobic segment from the N-terminal half. It, however, showed extremely low affinity with PtdIns(3,4,5)P 3 . This finding suggests that the binding is not simply attributed to electrostatic interactions between the basic residues and the charged head group of the phospholipid, and requires a cooperative action between hydrophobic and electrostatic motifs for interfacial recognition. This premise is further collaborated by the spectroscopic data. First, the CD spectra indicate that the peptide undergoes conformational change from a random coil to an ordered structure when it binds to micellar phosphoinositides. Secondly, a significant blue shift coupled with considerable attenuation of the fluorescence intensity shows that the N-terminal Trp physically interacts with the apolar environment at the interface. Consequently, it is reasonable to postulate that the PtdIns(3,4,5)P 3binding motifs consists of two contiguous segments, a polybasic region flanked by a hydrophobic segment for interfacial recognition. Previously, Cantley and co-workers reported that PtdIns(3,4,5)P 3 interacted with the SH2 domain on the regulatory subunit of PI 3-kinase (18). It is interesting that the N-terminal SH2 domain on the p85 subunit of human PI 3-kinase (35) contains an internal peptide sequence that bears some resemblance to W-NG 28 -42 in the terms of the spacing of the stretch of basic residues (Scheme 1). Moreover, in the inter-SH2 region, there is a partial sequence 506 KEYIEKFKR 514 corresponding to the consensus sequence of KXXXXKX(K/R)(K/R) (36). Investigations on the potential interaction between these SH2 internal peptides and PtdIns(3,4,5)P 3 are currently in progress.
As mentioned, W-NG 28 -43 corresponds to the phosphorylation domain of neurogranin and closely resembles that of neuromodulin (ATKIQASFRGHITRKK) (20). Both proteins are physiological relevant PKC substrates in the hippocampal region of central nerve system, and are not known to bind phospholipids. It is worthy to note that besides this homology, these two brain-specific proteins are not related over the rest of their sequences. Here, we provide evidence that neurogranin binds PtdIns(3,4,5)P 3 with affinity comparable to that of the partial peptide, though with much larger binding stoichiometry. Thus, the physiological implication of this selective recognition is suggested by the dose-dependence of PKC ␣ activity on PtdIns(3,4,5)P 3 . This enhancement is thought to result from the substrate-concentrating effect of PtdIns(3,4,5)P 3 -containing micelles based on the following rationales. First, it is known that PKC ␣ is not susceptible to PtdIns(3,4,5)P 3 activation (15). Secondly, the observed stimulatory effect was specific for W-NG 28 -43 and was not noted with other PKC substrates that lacked affinity with PtdIns(3,4,5)P 3 . This finding raises a possibility that PtdIns(3,4,5)P 3 serves as a targeting site on the plasma membrane for neurogranin and neuromodulin to facilitate their phosphorylation. The investigation of this hypothesis is currently underway in this laboratory.   28 -43 and putative PtdIns(4,5)P 2 -binding peptides All listed peptide fragments have been synthesized and shown competitive PtdIns(4,5)P 2 binding with the parent proteins.