Journal of Neurochemistry Raven Press, Ltd., New York 0 1992 International Society for Neurochemistry
Differential Responses of the Phosphorylation of Ribosomal Protein S6 to Nerve Growth Factor and Epidermal Growth Factor in PC12 Cells Tatsuro Mutoh, Brian B. Rudkin, and Gordon Guroff Section on Growth Factors, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryfand, U.S.A.
Abstract: Previous studies from this laboratory have shown that the phosphorylation of the S6 protein of the ribosomes is catalyzed by at least two different and separable kinase activities in PC12 cells. One of these activities is increased by treatment of the cells with nerve growth factor, the other by treatment of the cells with epidermal growth factor. The present work shows that these two factors stimulatethe phosphorylation of S6 with quite different kinetics, and that both the number of phosphates incorporated into S6 and the phosphopeptide pattern of S6 are different in cells treated with nerve growth factor than in cells treated with epidermal growth factor. The characteristics of the nerve growth factorsensitive S6 kinase and of the epidermal growth factor-sensitive kinase were also clearly different. Substrate specificity
and inhibitor studies indicated that neither was identical to cyclic AMP-dependent kinase, kinase C, or the calcium/calmodulin-dependent kinases. However, two major phosphopeptides produced by S6 phosphorylation in nerve growth factor-treated cells were also seen on phosphorylation of S6 by cyclic AMP-dependent kinase in vitro. In addition, when rat liver 40s ribosomal subunits were pretreated with cyclic AMP-dependent kinase in vitro, the action of the nerve growth factor-sensitive S6 kinase was increased about twofold. Key Words: Nerve growth factor-PC12-S6. Mutoh T. et al. Differential responses of the phosphorylation of ribosomal protein S6 to nerve growth factor and epidermal growth factor in PC12 cells. J. Neurochem. 58, 175-185 (1992).
Ribosomal protein S6 is the major phosphoprotein of the 40s ribosomal subunit of eukaryotic cells (Traugh, 1981). There are data to suggest that the phosphorylation state of this protein has a significant effect on the functional properties of the ribosomes (Gressner and Wool, 1974; Thomas et al., 1979, 1980). In addition, it has been suggested that different phosphorylation states of S6 produced by different kinases have different effects on the translation of natural mRNAs, when measured in a reconstituted proteinsynthesizing system (Palen and Traugh, 1987). Although not all the data in the literature are consistent with the findings mentioned above, it is fair to say that a relationship between S6 phosphorylation and the rate or characteristics of protein synthesis has been seen in several different experimental situations. Several mitogens, including serum and a number of
growth factors, have been shown to increase S6 phosphorylation by activating protein kinases that phosphorylate it in animal cells. Phosphorylation of up to five serine residues in S6 can occur when Swiss 3T3 cells are exposed to serum, platelet-derived growth factor, acidic or basic fibroblast growth factor, epidermal growth factor (EGF), insulin, insulin-like growth factor I, prostaglandin Fz,, or phorbol myristate acetate (PMA) (Haselbacher et al., 1979; Nilsen-Hamilton et al., 1982; Wettenhall et al., 1983; Blackshear et al., 1985; Pelech et al., 1986; Pelech and Krebs, 1987). In chick embryo fibroblasts the Rous sarcoma virus transforming gene product, pp6OV+", and PMA also increase the phosphorylation of S6 by increasing the activity of an S6 kinase (Blenis and Erikson, 1985). The activity of an S6 kinase in Xenopus oocytes is stimulated on their maturation to unfertilized eggs
Received February 6 , 199 1; revised manuscript received May 13, 1991; accepted May 22, 1991. Address correspondence and reprint requests to Dr. G. Guroff at Section on Growth Factors, National Institute of Child Health and Human Development, National Institutes of Health, Building 6, Room 130, Bethesda, MD 20892, U S A .
Abbreviations used: dBcAMP, dibutyryl cyclic AMP; DMEM, Dulbecco's modified Eagle's medium; DTT, dithiothreitol; EGF, epidermal growth factor; NGF, nerve growth factor; PBS, phosphatebuffered saline; PMA, phorbol mynstate acetate; PMSF, phenylmethylsulfonyl fluoride; TPCK, N-tosyl-L-phenylalanyl chloromethyl ketone.
,
I75
176
T. MUTOH ET AL.
(Erikson and Maller, 1985). Kinases responsible for the increased phosphorylation of S6 in Xenopus oocytes and 3T3 cells have been purified in several laboratories and the properties of these kinases have been studied. However, the enzymes described in these various studies have different properties and different molecular weights (Lubben and Traugh, 1983; Novak-Hofer and Thomas, 1984; Erikson and Maller, 1985, 1986; Tabarini et al., 1985; Price et al., 1989; Jeno et al., 1989), so it appears that there are several different enzymes involved in S6 phosphorylation in different cells. The information available at present about the enzymatic basis of the action of nerve growth factor (NGF) on S6 phosphorylation is somewhat limited. It has been reported that NGF treatment of PC 12 cells leads to an increase in the phosphorylation of S6 (Halegoua and Patrick, 1980; Blenis and Erikson, 1986; Matsuda et al., 1986), as does treatment of the cells with dibutyryl cyclic AMP (dbcAMP) (Halegoua and Patrick, 1980). Further study has shown that a specific S6 kinase is activated in extracts from nerve growth factor-treated PC 12 cells (Matsuda and Gurof, 1987). PC12 cells, originally placed in culture by Greene and his co-workers, have the interesting property of being responsive to both NGF and EGF. When given NGF they stop dividing, elaborate neurites, become electrically excitable (Greene and Tischler, 1976; Dichter et al., 1977), and will synapse with appropriate muscle cells in culture (Schubert et al., 1977). EGF, on the contrary, is a mild mitogen for PC 12 cells and clearly stimulates cell division (Huff et al., 198I), and, equally clearly, has no differentiative actions on the cells. Because both NGF and EGF stimulate the phosphorylation of S6, this system offers the opportunity to explore the possible role of S6 phosphorylation in the fulfillment of these alternate and seemingly exclusive routes of cellular development. We have shown that cell-free extracts of PC12 cells contain two different S6 kinase activities, separable on heparin-Sepharose chromatography (Mutoh et al., 1988). NGF treatment of the cells increases the activity of one of these peaks; EGF treatment increases the activity of the other. These data permit the hypothesis that NGF and EGF stimulate different S6 kinases in the cells and, further, that they stimulate the phosphorylation of different sites on S6. It is possible to speculate that the different phosphorylations produce different functional states of the S6 protein and that these different functional states are important in the developmental fate of the cells themselves. To begin to test these possibilities, we have examined the characteristics of S6 phosphorylation in cells treated with each of these ligands and tried to identify and characterize the enzyme responsible for each. The results show that the phosphorylation of S6 protein is different in cells treated with NGF than it is in cells treated with EGF, and that the NGF-sensitive phosphorylation appears to be regulated by cyclic AMP-dependent mechanisms at two separate levels. J. Neurochem., Vol. 58. No. I , 1992
MATERIALS AND METHODS Materials NGF was prepared by the method of Bocchini and Angeletti (1969). Ribosomal subunits were prepared from the livers of fasted adult Sprague-Dawley rats according to the method of Thomas et al. (1978). EGF was a product of Collaborative Research (Waltham, MA, U.S.A.). [32P]Pi phosphate was obtained from New England Nuclear (Boston, MA, U.S.A.). N-Tosyl-L-phenylalanyl chloromethyl ketone (TPCK)-treated trypsin was a product of Worthington Biologicals (Freehold, NJ, U.S.A.). Phosphoamino acids were purchased from Sigma (St. Louis, MO, U.S.A.), as were the catalytic subunits of cyclic AMP-dependent protein kinase and the heat-stable inhibitor of cyclic AMP-dependent kinase. Protein kinase C (Singh et al., 1984) and protein kinase M (Huang and Huang, 1986) were provided by Dr. K.-P. Huang. H-7 and W-7 were obtained from Seikagaku America (St. Petersberg, FL, U.S.A.). Mono S and TSKG 3000 SW were purchased from Pharmacia-LKB Biotechnology (Piscataway, NJ, U.S.A.).
