CCK activates RhoA and Rac1 differentially through Gα13 and Gαq in mouse pancreatic acini

Maria E. Sabbatini, Yan Bi, Baoan Ji, