EGF inhibits muscarinic receptor-mediated calcium signaling in a human salivary cell line

Bin-Xian Zhang, Chih-Ko Yeh, Tazuko K. Hymer, Meyer D. Lifschitz, Michael S. Katz

Abstract

The effects of epidermal growth factor (EGF) on intracellular calcium ([Ca2+]i) responses to the muscarinic agonist carbachol were studied in a human salivary cell line (HSY). Carbachol (10−4 M)-stimulated [Ca2+]i mobilization was inhibited by 40% after 48-h treatment with 5 × 10−10 M EGF. EGF also reduced carbachol-induced [Ca2+]i in Ca2+-free medium and Ca2+ influx following repletion of extracellular Ca2+. Under Ca2+-free conditions, thapsigargin, an inhibitor of Ca2+ uptake to internal stores, induced similar [Ca2+]i signals in control and EGF-treated cells, indicating that internal Ca2+ stores were unaffected by EGF; however, in cells exposed to thapsigargin, Ca2+influx following Ca2+ repletion was reduced by EGF. Muscarinic receptor density, assessed by binding of the muscarinic receptor antagonistl-[benzilic-4,4′-3HCN]quinuclidinyl benzilate ([3H]QNB), was decreased by 20% after EGF treatment. Inhibition of the carbachol response by EGF was not altered by phorbol ester-induced downregulation of protein kinase C (PKC) but was enhanced upon PKC activation by a diacylglycerol analog. Phosphorylation of mitogen-activated protein kinase (MAP kinase) and inhibition of the carbachol response by EGF were both blocked by the MAP kinase pathway inhibitor PD-98059. The results suggest that EGF decreases carbachol-induced Ca2+ release from internal stores and also exerts a direct inhibitory action on Ca2+ influx. A decline in muscarinic receptor density may contribute to EGF inhibition of carbachol responsiveness. The inhibitory effect of EGF is mediated by the MAP kinase pathway and is potentiated by a distinct modulatory cascade involving activation of PKC. EGF may play a physiological role in regulating muscarinic receptor-stimulated salivary secretion.

  • carbachol
  • signal transduction
  • protein kinase C
  • mitogen-activated protein kinase
  • epidermal growth factor

fluctuations of intracellular calcium ([Ca2+]i) mediate the cellular actions of many neurotransmitters, hormones, and growth factors. In salivary gland cells, activation of G protein-coupled muscarinic-cholinergic receptors results in hydrolysis of phosphatidylinositol 4,5-bisphosphate, yielding diacylglycerol and inositol 1,4,5-trisphosphate (IP3). Diacylglycerol is an endogenous activator of protein kinase C (PKC). IP3 induces calcium release from internal calcium stores, followed by the entry of extracellular Ca2+ and a sustained elevation of [Ca2+]i (2). The mobilization of [Ca2+]i in salivary cells is directly related to the level of fluid secretion (30). Currently, factors modulating the muscarinic receptor-coupled calcium signaling pathway in salivary cells have not been well characterized.

Epidermal growth factor (EGF), which is produced in abundance by salivary glands, is a multifunctional factor known to influence the proliferation, differentiation, and physiological function of a wide variety of cell types (reviewed in Ref. 5). The actions of EGF are mediated via a receptor tyrosine kinase cascade, leading to a number of signaling events including activation of PKC and the mitogen-activated protein kinase (MAP kinase) pathway (7, 21). Previous studies implicate interaction, or “cross talk,” between EGF-induced signals and other signaling pathways. For example, EGF modulates G protein-coupled receptor activation of adenylyl cyclase in a number of tissues (10); purinergic receptor-induced calcium signaling also appears to be influenced by EGF (13). In salivary glands, EGF modifies muscarinic agonist-mediated amylase secretion (29). Although this finding suggests a role for EGF in the regulation of muscarinic receptor function in salivary cells, to our knowledge no previous investigations have demonstrated an interaction between EGF and muscarinic signaling pathways. In the present study, we have used HSY cells, a ductal cell line from human parotid (42), as a model system to determine whether the EGF receptor tyrosine kinase cascade modulates muscarinic receptor signaling in salivary cells. Our results indicate that EGF inhibits muscarinic receptor-mediated [Ca2+]imobilization in HSY cells by activation of the MAP kinase pathway.

MATERIALS AND METHODS

Materials.

Recombinant human EGF (rhEGF), insulin-like growth factor I (rhIGF-I), and transforming growth factor-α (rhTGF-α) were purchased from Promega (Madison, WI); platelet-derived growth factor (rhPDGF-AB) was from Peprotech (Rocky Hill, NJ). Fura 2-AM was from Molecular Probes (Eugene, OR). Thapsigargin, 2-(2′-amino-3′-methoxyphenyl)-oxanaphthalen-4-one (PD-98059), and digitonin were purchased from RBI (Natick, MA). Monoclonal anti-PKC antibody (which recognizes α- and β-isoforms of PKC) was from Amersham (Arlington Heights, IL). Antibodies against p44/42 MAP kinase and phospho-p44/42 MAP kinase were purchased from New England Biolabs (Beverly, MA).l-[Benzilic-4,4′-3HCN]quinuclidinyl benzilate ([3H]QNB; 30 Ci/mmol) was obtained from New England Nuclear (Boston, MA). Trypsin-EDTA and DMEM were from Life Technologies (Gaithersburg, MD). Aprotinin, leupeptin, phorbol 12-myristate 13-acetate (PMA), carbachol, atropine, 1,2-dioctanoyl-sn-glycerol, and other chemicals were purchased from Sigma (St. Louis, MO).

Cell culture.

The HSY cell line was originally established by Yanagawa et al. (42) and was kindly provided by Dr. James Turner (NIDCR/NIH, Bethesda, MD). Cells were plated at a density of about 2 × 104 cells/cm2 in 60- or 100-mm culture dishes and cultured in DMEM supplemented with 10% fetal calf serum and penicillin (100 U/ml)/streptomycin (100 μg/ml) at 37°C in a humidified 5% CO2 atmosphere incubator. Cells were grown to near confluence at 72 h and were then harvested with trypsin (0.05%)-EDTA (0.02%) for [Ca2+]i and receptor binding measurements. Unless otherwise specified, EGF and other growth factors (TGF-α, PDGF, or IGF-I) were added 48 h prior to confluence, i.e., 24 h after plating. Addition of EGF caused a small but significant increase in cell number (∼9% increase; P < 0.001) at 72 h.

Measurement of intracellular calcium.

[Ca2+]i was determined by spectrofluorometric measurements in cell suspensions. HSY cells were loaded with 1.2 μM fura 2-AM in a high-salt glucose (HNG) buffer containing 140 mM NaCl, 5 mM KCl, 1 mM CaCl2, 1 mM MgSO4, 1.2 mM KH2PO4, 10 mM glucose, and 10 mM HEPES, pH 7.4, at 37°C for 30–45 min with gentle shaking (43). Loaded cells were washed with fresh buffer to remove the extracellular dye and resuspended in HNG or Ca2+-free HNG, as required for individual experiments, at a density of 106 cells/ml. The Ca2+-free HNG medium contained 100 μM EGTA instead of 1 mM CaCl2. Fura 2-loaded cells were then exposed to the muscarinic agonist carbachol or other stimulatory agents at 37°C.

Excitation ratio of 340/380 was measured with a PTI Delta Scan spectrofluorometer (Photon Technology International, South Brunswick, NJ) using 340-nm and 380-nm wavelengths for excitation and 510 nm for emission (25). [Ca2+]i values were calculated according to the equation [Ca2+]i (nM) = K d(R − Rmin)/(Rmax − R), where R is the fluorescent ratio of 340/380; Rmin and Rmaxare minimal and maximal fluorescent ratios, respectively; and theK d is the dissociation constant (taken as 224 nM) of fura 2 for Ca2+ (9). At the end of each experiment, Rmax was determined by adding 5 μM digitonin to the cell suspension, and Rmin was achieved by chelating Ca2+ with 10 mM EGTA.

Muscarinic receptor binding assay.

Muscarinic receptor binding in HSY membrane preparations was measured by an equilibrium binding assay, using the radiolabeled muscarinic receptor antagonist [3H]QNB, essentially as described by He et al. (12). Confluent HSY cells were washed twice with PBS and once with Tris buffer (50 mM, pH 7.4) at 4°C, scraped from culture dishes in 50 mM Tris, and centrifuged at 10,000 rpm for 15 min at 4°C in a Sorvall RT7 centrifuge. The cell pellet was resuspended in Tris and homogenized on ice for two 10-s periods using a Polytron homogenizer (Brinkmann) at a setting of 5. The homogenate was centrifuged twice at 43,000 g for 15 min at 4°C, and the resultant pellet was resuspended in Tris. The membrane preparation (about 180 μg protein) was incubated with [3H]QNB in 150 μl Tris (50 mM, pH 7.4) for 2 h at 37°C. Specific binding of [3H]QNB was measured as the difference between total binding and nonspecific binding, which was defined as radioligand bound in the presence of an excess (10 μM) of the muscarinic antagonist atropine. Saturation binding curves were constructed by measuring specific binding of [3H]QNB at six concentrations of radioligand in the approximate range of 0.01–6.0 nM. Nonspecific binding was about 38% of total binding at [3H]QNB concentrations approximating the dissociation constant,K d.

Immunoblot analysis of MAP kinase.

Immunoblot analysis of MAP kinase was performed essentially as described previously (18). Briefly, confluent HSY cells were washed twice with PBS at 4°C and scraped in 1–2 ml homogenization buffer containing 20 mM Tris, 10 mM KCl, 2 mM EDTA, 0.5 mM EGTA, 2.5 μg/ml leupeptin, 10 μg/ml aprotinin, 0.2 mM phenylmethylsulfonyl fluoride (PMSF), and 1 mM sodium orthovanadate (Na3VO4), pH 7.5. The cells were homogenized in a 1.5-ml tissue grinder (Kontes Duall, from Fisher Scientific) with 30 strokes on ice. The homogenate was centrifuged (1,000 g at 4°C) for 10 min, and supernatant proteins (20 μg) were separated on 10% SDS-PAGE gels and transferred to polyvinylidene difluoride membranes (Schleicher & Schuell, Keene, NH). The membranes were immunoblotted with p44/p42 MAP kinase or phospho-p44/p42 MAP kinase primary antibody (1:1,000) and secondary antibody (1:1,000) conjugated to horseradish peroxidase (Amersham). MAP kinases were visualized by an enhanced chemiluminescence system (ECL Plus, Amersham) and quantified by a phosphorimager system (Storm 860; Molecular Dynamics, Sunnyvale, CA).

Data analysis.

Agonist-induced increases in [Ca2+]i were calculated by subtracting unstimulated (basal) values from maximal agonist responses. Data from multiple experiments are means ± SE. Statistical significance of single comparisons was determined using Student's t-test. Multiple comparisons were performed by either ANOVA followed by the Dunnett multiple comparison test or two-factor ANOVA with the Sidak multiple comparison test. Scatchard analysis of [3H]QNB saturation binding curves was used to determine muscarinic receptor density (Bmax) andK d (34).

RESULTS

Attenuation of carbachol-induced [Ca2+]i mobilization by EGF.

Figure 1 A shows [Ca2+]i responses to carbachol (10−4 M) in suspensions of EGF-treated and untreated HSY cells. In both groups of cells, carbachol stimulated a rapid increase in [Ca2+]i followed by a sustained plateau phase. However, the [Ca2+]i response to carbachol was markedly reduced by treatment of cells with EGF. EGF (5 × 10−10 M, 48 h) caused a 40% decrease of carbachol-stimulated [Ca2+]i mobilization (+EGF, 381 ± 30 nM vs. −EGF, 629 ± 42 nM;P < 0.001) (Fig. 1 B). Treatment with EGF had no apparent effect on basal levels of [Ca2+]i or the reversal of the carbachol-induced [Ca2+]i signal following addition of atropine, a muscarinic receptor antagonist (Fig.1 A). Inhibition of carbachol-mediated [Ca2+]i mobilization was dependent on EGF concentration and time of incubation with the growth factor (Fig.2). The concentration of EGF producing half-maximal inhibition of carbachol response was 1.9 ± 0.2 × 10−10 M. Inhibition of the carbachol-induced [Ca2+]i signal was observed within about 2 h of EGF treatment, whereas the time of incubation required for maximal inhibition was 4–8 h.

Fig. 1.

Effect of EGF on carbachol-stimulated intracellular Ca2+ ([Ca2+]i) mobilization. Mobilization of [Ca2+]i in response to a maximally stimulatory concentration of carbachol (10−4 M) was determined in HSY cells either treated with EGF (+EGF, 5 × 10−10 M) or untreated (−EGF) for 48 h, as described in materials and methods. A: representative experiment showing carbachol-responsive [Ca2+]i signals in EGF-treated and untreated cells. Arrows indicate time of addition of carbachol followed by the muscarinic receptor antagonist atropine (10−5 M).B: values are means ± SE for carbachol-induced increases in [Ca2+]i from 12 experiments. *P < 0.001 vs. −EGF.

Fig. 2.

Concentration dependence and time course of EGF effect on carbachol-induced [Ca2+]i mobilization.A: HSY cells were treated with increasing concentrations of EGF for 48 h. Values are means ± SE for carbachol (10−4 M)-induced increases in [Ca2+]i from 6 experiments; the [Ca2+]i response to carbachol in cells not treated with EGF is shown on the ordinate. B: HSY cells were incubated with 5 × 10−10 M EGF for 1–48 h as indicated. Means ± SE of carbachol (10−4 M)-induced [Ca2+]i signals are from 3–4 experiments.

To determine whether growth factor inhibition of carbachol-induced [Ca2+]i mobilization is specific to EGF, we compared carbachol responses in HSY cells treated with EGF and other growth factors (TGF-α, PDGF, and IGF-I) known to act through receptor tyrosine kinase pathways. TGF-α binds and activates EGF receptors (5), whereas PDGF and IGF-I activate tyrosine kinase signaling via their own distinct receptors. Like EGF, TGF-α also decreased carbachol-induced [Ca2+]imobilization in HSY cells. In contrast, PDGF and IGF-I did not alter carbachol responsiveness (Fig. 3). These data implicate a specific inhibitory effect of EGF receptor activation on carbachol-induced [Ca2+]i mobilization in HSY cells.

Fig. 3.

