Roles of Protein Kinase C, Ca²⁺, Pyk2, and c-Src in Agonist Activation of Rat Lacrimal Gland p42/p44 MAPK

Abstract

purpose. Although p42/p44 mitogen-activated protein kinase (MAPK) negatively modulates protein secretion stimulated by cholinergic and α_1D-adrenergic agonists, it does not play a role in epidermal growth factor (EGF)–stimulated protein secretion. Therefore, this study was conducted to determine the roles that protein kinase C (PKC), intracellular Ca²⁺ ([Ca²⁺]_i), and nonreceptor tyrosine kinases Pyk2 and Src play in the activation of agonist- and EGF-stimulated MAPK activation.

methods. Lacrimal gland acini were isolated by collagenase digestion and incubated with phorbol 12-myristate 13-acetate (PMA) to activate PKC or ionomycin, a Ca²⁺ ionophore. Acini were preincubated with the PKC inhibitors calphostin C or Ro-31-8220, EGTA to chelate Ca²⁺, or the c-Src inhibitor PP1 before stimulation with the cholinergic agonist carbachol, the α_1D-adrenergic agonist phenylephrine, or EGF. Activated MAPK, Pyk2, and c-Src amounts were measured by Western blot analysis.

results. PMA and ionomycin significantly increased the activation of MAPK in a time- and concentration-dependent manner. Inhibition of PKC partially inhibited carbachol-stimulated MAPK activation while completely inhibiting phenylephrine- and EGF-stimulated MAPK activation. Chelation of Ca²⁺ also partially inhibited carbachol-stimulated MAPK with no effect on phenylephrine- and EGF-stimulated MAPK activation. Carbachol increased the phosphorylation of Pyk2 on tyrosine 402 and c-src on tyrosine 416 in a time-dependent manner. The c-src inhibitor PP1 inhibited carbachol-stimulated phosphorylation of Pyk2.

conclusions. It was concluded that cholinergic agonists use Ca²⁺ and PKC to phosphorylate Pyk2 and c-Src, which subsequently stimulate MAPK activity. In contrast, α_1D-adrenergic agonists and EGF do not use Pyk2 and Src but do use PKC to activate MAPK.

The lacrimal gland is the major contributor to the aqueous layer of the tear film, secreting water, electrolytes, and protein onto the ocular surface. Regulation of secretion is tightly controlled, and any disruption in quantity or quality can have deleterious effects on the ocular surface. This control is dependent on the stimulation of nerves in the cornea activating an afferent pathway to stimulate efferent pathways in the lacrimal gland, causing the gland to secrete. ¹  

Acinar cells are the main cell type in the lacrimal gland; they constitute approximately 80% of the gland and are the main cell type involved in secretion. ² These cells are highly polarized and have tight junctions at the luminal membranes, creating distinct basolateral and apical membranes. Neurotransmitters released from nerves bind to the appropriate receptor on the basolateral membrane of the acinar cell, initiating the signal that leads to the secretion of water, electrolytes, and protein across the apical membrane. 

Norepinephrine released from sympathetic nerves and acetylcholine released from parasympathetic nerves are major stimuli of lacrimal gland protein secretion. Norepinephrine binds to α_1D-adrenergic receptors to activate endothelial nitric oxide (NO) synthase to generate NO. ³ NO increases cGMP through the activation of guanylate cyclase, leading to protein secretion. This agonist also activates protein kinase C (PKC)-ε ⁴ to stimulate secretion. Acetylcholine binds to M₃ muscarinic receptors to stimulate protein secretion by activating Gαq/11 and phospholipase C (PLC)-β. ^{5 6} PLC-β hydrolyzes phosphatidylinositol bisphosphate into diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP₃). DAG activates PKC-α, PKC-δ, and PKC-ε in the lacrimal gland, whereas IP₃ releases Ca²⁺ from intracellular stores. ^{4 7} Ca²⁺, either alone or in conjunction with calmodulin, leads to protein secretion. 

Growth factors of the epidermal growth factor (EGF) family are also known to stimulate lacrimal gland protein secretion. ⁸ Although it is known that many growth factors in this family increase intracellular Ca²⁺ ([Ca²⁺]_i), only the signal transduction pathway used by EGF has been studied extensively. ^{8 9} EGF stimulates secretion through the recruitment of PLC-γ, which in turn activates PKC and increases [Ca²⁺]_i. ⁹  

In addition to the pathways described, G-protein–linked agonists and EGF also activate p42/p44 mitogen-activated protein kinase (MAPK), albeit by different mechanisms. As do the other G-protein–coupled receptors, the α_1D-adrenergic receptors in the lacrimal gland transactivate the EGF receptor (EGFR) through activation of a matrix metalloproteinase and ectodomain shedding of EGF (Dartt DA, unpublished observations, April 2003). The activated EGFR phosphorylates Shc, leading to the recruitment of Grb2. This activates a cascade of protein kinases, namely Ras, Raf, MEK, and ultimately MAPK. 

