Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The continuous agonist occupancy of G protein-coupled receptors (GPCRs) can trigger adaptive responses whereby the stimulatory effect of the ligand is subsequently reduced. Two such events include: 1) phosphorylation of the GPCR, which results in its uncoupling from the effector enzyme (Bohm et al., 1997); and 2) internalization of GPCRs initially present in a hydrophilic environment at the cell surface into an endosomal cell compartment (Koenig and Edwardson, 1997). Receptor endocytosis not only may serve to limit the duration of ligand activation of the GPCR (Sorensen et al., 1997), but may also permit the resensitization of the receptor (Pippig et al., 1995).

Despite the fact that receptor endocytosis has been demonstrated frequently, insight into the underlying molecular mechanism(s) is relatively limited. In this context, several GPCRs, including muscarinic cholinergic receptors (mAChRs), -adrenergic receptors (-ARs), serotonergic receptors, and Substance P receptors, appear to undergo endocytosis via a clathrin-coated pit mechanism (Chuang et al., 1986; Von Zastrow and Kobilka, 1996; Garlands et al., 1994; Slowiejko et al., 1996) in a process that involves the GTPase dynamin (Zhang et al., 1996). In addition, inositol lipids have been proposed to play an essential role in membrane trafficking events (for reviews, see De Camilli et al., 1996; Martin, 1997). The 3'-phosphoinositides have been implicated in late endocytic trafficking (Joly et al., 1995), whereas the quantitatively major inositol lipids, i.e., phosphatidylinositol 4-phosphate (PIP) and phosphatidylinositol 4,5-bisphosphate (PIP₂), are required for Ca²⁺-stimulated exocytosis (Eberhard et al., 1990; De Camilli et al., 1996). We recently demonstrated that these two lipids may also play a key role in the endocytosis of GPCRs (Sorensen et al., 1998). Inhibition of phosphatidylinositol 4-kinase (PI4K) activity in SH-SY5Y cells by three chemically distinct agents, namely wortmannin (WT), LY-294002, or phenylarsine oxide (PAO), resulted in both a selective inhibition of the synthesis of PIP (and PIP₂) and prevention of the agonist-induced endocytosis of the M₃ subtype of mAChRs. This inhibition of receptor endocytosis was demonstrated to be independent of phosphatidylinositol 3-kinase (PI3K) activity, the integrity of the actin cytoskeleton, and the formation of phosphoinositide-derived second messengers (Sorensen et al., 1998).

The objectives of the present study were to determine whether the dependence of mAChR endocytosis on phosphoinositide synthesis observed for SH-SY5Y cells could be extended to other cell types and to investigate whether inositol lipids also play a role in the endocytosis of those GPCRs that are coupled to the activation of adenylyl cyclase rather than to phospholipase C (PLC). The cell system chosen to address these issues was the human 1321N1 astrocytoma. The use of the 1321N1 astrocytoma offers the advantage that these cells express M₃ mAChRs (coupled to PLC) and ₂-ARs (linked to adenylyl cyclase), both of which readily undergo agonist-induced endocytosis. Furthermore, the characteristics of receptor internalization have been fully evaluated in these cells (Waldo et al., 1983, 1984; Harden et al., 1985). To modulate the phosphoinositide content of 1321N1 astrocytoma, we have used the fungal metabolite WT, which, at micromolar concentrations, inhibits PI4K activity and thereby reduces the synthesis of agonist-sensitive pools of both PIP and PIP₂ (Nakanishi et al., 1995; Downing et al., 1996; Sorensen et al., 1998; Willars et al., 1998). The results indicate that when PIP and PIP₂ synthesis in 1321N1 cells is inhibited by WT, the endocytosis of mAChRs is curtailed in a similar way. Inclusion of WT alone also partially inhibits the endocytosis of ₂-ARs. However, if PIP₂ concentrations are further reduced in these cells by the prior activation of mAChRs in the presence of WT, conditions under which phosphoinositide breakdown proceeds in the absence of lipid resynthesis, a more pronounced inhibition of ₂-AR endocytosis is observed. Taken collectively, the results suggest that phosphoinositide synthesis, in particular the availability of PIP₂, is a general prerequisite for the internalization of GPCRs regardless of the effector enzyme to which they couple.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. [³²P]Orthophosphoric acid (10 mCi/ml), [³H]N-methylscopolamine (87 Ci/mmol), myo-[³H]inositol (80 Ci/mmol), and detection reagents for enhanced chemiluminescence were obtained from Amersham (Arlington Heights, IL). ¹²⁵I-labeled cyanopindolol (CYP; 2200 Ci/mmol), [³H]quinuclidinylbenzilate (QNB; 45.4 Ci/mmol), and [-³²P]ATP (6,000 Ci/mmol) were obtained from New England Nuclear Research Products (Boston, MA). Isoproterenol (Iso), propranolol, phosphatidylinositol, ATP, atropine, WT, and PAO were obtained from Sigma Chemical Co. (St. Louis, MO). Oxo-M [2-butyn-1-ammonium, N,N,N-trimethyl-4-(2-oxo-1-pyrrolidinyl)iodide] was purchased from Research Biochemicals Inc. (Natick, MA). Tissue culture supplies were purchased from Corning Glass Works (Corning, NY) and Sarstedt, Inc. (Newton, NC). Powdered Dulbecco's modified Eagle's medium (DMEM) was obtained from GIBCO (Grand Island, NY). Fetal calf serum (FCS) was obtained from Summit Biotechnology (Fort Summit, CO). Protein A/G-agarose, peroxidase-conjugated anti-rabbit IgG and anti-goat IgG, and polyclonal anti-PI4K were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Polyclonal antibodies to PI4K and the p85 subunit of PI3K were obtained from Upstate Biotechnology (Lake Placid, NY). Fura-2-acetoxymethyl ester (fura-2/AM) was purchased from Molecular Probes (Eugene, OR). Silica gel 60 thin-layer chromatography plates were obtained from Merck (Darmstadt, Germany). 1321N1 astrocytoma cells were obtained from Dr. T. Kendall Harden (University of North Carolina, Chapel Hill, NC).