Cell culture PC12 cells were cultured as monolayers in 150-cm2culture flasks in Dulbecco’s modified Eagle’s medium (DMEM) containing 6% fetal bovine serum, 6% horse serum, and 100 pg of streptomycin and 100 U of penicillin per milliliter. The cells were kept at 37°C in 6% COz. The cells were split in a ratio of 1:3 or 1:5 each week and the medium changed once between splitting. Usual treatments involved the addition of NGF (50 ng/ml) for 30 min or EGF (30 ng/ml) for 20 min. Control cultures were kept under similar conditions. After treatment the cells were dislodged by shaking, and then collected and washed by centrifugation (1,100 g, 5 min).
isolation of 32P-labeled40s ribosomal subunits from PC12 cells PC12 cells were washed once with phosphate-free DMEM and preincubated in phosphate-free DMEM containing 0.1 % fetal bovine serum and 0.1% horse serum for 60 min at 37°C. t3’P]Pi (0.25 mCi/ml) was added and incubated with the cells for an additional 60 min. Following treatment with the appropriate ligand, the cells were washed twice with ice-cold phosphate-buffered saline (PBS), and 40s ribosomal subunits were isolated according to the method of Decker (198 1) with minor modifications. The cells were lysed with 5 mM TrisHCI (pH 7.4), 10 mM MgC12, 80 mM KCl (TMK buffer) containing 1% sodium deoxycholate, 1% Triton X-100, 2 mM EGTA, 1 mMsodium vanadate, 0.05 mM sodium fluoride, 10 mM KH2P04, and 1 mM phenylmethylsulfonyl fluoride (PMSF). The lysate was centrifuged at 30,000 g for 10 min, the supernatant fraction was layered onto TMK buffer containing 1.5 M sucrose, and the preparation centrifuged at 100,000g for 3 h at 2°C. The pellet was suspended in the same buffer, centrifuged again at 100,000 g for 1 h, and resuspended in fresh TMK buffer containing phosphatase inhibitors. The yield of ribosomes was estimated by measuring the absorbance at 260 nm and the 260/280 ratio was used as an index of the purity of the preparation.
Phosphopeptide mapping of ribosomal protein S6 The S6 protein was obtained from one-dimensional sodium dodecyl sulfate (SDS)-polyacrylamidegels. The gels were dried and stained, and the bands corresponding to S6 excised. The bands were bathed in I-ml portions of 25% isopropanol for
GROWTH FACTORS AND S6 PHOSPHOR YLATION IN PCI 2 6 h, removed from the solution, and crushed into small pieces. The crushed gels were digested with TPCK-treated trypsin ( 1 pg of trypsin/4 pg of S6 protein) in 500 p1 of 50 mM (NH4)HC03, pH 8.3, containing 10 mM CaCI2 for 24 h. Another equal portion of TPCK-treated trypsin was added after approximately 12 h of digestion. The samples were then lyophilized to remove (NH4)HC03and reconstituted in water. The samples were applied to precoated cellulose plates and analyzed in two dimensions. The plate was subjected to electrophoresis at 1.0 kV for 30 min in acetic acid/formic acid/water (15:5:80)at pH 1.9 and then to ascending TLC at a 90" angle in n-butanollacetic acid/pyridine/water (1 5:3: 10:12). The plates were dried and the separated peptides detected by autoradiography.
Phosphoamino acid analysis Portions of the lyophilized samples prepared for phosphopeptide analysis were used for phosphoamino analysis. These portions were hydrolyzed in constant boiling 6 M HCI under nitrogen at 1 10°C for 90 min. Five micrograms each of phosphoserine, phosphothreonine, and phosphotyrosine were added before the hydrolysis. After hydrolysisthe samples were lyophilized and resuspended in 10 pl of water. The samples were then applied to 20 X 20 cellulose thin-layer plates and the phosphoamino acids separated by the method of Nakabayashi et al. (1987). The phosphoamino acids were visualized with ninhydrin and analyzed by autoradiography.
Analysis of multiphosphorylated forms of S6 Ribosomal 40s subunits from 32P-labeledPC I2 cells were isolated as described above. Ribosomal proteins from these subunits and from 40s ribosomal subunits from rat liver that had been phosphorylated in vitro were separated in two dimensions by gel electrophoresis by the method of Siegmann and Thomas (1 987). The gels were dried, stained, and autoradiograms prepared. The multiphosphorylated forms of S6 were identified by comparison of their electrophoretic mobilities with those of the S2, the S4, and the unphosphorylated S6 ribosomal proteins.
Assay of S6 kinase activity S6 kinase activity was measured using 40s ribosomal subunits from the livers of fasted rats as substrate. The incubation was carried out in a total volume of 60 p1 containing 50 mM 4-morpholinepropanesulfonic acid, pH 7.2, 10 mM MgCI2; I Mdithiothreitol (DTT); 0.05% Triton X-100; 16 m M p nitrophenylphosphate; 11 pg of heat-stable inhibitor of cyclic AMP-dependent kinase; 55 pM ATP; [y-32P]ATP5 pg of 40s ribosomal subunit; and a cell-free extract from PC12 cells containing S6 kinase, prepared as previously described by Mutoh et al. (1988), or the partially purified S6 kinases described below. The reaction was initiated by the addition of a mixture of ATP and [y3*P]ATP, and the incubation was camed out usually for 30 min at 30°C. The incorporation of radioactive phosphate into S6 was linear with time for at least 60 min and with enzyme concentration up to at least 50 ng of protein. The reaction was stopped by the addition of 100 pl of concentrated SDS sample buffer (125 mMTrisHCI buffer, pH 6.7; 2% SDS; 20% glycerol; and 5% p-mercaptoethanol). The samples were boiled for 10 min and analyzed by SDS-polyacrylamidegel electrophoresis on 10%or 12%polyacrylamide gels according to the method of Laemmli (1970). The gels were stained with Coomassie Brilliant Blue, dried, and autoradiograms prepared. The autoradiograms were scanned with a Zeineh Soft Laser Densitometer and
177
additional quantitation was obtained by excising the relevant bands and estimating their contained radioactivity by Cerenkov counting.
Separation of S6 kinases from extracts of PC12 cells All procedures were camed out at 0-4°C; samples were never frozen. The cell-free extracts, prepared by both the alkaline lysis method as previously described by Mutoh et al. (1988) and by the method of Novak-Hofer and Thomas (1984) in the presence or absence of @-glycerophosphatein the extraction buffer, from control and from treated cells, were filtered through Millex-GS filters (0.22 Mm; Millipore) and applied directly onto heparin-Sepharose columns as previously described by Mutoh et al. (1988) with minor modifications. The buffer, 50 W Tris-HC1, pH 8.0, contained 2 mM EGTA, 1 mM EDTA, 0.1 pg/ml of leupeptin, and 1 mM benzamidine-HC1 to inhibit phosphatases and proteases. Fractions (0.5 ml each) were collected and assayed for S6 kinase activity, using 40s ribosomal subunits from rat liver as substrate. The active fractions were applied to PD10 columns (Pharmacia) equilibrated with Mono S buffer A (10 mMsodium phosphate, pH 6.8, containing 1 M E G T A , 0.5 mM EDTA, 2 mMDTT, 0.1 pg/ml of leupeptin, 1 M benzamidine-HCI, and 2 mMp-nitrophenylphosphate). The columns were eluted with the same buffer and fractions (20 drops; 0.7 ml) were collected. The A280of each fraction was measured and the peak fractions were pooled and applied to Mono S columns (Pharmacia) that had been equilibrated with Mono S buffer A. The Mono S columns were washed with 8 ml of the same buffer and eluted with Mono S buffer B (Mono S buffer A containing 1.O M NaCl). The active fractions were pooled and concentrated with Centricon 10 filters (Amicon). The concentrated samples were applied to TSK G-3000SW and eluted with 50 M i m i d a z o l e buffer, pH 7.0, containing 2 mMEGTA, 1 mMEDTA, 1 Wbenzamidine, and 0.1 pg/ml of leupeptin.