In contrast to α_1D-adrenergic receptors, muscarinic receptors in the lacrimal gland do not transactivate the EGFR to activate MAPK. Rather, these receptors activate the nonreceptor tyrosine kinase proline-rich tyrosine kinase (Pyk)2 and c-Src. Pyk2, also known as RAFTK, is a Ca²⁺-dependent member of the focal adhesion kinase (FAK) family. ^{10 11} Cholinergic agonist stimulation of Pyk2 and c-Src leads to the activation of MAPK in lacrimal gland because the inhibition of c-Src with a specific inhibitor decreases cholinergic agonist-stimulated MAPK activation while it increases protein secretion. ¹²  

Long-term activation of MAPK leads to cell growth and proliferation. ¹³ Short-term activation has been shown to be involved in secretion. ^{12 14} In the lacrimal gland, the activation of MAPK negatively modulates cholinergic- and α_1D-adrenergic agonist–stimulated protein secretion and thus may control agonist-stimulated protein secretion. Interestingly, EGF-stimulated MAPK does not affect EGF-stimulated protein secretion. ⁹  

In the present study, we investigated the role of PKC and Ca²⁺ on agonist- and growth factor–stimulated MAPK and the role of Pyk2 and c-Src in the activation of MAPK in the lacrimal gland. We demonstrated that MAPK is activated by the stimulation of PKC and is dependent on intracellular Ca²⁺ and that cholinergic agonists use Ca²⁺, PKC, Pyk2, and c-Src to activate MAPK. In contrast, α_1D-adrenergic agonists and EGF use PKC to activate MAPK. 

Materials and Methods

Materials used were as follows: antibodies against phosphorylated (active) p42/p44 MAPK (Tyr204), total MAPK, EGFR, and total Src (Santa Cruz Biotechnology, Santa Cruz, CA); antibodies against phosphorylated Pyk2 (Tyr402), total Pyk2, and phosphorylated Src (Tyr416) (Cell Signaling, Beverly, MA); antibodies against phosphorylated Pyk2 on Tyr580 and Tyr881 (Biosource International, Camarillo, CA); EGF (Peprotech Inc., Rocky Hill, NJ); anti–tyrosine antibody (4G10; Upstate Biotechnology, Chicago, IL); ionomycin (Alexis Biochemicals, San Diego, CA); and collagenase type III (Worthington Biochemicals, Freehold, NJ). All other reagents (Sigma Chemical, St. Louis, MO) were the highest quality available. 

Preparation of Rat Lacrimal Gland Acini

All animals were handled in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Acini were prepared using a modification of a method developed by Herzog et al. ¹⁵ In brief, lacrimal glands were trimmed, fragmented, and incubated with collagenase CLSIII (100 U/mL) in Krebs-Ringer bicarbonate (KRB) buffer (119 mM NaCl, 4.8 mM KCl, 1 mM CaCl₂, 1.2 mM MgSO₄, 1.2 mM KH₂PO₄, and 25 mM NaHco ₃) supplemented with 10 mM HEPES, 5.5 mM glucose (KRB-HEPES), and 0.5% BSA, pH 7.4, at 37°C. Lobules were subjected to gentle pipetting through tips of decreasing diameter, filtered through nylon mesh (150-μm pore size), and briefly centrifuged. The pellet was washed twice by centrifugation (50g, 2 minutes) through a 4% BSA solution made in KRB-HEPES buffer. Dispersed acini were allowed to recover for 30 minutes in fresh KRB-HEPES buffer containing 0.5% BSA. 