Effects of various growth factors on carbachol-induced [Ca2+]i mobilization. Carbachol (10−4 M)-stimulated [Ca2+]imobilization was determined in HSY cells either untreated (control) or treated with EGF (3 ng/ml, i.e., 5 × 10−10 M), transforming growth factor-α (TGF-α, 3 ng/ml), insulin-like growth factor I (IGF-I, 50 ng/ml), or platelet-derived growth factor (PDGF, 10 ng/ml) for 48 h, as described in materials and methods. Growth factors were tested at concentrations commonly found to elicit a variety of cellular responses. Values are means ± SE for carbachol-induced increases in [Ca2+]i from 9 experiments. *P < 0.02 vs. control.

Effects of EGF on carbachol-induced [Ca2+]i signaling in the absence of extracellular Ca2+ and during Ca2+ repletion.

In nonexcitable cells, G protein-coupled receptor-activated [Ca2+]i mobilization consists of initial Ca2+ release from IP3-sensitive internal stores, followed by influx of extracellular Ca2+ across the plasma membrane. The influx of Ca2+ is thought to be activated by depletion of the internal Ca2+ stores via a process of “capacitative calcium entry” (31). In a Ca2+-free medium, Ca2+ influx does not occur and [Ca2+]i signals represent mainly release of internal Ca2+ stores. To investigate the effect of EGF on carbachol-induced Ca2+ release from internal stores, we examined [Ca2+]i responsiveness of HSY cells to carbachol under Ca2+-free conditions. Figure4 shows that in the absence of extracellular Ca2+, EGF reduced the carbachol-induced [Ca2+]i signal by about one-half (+EGF, 146 ± 20 nM vs. −EGF, 286 ± 47 nM; P < 0.05). Subsequent repletion of extracellular Ca2+ (1 mM) after carbachol stimulation induced a rapid rise of [Ca2+]i, presumably as a result of Ca2+ influx (Fig. 4); this apparent Ca2+ influx was also significantly reduced by EGF (+EGF, 586 ± 111 nM vs. −EGF, 885 ± 106 nM; P < 0.05) (Fig. 4). These findings suggest that EGF inhibition of carbachol-induced [Ca2+]i mobilization in HSY cells (Fig. 1) involves reduction in both Ca2+ release and Ca2+ influx.

Fig. 4.

Effects of EGF on carbachol-induced [Ca2+]i signaling in the absence of extracellular Ca2+ and during Ca2+ repletion. HSY cells were either treated with EGF (+EGF, 5 × 10−10 M) or untreated (−EGF) for 48 h and then stimulated with carbachol (10−4 M) in Ca2+-free high-salt glucose (HNG) medium as described inmaterials and methods. When the carbachol-induced [Ca2+]i signal returned to basal level, 1 mM Ca2+ was added to the medium. A: representative experiment showing [Ca2+]i signals in response to carbachol and repletion of Ca2+ in EGF-treated and untreated cells. B: values are means ± SE for carbachol-induced increases of [Ca2+]i in response to carbachol and repletion of Ca2+ in EGF-treated and untreated cells. *P < 0.05 vs. −EGF;n = 3 experiments.

To explore further the action(s) of EGF on Ca2+ release and Ca2+ influx, we compared the [Ca2+]i responses of EGF-treated and untreated HSY cells to thapsigargin, a specific inhibitor of endoplasmic reticulum Ca2+-ATPase (36). Thapsigargin inhibits Ca2+ uptake to the internal stores, resulting in depletion of the internal pool and elevation of [Ca2+]i. In the absence of extracellular Ca2+, thapsigargin-induced elevation of [Ca2+]i provides an indirect measure of Ca2+ in the internal stores. Under Ca2+-free conditions, thapsigargin produced equivalent increases in [Ca2+]i in both EGF-treated and untreated HSY cells (Fig. 5), suggesting that the growth factor does not alter the internal pool. Interestingly, whereas subsequent addition of Ca2+ to the medium increased [Ca2+]i, this response to extracellular Ca2+ was significantly diminished in EGF-treated cells (+EGF, 702 ± 55 nM vs. −EGF, 908 ± 42 nM;P < 0.03) (Fig. 5). The observation that EGF decreases the [Ca2+]i response to extracellular Ca2+ repletion without affecting the internal stores suggests a direct inhibitory effect of EGF on Ca2+ influx.

Fig. 5.

Effects of EGF on thapsigargin-induced [Ca2+]i signaling in the absence of extracellular Ca2+ and during Ca2+ repletion. HSY cells were either treated with EGF (+EGF, 5 × 10−10 M) or untreated (−EGF) for 48 h and then stimulated with thapsigargin (5 × 10−7 M) in Ca2+-free HNG medium as described in materials and methods. When the thapsigargin-induced [Ca2+]i signal returned to basal level, 1 mM Ca2+ was added to the medium. A: representative experiment showing [Ca2+]i signals in response to thapsigargin and repletion of Ca2+ in EGF-treated and untreated cells. B: values are means ± SE for increases of [Ca2+]i in response to thapsigargin and repletion of Ca2+ in EGF-treated and untreated cells. *P < 0.05 vs. −EGF;n = 7 experiments.