Cell Culture Conditions. 1321N1 astrocytoma cells (passage unknown) were grown (in the absence of antibiotics) in tissue culture flasks (75 cm²/250 ml) in 20 ml of DMEM supplemented with 10% (by volume) FCS. Confluent flasks of cells were subcultured 1:5 and grown for 7 days at 37°C in a humidified atmosphere containing 10% CO₂. Cells were isolated after aspiration of the medium and incubation with a modified Pucks D₁/trypsin/EDTA solution (Sorensen et al., 1998). Unless otherwise noted, cells then were resuspended in buffer A (142 mM NaCl, 5.6 mM KCl, 2.2 mM CaCl₂, 3.6 mM NaHCO₃, 1 mM MgCl₂, 5.6 mM D-glucose, and 30 mM HEPES, pH 7.4).

Phospholipid Labeling. Confluent 75-cm² flasks of 1321N1 cells were routinely prelabeled for 2 to 4 h in the presence of [³²P]orthophosphoric acid (100 to 250 µCi/flask) in 10 ml of buffer A. Alternatively, cells were incubated in the presence of 10 µCi/ml of [³H]inositol for 72 h. Cells were detached, washed in buffer A, and resuspended in 3.0 ml of buffer A/flask, and 0.2-ml aliquots were removed. Cells then were incubated in the presence of WT and/or agonist as indicated, at 37°C in a final volume of 0.5 ml. Reactions were terminated by the addition of 1.5 ml of chloroform/methanol (1:2). Lipids were extracted, separated by thin-layer chromatography, and quantitated as described previously (Thompson and Fisher, 1990).

Measurement of Total Inositol Phosphate Release. Cells were allowed to prelabel in DMEM/10% FCS containing 10 µCi/ml [³H]inositol at 37°C for 48 h to achieve isotopic equilibrium. Cells were detached, washed in buffer A, and incubated with WT and/or Oxo-M as indicated in a final volume of 0.5 ml of buffer A. Reactions were terminated by the addition of 1.5 ml of chloroform/methanol (1:2), and the accumulation of [³H]inositol phosphates in the presence of Li⁺ was monitored as described previously (Thompson and Fisher, 1990).

Measurement of Cytoplasmic Ca²⁺ Concentrations ([Ca²⁺_i]). Cells were resuspended in buffer A and incubated with 2 µM fura-2/AM for 15 min at 37°C. [Ca²⁺]_i were determined by monitoring fura-2 fluorescence in a Shimadzu RF-5000 spectrofluorometer (Shimadzu Scientific Instruments, Columbia, MD) by the dual wavelength method of Grynkiewicz et al. (1985). Under these conditions, [Ca²⁺]_i = (RR_min/R_maxR) B · K_d, where R, R_min, and R_max are the ratios of the fluorescence obtained at excitation wavelengths of 340 and 380 nm ( emission = 505 nm), B is the ratio of the fluorescence of Ca²⁺-free/Ca²⁺-saturated signals at 380 nm, and K_d is the affinity constant of fura-2 for Ca²⁺ (224 nM).