Pretreatment of 40s ribosomal subunits with cyclic AMP-dependent protein kinase Fifteen micrograms of 40s ribosomal subunits prepared from the livers of fasted rats were treated with catalytic subunits of cyclic AMP-dependent kinase in a 30-pl reaction mixture that contained 25 mM MOPS buffer, pH 7.0; 10 mM MgC12; 1 mM DTT; 2 U of catalytic subunits of cyclic AMP-dependent protein kinase; and 20 pM unlabeled ATP at 30°C for 20 min. The catalytic subunits of cyclic AMPdependent protein kinase were reconstituted in 50 mMDTT before use. In separate experiments it was shown that, under these conditions, a maximum of 2 mol of phosphate were incorporated per mole of S6 and that that incorporation was complete within 10 min. In control reaction mixtures the catalytic subunits were omitted and only the 50 mM DTT vehicle was added. The reaction was terminated by the addition of 20 pg of the heat-stable inhibitor of cyclic AMPdependent protein kinase. These (pre)phosphorylated ribosomal subunits were used as substrate for subsequent phosphorylation of the S6 protein by partially purified S6 kinases from PC12 cells.
RESULTS Phosphorylation of S6 protein in intact cells Experiments were designed by which to compare and contrast the phosphorylation of S6 in PC 12 cells treated with NGF with that in PC12 cells treated with J. Neurochem.. Val. 58. No. 1, 1992
178
T. MUTOH ET AL.
EGF. The time course of S6 phosphorylation under these two conditions was examined. PC12 cells were equilibrated with [32P]orthophosphateand stimulated with either NGF (50 ng/ml) or EGF (30 ng/ml). These concentrations gave maximal stimulation of their respective S6 kinase activities (Mutoh et al., 1988). NGF and EGF stimulated the phosphorylation of S6 protein with quite different time courses (Fig. 1). Half-maximal phosphorylation was observed approximately 20 rnin after the addition of NGF and 40 rnin &er the addition of EGF. Maximal phosphorylation of S6 occurred within 30 rnin after the addition of NGF and stayed at the maximal level for at least 2 h. In contrast, EGF treatment caused a gradual increase in the phosphorylation of the S6 protein which reached a maximal level 2 h after stimulation. It has been shown in previous studies (Yu et al., 1980) that no substantial increase in either the uptake of [32P]orthophosphateinto PC12 cells, or its incorporation into total cellular protein, results from treatment with NGF or EGF under these conditions. Multiphosphorylation of S6 in intact cells The extent of S6 phosphorylation produced by stimulation with these two different ligands was studied by two-dimensionalgel electrophoresis (Fig. 2). In control cells most of the S6 protein was in the native non-
30 Min
-
120 Min S6
edcb a
S6
e dcba
edcba
S6
S6
EGF
J
t
‘4,
edcba
edcba
FIG. 2. Effect of stimulation of PC12 cells with NGF or EGF on
the extent of S6 phosphorylation. PC12 cells,grown as monolayers in 150-cm2 flasks, were treated with [32P]orthophosphate (0.25 mCi/ml) for 60 min. The cells were then treated with either NGF (50 ng/ml) or EGF (30 ng/ml) for either 30 rnin or 120 min. Control cells were incubatedwith [32P]orthophosphate for 60 min and then for another 30 min before harvesting. Ribosomes were prepared from the cells and the ribosomal proteins were analyzed in two dimensions, as described in Materials and Methods. Oblique arrows indicate the position of unphosphorylatedS6 protein as seen on the stained gels; vertical arrows indicate the multiphosphorylated forms of S6. Counts applied: NGF, 30 min, 400 cpm; NGF, 120 min, 420 cprn; EGF, 30 min, 180 cpm; EGF, 120 min, 550 cpm; Control, 30 rnin. 220 cpm.
phosphorylated form; a small amount was present as the monophosphorylated derivative, S6a. In cells treated with NGF, S6 became highly phosphorylated, appeared largely as the tetraphosphorylated (S6d) and pentaphosphorylated (S6e) forms after 30 rnin of treatment, and stayed in these forms for at least 2 h. In contrast, in EGF-treated cells a maximum of 3 mol of phosphate were incorporated into S6 in 30 rnin (S6a, S6b, and S6c);after 2 h pentaphosphorylated S6 could be seen (S6e), but most of the S6 was still in the less phosphorylated forms.
0
60
30
90
120
TIME imiol
FIG. 1. Phosphorylationof the S6 protein following stimulationof
PC12 cells with either NGF or EGF. PC12 cells, grown in monolayer cultures in 150cm2 flasks, were treated with [32P]orthophosphate (0.25 mCi/ml)for 60 min and then for the indicated periods of time with either NGF (a) (50 ng/ml) or EGF (b) (30 ng/ml). Control cells (c) were treated in a similar fashion, but without the addition of growth factors. S6 protein was preparedas describedin Materials and Methods. Samples (0.7 Am units each) were subjected to one-dimensional SDS-polyacrylamide gel electrophoresisand the band corresponding to S6 excised. The radioactivity in these bands was estimated by Cerenkov counting. The experiment was done three times with similar results.
J. Neurochcm.. Vol. 58, No. I . 1992
Phosphopeptide maps of S6 protein phosphorylated in intact cells To determine if the differences produced by these two ligands were simply quantitative, or, alternatively, if there was a qualitative difference in the phosphorylation patterns, the S6 protein from NGF-treated and from EGF-treated cells was cleaved with trypsin and the phosphopeptides produced were examined by twodimensional thin-layer electrophoresis and chromatography (Fig. 3). In unstimulated cells, only three lightly labeled phosphopeptides were seen (Fig. 3C). In contrast, at least nine major phosphopeptides were found in peptide maps of S6 from both NGF-treated and EGF-treated cells. Eight of the phosphopeptides showed identical mobilities;each of the maps contained a unique peptide [peptide “a” on the map from NGF-
GROWTH FACTORS AND S6 PHOSPHORYLATION IN PC12
179
FIG. 3. Two-dimensionalanalysis of tryptic digests of S6 from intact PC12 cells stimulated with either NGF or EGF. S6 protein phosphorylated in intact PC12 cells for 90 min was subjected to complete tryptic digestion. The resulting peptides were separated by two-dimensional thin-layer electrophoresis and chromatography and analyzed by autoradiography. A S6 peptides from cells stimulated with NGF (50 ng/ ml) (1,300 cpm). B: S6 peptides from cells stimulated with EGF (30 ng/ml) (1,100 cpm). C: S6 peptides from control cells (350 cpm). D: Combination of samples A and B (600 cpm each). E: Tracing of superimposed radioautograms from A and B. Although the resolution between peptides a and b in the combined sample (D) in the experiment presented was not complete, this study was done four times with four different preparations and partial or complete resolutionbetween these peptides was obtained each time.