Influence of EGF on muscarinic receptor binding.

Because EGF had no apparent effect on internal Ca2+ stores, EGF inhibition of carbachol-induced Ca2+ release is likely to occur at one or more proximal steps of the muscarinic receptor-activated signaling pathway. In the present study, we examined the effect of EGF on muscarinic receptors assessed by binding of the muscarinic receptor antagonist [3H]QNB to HSY cell membrane preparations. Scatchard analysis of [3H]QNB saturation binding curves revealed a single class of high-affinity binding sites. EGF caused a 20% decrease of receptor density (Bmax) without affecting receptor binding affinity (K d) (Fig. 6). Thus EGF inhibition of carbachol-induced [Ca2+]i mobilization could be mediated at least partly by a reduction in muscarinic receptor density.

Fig. 6.

Effect of EGF on muscarinic receptor binding. Cell membranes were obtained from HSY cells either treated with EGF (+EGF, 5 × 10−10 M) or untreated (−EGF) for 48h. Membrane receptor binding was measured using the muscarinic receptor antagonistl-[benzilic-4,4′-3HCN]quinuclidinyl benzilate ([3H]QNB) as the radioligand, as described inmaterials and methods. Saturation binding curves with or without EGF from a representative experiment are shown. Receptor density (Bmax) and dissociation constant (K d) were determined by Scatchard analysis of saturation binding curves (see inset). Means ± SE of Bmax and K d from 7 experiments are presented at bottom. *P < 0.007 vs. −EGF.

EGF-responsive signals involved in modulating carbachol-induced [Ca2+]i mobilization.

The EGF receptor tyrosine kinase is coupled to activation of phospholipase C-γ (PLC-γ), which in turn generates IP3and the endogenous PKC activator diacylglycerol (23). Activation of PKC has been linked to modulation of muscarinic agonist-sensitive [Ca2+]i responses and also muscarinic receptor expression (3, 11, 26, 27, 32, 33). We therefore examined the role of PKC in EGF inhibition of carbachol-responsive [Ca2+]i mobilization. PKC in HSY cells was downregulated by prolonged (18–20 h) exposure to the phorbol ester PMA (10−5 M) prior to EGF treatment (18). Complete downregulation of PKC α- and β-isoforms in PMA-treated cells was confirmed by immunoblot analysis as described previously (Ref. 18, data not shown). EGF reduced the carbachol response by one-third with or without PMA pretreatment (Fig.7). In contrast, pretreatment with 1,2-dioctanoyl-sn-glycerol (10−5 M), a cell-permeable diacylglycerol analog that causes initial stimulation without subsequent downregulation of PKC (Ref. 18, data not shown), markedly enhanced the inhibitory effect of EGF. As shown in Fig. 7, EGF completely abolished the carbachol-induced [Ca2+]i signal in cells pretreated with the diacylglycerol analog. Taken together, these data suggest that activation of PKC, although not required for the inhibitory effect of EGF, does potentiate EGF inhibition of the carbachol response.

Fig. 7.

EGF inhibition of carbachol-induced [Ca2+]i mobilization: effects of pretreatment with phorbol 12-myristate 13-acetate (PMA) and 1,2-dioctanoyl-sn-glycerol (DG). Carbachol (10−4 M)-stimulated [Ca2+]imobilization was determined in HSY cells incubated for 18–20 h with PMA (10−5 M) or DG (10−5 M), followed by addition of EGF (+EGF, 5 × 10−10 M) or vehicle (−EGF) for 6–8 h in the continued presence of PMA or DG. Values are means ± SE for carbachol-induced increases in [Ca2+]i from 9 experiments. *P < 0.001, −EGF vs. +EGF (control) and PMA vs. PMA+EGF. **P < 0.0002, DG vs. DG+EGF. In the absence of EGF, mean values from PMA- and DG-treated cells were not significantly different from control (P > 0.05).

EGF stimulation of IP3 formation causes a rapid increase of [Ca2+]i in a number of tissues (14). In HSY cells we have found that EGF causes an acute elevation of [Ca2+]i (i.e., within seconds of EGF exposure) in addition to the decrease in carbachol-responsive [Ca2+]i mobilization occurring after more prolonged exposure to EGF. However, the acute effect of EGF was observed only at much higher concentrations than those required to inhibit the carbachol response (Fig. 8). At a concentration of 5 × 10−10 M, EGF produced near maximal inhibition of carbachol-induced [Ca2+]i mobilization without any detectable acute change in [Ca2+]i (cf. Figs.2 A and 8). Therefore, although EGF causes a rapid increase of [Ca2+]i in HSY cells, this acute [Ca2+]i signal is not likely to mediate inhibition of the carbachol response observed after prolonged EGF treatment.

Fig. 8.