Subcellular Fractionation. A differential centrifugation technique was used to obtain "light" membrane, vesicular (V₁) fractions containing endocytosed mAChRs and ₂-ARs. Cell suspensions containing ~40 to 50 mg of cell protein per condition were incubated with WT and/or agonist as indicated. Subcellular fractions were isolated as previously described (Slowiejko et al., 1996), with the exception that the initial centrifugation was carried out at 14,500g for 10 min instead of at 30,000g for 10 min (Waldo et al., 1983). V₁ fractions were resuspended in Tris-EDTA buffer (10 mM Tris-HCl, pH 7.4, and 2 mM EDTA) containing 2 µg/ml aprotinin, 1 µg/ml leupeptin, and 1 mM Pefabloc (Boehringer Mannheim Corp., Indianapolis, IN) at a final protein concentration of 3 to 6 mg/ml. The conditions chosen for monitoring the agonist-induced endocytosis of mAChRs and ₂-ARs (e.g., duration, concentration of agonist) were based on previous studies in which optimal conditions had been established (Waldo et al., 1983, 1984; Harden et al., 1985).

Radioligand Binding. Agonist-induced endocytosis of mAChRs and -ARs was monitored by the appearance of receptors in the V₁ fraction, as determined by an increase in [³H]QNB binding (mAChRs) or ¹²⁵I-labeled CYP binding (₂-ARs) sites, respectively. For detection of mAChRs, 200-µl aliquots of V₁ fractions were incubated in Tris-EDTA buffer with 1 nM [³H]QNB at 37°C for 90 min, as described previously (Sorensen et al., 1998). Nonspecific binding was determined as that unaffected by inclusion of 25 µM atropine. In some experiments, the agonist-induced sequestration of mAChRs in intact cells was monitored by a loss of cell-surface [³H]N-methylscopolamine binding sites, as described previously (Slowiejko et al., 1996). For detection of -ARs, 200-µl aliquots of V₁ fractions were incubated in buffer B (154 mM NaCl, 20 mM Tris-HCl, pH 7.4, and 5 mM MgCl₂) containing 8 pM ¹²⁵I-labeled CYP at 37°C for 90 min (final volume = 2 ml), as described previously (Waldo et al., 1984). Nonspecific binding was determined as that unaffected by inclusion of 1 µM propranolol. Reactions were rapidly terminated by filtration through Whatman GF/B glass-fiber filters, and radioactivity was determined after the addition of 5 ml of Universol scintillation fluid.

SDS-Polyacrylamide Gel Electrophoresis. 1321N1 cells or rat brain tissues were homogenized in 1 ml of lysis buffer containing 20 mM HEPES (pH 7.4), 1% Triton X-100, 50 mM NaCl, 1 mM EGTA, 5 mM -glycerophosphate, 30 mM sodium pyrophosphate, 100 µM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride, 10 µg/ml aprotinin, and 10 µg/ml leupeptin. Cell debris was removed by centrifugation at 12,000g for 5 min at 4°C. Aliquots of whole-cell lysates were boiled in SDS-polyacrylamide gel electrophoresis sample buffer for 5 min and electrophoresed through 7.5% SDS-polyacrylamide gels. Proteins were transferred to polyvinylidene fluoride membranes (Millipore, Bedford, MA) and processed for immunoblot analysis.

Immunoblot Analysis. Nonspecific binding sites were blocked in PBS (pH 7.4) containing 0.1% Tween 20 (PBS-T) and 1% BSA for 1 h at room temperature. Primary antibody was diluted in blocking solution (final concentration, 0.5-1.0 µg/ml) and incubated with the membranes for 1 h. Excess primary antibody was removed by washing the membranes three times in PBS-T. The blots then were incubated in the appropriate peroxidase-conjugated secondary antibody diluted in PBS-T (1:10,000) for 1 h and subsequently washed three more times in PBS-T. Immunoreactive proteins were detected by enhanced chemiluminescence.