Measurement of MAPK, Pyk2, and c-Src Activity

The activation of p42/44 MAPK, Pyk2, and c-Src was examined with Western blot techniques. Acini were incubated for 5 minutes with the cholinergic, muscarinic agonist carbachol (10^–4 M), the α_1D-adrenergic agonist phenylephrine (10^–4 M), or EGF (10^–7 M). Inhibitors were added 10 minutes before carbachol was added. To terminate incubation, ice-cold KRB buffer was added and centrifuged briefly. The pellet was homogenized in RIPA buffer (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% sodium deoxycholate, 1% Triton X-100, 0.1% SDS, 1 mM EDTA, 100 μg/mL phenylmethylsulfonyl fluoride, 30 μL/mL aprotinin, 1 mM Na₃VO₃), sonicated, and centrifuged at 20,000g for 30 minutes. Proteins in the supernatant were separated by SDS-PAGE on a 10% gel and transferred onto nitrocellulose membranes that had been blocked overnight at 4°C in 5% nonfat dried milk in buffer containing 10 mM Tris-HCl, pH 8.0, 150 mM NaCl, and 0.05% Tween-20 (TBST). Blots were then probed with antibodies directed against the nonphosphorylated form of the enzyme (total) or the phosphorylated form of the enzyme (activated), followed by horseradish peroxidase (HRP)–conjugated secondary antibody. Immunoreactive bands were digitally scanned and analyzed using ImageJ (National Institutes of Health). The amount of phosphorylated enzyme in each sample was standardized to the amount of either total enzyme or total MAPK. 

Immunoprecipitation Experiments

Lacrimal gland acini were incubated for 5 minutes with phorbol 12-myristate 13-acetate (PMA; 10^–6 M) or EGF (10^–7 M) as a positive control. To terminate incubation, the acini were centrifuged, the supernatant was discarded, and ice-cold RIPA buffer was added. The homogenate was centrifuged at 20,000g for 30 minutes, and the supernatant was incubated overnight at 4°C on a rocker platform in the presence of an antibody against EGFR. Protein A-agarose was added for 2 hours and incubated at 4°C. The immunoprecipitate was collected by brief centrifugation, washed three times with RIPA buffer, and resuspended in Laemmli sample buffer. Proteins were separated by SDS-PAGE and transferred to nitrocellulose membranes. Western blot analysis was performed using an antibody against phosphorylated tyrosine or the immunoprecipitating antibody EGFR. 

Data Presentation and Statistical Analysis

Data are expressed as fold increase above basal value, which was standardized to 1.0. Results are expressed as mean ± SEM. Data were analyzed by Student t test. P < 0.05 was considered statistically significant. 

Results

Effect of Time and Concentration of Phorbol Esters on p42/p44 MAPK Activation

We previously showed that the cholinergic agonist carbachol activates PKC-α, PKC-δ, and PKC-ε, that the α_1D-adrenergic agonist phenylephrine activates PKC-ε, and that EGF activates PKC-α and PKC-δ to stimulate lacrimal gland protein secretion. ⁴ Because PKC is known to activate MAPK, ^{16 17 18} we sought to determine whether PKC plays a role in the activation of p42/p44 MAPK in the lacrimal gland. Lacrimal gland acini were incubated with the phorbol ester PMA, which activates classic and novel isoforms of PKC, for varying times and concentrations. The amount of phosphorylated MAPK was then determined through Western blot analysis. PMA (10^–6 M) significantly increased the activation of MAPK in a time-dependent manner, with a maximum stimulation at 5 minutes of 3.5 ± 0.8-fold above basal value (Fig. 1A) . This increase persisted for at least 30 minutes. 

Lacrimal gland acini were also incubated with increasing concentrations of PMA for 5 minutes, and the amount of phosphorylated MAPK was determined. As shown in Figure 1B , PMA (10^–6 M) significantly increased MAPK activation 2.3 ± 0.5-fold above basal value. 

To determine whether PMA activates MAPK through activation of the EGFR, lacrimal gland acini were incubated with PMA (10^–6 M) or EGF (10^–7 M) as a positive control for 5 minutes Proteins were subjected to immunoprecipitation experiments using an anti–EGFR antibody. Immunoprecipitated proteins were analyzed by Western blot analysis with either an antiphosphotyrosine antibody or the same anti–EGFR antibody used for immunoprecipitation to control for the amount of proteins in each sample. As shown in Figure 1C , EGF caused a substantial increase in the amount of tyrosine phosphorylation of the EGFR, whereas PMA did not increase the amount of tyrosine phosphorylation of the EGFR. These data indicate that PKC activation stimulates MAPK in rat lacrimal gland acini but not by activating the EGFR. 

