Phosphorylation of chromogranin A and catecholamine secretion stimulated by elevation of intracellular Ca2+ in cultured bovine adrenal medullary cells.

We have recently isolated a new endogenous substrate of 70 kDa for Ca2+/calmodulin-dependent protein kinase II (CaM kinase II) from bovine adrenal medullary cells (Yanagihara, N., Toyohira, Y., Yamamoto, H., Ohta, Y., Tsutsui, M., Miyamoto, E., and Izumi, F. (1994) Mol. Pharmacol. 46, 423-430). Here we report the sequence analysis of the 70-kDa protein and examine its phosphorylation by various protein kinases in vitro and by depolarization of the cultured cells. Protein sequencing and immunoblotting revealed that the 70-kDa protein is chromogranin A (CgA) or a closely related protein. Partially purified CgA was phosphorylated by cyclic AMP-dependent protein kinase and protein kinase C as well as CaM kinase II. Tryptic phosphopeptide mapping patterns of CgA differed among these protein kinases. In 32P-labeled bovine adrenal medullary cells, 56 mM K+ increased the phosphorylation of CgA and catecholamine secretion in similar time- and concentration-dependent manners, both of which were inhibited by 20 mM MgSO4, an inhibitor of voltage-dependent Ca2+ channels. These findings suggest that CgA serves as a substrate for several multifunctional protein kinases and that the elevation of the intracellular Ca2+ stimulates the phosphorylation of CgA associated with catecholamine secretion in cultured adrenal medullary cells.

Stimulation of cell-surface receptors leads to the generation of intracellular second messengers such as cyclic AMP, Ca 2ϩ , and diacyl glycerol, which cause the activation of a variety of protein kinases (1). These protein kinases, including cyclic AMP-dependent protein kinase (PKA), 1 protein kinase C (PKC), and Ca 2ϩ /calmodulin-dependent protein kinase II (CaM kinase II) (see reviews, Refs. [2][3][4], have broad substrate specificities and, therefore, are considered as multifunctional pro-tein kinases that are involved in diverse cellular processes such as synaptic transmission, metabolism, synaptic plasticity, gene expression, stimulus-secretion coupling, and cell growth and differentiation. In adrenal medullary cells, Ca 2ϩ which enters the cells by activation of the acetylcholine receptor plays a critical role in the stimulation of catecholamine secretion (5) and synthesis (6). Accumulating evidence has indicated that protein phosphorylation is associated with stimulus-secretion coupling (7,8) and with the regulation of catecholamine synthesis (6,9). Cholinergic stimulation of bovine adrenal medullary cells produces an increase in cyclic AMP production (10,11) and activates PKC (12), which modulate the secretion and synthesis of catecholamines. Furthermore, several studies have shown that CaM kinase II is a possible candidate to mediate the phosphorylation of tyrosine hydroxylase by stimulation with carbachol or depolarization in rat pheochromocytoma PC12 cells (13,14) and bovine adrenal medullary cells (15,16).
Recently, we have reported the presence of an isoform of CaM kinase II and its new endogenous substrate of 70 kDa in bovine adrenal medullary cells (17). Furthermore, we have demonstrated a close relationship among CaM kinase II activation, catecholamine secretion, and tyrosine hydroxylase activation in cultured adrenal medullary cells (18). In the present study, we have analyzed the substrate of 70 kDa for CaM kinase II and found that the protein is chromogranin A (CgA) or a closely related protein. The protein was phosphorylated in vitro by multifunctional protein kinases, and the phosphorylation of CgA was associated with catecholamine secretion in cultured adrenal medullary cells stimulated by 56 mM K ϩ .
Preparation of Cultured Adrenal Medullary Cells and Purification of the 70-kDa Protein from the Cells-Bovine adrenal medullary cells were isolated by collagenase digestion, as described previously (22). The isolated cells were purified by a selective plating method (23) and maintained in a CO 2 incubator under 5% CO 2 /95% air (17). The 70-kDa protein was purified from the cultured bovine adrenal medullary cells on DEAE-cellulose, CaM affinity, and Sephacryl S-300 columns (17).