Phosphatidylinositol (PI) Kinase Activity in Immunoprecipitates. 1321N1 cells (two 75-cm² flasks for each immunoprecipitation) were homogenized in 4 ml of lysis buffer. Cell debris was removed by centrifugation at 12,000g for 5 min at 4°C. Cell lysates then were incubated with 50 µg of anti-PI4K or 50 µl of anti-p85 subunit of PI3K at 4°C for 2 h with continuous mixing. Protein A/G-agarose (400 µl) was added for an additional 1 h with mixing. Immune complexes were pelleted by centrifugation and washed three times with the respective assay buffer and resuspended in either 1.0 ml (PI4K) or 1.6 ml (p85) of assay buffer. PI4K activity in 50-µl aliquots and PI3K activity in 100-µl aliquots was determined in the absence or presence of WT as described previously (Sorensen et al., 1998).

Protein. Protein content was measured with a Pierce BCA protein assay reagent (Pierce Chemical Co., Rockford, IL).

Data Analysis. Data are expressed as the mean ± S.E. for the number of experiments performed. Student's two-tailed t tests were used to evaluate the statistical differences between the mean values of paired or unpaired sets of data.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

WT Inhibits Basal Polyphosphoinositide Synthesis in 1321N1 Astrocytoma Cells. Because the effects of WT on inositol lipid turnover in 1321N1 astrocytoma cells had not been reported previously, initial experiments were directed toward the establishment of conditions under which WT could be demonstrated to modulate phosphoinositide synthesis. 1321N1 cells were first allowed to prelabel in buffer A in the presence of ³²P_i (inorganic orthophosphate) for 2 h before resuspension of cells in buffer alone. ³²P-label associated with the inositol lipids then was monitored as a function of time (0-20 min) in the presence or absence of WT (Fig. 1). In the absence of WT, a slow time-dependent loss of label from the polyphosphoinositides, in particular that of PIP, was observed. When cells were incubated with WT at a concentration (10 µM) that is known to inhibit PI4K activity (see Downing et al., 1996; Balla et al., 1997; Meyers and Cantley, 1997; Sorensen et al., 1998), the loss of label from both PIP and PIP₂ was accelerated, such that after 20 min of WT treatment, ³²P-labeling of the two lipids was reduced by approximately 30 to 40%. No further reduction in either PIP or PIP₂ labeling was observed in more extended incubations (data not shown). The WT-induced loss of label from PIP was maximal at 10 min and occurred before that observed for PIP₂. In contrast, neither [³²P]PI labeling nor label associated with the total phospholipid fraction was altered by the inclusion of WT at any time point examined. Taken collectively, these results indicate that inclusion of WT results in a selective inhibition of the synthesis of PIP in 1321N1 cells, and as a consequence, that of PIP₂. Unless stated otherwise, in all subsequent experiments, cells were preincubated in the presence of WT for 20 min.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   WT inhibits polyphosphoinositide synthesis in 1321N1 astrocytoma cells. Cells were prelabeled with 32Pi for 2 to 4 h before treatment with vehicle (DMSO) alone () or 10 µM WT () for the times indicated. Reactions were terminated by the addition of chloroform/methanol; lipids then were extracted and separated by thin-layer chromatography as described in Materials and Methods. Results are reported as radioactivity present in PIP2, PIP, or PI as a function of time. Values are expressed as the mean ± S.E. of triplicate replicates obtained from one experiment and are representative of three experiments performed.

Activation of mAChRs in the Presence of WT Depletes PIP₂ Stores. Preincubation of 1321N1 astrocytoma cells for 20 min in the presence of 10 µM WT resulted in a reduction in the labeling of PIP (52 ± 5% of control) and PIP₂ (57 ± 6% of control; n = 5). The addition of a 1 mM concentration of Oxo-M, a muscarinic agonist, to vehicle-pretreated cells resulted in only a small reduction in the labeling of either lipid (<15%; Fig. 2). However, when Oxo-M was added to WT-pretreated cells during the final 5 min of the 20-min preincubation period, there was a significant further loss of label from PIP₂ (35 ± 3% versus 57 ± 6% of control; P < .05; n = 4), but not from PIP (53 ± 6% versus 52 ± 5%; n = 4). Incubation of WT-pretreated cells with Oxo-M for time periods >5 min did not result in any further loss of label from PIP₂, indicating that PIP₂ was maximally depleted (data not shown). The inhibitory effects of WT on PIP and PIP₂ labeling were specific because no alterations in the ³²P-labeling of PI or of other quantitatively major lipids were observed in the presence of WT. No significant agonist-induced loss of label from either PIP or PIP₂ was observed in the presence of 100 nM WT (Fig. 2), a condition under which PI3K activity is fully inhibited (see Fig. 6).