Effect of Inhibition of PKC on Agonist-Stimulated MAPK Activation

Because PKC can activate p42/p44 MAPK, we sought to determine the role of PKC in cholinergic-, α_1D-adrenergic-, and EGF-stimulated activated MAPK. Acini were preincubated for 10 minutes with either of two PKC inhibitors, calphostin C (10^–10-10^–6 M) or Ro-31-8220 (10^–9-10^–5 M), before incubation for 5 minutes with carbachol (Cch; 10^–4 M), phenylephrine (Ph; 10^–4 M), or EGF (10^–7 M). Cch (10^–4 M) significantly increased the activation of MAPK 4.55 ± 0.77-fold over the basal value (data not shown). Calphostin C inhibited this activation in a concentration-dependent manner with a maximum decrease, which was significant, of 73.5% ± 10.4% at 10^–6 M. In another set of experiments, Ro-31-8220 also significantly decreased carbachol-stimulated MAPK activation in a biphasic manner, with a maximum of 77.8% ± 20.3% at 10^–8 M and 67.8% ± 23.3% at 10^–5 M, each of which was significantly decreased from carbachol alone (Fig. 2) . In these experiments, Cch significantly increased activation of MAPK 2.00 ± 1.23-fold over the basal value (data not shown). 

Ph (10^–4 M), an α_1D-adrenergic receptor agonist, significantly increased MAPK 2.27 ± 0.91-fold above basal value (data not shown). This activation was significantly inhibited by calphostin C a maximum of 92.6% ± 7.3% at 10^–7 M. In another set of experiments, Ro-31-8220 also significantly inhibited Ph-stimulated p42/p44 MAPK activation a maximum of 84.7% ± 10.3% at 10^–9 M (Fig. 2) . In these experiments, Ph significantly increased the activation of MAPK 1.82 ± 0.20-fold over the basal value (data not shown). 

The growth factor EGF (10^–7 M) increased MAPK activation by 2.32 ± 0.21-fold above basal value (data not shown). Calphostin C significantly decreased EGF-stimulated MAPK activation in a concentration-dependent manner, with a maximum inhibition of 99.7% ± 0.3% at 10^–6 M. In separate experiments, Ro-31-8220 also significantly decreased EGF-stimulated MAPK a maximum of 85.7% ± 9.8% at 10^–5 M (Fig. 2) . In these experiments, EGF significantly increased the activation of MAPK 3.01 ± 0.57-fold over the basal value (data not shown). 

These results suggest that PKC is necessary for cholinergic-, α_1D-adrenergic–, and EGF-stimulated activation of MAPK though calphostin C–inhibited α_1D-adrenergic agonist- and EGF-stimulated MAPK at lower concentrations than it inhibited cholinergic agonists. 

Effect of Ca²⁺ on MAPK Activation

Because cholinergic and α_1D-adrenergic agonists and EGF increase intracellular Ca²⁺, ^{7 9 19} we sought to determine the role that Ca²⁺ plays in the activation of p42/p44 MAPK. Acini were incubated with the calcium ionophore ionomycin (Fig. 3) . Lacrimal gland acini were incubated with ionomycin 10^–9 to 10^–5 M for 5 minutes, followed by Western blot analysis for activated MAPK. Ionomycin increased the amount of activated MAPK by 2.83 ± 0.33-, 3.84 ± 1.08-, 3.66 ± 0.86-, 4.03 ± 0.99-, and 2.55 ± 0.55-fold increases above basal value at 10^–9, 10^–8, 10^–7, 10^–6, and 10^–5 M, respectively. These values were significantly increased above basal value. 

To explore the role of Ca²⁺ in agonist-stimulated MAPK activation, Ca²⁺-free buffer was made by omitting Ca²⁺ and adding EGTA (2 mM) to the KRB-BSA. Acini were resuspended in the Ca²⁺-free buffer before stimulation with Cch, Ph, or EGF (Fig. 4) . Activated MAPK was significantly increased by 2.79 ± 0.57-, 1.78 ± 0.26-, and 2.24 ± 0.47-fold above basal value by Cch (10^–4 M), Ph (10^–4 M), and EGF (10^–7 M), respectively (data not shown). Cch-stimulated MAPK activation was significantly decreased by 54.0% ± 19.1% by chelation of Ca²⁺. In contrast, chelation of Ca²⁺ did not significantly decrease Ph- and EGF-stimulated MAPK activation, which were decreased by 17.7% ± 13.6% and 27.8% ± 13.8%, respectively. 

These results indicate that Ca²⁺ alone can activate MAPK. In addition, Ca²⁺ plays a pivotal role in cholinergic-stimulated, but not α_1D-adrenergic agonist– or EGF-stimulated, MAPK activation. 