EGF-induced [Ca2+]imobilization. Previously untreated HSY cells were stimulated with EGF at concentrations of 0.5–16.7 × 10−9 M as indicated in each trace. [Ca2+]i was measured as described in materials and methods. The arrow indicates time of addition of EGF. Each trace is representative of 3 experiments.

Another important signaling cascade activated by the EGF receptor tyrosine kinase is the MAP kinase pathway, which plays an essential role in mediating cellular responses to EGF and a variety of other growth factors (21). In HSY cells, phosphorylation of MAP kinase by EGF was determined using immunoblot analysis. EGF induced the phosphorylation of the p44 isoform of MAP kinase without affecting the protein level. Treatment of HSY cells with PD-98059 (5 × 10−5 M), a synthetic inhibitor of the MAP kinase pathway (1), blocked EGF-induced phosphorylation of p44 MAP kinase (Fig. 9 A). As shown in Fig.9 B, inhibition of carbachol-stimulated [Ca2+]i mobilization by EGF was prevented by PD-98059. These results suggest that the inhibitory effect of EGF on the carbachol response in HSY cells is mediated by activation of the MAP kinase pathway.

Fig. 9.

Role of mitogen-activated protein kinase (MAP kinase) pathway in EGF inhibition of carbachol-induced [Ca2+]i mobilization. A: phosphorylation of MAP kinase by EGF. Confluent HSY cells were either untreated or treated with EGF (5 × 10−10 M) for 10 min following 30-min preincubation with or without PD-98059 (5 × 10−5 M). Phospho-p44 MAP kinase and p44 MAP kinase were measured by immunoblot analysis as described in materials and methods. Identity of the p44 isoform of MAP kinase was determined by molecular weight markers. The p42 isoform of MAP kinase was not detectable in HSY cells, presumably because of low levels of expression; both p44 and p42 isoforms were detected in the C6 rat glioma cell line (not shown). EGF increased phosphorylation of p44 MAP kinase by 2,300 ± 80% (P < 0.0001,n = 5 experiments); PD-98059 inhibited EGF-induced phosphorylation by 53 ± 5% (P < 0.0001).B: effect of PD-98059 on EGF inhibition of carbachol-induced [Ca2+]i mobilization. Carbachol (10−4 M)-stimulated [Ca2+]imobilization was determined in HSY cells either untreated (−PD-98059) or treated with PD-98059 (+PD-98059, 5 × 10−5 M) for 30 min, followed by continuous culture ± EGF (5 × 10−10 M) for 48 h. Values are means ± SE for carbachol-induced increases in [Ca2+]i from 8 experiments. *P < 0.02 vs. −EGF.

DISCUSSION

The results of the present study demonstrate a novel role for EGF in the regulation of muscarinic receptor-coupled [Ca2+]i signaling in a human salivary cell line (HSY). We found that EGF inhibited carbachol-responsive [Ca2+]i mobilization in a concentration- and time-dependent manner (Figs. 1 and 2). The [Ca2+]i response to carbachol was decreased by both EGF and TGF-α, which activate the EGF receptor, but not by growth factors (IGF-I and PDGF) acting on other receptor tyrosine kinases (Fig. 3).

Inhibition of the carbachol response by EGF was observed in the absence of extracellular Ca2+ (Fig. 4), suggesting a modulatory action of the growth factor on release of Ca2+ from carbachol-sensitive internal stores. The internal Ca2+pool, as reflected by the [Ca2+]i response to thapsigargin in cells incubated without external Ca2+, was intact after EGF treatment (Fig. 5). EGF has also been reported to have no effect on thapsigargin-sensitive Ca2+stores in a rat pheochromocytoma cell line (PC12) (13). Therefore, inhibition of Ca2+ release by EGF is likely to be exerted on one or more elements of the signaling pathway coupling activation of the muscarinic receptor to mobilization of the internal Ca2+ store. Measurement of [3H]QNB binding to HSY cell membranes revealed a small but significant decrease in muscarinic receptor density (Bmax) after EGF treatment (Fig. 6). The loss of muscarinic receptor binding sites could be the result of alterations in receptor expression, posttranslational modifications, or desensitization mechanisms. Our data do not exclude actions of EGF on other elements of muscarinic receptor signaling, including receptor/G protein coupling, PLC-mediated generation of IP3, or activation of IP3-sensitive Ca2+ release channels (IP3 receptors) in the endoplasmic reticulum. In a variety of cell types, IP3 receptor activity is regulated by tyrosine and serine/threonine phosphorylation (8, 15, 35,44). Whether EGF exerts direct or indirect effects on the phosphorylation characteristics of IP3 receptors in HSY cells is not known.