View larger version (67K): [in this window] [in a new window]   Fig. 2.   Interactive effects of WT and Oxo-M on 32P-phosphoinositide labeling. Cells were prelabeled with 32Pi for 2 to 4 h before treatment with either 0.1% DMSO, 100 nM WT, or 10 µM WT for 15 min at 37°C. Cells then were incubated with buffer A or 1 mM Oxo-M for an additional 5 min before the addition of chloroform/methanol to terminate the reactions. Lipids were extracted, separated, and quantitated as described in the text. Values are expressed as lipid labeling relative to control cells (vehicle alone) for PIP2, PIP, and PI. The results shown are the mean ± S.E. for three to four separate experiments. **different from control; p < .01. different from 10 µM WT; p < .05.

WT Inhibits mAChR-Stimulated Inositol Phosphate Formation and Ca²⁺ Signaling. As a consequence of its ability to inhibit inositol lipid synthesis, WT has been demonstrated to prevent the sustained receptor-mediated production of phosphoinositide-derived second messengers in some cell types (Nakanishi et al., 1995; Linseman et al., 1998; Willars et al., 1998). The addition of 1 mM Oxo-M to 1321N1 cells that had been prelabeled to isotopic equilibrium with [³H]inositol resulted in a substantial accumulation (10- to 30-fold) of labeled inositol phosphates. Pretreatment of the cells with WT resulted in a concentration-dependent inhibition of inositol phosphate release (IC₅₀ = 1 µM; Fig. 3A). Inhibition was observed at WT concentrations >100 nM and was maximal at 10 µM WT. The inclusion of WT also resulted in a small reduction in the basal release of inositol phosphates.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Inhibition of mAChR-stimulated inositol phosphate release and Ca2+ signaling by WT. A, cells prelabeled with [3H]inositol for 48 h were treated with the indicated concentrations of WT for 20 min before the addition of buffer A alone () or 1 mM Oxo-M () for an additional 30 min. Reactions were terminated by the addition of chloroform/methanol, and a total inositol phosphate fraction was isolated as described in Experimental Procedures. Results shown are the mean ± S.E. of triplicate replicates obtained from one experiment and are representative of three experiments performed. B, [Ca2+]i responses were monitored in fura-2-loaded cells. Cells were pretreated with vehicle (solid line) or 10 µM WT (dotted line) before the addition of 1 mM Oxo-M, as indicated by the arrow. Traces shown are representative of seven to nine separate experiments performed.

WT Inhibits mAChR Endocytosis in 1321N1 Cells. To determine whether the endocytosis of mAChRs in 1321N1 cells exhibits the same sensitivity to WT as that previously observed for SH-SY5Y cells, hypotonic cell lysates of 1321N1 cells were subjected to differential centrifugation and the agonist-induced appearance of [³H]QNB binding sites in a V₁ fraction was quantitated. The addition of Oxo-M for 30 min resulted in an increased translocation of mAChRs into the V₁ fraction (287 ± 22% of control; n = 4; Fig. 4). Pretreatment of the cells with 10 µM WT for 20 min had little or no effect on mAChR densities in V₁ fractions obtained from control cells, but inhibited the agonist-induced internalization of mAChRs by >85%, a result similar to that previously obtained for mAChR endocytosis in SH-SY5Y neuroblastoma cells (Sorensen et al., 1998). In contrast, preincubation of cells with 100 nM WT, a concentration that has little or no effect on phosphoinositide synthesis (see Fig. 2), did not inhibit mAChR endocytosis. Simultaneous activation of ₂-ARs by the addition of 1 µM Iso also had no effect on the extent of mAChR endocytosis (Fig. 4).

View larger version (48K): [in this window] [in a new window]   Fig. 4.   WT inhibits the agonist-induced endocytosis of mAChRs. Cells were pretreated with either 0.1% DMSO, 100 nM WT, or 10 µM WT for 15 min before the addition of 1 µM Iso or buffer A for an additional 5 min as indicated. Cells then were incubated with either buffer A or 1 mM Oxo-M for an additional 30 min. Reactions were terminated by the addition of ice-cold buffer A, and hypotonic cell lysates were subjected to differential centrifugation, as described in Experimental Procedures. mAChRs present in the V1 fractions were monitored by means of [3H]QNB binding. Values are expressed as the specific binding of [3H]QNB relative to control cells (vehicle alone). Results shown are the mean ± S.E. for three to four separate experiments. **different from Oxo-M alone; p < .01. Control V1 fractions contained 4.1 ± 0.6 fmol receptor/mg of protein.