Sequence Analysis of the 70-kDa Protein-The N-terminal sequence of the 70-kDa protein was determined as follows. The 70-kDa protein was separated by SDS-PAGE, and the protein was electroblotted onto polyvinylidene difluoride membrane (Immobilon, Millipore) using a semidry blotting apparatus (Biometra-Fast-Blot, Biometrabiomedizinische Analytik). After staining with Coomassie Brilliant Blue, the band of 70-kDa protein was cut off and directly analyzed by using an automated gas-phase sequencer (Shimadzu, PPSQ-10) (24). The partial amino acid sequences were determined with four peptides obtained by subfragmentation of the 70-kDa protein with Achromobacter protease 1 (25), using an automated gas-phase sequencer (Applied Biosystems, model 470A). The determined amino acid sequences were analyzed, using the National Biomedical Research Foundation data base.
Partial Purification of CgA from Bovine Adrenal Chromaffin Granules-Intact chromaffin granules were isolated from fresh bovine adrenal medulla (26). After destruction of granules by hypotonic shock, the soluble fraction of chromaffin granules was subjected to a series of column chromatography of DEAE-cellulose and Sephacryl S-300, with a 50% ammonium sulfate fractionation of the DEAE-cellulose column eluate. The CgA fraction was collected and checked by SDS-PAGE (27).
Preparation of Antisera-Antisera to CgA was prepared in female New Zealand White rabbits. One mg of CgA separated from SDS-PAGE gels was emulsified with complete Freund's adjuvant and was injected into multiple intradermal sites, on four occasions at 3-week intervals. After separation of the sera by centrifugation, the IgG fraction was precipitated by 40% ammonium sulfate and separated by application to a DEAE-cellulose column (1.1 ϫ 8 cm). The unretained fraction was collected and used as the anti-CgA antibody.
Immunoblotting of the 70-kDa Protein with the Anti-CgA Antibody-The 70-kDa protein (1.9 g) and partially purified CgA (2.6 g) were separated by SDS-PAGE in 10% acrylamide. After the electrophoresis, proteins were transferred to a polyvinylidene difluoride membrane, as described above. The membrane was incubated at 4°C overnight with the anti-CgA antibody. The bound antibody was treated with goat anti-rabbit IgG conjugated to horseradish peroxidase (Medical & Biological Laboratories). The enzyme color was developed with O-phenylenediamine as a chromogen.
Assay of CgA in the Cell Supernatant-The amount of CgA in the supernatant and the particulate fractions of cell homogenates was measured by the enzyme-linked immunoassay (28). In brief, cultured adrenal medullary cells (4 ϫ 10 6 cells/dish) were harvested and homogenized with 200 l of an isotonic 0.27 M sucrose buffer, containing 50 mM Tris-HCl, pH 7.5, 10 mM EDTA, 25 mM NaF, 4 mM EGTA, 0.43 mM phenylmethylsulfonyl fluoride, 0.05 mM leupeptin, and 50 mg/liter trypsin inhibitor. After centrifugation at 15,000 ϫ g for 10 min, the resultant supernatant was reserved at 4°C. Chromaffin granules in the precipitated fraction were disrupted by hypotonic shock and homogenization with 500 l of 25 mM Tris-HCl buffer, pH 7.5. Polystyrene plates with 96 wells (Immunomodule, Maxisorp F16, Nunc) were coated with 100 l of diluted samples (the supernatant or the particulate fraction). The CgA attached in the wells was assayed by the enzyme-linked immunoassay (28) using the partially purified anti-CgA antiserum and a peroxidase-conjugated goat affinity purified antibody to rabbit IgG F(ABЈ) 2 (Organo Teknika). The enzyme activity was measured by O-phenylenediamine/H 2 O 2 method (28). The amount of the supernatant CgA was expressed as percent of the total CgA (the supernatant and the particulate CgA).