Depletion of internal Ca2+ stores activates influx of extracellular Ca2+ across the plasma membrane in a process known as capacitative calcium entry (31). This process is thought to be mediated by a plasma membrane Ca2+ channel termed the “Ca2+ release-activated Ca2+channel” (CRAC channel) or “store-operated Ca2+channel” (SOC channel) (45). In HSY cells exposed to carbachol in the absence of extracellular Ca2+, we found that EGF decreased Ca2+ influx following Ca2+repletion (Fig. 4). Inhibition of Ca2+ influx under these conditions could reflect either reduced Ca2+ release, leading to less depletion of internal stores and a secondary decline in capacitative Ca2+ entry, or a direct effect of EGF on Ca2+ influx. Growth factor inhibition of Ca2+influx was also observed in cells treated with thapsigargin, even though EGF had no effect on thapsigargin-induced depletion of internal stores (Fig. 5). Our findings therefore suggest that EGF not only inhibits muscarinic receptor-coupled Ca2+ release but also modulates Ca2+ influx independently of internal store depletion.

An inhibitory effect of EGF on Ca2+ influx could be exerted directly on the Ca2+ influx channel, although growth factor modification of as yet undefined signals coupling internal store depletion to Ca2+ influx cannot be excluded. The CRAC channel protein(s) has yet to be identified but may be a mammalian homolog of one or more products of the trp (transient receptor potential) gene family in Drosophilaphotoreceptors. Six trp homologs (trp1–trp6) have been identified in a variety of mammalian tissues (45), and a trp1 homologous sequence has been detected in rat submandibular gland RNA (40). Calcium influx mediated by a mouse Trp6 protein appears to be stimulated by G protein-coupled receptors independently of internal Ca2+ store depletion (4). Interestingly, a muscarinic receptor-activated Ca2+ influx pathway resembling Ca2+ entry through Trp6 has been described in a human submandibular gland cell line (HSG) (16). Whether one or more Trp homologs mediate Ca2+ influx in HSY cells has not yet been determined, and a role for EGF in regulating Trp channels in salivary or other cell types remains speculative.

Agonists binding to the EGF receptor modulate cellular functions by a complex series of signaling events, including activation of the MAP kinase and PKC pathways (7, 21). In the present study EGF was found to induce phosphorylation of MAP kinase in HSY cells. Phosphorylation of MAP kinase and inhibition of carbachol-induced [Ca2+]i mobilization by EGF were both blocked by PD-98059, a selective inhibitor of the MAP kinase pathway (Fig. 9). In contrast, downregulation of PKC by prolonged exposure of HSY cells to phorbol ester had no effect on EGF inhibition of carbachol-sensitive [Ca2+]i mobilization (Fig. 7). Thus activation of the MAP kinase pathway, but not PKC, appears to play an essential role in the inhibitory effect of EGF on muscarinic receptor-coupled [Ca2+]i signaling.

Although our data do not support the involvement of PKC activation as a signaling intermediate in modulation of the muscarinic response by EGF, the level of PKC activity may influence carbachol-induced [Ca2+]i signaling in HSY cells. Our results revealed an increase in the [Ca2+]i response to carbachol following downregulation of PKC by phorbol ester, as well as a reduced [Ca2+]i response after PKC activation by a diacylglycerol analog (Fig. 7). These changes in [Ca2+]i mobilization, although relatively small in magnitude and not statistically significant, suggest that PKC may exert an inhibitory effect on [Ca2+]i signaling independent of the modulatory action of EGF. Activation of PKC in a number of cell types, including human salivary cells (HSG and HSY), is known to exert differential effects, both stimulatory and inhibitory, on Ca2+ release responsive to muscarinic and other G protein-coupled receptors and also on capacitative Ca2+ entry (3, 11, 19, 26, 27, 32,41). In addition, downregulation of several muscarinic receptor subtypes, occurring via phosphorylation-dependent receptor internalization or altered receptor gene transcription, is enhanced by PKC stimulation in various cell lines (33). A particularly striking result of our own studies was the finding that the inhibitory effect of EGF on carbachol-stimulated [Ca2+]imobilization was greatly accentuated in HSY cells pretreated with the diacylglycerol analog (Fig. 7). Whereas EGF alone inhibited the carbachol-responsive [Ca2+]i signal by one-third, the carbachol response was completely blocked by the combination of growth factor and diacylglycerol analog (Fig. 7). This marked synergy between the regulatory actions of an EGF receptor-activated, MAP kinase-dependent pathway and a distinct PKC activation cascade has not, to our knowledge, been described previously. Studies are currently underway in our laboratory to determine the cross talk mechanisms by which PKC potentiates EGF inhibition of carbachol-induced [Ca2+]isignaling in HSY cells.