treated cells (Fig. 3A), peptide “b” on the map from EGF-treated cells (Fig. 3B)]. To ensure that these two peptides were, indeed, different, portions of each sample were mixed and the combined sample was analyzed. The result (Fig. 3D) showed that peptide 9a was different than peptide 9b, and that the other eight major peptides were probably identical. Phosphoamino acid analysis The phosphoamino acid analysis of the S6 protein phosphorylated in intact cells (Fig. 4)or by the partially purified NGF-sensitive and EGF-sensitivekinases (data not shown) demonstrated that the phosphorylation in all cases was solely on serine residues. Separation and characterization of S6 kinases To characterize the enzymes responsible for these two, apparently different, phosphorylations, attempts were made to separate and purify the relevant kinases. Initial studies were designed to determine if the alkaline-lysis method for preparing cell-free extracts was comparable to methods used by other groups. Cell-free extracts prepared by the method of Novak-Hofer and Thomas (1984) from both NGF-treated and EGFtreated cells were applied to heparin-Sepharose columns as described in Materials and Methods. The re-
FIG. 4. Phosphoamino acid analysis of
S6 from intact PC12 cells stimulated with either NGF or EGF. Portions of the lyophilized samples prepared for the phosphopeptide mapping shown in Fig. 3 were hydrolyzed in 6 M HCI for 90 min at llO°C in a nitrogenfilled, sealed tube. Appropriatestandards were added and the hydrolysate analyzed by thin-layer electrophoresis on cellulosecoated plates as described in Materials and Methods. The autoradiogram from such a plate is shown. The standards were visualized with ninhydrin and their positionsare indicated by the dotted lines. Left S6 phosphoaminoacids from cells stimulated with NGF. Right: S6 phosphoamino acids from cells stimulated with EGF. P-Ser, phosphoserine; P-Thr, phosphothreonine; P-Tyr, phosphotyrosine.
J.
Neurochem., Vol. 58, No. I, 1992
I80
T. MUTOH ET AL.
sults showed that NGF- and EGF-sensitive kinases were eluted at the same positions as they were when the extracts were prepared by alkaline lysis (Mutoh et al., 1988) and, further, that the inclusion of P-glycerophosphate in the extraction buffer resulted in the loss of the peak representing the NGF-sensitive lunase from either type of extract. When cells were treated with both NGF and EGF, the S6 kinase activity in both peaks was increased (data not shown). The alkaline lysis method did have the advantage that the specific activities of the kinases were at least twice as high as those found with the other method, so subsequent studies were done using alkaline lysis. The active fractions from the heparin-Sepharose columns were desalted on PD-10 columns and applied to Mono S (Fig. 5). The NGF-sensitive kinase activity was eluted at about 200 mM NaCl; the EGF-sensitive kinase was eluted at about 250 mM NaCl. When these peak fractions were concentrated and analyzed by TSKG 3000 SW gel filtration (Fig. 6), the NGF-sensitive kinase eluted with an apparent molecular weight of about 50,000 and the EGF-sensitive kinase with an apparent molecular weight of about 90,000. After Mono S chromatography, each enzyme had been purified at least 60-fold (activity > 1,OOO pmol/min/mg of protein) with recoveries of approximately 20%. Additional characterization of these two enzyme activities was done by inspecting the substrate specificity
NGF
FRACTION NUMBER
FIG. 5. Mono S chromatography of partially purified S6 kinases from PC12 cells. The partially purified NGF-sensitive S6 kinase (fractions 35-40 from heparin-Sepharose) and EGF-sensitive kinase (fractions45-49 from heparin-Sepharose) were pooled separately and desalted on PD-10 columns (PharmaciaBiotechnology) that had been equilibrated with Mono S buffer A (see Materials and Methods). The desalted samples were applied to Mono S columns equilibrated with the same buffer. The cdumns were washed with 8 ml of the same buffer and eluted with a linear gradient made up of 8-ml portions of Mono S buffer A and Mono S buffer B (buffer A containing I M NaCI) at a flow rate of 0.5 ml/min. Portions (4 pl) of each fraction were assayed for S6 kinase activity as described in Materials and Methods. An extract from control cells was applied to heparin-Sepharose and fractions 35-40 and 45-49 combined for chromatography on Mono S. S6 kinase activity from NGFtreated cells (0);S6 kinase activity from EGF-treated cells (m); S6 kinase activity from control cells (A).
J. Neurochem.. Vol. 58. No. 1. 1992
WO
i
800 -
-
6
fr L 4 %
i 2 w
600-
400-
200 -
i
0
I
d
5
10
15
6
20
L
A
25
30
FRACTION NUMBER
FIG. 6. TSKG 3000SW chromatography of partially purified S6
kinases from PC12 cells. The partially purified NGF-sensitive S6 kinase (fractions 21-24 from Mono S;Fig. 6) and EGF-sensitive kinase (fractions 26-29 from Mono S; Fig. 6) were pooled separately and concentrated as described in Materials and Methods. Fractions 21-24 and 26-29 were collected from the chromatography of an extract from control cells. The samples were applied to a TSKG 3000SW column equilibrated with 50 mM imidazole buffer pH 7.0, containing 2 mM EGTA, 1 mM EDTA, 0.1 pg/ml of leupeptin, and 1 mM benzamidine. The column was eluted at a flow rate of 0.5 ml/min with the same buffer and fractions of 0.5 ml each were collected. Portions of each fraction(4 PI) were assayed for S6 kinase activity as described in Materials and Methods, S6 kinase activity from NGF-treatedcells (0);S6 kinase activity from EGF-treatedcells (0); S6 kinase activity from control cells (A). Wo indicates the position at which blue dextran 2000 (Pharmacia P-L Biochemicals)was eluted; the other numbers indicate the molecular masses and positions of elution of a series of marker proteins.
of the activity in each preparation. For both kinase preparations, S6 protein was the best substrate (Table 1). Various kinase inhibitors and activators were also studied (Table 2). Neither H-7, Ca2+,chlorpromazine, nor the heat-stable inhibitor of cyclic AMP-dependent kinase had an effect on either of the kinase preparations. P-Glycerophosphate,however, had a strong inhibitory effect on the NGF-sensitive kinase activity and the presence of NaF resulted in a slight inhibition. P-Glycerophosphate had little or no effect, under the same conditions, on the EGF-sensitive S6 kinase. These results supported the postulate that the NGF-sensitive and the EGF-sensitive S6 kinases, prepared in these experiments, are, indeed, different enzymes and that both are different from cyclic AMP-dependent kinase, protein kinase C, or calcium/calmodulin-dependent protein kinase. Phosphopeptide maps of S6 protein phosphorylated in vitro Further support for the postulate that these kinases are different than the major, well-characterized kinases in the cell was obtained from experiments in which the purified enzymes from PC 12 cells, and preparations of cyclic AMP-dependent kinase, kinase C, and kinase M (protease-cleaved fragment of kinase C), were used
GROWTH FACTORS AND S6 PHOSPHORYLATION IN PC12 TABLE 1. Substrate specificity of the NGF-sensitive and EGF-sensitive S6 kinases from PC12 cells Concentration (mdml)
Substrate ~~~
NGF-sensitive EGF-sensitive kinase kinase (cpm 32Pincorporated)
~
40s subunits Casein Histone HI Histone IIa Histone 111s Phosvitin Phosphorylase a Phosphorylase b Glycogen synthetase ~~
0.08 1.14 0.28 0.28 0.28 0.57 0.72 0.12
712 22 12 0 0 26 0 0
523 11 0 0 0 15 0 0
0.57
0
0
~
The concentrated samples from Mono S column chromatography (
Differential Responses of the Phosphorylation of Ribosomal Protein S6 to Nerve Growth Factor and Epidermal Growth Factor in PC12 Cells Tatsuro Mutoh, Brian B. Rudkin, and Gordon Guroff Section on Growth Factors, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryfand, U.S.A.