WT-Mediated Inhibition of ₂-AR Endocytosis Is Potentiated after mAChR Activation. The endocytosis of ₂-ARs was monitored as the agonist-induced appearance of ¹²⁵I-labeled CYP binding sites in the V₁ fraction after a 20-min exposure of the cells to 1 µM Iso. The addition of 1 µM Iso to 1321N1 cells resulted in an increased translocation of ₂-ARs into the V₁ fraction (273 ± 21% of control; n = 5). Pretreatment of the cells with 10 µM WT for 20 min resulted in a partial inhibition (~30%) of ₂-AR endocytosis (Fig. 5). However, if cells that had been pretreated with WT for 15 min were incubated in the presence of 1 mM Oxo-M for the final 5 min of the preincubation period, thereby resulting in a further loss of PIP₂, a more pronounced inhibition (>70%) of Iso-mediated ₂-AR endocytosis was observed (Fig. 5). No inhibition of Iso-stimulated ₂-AR endocytosis was observed after mAChR activation when cells were preincubated in either the absence of WT or at a lower concentration of WT (100 nM).

View larger version (55K): [in this window] [in a new window]   Fig. 5.   Inhibition of 2-AR endocytosis by WT. Cells were pretreated with either 0.1% DMSO, 100 nM WT, or 10 µM WT for 15 min before the addition of 1 mM Oxo-M or buffer A for an additional 5 min as indicated. Cells then were incubated with either buffer A or 1 µM Iso for an additional 20 min. Reactions were terminated by the addition of ice-cold buffer A, and V1 fractions were isolated as described in the legend to Fig. 4. -ARs present in the V1 fraction were monitored by 125I-labeled CYP binding. Values are expressed as the specific binding of 125I-labeled CYP relative to control cells (vehicle alone). The inclusion of WT at either 100 nM or 10 µM had no effect on 125I-labeled CYP binding in cell lysates. Results shown are the mean ± S.E. for three to six separate experiments. *different from Iso alone; p < .05. different from Iso and 10 µM WT; p < .01. Control V1 fractions contained 0.3 ± 0.08 fmol of receptor/mg of protein.

PI4K, a WT-Sensitive Isoform, Is Present in 1321N1 Cells. Of the known isoforms of PI4K, only two, namely PI4K and a 230-kDa enzyme, have been shown to be WT-sensitive (Nakagawa et al., 1996a,b; Meyers and Cantley, 1997). With an antibody raised to the carboxyl terminus of PI4K (which also recognizes the 230-kDa isoform, a splice variant of PI4K), we were unable to detect the 230-kDa enzyme in 1321N1 cells but could readily detect the same isoform in rat brain lysates (data not shown). However, PI4K was readily detectable in both 1321N1 and rat brain lysates by Western blot analysis (Fig. 6A). Furthermore, PI4K activity monitored in immunoprecipitates of PI4K obtained from 1321N1 cell lysates was sensitive to WT with an IC₅₀ of ~300 nM (Fig. 6B). This value is similar to that previously observed for WT-mediated inhibition of inositol phosphate release (1 µM; Fig. 3A). In contrast, the IC₅₀ for WT-mediated inhibition of PI3K activity was ~3 nM. PI3K activity was completely inhibited by 100 nM WT, a concentration that has no effect on either phosphoinositide synthesis or receptor endocytosis.

View larger version (26K): [in this window] [in a new window]   Fig. 6.   WT-sensitive PI4K activity is present in lysates of 1321N1 cells. A, lysates of either 1321N1 cells (5, 10, or 25 µl) or rat brain (50 µl) were electrophoresed through 7.5% SDS-polyacrylamide gels and transferred to polyvinylidene fluoride membranes. Membranes then were immunoblotted for PI4K as described in the text. B, anti-p85 () or anti-PI4K () immunoprecipitates of cell lysates from 1321N1 astrocytoma were treated with vehicle alone or the indicated concentration of WT before determination of enzyme activity as described in Experimental Procedures. Values are expressed as a percentage of PI kinase activity in control cells (vehicle alone) and are the mean ± S.E. for triplicates replicates. Where error bars are not shown, they fell within the symbol.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