Effect of Time on Agonist Activation of Pyk2

We previously showed that Cch activated Pyk2 on Tyr402 after a 5-minute stimulation but that Ph and EGF did not. ¹² Therefore, we examined the time dependence of this activation and determined whether Ph or EGF activated Pyk2 at a wide range of times. Lacrimal gland acini were incubated with Cch (10^–4 M), Ph (10^–4 M), or EGF (10^–7 M) for 0 to 10 minutes Acini were homogenized, and the amounts of phosphorylation at Tyr402 (active) Pyk2 and total Pyk2 were determined by Western blot analysis. Cch increased the amount of phospho-Pyk2 with a maximum response after 1 minute, whereas Ph and EGF did not increase phospho-Pyk2 at any time tested (Fig. 5A) . When the results from three independent experiments were analyzed, Cch increased phospho-Pyk2 to 1.52 ± 0.01-fold above basal value after 30 seconds, 1.75 ± 0.19-fold after 1 minute, 1.31 ± 0.04-fold after 5 minutes, and 1.13 ± 0.04-fold after 10 minutes (Fig. 5B) . The increase was significantly different from the basal value at 30 seconds and at 5 minutes. Neither Ph nor EGF increased phosphorylated Pyk2 (Fig. 5B) . 

In addition to phosphorylation on Tyr402, it is known that Pyk2 can be phosphorylated on Tyr580 and Tyr881. ²⁰ Samples stimulated with Cch were also analyzed with the use of antibodies to Pyk2 phosphorylated on these tyrosine residues. Cch did not increase the amount of phosphorylation on Tyr580 or Tyr881, indicating that Tyr402 phosphorylation is responsible for Pyk2 activation in the lacrimal gland (data not shown). 

These results indicate that cholinergic agonists, but not α_1D-adrenergic agonists or EGF, activate Pyk2 in a time-dependent manner. Because Pyk2 stimulation is Ca²⁺ dependent, it is consistent that the Ca²⁺-dependent cholinergic agonist is the only agonist to activate Pyk2. 

Effect of Time on Agonist Activation of c-Src

We previously showed that Cch activates c-Src and that c-Src inhibition decreases Cch-stimulated MAPK while it increases Cch-stimulated protein secretion. ¹² In addition, given that Pyk2 and c-Src interact to be active, ²⁰ we examined the time dependence of c-Src activation and whether Ph or EGF activated c-Src. Acini were incubated with Cch (10^–4 M), Ph (10^–4 M), or EGF (10^–7 M) for 0 to 10 minutes, and the amount of phosphorylated (activated) c-Src was measured by Western blot analysis with an antibody against phosphorylated c-Src (Tyr416). Each condition was standardized for the number of cells using total c-Src. As shown in Figure 6A , Cch stimulated the phosphorylation of c-Src by 30 seconds and 1 minute, whereas Ph and EGF did not. When four to five independent experiments were analyzed, Cch significantly increased the amount of phosphorylated c-Src to 1.73 ± 0.19-fold above basal value after 30 seconds (Fig. 6B) . Cch further increased the amount of phosphorylated c-Src, though the increase was not significant, to 2.51 ± 0.59-fold above basal value after 1 minute, 2.61 ± 0.64-fold after 5 minutes, and 1.68 ± 0.32-fold after 10 minutes. Neither Ph nor EGF increased the amount of phosphorylated c-Src at any time tested (Fig. 6B) . These results indicated that cholinergic agonists increased the amount of phosphorylated c-Src in a time-dependent manner but that α_1D-adrenergic agonists and EGF did not. 

Effect of PKC and Ca²⁺ on Activation of Pyk2

Given that Cch activates Pyk2 and c-Src and that MAPK activation is dependent on PKC, we determined the role of PKC activation with PMA and Ca²⁺ in Cch-stimulated Pyk2 and c-Src activation. Lacrimal gland acini were incubated with Cch (10^–4 M) and the PKC activator PMA (10^–6 M), acini were homogenized, and the amounts of phosphorylated Pyk2 on Tyr402 and c-Src and Tyr416 were determined by Western blot analysis. Each sample was normalized to the amount of total p42/p44 MAPK to correct for the number of cells in each condition. Cch and PMA significantly increased the amount of phosphorylated Pyk2 1.88 ± 0.22-fold and 1.88 ± 0.20-fold above basal value, respectively (Fig. 7A) . 