Abstract: Previous studies from this laboratory have shown that the phosphorylation of the S6 protein of the ribosomes is catalyzed by at least two different and separable kinase activities in PC12 cells. One of these activities is increased by treatment of the cells with nerve growth factor, the other by treatment of the cells with epidermal growth factor. The present work shows that these two factors stimulatethe phosphorylation of S6 with quite different kinetics, and that both the number of phosphates incorporated into S6 and the phosphopeptide pattern of S6 are different in cells treated with nerve growth factor than in cells treated with epidermal growth factor. The characteristics of the nerve growth factorsensitive S6 kinase and of the epidermal growth factor-sensitive kinase were also clearly different. Substrate specificity
and inhibitor studies indicated that neither was identical to cyclic AMP-dependent kinase, kinase C, or the calcium/calmodulin-dependent kinases. However, two major phosphopeptides produced by S6 phosphorylation in nerve growth factor-treated cells were also seen on phosphorylation of S6 by cyclic AMP-dependent kinase in vitro. In addition, when rat liver 40s ribosomal subunits were pretreated with cyclic AMP-dependent kinase in vitro, the action of the nerve growth factor-sensitive S6 kinase was increased about twofold. Key Words: Nerve growth factor-PC12-S6. Mutoh T. et al. Differential responses of the phosphorylation of ribosomal protein S6 to nerve growth factor and epidermal growth factor in PC12 cells. J. Neurochem. 58, 175-185 (1992).
Ribosomal protein S6 is the major phosphoprotein of the 40s ribosomal subunit of eukaryotic cells (Traugh, 1981). There are data to suggest that the phosphorylation state of this protein has a significant effect on the functional properties of the ribosomes (Gressner and Wool, 1974; Thomas et al., 1979, 1980). In addition, it has been suggested that different phosphorylation states of S6 produced by different kinases have different effects on the translation of natural mRNAs, when measured in a reconstituted proteinsynthesizing system (Palen and Traugh, 1987). Although not all the data in the literature are consistent with the findings mentioned above, it is fair to say that a relationship between S6 phosphorylation and the rate or characteristics of protein synthesis has been seen in several different experimental situations. Several mitogens, including serum and a number of
growth factors, have been shown to increase S6 phosphorylation by activating protein kinases that phosphorylate it in animal cells. Phosphorylation of up to five serine residues in S6 can occur when Swiss 3T3 cells are exposed to serum, platelet-derived growth factor, acidic or basic fibroblast growth factor, epidermal growth factor (EGF), insulin, insulin-like growth factor I, prostaglandin Fz,, or phorbol myristate acetate (PMA) (Haselbacher et al., 1979; Nilsen-Hamilton et al., 1982; Wettenhall et al., 1983; Blackshear et al., 1985; Pelech et al., 1986; Pelech and Krebs, 1987). In chick embryo fibroblasts the Rous sarcoma virus transforming gene product, pp6OV+", and PMA also increase the phosphorylation of S6 by increasing the activity of an S6 kinase (Blenis and Erikson, 1985). The activity of an S6 kinase in Xenopus oocytes is stimulated on their maturation to unfertilized eggs
Received February 6 , 199 1; revised manuscript received May 13, 1991; accepted May 22, 1991. Address correspondence and reprint requests to Dr. G. Guroff at Section on Growth Factors, National Institute of Child Health and Human Development, National Institutes of Health, Building 6, Room 130, Bethesda, MD 20892, U S A .
Abbreviations used: dBcAMP, dibutyryl cyclic AMP; DMEM, Dulbecco's modified Eagle's medium; DTT, dithiothreitol; EGF, epidermal growth factor; NGF, nerve growth factor; PBS, phosphatebuffered saline; PMA, phorbol mynstate acetate; PMSF, phenylmethylsulfonyl fluoride; TPCK, N-tosyl-L-phenylalanyl chloromethyl ketone.
,
I75
176
T. MUTOH ET AL.
(Erikson and Maller, 1985). Kinases responsible for the increased phosphorylation of S6 in Xenopus oocytes and 3T3 cells have been purified in several laboratories and the properties of these kinases have been studied. However, the enzymes described in these various studies have different properties and different molecular weights (Lubben and Traugh, 1983; Novak-Hofer and Thomas, 1984; Erikson and Maller, 1985, 1986; Tabarini et al., 1985; Price et al., 1989; Jeno et al., 1989), so it appears that there are several different enzymes involved in S6 phosphorylation in different cells. The information available at present about the enzymatic basis of the action of nerve growth factor (NGF) on S6 phosphorylation is somewhat limited. It has been reported that NGF treatment of PC 12 cells leads to an increase in the phosphorylation of S6 (Halegoua and Patrick, 1980; Blenis and Erikson, 1986; Matsuda et al., 1986), as does treatment of the cells with dibutyryl cyclic AMP (dbcAMP) (Halegoua and Patrick, 1980). Further study has shown that a specific S6 kinase is activated in extracts from nerve growth factor-treated PC 12 cells (Matsuda and Gurof, 1987). PC12 cells, originally placed in culture by Greene and his co-workers, have the interesting property of being responsive to both NGF and EGF. When given NGF they stop dividing, elaborate neurites, become electrically excitable (Greene and Tischler, 1976; Dichter et al., 1977), and will synapse with appropriate muscle cells in culture (Schubert et al., 1977). EGF, on the contrary, is a mild mitogen for PC 12 cells and clearly stimulates cell division (Huff et al., 198I), and, equally clearly, has no differentiative actions on the cells. Because both NGF and EGF stimulate the phosphorylation of S6, this system offers the opportunity to explore the possible role of S6 phosphorylation in the fulfillment of these alternate and seemingly exclusive routes of cellular development. We have shown that cell-free extracts of PC12 cells contain two different S6 kinase activities, separable on heparin-Sepharose chromatography (Mutoh et al., 1988). NGF treatment of the cells increases the activity of one of these peaks; EGF treatment increases the activity of the other. These data permit the hypothesis that NGF and EGF stimulate different S6 kinases in the cells and, further, that they stimulate the phosphorylation of different sites on S6. It is possible to speculate that the different phosphorylations produce different functional states of the S6 protein and that these different functional states are important in the developmental fate of the cells themselves. To begin to test these possibilities, we have examined the characteristics of S6 phosphorylation in cells treated with each of these ligands and tried to identify and characterize the enzyme responsible for each. The results show that the phosphorylation of S6 protein is different in cells treated with NGF than it is in cells treated with EGF, and that the NGF-sensitive phosphorylation appears to be regulated by cyclic AMP-dependent mechanisms at two separate levels. J. Neurochem., Vol. 58. No. I , 1992
MATERIALS AND METHODS Materials NGF was prepared by the method of Bocchini and Angeletti (1969). Ribosomal subunits were prepared from the livers of fasted adult Sprague-Dawley rats according to the method of Thomas et al. (1978). EGF was a product of Collaborative Research (Waltham, MA, U.S.A.). [32P]Pi phosphate was obtained from New England Nuclear (Boston, MA, U.S.A.). N-Tosyl-L-phenylalanyl chloromethyl ketone (TPCK)-treated trypsin was a product of Worthington Biologicals (Freehold, NJ, U.S.A.). Phosphoamino acids were purchased from Sigma (St. Louis, MO, U.S.A.), as were the catalytic subunits of cyclic AMP-dependent protein kinase and the heat-stable inhibitor of cyclic AMP-dependent kinase. Protein kinase C (Singh et al., 1984) and protein kinase M (Huang and Huang, 1986) were provided by Dr. K.-P. Huang. H-7 and W-7 were obtained from Seikagaku America (St. Petersberg, FL, U.S.A.). Mono S and TSKG 3000 SW were purchased from Pharmacia-LKB Biotechnology (Piscataway, NJ, U.S.A.).