In addition to their established role as precursor molecules for the generation of phosphoinositide-derived second messengers, it is now evident that inositol lipids also are essential components in the operation of a diverse array of cell functions. The latter include the regulation of cell adhesion, cytoskeletal assembly, ion channel permeability, and membrane trafficking (De Camilli et al., 1996; Martin, 1997; Toker and Cantley, 1997; Baukrowitz et al., 1998). However, investigations into the role played by these lipids in cell biology have been hampered by the absence of suitable experimental paradigms in which the concentrations of inositol lipids in intact cells can be modulated. One such pharmacological approach that has been described recently is the inhibition of phosphoinositide synthesis by the administration of WT. Although originally identified as an inhibitor of PI3K, it is now evident that higher concentrations of WT inhibit specific isoforms of PI4K and thereby inhibit the synthesis of agonist-sensitive pools of inositol lipids. The isozymes of PI4K inhibited by WT (type III) include PI4K and a 230-kDa enzyme (Nakagawa et al., 1996a,b; Balla et al., 1997; Meyers and Cantley, 1997). In contrast, the type II forms of PI4K are insensitive to WT (Wong and Cantley, 1994; Nakagawa et al., 1996a), as is PIP 5-kinase (Nakanishi et al., 1995). The ability of WT to block the resynthesis of inositol lipids, and thereby the sustained production of phosphoinositide-derived second messengers, was first demonstrated by Nakanishi et al. (1995) for adrenal glomerulosa cells. These findings recently have been extended to SH-SY5Y neuroblastoma cells (Linseman et al., 1998; Willars et al., 1998), to PC-12 cells (D.A.L., S.D.S., and S.K.F., unpublished data), and, in the present study, to 1321N1 astrocytoma. Exposure of 1321N1 cells to WT results in a selective inhibition of both PIP and PIP₂ synthesis, an effect that results from the combination of an inhibition of inositol lipid kinase activity in the presence of continuing PIP- and PIP₂-phosphatase action. The ability of WT to block phospholipid synthesis was specific to the polyphosphoinositides and occurred in a concentration range (100 nM-10 µM) that has been demonstrated to inhibit PI4K activity (Downing et al., 1996; Nakagawa et al., 1996b; Balla et al., 1997; Meyers and Cantley, 1997; Sorensen et al., 1998). Furthermore, these concentrations are well in excess of those required for inhibition of PI3K activity. From the kinetics of loss of label from PIP and PIP₂ (Fig. 1), it is evident that WT primarily inhibits PI4K activity and PIP synthesis, but that as a consequence, PIP₂ concentrations are reduced. The observation that activation of a PLC-coupled receptor, such as the mAChR, can further deplete PIP₂ stores provides additional support for the utility of this experimental paradigm. Under the latter conditions, polyphosphoinositide hydrolysis is accelerated, whereas lipid resynthesis is prevented. As a result, this experimental approach can provide an estimate of the size of the agonist-sensitive pool of inositol lipids in a given tissue. For example, in SH-SY5Y neuroblastoma cells, >85% of the PIP₂ pool is susceptible to depletion by a muscarinic agonist in the presence of WT (Linseman et al., 1998; Sorensen et al., 1998; Willars et al., 1998). The comparable value for bradykinin-stimulated PC-12 cells is 55% (D.A.L., S.D.S., and S.K.F., unpublished data), whereas for either angiotensin-stimulated glomerulosa cells (Nakanishi et al., 1995) or mAChR-activated 1321N1 astrocytoma, approximately 65% of the PIP₂ pool appears to be agonist-sensitive (Fig. 2).