To determine the effects of Ca²⁺ on Cch-stimulated Pyk2, acini were incubated with ionomycin (10^–7 M) for 5 minutes or in KRB buffer containing 2 mM EGTA and were stimulated with Cch (10^–4 M) for 5 minutes. Cch significantly increased the phosphorylation of Pyk2 1.88 ± 0.22-fold above basal value. This increase was completely inhibited with the chelation of Ca²⁺. Ionomycin also significantly increased the amount of phosphorylated Pyk2 1.82 ± 0.29-fold above basal value (Fig. 7A) . 

These results indicate that PKC and Ca²⁺ can play a role in the activation of Pyk2. Furthermore, Ca²⁺ plays a major role in cholinergic agonist activation of Pyk2. 

Effect of c-Src on Cholinergic Agonist-Stimulated Pyk2 Activation

To determine whether Pyk2 activates with c-Src during cholinergic agonist stimulation, acini were preincubated with the c-Src inhibitor PP1 (10^–5 M) ²¹ for 15 minutes before stimulation with Cch (10^–4 M) for 5 minutes The amount of phosphorylated Pyk2 was determined by Western blot analysis. PP1 significantly inhibited phosphorylation of Pyk2 on Tyr402 56.12% ± 19.48% from 1.41 ± 0.16-fold increase above basal value with Cch alone to 1.08 ± 0.18-fold increase above basal value in the presence of Cch plus PP1 (Fig. 7B) . These results indicated that Pyk2 and c-Src are both necessary for Cch to activate MAPK in the lacrimal gland. 

Discussion

The activation of MAPK by EGF is central to the control of short- and long-term processes in multiple cell types. ¹³ In addition, G-protein–linked agonists can activate MAPK by transactivating EGFR with the triple-membrane passing mechanism, which releases the EGF family of growth factors by ectodomain shedding. ²² G-protein–linked agonists can also stimulate MAPK activity by other mechanisms, depending on the cell type and the stimulus. Three different known agonists in the lacrimal gland—EGF, cholinergic agonists, and α_1D-adrenergic agonists—activate MAPK, and each uses a distinct cellular pathway (Fig. 8) . ² EGF activates MAPK by the classical pathway through phosphorylation of the EGFR, recruitment of Shc and Grb2, and stimulation of Sos, which induces Ras to activate Raf (MAPKKK), MEK (MAPKK), and ERK (MAPK). ¹² In this study, we found that EGF does not phosphorylate Pyk2 or Src. EGF activation of MAPK does not stimulate protein secretion, though EGF can stimulate protein secretion with the use of PLC-γ, intracellular Ca²⁺, and PKC. ⁹ Cholinergic agonists stimulate PLC-β with the use of Gαq/11 to increase the intracellular Ca²⁺ and to activate the PKC isoforms α, ε, and δ (in rank order of importance). ⁴ Cholinergic agonists also activate MAPK but do not transactivate the EGFR. ¹² In this study, we demonstrated that these agonists activate MAPK by increasing intracellular Ca²⁺ and activating PKC, which phosphorylates Pyk2 and Src. In contrast to cholinergic agonists, α₁-adrenergic agonists cause a small increase in intracellular Ca²⁺. In addition, they activate PKC-ε to stimulate secretion and PKC-α and PKC -δ to inhibit secretion. ⁴ α_1D-Adrenergic agonists transactivate the EGFR through a matrix metalloproteinase. In this study, we found that α₁-adrenergic agonists do not activate Pyk2 or Src. Activation of MAPK by cholinergic and α₁-adrenergic agonists attenuates stimulated protein secretion. Thus EGF, cholinergic agonists, and α_1D-adrenergic agonists all activate MAPK in the lacrimal gland, causing the same functional result, but they use distinct signaling components. 

In spite of the fact that different agonists activate MAPK in the lacrimal gland, this enzyme can be stimulated in an agonist-independent fashion. Increased intracellular Ca²⁺ and PKC activation of PKC cause the phosphorylation of MAPK. Thus the calcium ionophore ionomycin and the phorbol esters activate lacrimal gland MAPK in time- and concentration-dependent manners. However, the three agonists that activate MAPK in the lacrimal gland depend on Ca²⁺ and PKC differently. Cholinergic agonist phosphorylation of MAPK was blocked 54% by chelating extracellular Ca²⁺ with EGTA. In contrast, α_1D-adrenergic agonist and EGF stimulation were not significantly inhibited by chelating extracellular Ca²⁺. Thus only cholinergic agonist activation of MAPK was Ca²⁺ dependent, which is consistent with our previous findings that cholinergic agonists cause a substantial increase in intracellular Ca²⁺, but α_1D-adrenergic agonists and EGF cause only small, though significant, increases. 