Cell culture PC12 cells were cultured as monolayers in 150-cm2culture flasks in Dulbecco’s modified Eagle’s medium (DMEM) containing 6% fetal bovine serum, 6% horse serum, and 100 pg of streptomycin and 100 U of penicillin per milliliter. The cells were kept at 37°C in 6% COz. The cells were split in a ratio of 1:3 or 1:5 each week and the medium changed once between splitting. Usual treatments involved the addition of NGF (50 ng/ml) for 30 min or EGF (30 ng/ml) for 20 min. Control cultures were kept under similar conditions. After treatment the cells were dislodged by shaking, and then collected and washed by centrifugation (1,100 g, 5 min).
isolation of 32P-labeled40s ribosomal subunits from PC12 cells PC12 cells were washed once with phosphate-free DMEM and preincubated in phosphate-free DMEM containing 0.1 % fetal bovine serum and 0.1% horse serum for 60 min at 37°C. t3’P]Pi (0.25 mCi/ml) was added and incubated with the cells for an additional 60 min. Following treatment with the appropriate ligand, the cells were washed twice with ice-cold phosphate-buffered saline (PBS), and 40s ribosomal subunits were isolated according to the method of Decker (198 1) with minor modifications. The cells were lysed with 5 mM TrisHCI (pH 7.4), 10 mM MgC12, 80 mM KCl (TMK buffer) containing 1% sodium deoxycholate, 1% Triton X-100, 2 mM EGTA, 1 mMsodium vanadate, 0.05 mM sodium fluoride, 10 mM KH2P04, and 1 mM phenylmethylsulfonyl fluoride (PMSF). The lysate was centrifuged at 30,000 g for 10 min, the supernatant fraction was layered onto TMK buffer containing 1.5 M sucrose, and the preparation centrifuged at 100,000g for 3 h at 2°C. The pellet was suspended in the same buffer, centrifuged again at 100,000 g for 1 h, and resuspended in fresh TMK buffer containing phosphatase inhibitors. The yield of ribosomes was estimated by measuring the absorbance at 260 nm and the 260/280 ratio was used as an index of the purity of the preparation.
Phosphopeptide mapping of ribosomal protein S6 The S6 protein was obtained from one-dimensional sodium dodecyl sulfate (SDS)-polyacrylamidegels. The gels were dried and stained, and the bands corresponding to S6 excised. The bands were bathed in I-ml portions of 25% isopropanol for
GROWTH FACTORS AND S6 PHOSPHOR YLATION IN PCI 2 6 h, removed from the solution, and crushed into small pieces. The crushed gels were digested with TPCK-treated trypsin ( 1 pg of trypsin/4 pg of S6 protein) in 500 p1 of 50 mM (NH4)HC03, pH 8.3, containing 10 mM CaCI2 for 24 h. Another equal portion of TPCK-treated trypsin was added after approximately 12 h of digestion. The samples were then lyophilized to remove (NH4)HC03and reconstituted in water. The samples were applied to precoated cellulose plates and analyzed in two dimensions. The plate was subjected to electrophoresis at 1.0 kV for 30 min in acetic acid/formic acid/water (15:5:80)at pH 1.9 and then to ascending TLC at a 90" angle in n-butanollacetic acid/pyridine/water (1 5:3: 10:12). The plates were dried and the separated peptides detected by autoradiography.
Phosphoamino acid analysis Portions of the lyophilized samples prepared for phosphopeptide analysis were used for phosphoamino analysis. These portions were hydrolyzed in constant boiling 6 M HCI under nitrogen at 1 10°C for 90 min. Five micrograms each of phosphoserine, phosphothreonine, and phosphotyrosine were added before the hydrolysis. After hydrolysisthe samples were lyophilized and resuspended in 10 pl of water. The samples were then applied to 20 X 20 cellulose thin-layer plates and the phosphoamino acids separated by the method of Nakabayashi et al. (1987). The phosphoamino acids were visualized with ninhydrin and analyzed by autoradiography.
Analysis of multiphosphorylated forms of S6 Ribosomal 40s subunits from 32P-labeledPC I2 cells were isolated as described above. Ribosomal proteins from these subunits and from 40s ribosomal subunits from rat liver that had been phosphorylated in vitro were separated in two dimensions by gel electrophoresis by the method of Siegmann and Thomas (1 987). The gels were dried, stained, and autoradiograms prepared. The multiphosphorylated forms of S6 were identified by comparison of their electrophoretic mobilities with those of the S2, the S4, and the unphosphorylated S6 ribosomal proteins.
Assay of S6 kinase activity S6 kinase activity was measured using 40s ribosomal subunits from the livers of fasted rats as substrate. The incubation was carried out in a total volume of 60 p1 containing 50 mM 4-morpholinepropanesulfonic acid, pH 7.2, 10 mM MgCI2; I Mdithiothreitol (DTT); 0.05% Triton X-100; 16 m M p nitrophenylphosphate; 11 pg of heat-stable inhibitor of cyclic AMP-dependent kinase; 55 pM ATP; [y-32P]ATP5 pg of 40s ribosomal subunit; and a cell-free extract from PC12 cells containing S6 kinase, prepared as previously described by Mutoh et al. (1988), or the partially purified S6 kinases described below. The reaction was initiated by the addition of a mixture of ATP and [y3*P]ATP, and the incubation was camed out usually for 30 min at 30°C. The incorporation of radioactive phosphate into S6 was linear with time for at least 60 min and with enzyme concentration up to at least 50 ng of protein. The reaction was stopped by the addition of 100 pl of concentrated SDS sample buffer (125 mMTrisHCI buffer, pH 6.7; 2% SDS; 20% glycerol; and 5% p-mercaptoethanol). The samples were boiled for 10 min and analyzed by SDS-polyacrylamidegel electrophoresis on 10%or 12%polyacrylamide gels according to the method of Laemmli (1970). The gels were stained with Coomassie Brilliant Blue, dried, and autoradiograms prepared. The autoradiograms were scanned with a Zeineh Soft Laser Densitometer and
177
additional quantitation was obtained by excising the relevant bands and estimating their contained radioactivity by Cerenkov counting.
Separation of S6 kinases from extracts of PC12 cells All procedures were camed out at 0-4°C; samples were never frozen. The cell-free extracts, prepared by both the alkaline lysis method as previously described by Mutoh et al. (1988) and by the method of Novak-Hofer and Thomas (1984) in the presence or absence of @-glycerophosphatein the extraction buffer, from control and from treated cells, were filtered through Millex-GS filters (0.22 Mm; Millipore) and applied directly onto heparin-Sepharose columns as previously described by Mutoh et al. (1988) with minor modifications. The buffer, 50 W Tris-HC1, pH 8.0, contained 2 mM EGTA, 1 mM EDTA, 0.1 pg/ml of leupeptin, and 1 mM benzamidine-HC1 to inhibit phosphatases and proteases. Fractions (0.5 ml each) were collected and assayed for S6 kinase activity, using 40s ribosomal subunits from rat liver as substrate. The active fractions were applied to PD10 columns (Pharmacia) equilibrated with Mono S buffer A (10 mMsodium phosphate, pH 6.8, containing 1 M E G T A , 0.5 mM EDTA, 2 mMDTT, 0.1 pg/ml of leupeptin, 1 M benzamidine-HCI, and 2 mMp-nitrophenylphosphate). The columns were eluted with the same buffer and fractions (20 drops; 0.7 ml) were collected. The A280of each fraction was measured and the peak fractions were pooled and applied to Mono S columns (Pharmacia) that had been equilibrated with Mono S buffer A. The Mono S columns were washed with 8 ml of the same buffer and eluted with Mono S buffer B (Mono S buffer A containing 1.O M NaCl). The active fractions were pooled and concentrated with Centricon 10 filters (Amicon). The concentrated samples were applied to TSK G-3000SW and eluted with 50 M i m i d a z o l e buffer, pH 7.0, containing 2 mMEGTA, 1 mMEDTA, 1 Wbenzamidine, and 0.1 pg/ml of leupeptin.