Phosphorylation of the 70-kDa Protein or CgA by CaM Kinase II and Other Protein Kinases in Vitro-The standard assay system for protein kinases contained, in 25 l of a final volume, the following constituents: 50 mM HEPES, pH 7.5, 10 mM MgCl 2 , 1 mM EGTA, 0.1 or 0.5 mM [␥-32 P]ATP (3000 -5000 cpm/pmol), the 70-kDa protein (or CgA), and the indicated amount of the catalytic subunit of PKA. In addition, the assay mixtures contained 1 mM CaCl 2 and 1.5 M calmodulin for CaM kinase II and 1 mM CaCl 2 , 50 g/ml phosphatidylserine, and 5 g/ml 1,2-diolein for PKC, instead of EGTA. After incubation, 15 l of each sample was spotted on a phosphocellulose paper square (Whatman) and processed, as described (29).
Peptide Mapping-Two-dimensional tryptic peptide mapping was carried out as reported (30). Briefly, the band of CgA was excised from the SDS-PAGE gel and incubated at 37°C overnight with 100 g of TPCK-trypsin in 50 M NH 4 HCO 3 , pH 8.0. After centrifugation, the supernatant was incubated with another 100 g of TPCK-trypsin for 36 h. The supernatant was lyophilized by a "Speed Vac" concentrater and subjected to two-dimensional thin layer chromatography (TLC).
Labeling of Cells with [ 32 P]P i and Immunoprecipitation with the Anti-CgA Antibody-Cultured adrenal medullary cells (4 ϫ 10 6 /dish, Falcon, 35 mm) were labeled with [ 32 P]P i (0.2 mCi/ml) in phosphate-free Eagle's MEM medium for 6 h (18). Then, the cells were washed with 1 ml of oxygenated Krebs-Ringer/HEPES buffer, containing 125 mM NaCl, 5.6 mM KCl, 1.1 mM MgSO 4 , 2.2 mM CaCl 2 , 25 mM HEPES, pH 7.4, and 10 mM glucose. The cells were stimulated with or without high concentrations (25,56, and 75 mM) of K ϩ at 37°C for the indicated periods. In the high K ϩ medium, NaCl was reduced to maintain the isotonicity of the medium. After incubation, the cells were harvested and homogenized in 160 l of the isotonic 0.27 M sucrose buffer (see above). After centrifugation at 15,000 ϫ g for 10 min, the resultant supernatants were incubated with the anti-CgA antibody (160 g). The antigen-antibody complex was immobilized on Protein A-Sepharose gel (Pharmacia), and the phosphorylation of CgA was analyzed by SDS-PAGE. In some experiments, the phosphorylation of CgA in the supernatant and the particulate fractions of cell homogenates and the phosphorylation of CgA released from cultured cells were also analyzed by the same method as described above.
Catecholamine Secretion-Cultured cells (4 ϫ 10 6 /dish) were incubated with or without various concentrations (25,56, and 75 mM) of K ϩ at 37°C for the indicated periods. The catecholamines secreted into the medium were adsorbed to aluminum hydroxide and estimated by the ethylenediamine condensation method (31).
Assay of Dopamine ␤-Hydroxylase Activity-The activity of dopamine ␤-hydroxylase activity in the supernatant and the particulate fractions of cultured cell homogenates was measured with tyramine as the substrate (32). The ␤-hydroxylated product octopamine was separated by a Dowex 50 ϫ 8 column (0.4 ϫ 4 cm) and determined spectrophotometrically by the periodate method (32).
Other Procedures-SDS-PAGE was performed by the method of Laemmli (33). Phosphoamino acid analysis was performed, as described previously (34). The protein concentration was determined by the method of Bradford (35) with bovine serum albumin as standard.
Statistical Analysis-Data are expressed as means Ϯ standard deviation (S.D.). The vertical bar in Figs. 7 and 8 represents the standard deviation. The statistical evaluation of the data was performed with analysis of variance. If a significant F value was found, Scheffe's test for multiple comparisons was carried out to identify differences among groups. In Fig. 7, a relationship between CgA phosphorylation and catecholamine secretion was assessed by a linear regression analysis (y axis, CgA phosphorylation; x axis, catecholamine secretion). Phosphorylation of the 70-kDa Protein and CgA by CaM Kinase II-Since the 70-kDa protein is phosphorylated by CaM kinase II (17), we examined whether CgA partially purified from chromaffin granules was phosphorylated by CaM kinase II. As shown in Fig. 2, CgA was phosphorylated by CaM kinase II (Fig. 2, lane 5) and comigrated with the 70-kDa protein (Fig.  2, lane 4).