PKC activation is the second mechanism by which MAPK is activated in the lacrimal gland. We used two different PKC inhibitors, Ro-31-8220 and calphostin C, and found that cholinergic activation of MAPK was only partially blocked 59% and 68% by Ro-31-8220 and calphostin C, respectively. This is consistent with our finding that cholinergic activation of MAPK is also Ca²⁺ dependent. In contrast to cholinergic agonists, the effect of α₁-adrenergic agonists and EGF on MAPK was completely blocked by both PKC inhibitors. Thus cholinergic agonists activate MAPK by increasing intracellular Ca²⁺ and activating PKC, whereas α₁-adrenergic agonists and EGF use only PKC. 

We found that cholinergic agonists, but not α₁-adrenergic agonists or EGF, use Pyk2 and Src to increase MAPK activity. Pyk2 and Src are well known to activate MAPK, though the mechanism of action differs between tissues. In some tissues, Pyk2 and Src play roles in the transactivation of the EGFR. ^{23 24} In other tissues, Pyk2 and Src bypass the EGFR and interact with MAPK directly or with other components of the signaling pathway. ¹¹ In the lacrimal gland, cholinergic agonists do not transactivate the EGFR, so Pyk2 and Src must use an alternative mechanism. Pyk2 has an SH1 domain that allows it to interact with a variety of proteins, including Src family kinases, Grb2/Sos complex, p130Cas, paxillin, Hic-5, and Graf. ²⁵ In PC12 cells, H₂O₂ stimulation causes Pyk2 and Src to interact and to be coimmunoprecipitated. ²⁶ An increase in intracellular Ca²⁺ causes Pyk2 to be autophosphorylated on Tyr402 ^{11 20 25} The phosphorylated Pyk2 binds to and phosphorylates Src, which then phosphorylates Pyk2 on Tyr580 and Tyr881 and fully activates it. We studied cholinergic activation of Pyk2 and c-Src in the lacrimal gland and found that Pyk2 could be activated by increasing the intracellular Ca²⁺ with ionomycin and by activating PKC with phorbol esters. Furthermore, chelating extracellular Ca²⁺ with EGTA blocked the activation of Pyk2 by cholinergic agonists. We also found that inhibiting c-Src with PP1 blocked cholinergic agonist activation of Pyk2. This suggests that Pyk2 and c-Src act together to mediate the effects of cholinergic agonists on MAPK in the lacrimal gland. 

Interestingly, in rat lacrimal gland, cholinergic agonists caused phosphorylation on only Tyr402 of Pyk2 and not Tyr580 or Tyr881. This is in contrast to H₂O₂ stimulation of PC12 cells, by which H₂O₂ increases the phosphorylation of Pyk2 on Tyr580, Tyr881, and Tyr402, ²⁶ or of C6 glioma cells, by which activation of P2Y₁₂ receptors and noradrenaline stimulation of mesenteric small arteries cause phosphorylation on Tyr402 and Tyr597 (phosphorylation of Tyr881 was not measured in these studies). ^{27 28} In PC12 cells, the phosphorylation of Tyr580 and Tyr881 requires phospholipase D2 (PLD2) activity, ²⁶ whereas the phosphorylation of Tyr881 in astrocytes assists with the binding of Grb2. ²⁹ It is known that cholinergic agonists do not require Grb2 stimulation for MAPK activation in the lacrimal gland, ¹² which is consistent with the lack of phosphorylation of Tyr881. However, the role of PLD2 in Pyk2 and MAPK activation in the rat lacrimal gland are unknown. 

Although we know that Pyk2 and c-Src do not interact with the EGFR on cholinergic stimulation, ¹² we do not know which signaling proteins these nonreceptor tyrosine kinases use to activate MAPK. We previously showed that cholinergic agonists do not activate Grb2 or Shc, so potential target proteins are Sos, Ras, Raf, MEK, MAPK, and proteins that interact with these components. ¹²  

In conclusion our study demonstrates that although MAPK activation in the lacrimal gland is Ca²⁺ and PKC dependent, three stimuli of lacrimal gland secretion differentially activate MAPK. Cholinergic agonists use Ca²⁺ and PKC to phosphorylate Pyk2 and c-Src, which activate a target protein distal to the EGFR, Shc, and Grb2, subsequently stimulating MAPK activity. In contrast α_1D-adrenergic agonists and EGF do not use Pyk2 or Src but do activate PKC. α_1D-Adrenergic agonists activate PKC to transactivate the EGFR, which in turn activates Shc, Grb2, Sos, Ras, Raf, MEK, and finally MAPK. The activation of MAPK attenuates stimulated protein secretion. 