Pretreatment of 40s ribosomal subunits with cyclic AMP-dependent protein kinase Fifteen micrograms of 40s ribosomal subunits prepared from the livers of fasted rats were treated with catalytic subunits of cyclic AMP-dependent kinase in a 30-pl reaction mixture that contained 25 mM MOPS buffer, pH 7.0; 10 mM MgC12; 1 mM DTT; 2 U of catalytic subunits of cyclic AMP-dependent protein kinase; and 20 pM unlabeled ATP at 30°C for 20 min. The catalytic subunits of cyclic AMPdependent protein kinase were reconstituted in 50 mMDTT before use. In separate experiments it was shown that, under these conditions, a maximum of 2 mol of phosphate were incorporated per mole of S6 and that that incorporation was complete within 10 min. In control reaction mixtures the catalytic subunits were omitted and only the 50 mM DTT vehicle was added. The reaction was terminated by the addition of 20 pg of the heat-stable inhibitor of cyclic AMPdependent protein kinase. These (pre)phosphorylated ribosomal subunits were used as substrate for subsequent phosphorylation of the S6 protein by partially purified S6 kinases from PC12 cells.
RESULTS Phosphorylation of S6 protein in intact cells Experiments were designed by which to compare and contrast the phosphorylation of S6 in PC 12 cells treated with NGF with that in PC12 cells treated with J. Neurochem.. Val. 58. No. 1, 1992
178
T. MUTOH ET AL.
EGF. The time course of S6 phosphorylation under these two conditions was examined. PC12 cells were equilibrated with [32P]orthophosphateand stimulated with either NGF (50 ng/ml) or EGF (30 ng/ml). These concentrations gave maximal stimulation of their respective S6 kinase activities (Mutoh et al., 1988). NGF and EGF stimulated the phosphorylation of S6 protein with quite different time courses (Fig. 1). Half-maximal phosphorylation was observed approximately 20 rnin after the addition of NGF and 40 rnin &er the addition of EGF. Maximal phosphorylation of S6 occurred within 30 rnin after the addition of NGF and stayed at the maximal level for at least 2 h. In contrast, EGF treatment caused a gradual increase in the phosphorylation of the S6 protein which reached a maximal level 2 h after stimulation. It has been shown in previous studies (Yu et al., 1980) that no substantial increase in either the uptake of [32P]orthophosphateinto PC12 cells, or its incorporation into total cellular protein, results from treatment with NGF or EGF under these conditions. Multiphosphorylation of S6 in intact cells The extent of S6 phosphorylation produced by stimulation with these two different ligands was studied by two-dimensionalgel electrophoresis (Fig. 2). In control cells most of the S6 protein was in the native non-
30 Min
-
120 Min S6
edcb a
S6
e dcba
edcba
S6
S6
EGF
J
t
‘4,
edcba
edcba
FIG. 2. Effect of stimulation of PC12 cells with NGF or EGF on
the extent of S6 phosphorylation. PC12 cells,grown as monolayers in 150-cm2 flasks, were treated with [32P]orthophosphate (0.25 mCi/ml) for 60 min. The cells were then treated with either NGF (50 ng/ml) or EGF (30 ng/ml) for either 30 rnin or 120 min. Control cells were incubatedwith [32P]orthophosphate for 60 min and then for another 30 min before harvesting. Ribosomes were prepared from the cells and the ribosomal proteins were analyzed in two dimensions, as described in Materials and Methods. Oblique arrows indicate the position of unphosphorylatedS6 protein as seen on the stained gels; vertical arrows indicate the multiphosphorylated forms of S6. Counts applied: NGF, 30 min, 400 cpm; NGF, 120 min, 420 cprn; EGF, 30 min, 180 cpm; EGF, 120 min, 550 cpm; Control, 30 rnin. 220 cpm.
phosphorylated form; a small amount was present as the monophosphorylated derivative, S6a. In cells treated with NGF, S6 became highly phosphorylated, appeared largely as the tetraphosphorylated (S6d) and pentaphosphorylated (S6e) forms after 30 rnin of treatment, and stayed in these forms for at least 2 h. In contrast, in EGF-treated cells a maximum of 3 mol of phosphate were incorporated into S6 in 30 rnin (S6a, S6b, and S6c);after 2 h pentaphosphorylated S6 could be seen (S6e), but most of the S6 was still in the less phosphorylated forms.
0
60
30
90
120
TIME imiol
FIG. 1. Phosphorylationof the S6 protein following stimulationof
PC12 cells with either NGF or EGF. PC12 cells, grown in monolayer cultures in 150cm2 flasks, were treated with [32P]orthophosphate (0.25 mCi/ml)for 60 min and then for the indicated periods of time with either NGF (a) (50 ng/ml) or EGF (b) (30 ng/ml). Control cells (c) were treated in a similar fashion, but without the addition of growth factors. S6 protein was preparedas describedin Materials and Methods. Samples (0.7 Am units each) were subjected to one-dimensional SDS-polyacrylamide gel electrophoresisand the band corresponding to S6 excised. The radioactivity in these bands was estimated by Cerenkov counting. The experiment was done three times with similar results.
J. Neurochcm.. Vol. 58, No. I . 1992
Phosphopeptide maps of S6 protein phosphorylated in intact cells To determine if the differences produced by these two ligands were simply quantitative, or, alternatively, if there was a qualitative difference in the phosphorylation patterns, the S6 protein from NGF-treated and from EGF-treated cells was cleaved with trypsin and the phosphopeptides produced were examined by twodimensional thin-layer electrophoresis and chromatography (Fig. 3). In unstimulated cells, only three lightly labeled phosphopeptides were seen (Fig. 3C). In contrast, at least nine major phosphopeptides were found in peptide maps of S6 from both NGF-treated and EGF-treated cells. Eight of the phosphopeptides showed identical mobilities;each of the maps contained a unique peptide [peptide “a” on the map from NGF-
GROWTH FACTORS AND S6 PHOSPHORYLATION IN PC12
179
FIG. 3. Two-dimensionalanalysis of tryptic digests of S6 from intact PC12 cells stimulated with either NGF or EGF. S6 protein phosphorylated in intact PC12 cells for 90 min was subjected to complete tryptic digestion. The resulting peptides were separated by two-dimensional thin-layer electrophoresis and chromatography and analyzed by autoradiography. A S6 peptides from cells stimulated with NGF (50 ng/ ml) (1,300 cpm). B: S6 peptides from cells stimulated with EGF (30 ng/ml) (1,100 cpm). C: S6 peptides from control cells (350 cpm). D: Combination of samples A and B (600 cpm each). E: Tracing of superimposed radioautograms from A and B. Although the resolution between peptides a and b in the combined sample (D) in the experiment presented was not complete, this study was done four times with four different preparations and partial or complete resolutionbetween these peptides was obtained each time.