In this study EGF was found to exert two time-dependent actions on [Ca2+]i levels in HSY cells. Exposure of cells to EGF for prolonged periods (i.e., ≥ 2 h) reduced carbachol-responsive [Ca2+]i mobilization (Fig. 2 B), whereas a rapid increase in [Ca2+]i occurred within seconds of EGF exposure in the absence of other [Ca2+]imobilizing agonists (Fig. 8). Although both prolonged and rapid actions of EGF were concentration dependent, inhibition of carbachol responsiveness was observed at much lower concentrations of EGF than those required to cause rapid [Ca2+]imobilization (cf. Figs. 2 A and 7). The two effects of EGF were further dissociated by treatment of cells with the MAP kinase inhibitor PD-98059, which totally blocked the modulatory action of EGF on carbachol-sensitive [Ca2+]i signaling (Fig. 9) but did not decrease the acute [Ca2+]i response to EGF (data not shown). Growth factor inhibition of carbachol-responsive Ca2+release also could not be attributed to depletion of internal Ca2+ stores following the rapid [Ca2+]i mobilizing action of EGF, because the internal stores were unaltered after prolonged EGF treatment. These observations indicate that the rapid and prolonged [Ca2+]i responses to EGF represent distinct and independent actions of the growth factor.

Our results suggest a potential role for EGF in regulation of salivary gland function. EGF modulation of secretory response to muscarinic receptor activation has been reported in salivary gland and other gastrointestinal epithelial tissues. In a previous study examining the effect of EGF on rat parotid gland secretory function, EGF treatment in vivo caused an increase in muscarinic agonist-induced amylase secretion, although no change in salivary flow rate was detected (29). Other investigations have demonstrated acute inhibitory effects of EGF in vitro on carbachol-stimulated secretory functions in rabbit and canine gastric parietal cells and in human T84 colonic epithelial cells (6, 20, 38, 39). Whereas EGF inhibition of carbachol responsiveness in parietal cells may be mediated via activation of PKC, the inhibitory actions of EGF in both parietal and colonic cells do not appear to involve modulation of [Ca2+]i transients (20, 38, 39). In contrast to the acute inhibitory effect of EGF on carbachol responsiveness, prolonged EGF treatment of rabbit parietal cells enhances carbachol-stimulated secretory activity. Like the modulatory action of EGF described in the present study of HSY cells, the prolonged effect of EGF in rabbit parietal cells appears to be mediated by the MAP kinase signaling pathway (6). EGF has also been found to produce acute inhibition, as well as prolonged enhancement, of G protein-coupled receptor-mediated secretory function in pancreatic acinar cells (28). Both inhibition and stimulation of G protein-coupled receptor-activated [Ca2+]imobilization have been reported in other cell types after prolonged EGF treatment (13, 22).

Insofar as EGF is produced and secreted by salivary cells, our findings also raise the possibility that EGF may regulate salivary secretory function through autocrine or paracrine mechanisms. An analogous regulatory role for EGF has been suggested in studies of kidney tubule cells, which produce and secrete EGF and at the same time exhibit a variety of differentiated functions subject to modulation by EGF. Although under baseline conditions renal EGF may be localized to the luminal membrane of tubular epithelial cells, the actions of EGF on sensitive tubule segments are exerted on receptors at the antiluminal, or basolateral, surface. Interestingly, autocrine or paracrine actions of renal EGF may, under some conditions, be facilitated by redistribution of intracellular EGF to the basolateral membrane (24). Whether the in vivo actions of EGF in salivary cells are subject to similar spatial considerations as exist in kidney epithelial cells is not known. Of note in this regard are recent studies demonstrating EGF-independent transactivation of EGF receptors by muscarinic agonists in human embryonic kidney 293 cells and T84 colonic epithelial cells (17, 37). In T84 cells, EGF receptor activation has been proposed to function as an inhibitory signal by which a muscarinic agonist negatively regulates its own secretory response (17). These findings, together with our own observations, suggest that HSY cells may be a useful model to explore the regulation of salivary secretion by spatial and functional interactions between EGF and muscarinic receptor signaling pathways.

In summary, the present study demonstrates that EGF inhibits carbachol-activated [Ca2+]i mobilization in the HSY salivary cell line. EGF decreases muscarinic agonist-induced release of Ca2+ from internal stores and also exerts a direct inhibitory action on influx of Ca2+ across the plasma membrane. A decline in muscarinic receptor density may contribute to EGF inhibition of carbachol responsiveness. The inhibitory effect of EGF is mediated by the MAP kinase signaling pathway and is potentiated by a distinct modulatory cascade involving activation of PKC. These data provide the basis for further investigation of possible mechanisms by which cross talk between EGF and muscarinic receptor signaling pathways may function in the physiological regulation of salivary secretion.

Acknowledgments

We thank Jeffrey L. Harrison and Irene Herrera for technical contribution to the initial stage of this study, and Michael F. Luther, for help with data analysis.

Footnotes

  • This work was supported by medical research funds from the Department of Veterans Affairs (to M. S. Katz) and by National Institutes of Health Grant DE-10756 (to C.-K. Yeh).

  • Address for reprint requests and other correspondence: M. S. Katz, Geriatric Research, Education and Clinical Center (182), South Texas Veterans Health Care System, Audie L. Murphy Division, 7400 Merton Minter Blvd., San Antonio, TX 78284 (E-mail katz{at}uthscsa.edu).

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

REFERENCES

View Abstract