In the current study, the use of the WT experimental paradigm has permitted two major conclusions to be drawn regarding GPCR endocytosis. The first is that the requirement for phosphoinositides in receptor internalization appears to be general and is neither receptor- nor tissue-specific. Not only is the endocytosis of M₃ mAChRs prevented by WT in the astrocytoma cells (as previously demonstrated for the SH-SY5Y neuroblastoma cells), but more significantly, the internalization of the ₂-AR, a non-PLC-coupled receptor, is similarly dependent on phosphoinositide availability. Thus, whereas PIP₂ concentrations and the endocytosis of ₂-ARs are only partially reduced by inclusion of WT alone, the further lowering of PIP₂ concentrations after the activation of mAChRs is accompanied by a corresponding reduction in the extent of ₂-AR internalization (Fig. 5). Because mAChR modulation of ₂-AR endocytosis is not observed either in the absence of WT or in the presence of a lower concentration of WT (100 nM) that has no effect on inositol lipid synthesis, a role for the production of phosphoinositide-derived second messengers in this mAChR-mediated inhibition of endocytosis can be discounted. Although it is recognized that WT is not a specific inhibitor of PI4K, its ability to attenuate receptor endocytosis is independent of either PI3K activity, disruption of the cytoskeleton, second messenger formation, inhibition of myosin light chain kinase activity, or perturbation of ligand binding to the receptor (Sorensen et al., 1998). Moreover, a pronounced inhibition of ₂-AR endocytosis is only observed in WT-pretreated cells after the agonist-induced activation of PLC with the attendant hydrolysis of PIP₂. To date, the only biochemical change observed to consistently parallel the attenuation of receptor endocytosis in WT-pretreated cells is a reduction in inositol lipid availability. A second conclusion to emanate from the present study is that although WT primarily targets PI4K, it is the availability of PIP₂ rather than PIP that is required for receptor endocytosis. This was particularly evident for the internalization of ₂-ARs, for which a maximum inhibition of receptor endocytosis was observed only after a substantial depletion of PIP₂ (but not of PIP; see Figs. 2 and 5). The results obtained suggest that maintenance of a minimum concentration of PIP₂ is required for the sustained operation of receptor endocytosis. In agreement with this conclusion, transfection of NIH 3T3 fibroblasts with a catalytically inactive form of type I PIP 5-kinase, which appears to act as a dominant negative, has been shown to result in the inhibition of the endocytosis of colony-stimulating-factor receptors (Davis et al., 1997).

A role for phosphoinositides in receptor trafficking events is supported also by results obtained with agents other than WT. We have demonstrated previously that incubation of SH-SY5Y neuroblastoma cells with two chemically distinct agents, namely LY-294002 or PAO, also results in an attenuation of phosphoinositide synthesis and the inhibition of receptor endocytosis (Sorensen et al., 1998). Furthermore, the inhibitory effects of PAO on both mAChR endocytosis and phosphoinositide synthesis could be reversed with the bifunctional thiol, 2,3-dimercaptopropanol. Both LY-294002 and PAO have been demonstrated to inhibit PI4K activity and thereby regulate PIP and PIP₂ availability in a variety of tissues (Downing et al., 1996; Wiedemann et al., 1996, 1998; Khvotchev and Südhof, 1998; Sorensen et al., 1998). Although the ability of PAO to inhibit PI4K activity has been appreciated only recently, it should be noted that this agent has been shown to prevent the agonist-induced internalization of ₂-ARs in 1321N1 astrocytoma cells (Hertel et al., 1985) and of angiotensin receptors in adrenal glomerulosa cells (Hunyady et al., 1991). Similarly, we have observed that the addition of 20 µM PAO results in a 99 ± 1% (n = 3) inhibition of the Oxo-M-induced sequestration of mAChRs in 1321N1 cells, as monitored by the loss of cell-surface [³H]N-methylscopolamine binding sites (see Slowiejko et al., 1996), which could be fully reversed by the addition of 2,3-dimercaptopropanol. Taken collectively, these observations support the existence of a mechanistic link between phosphoinositides and receptor endocytosis. However, establishment of a definitive relationship between these two parameters must await the development of more specific inhibitors of PI4K and PIP 5-kinase.

The precise function of inositol lipids in receptor endocytosis remains unknown. Conceivably, high concentrations of a phosphoinositide such as PIP₂ might be localized to sites of vesicle budding where, due to its highly negative polar head group, it may alter membrane curvature and thereby promote membrane budding. Another potential site of regulation is dynamin, a GTPase involved in the scission reaction of plasmalemma-derived clathrin-coated vesicles. PIP₂ can bind to the pleckstrin homology domain of dynamin and activate GTPase activity (Lin and Gilman, 1996). Alternatively, inositol lipids may serve to recruit, activate, or modulate other factors necessary for membrane function. In this context, it is relevant to note that phosphorylated derivatives of PI can bind to at least three distinct binding domains of proteins, namely the src-homology 2 binding domain, and the pleckstrin homology and the phosphotyrosine binding domains (Lemon et al., 1995; Rameh et al., 1995; Zhou et al., 1995). By means of binding to these domains, inositol lipids may serve to recruit proteins and thereby regulate the protein-protein interactions that underlie endocytic events.

In summary, the current study has provided additional evidence to support a role for the phosphoinositides, in particular PIP₂, in the endocytosis of GPCRs. Inositol lipids appear to be required not only for the endocytosis of PLC-coupled receptors, but also for receptors such as the ₂-AR that activate adenylyl cyclase.