Sequence
Analysis of the 70-kDa Protein by Immunoblotting with the Anti-CgA Antibody-To further confirm the idea that the 70-kDa protein is CgA, we prepared the antibody against bovine adrenal CgA. The samples were separated by SDS-PAGE and analyzed by immunoblotting. The anti-CgA antibody recognized the 70-kDa protein (Fig. 3, lane 3) as well as CgA (Fig. 3,  lane 4). These results indicate that the 70-kDa protein is CgA.
Phosphorylation of CgA by Various Protein Kinases-We examined the phosphorylation of CgA by other protein kinases. When the CgA was incubated with CaM kinase II, PKC, and PKA, the increased phosphorylation of CgA was observed on SDS-polyacrylamide gels (Fig. 4). Based on the calculation of CgA phosphorylation by these kinases, the amounts of phosphate incorporated into CgA were 0.9, 1.2, and 0.5 mol/mol of CgA by the incubation for 1 h with CaM kinase II, PKC, and PKA, respectively.
Two-dimensional TLC Separation of Tryptic [ 32 P]Phosphopeptides of CgA Incubated with Protein Kinases-Phosphopeptides of CgA were further examined by two-dimensional TLC after extensive digestion with TPCK-trypsin (Fig. 5).
When CgA was phosphorylated by CaM kinase II, several sites were phosphorylated (Fig. 5A). One phosphorylation site by CaM kinase II (phosphopeptide a) was strongly phosphorylated by PKC (Fig. 5B) and slightly phosphorylated by PKA (Fig. 5C). One minor phosphorylation site by CaM kinase II (phosphopeptide c) was a major phosphorylation site by PKA (Fig. 5C). Another phosphorylation site by CaM kinase II (phosphopeptide b) was not phosphorylated by PKC or PKA. By contrast, one phosphorylation site by PKA (phosphopeptide d) was not phosphorylated by CaM kinase II and PKC. These differences in phosphopeptide mapping patterns by the three protein kinases were clearly shown when two samples were mixed before separation (Fig. 5, D and E).
CgA phosphorylated by three protein kinases was cut out from the SDS-PAGE gel and subjected to partial acid hydrolysis, followed by phosphoamino acid analysis (Fig. 6). PKC and PKA phosphorylated only the serine residue, whereas CaM kinase II phosphorylated both serine and threonine residues.
Effects of High Concentrations of K ϩ on CgA Phosphorylation and Catecholamine Secretion in Cultured Adrenal Medullary Cells-32 P-Labeled adrenal medullary cells were stimulated with or without high K ϩ (25, 56, and 75 mM), and the supernatants were immunoprecipitated with the anti-CgA antibody. The immunoprecipitates were analyzed by SDS-PAGE, followed by autoradiography (Fig. 7). Incubation of cells with the control medium resulted in a small amount of 32 P incorporation into CgA (Fig. 7B, lane 1). Stimulation of cells with high K ϩ (25, 56, and 75 mM) increased the phosphorylation of CgA in a concentration-dependent manner (Fig. 7B, lanes 2-4 and Fig.  7C). The maximal effect (a 2.4-fold increase) was observed with 56 mM K ϩ . This concentration-dependent increase was correlated with that of catecholamine secretion (Fig. 7, D) (y ϭ 4.47x ϩ 43.2; r ϭ 0.998, p Ͻ 0.002). Fig. 8 shows the time courses of the increases in CgA phosphorylation (Fig. 8, A and B) and catecholamine secretion (Fig. 8C), respectively, produced by 56 mM K ϩ . The phosphorylation of CgA increased rapidly at 30 s after stimulation and reached a plateau at 1 min. The time dependence of the CgA phosphorylation was also similar to that of catecholamine secretion. When the cells were incubated with 20 mM MgSO 4 , an inhibitor of voltage-dependent Ca 2ϩ channels, the 56 mM K ϩ -stimulated phosphorylation of CgA and catecholamine secretion were significantly inhibited (Table I). In some experiments, the phosphorylation of CgA in the particulate (chromaffin granule) fraction of cell homogenates was measured. After incubation of cells with [ 32 P]P i for 6 h at 37°C, the phosphorylation of CgA in the chromaffin granules was observed (data not shown). However, 56 mM K ϩ did not increase the phosphorylation of CgA in the granule fraction. In the incubation medium, 32 P-phosphorylated CgA was spontaneously released from nonstimulated cells during incubation for 5 min. Stimulation with 56 mM K ϩ enhanced the release of The Level of CgA and Dopamine ␤-Hydroxylase Activity in the Supernatant of Cell Homogenates-The activity of dopamine ␤-hydroxylase, a marker of chromaffin granule lumen, and CgA in the supernatant fraction of cell homogenates were measured to check the destruction of chromaffin granules during the homogenization. The rations of dopamine ␤-hydroxylase activity and CgA in the supernatant fraction to those in the total fraction (the supernatant and the particulate fractions) were 2.97 Ϯ 1.02 and 8.93 Ϯ 0.37%, respectively (data are means Ϯ S.D. of three separate experiments).