 Supported by National Institutes of Health Grant EY06177.

 Submitted for publication January 10, 2006; revised March 15, 2006; accepted May 24, 2006.

 Disclosure: R.R. Hodges, None; J.D. Rios, None; J. Vrouvlianis, None; I. Ota, None; D. Zoukhri, None; D.A. Dartt, None

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

 

Figure 1.

 

Effect of PMA on p42/p44 MAPK activation. Lacrimal gland acini were incubated with PMA (10^–6 M) for 0 to 30 minutes (A) or with PMA (10^–10-10^–6 M) for 5 minutes (B), and the amount of activated p42/p44 MAPK/total p42 MAPK was measured. Inset: Blots are representative of three to seven independent experiments. Data are mean ± SEM. *Statistical significance from t = 0 or the basal value. (C) Acini were also incubated with EGF (10^–7 M) or PMA (10^–6 M) for 5 minutes, EGFR was immunoprecipitated, and samples were blotted for phosphotyrosine or EGFR.

Figure 2.

 

Effect of inhibition of PKC on agonist-stimulated p42/p44 MAPK activation. Acini were preincubated for 10 minutes with calphostin C (10^–10-10^–6 M) or Ro-31-8220 (10^–9-10^–5 M). Acini were then stimulated with (A) carbachol (10^–4 M), (B) phenylephrine (10^–4 M), and (C) EGF (10^–7 M) for 5 minutes, and the amount activated p42/p44 MAPK/total p42 MAPK was measured. Data are mean ± SEM from three independent experiments. *Statistical significance from the basal value. Blots depict representative results of the experiments.

Figure 3.

 

Effect of ionomycin on p42/p44 MAPK activation. Acini were incubated for 10 minutes with increasing concentrations of ionomycin (10^–9-10^–5 M). Activated p42/p44 MAPK/total p42 MAPK was analyzed by Western blot. Representative experiment is shown (A). Results of five independent experiments are shown (B). Data are mean ± SEM *Statistical significance from the basal value.

Figure 4.

 

Effect of chelation of Ca²⁺ on p42/p44 MAPK activation. Acini were incubated for 10 minutes with Cch (10^–4), Ph (10^–4 M), or EGF (10^–7 M) in the presence of extracellular Ca²⁺ or in the absence of Ca²⁺ plus 2 mM EGTA, and the amount of phosphorylated p42/p44 MAPK/total p42 MAPK was measured. (A) Representative blot. (B) Results of three to five independent experiments were analyzed. Data are mean ± SEM. *Statistical significance from the agonist alone.

Figure 5.

 

Effect of time on agonist-induced Pyk2 Activation. Acini were stimulated with Cch (10^–4 M), Ph (10^–4 M), or EGF (10^–7 M) for 0 to 10 minutes, and the amount of phosphoPyk2 (Tyr402)/total Pyk2 was measured. Representative blots are shown (A). Results from three independent experiments were analyzed and are shown (B). Data are mean ± SEM. *Statistical significance from the basal value.

Figure 6.

 

Effect of time on agonist-induced c-Src activation. Acini were stimulated with Cch (10^–4 M), Ph (10^–4 M), or EGF (10^–7 M) for 0 to 10 minutes, and the amount of phospho-c-Src(Tyr416)/total c-Src was measured. Representative blots are shown in (A). Results from three independent experiments were analyzed and are shown in (B). Data are mean ± SEM. *Statistical significance from the basal value.

Figure 7.

 

Effect of PKC and Ca²⁺ and c-Src on Pyk2 activation. Acini were stimulated for 5 minutes with Cch (10^–4 M), PMA (10^–6 M), Cch and in Ca²⁺-free buffer containing 2 mM EGTA, and ionomycin (10^–7 M). The amount of phosphorylated Pyk2 (Tyr402)/total Pyk2 is shown (A). Data are mean ± SEM of three to six independent experiments. Acini were also preincubated with PP1 for 15 minutes before stimulation with Cch. The amount of phosphorylated Pyk (Tyr402) is shown (B). Data are mean ± SEM from four independent experiments. *Statistical significance from the basal value. #Statistical significance from Cch alone. Blots are representative results of the experiments.

Figure 8.

 

Schematic diagram of the signal transduction pathways used by cholinergic and α₁-adrenergic agonists and EGF to activate MAPK and stimulate protein secretion. Norepi, norepinephrine; Ach, acetylcholine.