treated cells (Fig. 3A), peptide “b” on the map from EGF-treated cells (Fig. 3B)]. To ensure that these two peptides were, indeed, different, portions of each sample were mixed and the combined sample was analyzed. The result (Fig. 3D) showed that peptide 9a was different than peptide 9b, and that the other eight major peptides were probably identical. Phosphoamino acid analysis The phosphoamino acid analysis of the S6 protein phosphorylated in intact cells (Fig. 4)or by the partially purified NGF-sensitive and EGF-sensitivekinases (data not shown) demonstrated that the phosphorylation in all cases was solely on serine residues. Separation and characterization of S6 kinases To characterize the enzymes responsible for these two, apparently different, phosphorylations, attempts were made to separate and purify the relevant kinases. Initial studies were designed to determine if the alkaline-lysis method for preparing cell-free extracts was comparable to methods used by other groups. Cell-free extracts prepared by the method of Novak-Hofer and Thomas (1984) from both NGF-treated and EGFtreated cells were applied to heparin-Sepharose columns as described in Materials and Methods. The re-
FIG. 4. Phosphoamino acid analysis of
S6 from intact PC12 cells stimulated with either NGF or EGF. Portions of the lyophilized samples prepared for the phosphopeptide mapping shown in Fig. 3 were hydrolyzed in 6 M HCI for 90 min at llO°C in a nitrogenfilled, sealed tube. Appropriatestandards were added and the hydrolysate analyzed by thin-layer electrophoresis on cellulosecoated plates as described in Materials and Methods. The autoradiogram from such a plate is shown. The standards were visualized with ninhydrin and their positionsare indicated by the dotted lines. Left S6 phosphoaminoacids from cells stimulated with NGF. Right: S6 phosphoamino acids from cells stimulated with EGF. P-Ser, phosphoserine; P-Thr, phosphothreonine; P-Tyr, phosphotyrosine.
J.
Neurochem., Vol. 58, No. I, 1992
I80
T. MUTOH ET AL.
sults showed that NGF- and EGF-sensitive kinases were eluted at the same positions as they were when the extracts were prepared by alkaline lysis (Mutoh et al., 1988) and, further, that the inclusion of P-glycerophosphate in the extraction buffer resulted in the loss of the peak representing the NGF-sensitive lunase from either type of extract. When cells were treated with both NGF and EGF, the S6 kinase activity in both peaks was increased (data not shown). The alkaline lysis method did have the advantage that the specific activities of the kinases were at least twice as high as those found with the other method, so subsequent studies were done using alkaline lysis. The active fractions from the heparin-Sepharose columns were desalted on PD-10 columns and applied to Mono S (Fig. 5). The NGF-sensitive kinase activity was eluted at about 200 mM NaCl; the EGF-sensitive kinase was eluted at about 250 mM NaCl. When these peak fractions were concentrated and analyzed by TSKG 3000 SW gel filtration (Fig. 6), the NGF-sensitive kinase eluted with an apparent molecular weight of about 50,000 and the EGF-sensitive kinase with an apparent molecular weight of about 90,000. After Mono S chromatography, each enzyme had been purified at least 60-fold (activity > 1,OOO pmol/min/mg of protein) with recoveries of approximately 20%. Additional characterization of these two enzyme activities was done by inspecting the substrate specificity
NGF
FRACTION NUMBER
FIG. 5. Mono S chromatography of partially purified S6 kinases from PC12 cells. The partially purified NGF-sensitive S6 kinase (fractions 35-40 from heparin-Sepharose) and EGF-sensitive kinase (fractions45-49 from heparin-Sepharose) were pooled separately and desalted on PD-10 columns (PharmaciaBiotechnology) that had been equilibrated with Mono S buffer A (see Materials and Methods). The desalted samples were applied to Mono S columns equilibrated with the same buffer. The cdumns were washed with 8 ml of the same buffer and eluted with a linear gradient made up of 8-ml portions of Mono S buffer A and Mono S buffer B (buffer A containing I M NaCI) at a flow rate of 0.5 ml/min. Portions (4 pl) of each fraction were assayed for S6 kinase activity as described in Materials and Methods. An extract from control cells was applied to heparin-Sepharose and fractions 35-40 and 45-49 combined for chromatography on Mono S. S6 kinase activity from NGFtreated cells (0);S6 kinase activity from EGF-treated cells (m); S6 kinase activity from control cells (A).
J. Neurochem.. Vol. 58. No. 1. 1992
WO
i
800 -
-
6
fr L 4 %
i 2 w
600-
400-
200 -
i
0
I
d
5
10
15
6
20
L
A
25
30
FRACTION NUMBER
FIG. 6. TSKG 3000SW chromatography of partially purified S6
kinases from PC12 cells. The partially purified NGF-sensitive S6 kinase (fractions 21-24 from Mono S;Fig. 6) and EGF-sensitive kinase (fractions 26-29 from Mono S; Fig. 6) were pooled separately and concentrated as described in Materials and Methods. Fractions 21-24 and 26-29 were collected from the chromatography of an extract from control cells. The samples were applied to a TSKG 3000SW column equilibrated with 50 mM imidazole buffer pH 7.0, containing 2 mM EGTA, 1 mM EDTA, 0.1 pg/ml of leupeptin, and 1 mM benzamidine. The column was eluted at a flow rate of 0.5 ml/min with the same buffer and fractions of 0.5 ml each were collected. Portions of each fraction(4 PI) were assayed for S6 kinase activity as described in Materials and Methods, S6 kinase activity from NGF-treatedcells (0);S6 kinase activity from EGF-treatedcells (0); S6 kinase activity from control cells (A). Wo indicates the position at which blue dextran 2000 (Pharmacia P-L Biochemicals)was eluted; the other numbers indicate the molecular masses and positions of elution of a series of marker proteins.
of the activity in each preparation. For both kinase preparations, S6 protein was the best substrate (Table 1). Various kinase inhibitors and activators were also studied (Table 2). Neither H-7, Ca2+,chlorpromazine, nor the heat-stable inhibitor of cyclic AMP-dependent kinase had an effect on either of the kinase preparations. P-Glycerophosphate,however, had a strong inhibitory effect on the NGF-sensitive kinase activity and the presence of NaF resulted in a slight inhibition. P-Glycerophosphate had little or no effect, under the same conditions, on the EGF-sensitive S6 kinase. These results supported the postulate that the NGF-sensitive and the EGF-sensitive S6 kinases, prepared in these experiments, are, indeed, different enzymes and that both are different from cyclic AMP-dependent kinase, protein kinase C, or calcium/calmodulin-dependent protein kinase. Phosphopeptide maps of S6 protein phosphorylated in vitro Further support for the postulate that these kinases are different than the major, well-characterized kinases in the cell was obtained from experiments in which the purified enzymes from PC 12 cells, and preparations of cyclic AMP-dependent kinase, kinase C, and kinase M (protease-cleaved fragment of kinase C), were used
GROWTH FACTORS AND S6 PHOSPHORYLATION IN PC12 TABLE 1. Substrate specificity of the NGF-sensitive and EGF-sensitive S6 kinases from PC12 cells Concentration (mdml)
Substrate ~~~
NGF-sensitive EGF-sensitive kinase kinase (cpm 32Pincorporated)
~
40s subunits Casein Histone HI Histone IIa Histone 111s Phosvitin Phosphorylase a Phosphorylase b Glycogen synthetase ~~
0.08 1.14 0.28 0.28 0.28 0.57 0.72 0.12
712 22 12 0 0 26 0 0
523 11 0 0 0 15 0 0
0.57
0
0
~
The concentrated samples from Mono S column chromatography (