DISCUSSION
The 70-kDa Protein Is CgA or a Protein Closely Related to CgA-In our previous study (17), the 70-kDa protein was copurified with CaM kinase II from the soluble fraction treated with Triton X-100, followed by DEAE-cellulose, CaM-agarose, and Sephacryl S-300 column chromatography. The purification method of the protein differed from the method reported for CgA. Therefore, we considered that the 70-kDa protein was a new substrate for CaM kinase II. In the present study, we isolated CgA from bovine adrenal chromaffin granules and compared the two proteins. We demonstrated that (i) the amino acid sequences of five peptides cleft from the 70-kDa protein reveal a high homology with those of CgA (Fig. 1); (ii) CaM kinase II phosphorylates CgA as well as the 70-kDa protein (Fig. 2); and (iii) the 70-kDa protein is immunoblotted with the anti-CgA antibody (Fig. 3). From these findings, we concluded that the 70-kDa protein is CgA or a closely related protein.
CgA, an acid glycoprotein, first identified in chromaffin granules of the adrenal medulla, is the major member of the secretogranin/chromogranin class of proteins and has a widespread distribution in endocrine tissues and the brain (see reviews Refs. 37 and 38). CgA is comprised of 431 amino acid residues, corresponding to an unmodified protein of 48,000 of the molecular mass (36). The deduced molecular weight is considerably less than that we obtained (apparent molecular mass, 70 kDa) and other previous estimates based on SDS-PAGE (apparent molecular mass, 70 -80 kDa) (37). The discrepancy is well explained by Benedum et al. (36) that even the in vitro translation product of CgA has a highly abnormal mobility by SDS-PAGE (the major product of CgA has an apparent molecular mass of 72 kDa).
Phosphorylation of CgA by Several Multifunctional Protein Kinases-The cDNA encoding CgA (36) shows the presence of several potentially accessible consensus sites (39) for the action of PKA and PKC in addition to CaM kinase II. Indeed, in the present study, we directly demonstrated the phosphorylation of CgA by PKA and PKC (Fig. 4). Therefore, this is the first report to show that CgA is the substrate for these three protein kinases. By the two-dimensional peptide mapping, we identified at least four distinct 32 P-phosphopeptides derived from CgA phosphorylated by three protein kinases (Fig. 5). In these peptides, CaM kinase II and PKA phosphorylated three peptides a, b, and c and a, c, and d, respectively, whereas PKC phosphorylated only one peptide (a). The phosphoamino acid analysis demonstrated that the phosphorylated amino acid residue of CgA by three protein kinases is exclusively serine. The threonine residue was slightly phosphorylated by CaM kinase II (Fig. 6). A previous in situ study (40) indicated that CgA was phosphorylated on the serine residue and to a small extent on the threonine residue in nonstimulated adrenal medullary cells. Therefore, CgA seems to be phosphorylated by the endogenous protein kinase(s) in the cells.
In Situ Phosphorylation of CgA by Cell Depolarization in Cultured Adrenal Medullary Cells-CgA is the major secreted protein that is located in chromaffin granules of adrenal medullary cells (see reviews Refs. 37,38). Therefore, the question is  whether the phosphorylation of CgA increases by cell stimulation. In the present study, depolarization of cultured adrenal medullary cells with high K ϩ stimulated the phosphorylation of CgA in concentration-and time-dependent manners (Figs. 7 and 8). In the cell homogenizing buffer, we used various inhibitors for protein kinases, protein phosphatases, and proteases such as 10 mM EDTA, 4 mM EGTA, 25 mM NaF, 0.43 mM phenylmethylsulfonyl fluoride, 0.05 mM leupeptin, and 50 mg/ liter trypsin inhibitor. In order to attain further complete inhibition of CgA phosphorylation after cell homogenization, staurosporine (100 nM) and ␤-glycerophosphate (50 mM) were added to the homogenizing buffer as nonselective inhibitors for protein kinases (41,42) and protein phosphatases, respectively. The stimulation of CgA phosphorylation by 56 mM K ϩ was also observed even when these two inhibitors existed in the homogenizing buffer (data not shown). Therefore, it is unlikely that further phosphorylation of CgA occurs during the subsequent manipulation. Previously, Côté et al. (8) reported that acetylcholine caused an increase (about 30%) in phosphorylation of an 80-kDa protein in bovine adrenal medullary cells. They considered the 80-kDa protein as CgA, because it reacted with the antiserum against CgA, but they did not show the data.
Although the physiological significance of CgA has not been established, there are several proposals for it. They include roles (i) in the involvement in catecholamine or neuropeptides storages, (ii) in the binding of calcium and possible consequences for granule formation, (iii) as a regulatory protein after secretion, and (iv) as a precursor of peptide hormones and neuropeptides (37,38). Therefore, the phosphorylation of CgA might modify these possible functions. On the other hand, CgA immunoreactivity is most prominent in the cytosol rather than the synaptic vesicles in the brain (43). Somogyi et al. (44) reported that the distribution of CgA immunoreactivity resembled the location of the Golgi apparatus in the brain and that some neurons exhibited a homogeneous staining throughout the cytoplasm, suggesting that CgA in the brain has a cellular function independent of that in the vesicular storage. In the present study, it is important to know whether the phosphorylation of CgA stimulated by high K ϩ occurs in the cytosol or within the chromaffin granules. Stimulation of cells with 56 mM K ϩ did not increase the phosphorylation of CgA in the chromaffin granule fractions (see the "Results"). To check the disruption of chromaffin granules during homogenization, we measured the amount of CgA and the activity of dopamine ␤-hydroxylase in the supernatant fraction of the cell homogenates. The level of CgA was slightly higher than that of dopamine ␤-hydroxylase activity in the supernatant fraction, suggesting that 56 mM K ϩ stimulates the phosphorylation of CgA which is, at least in part, derived from the cytosol in the cells.
The present study demonstrated interesting evidence that the stimulatory effect of high K ϩ on the CgA phosphorylation has a good correlation with that on catecholamine secretion (Figs. 7 and 8). The phosphorylation of CgA by high K ϩ was inhibited by 20 mM MgSO 4 (Table I) which attenuates 56 mM K ϩ -evoked influx of 45 Ca 2ϩ (17). Therefore, the elevation of intracellular Ca 2ϩ stimulates the phosphorylation of CgA as well as catecholamine secretion. These findings give rise to the possibility that CgA has a role in Ca 2ϩ -dependent cellular functions in the adrenal medulla or in the brain. In our laboratory, the precise experiments such as the subcellular distribution and cellular functions of CgA, and the phosphopeptide mapping of CgA by in situ phosphorylation using a physiological secretagogue, acetylcholine, or activators of PKA and PKC, are now ongoing in adrenal medullary cells.
In conclusion, we have demonstrated that the 70-kDa protein is CgA and serves as the substrate for several multifunctional protein kinases. Furthermore, high K ϩ -evoked depolarization stimulates the phosphorylation of CgA which is associated with catecholamine secretion in cultured bovine adrenal medullary cells.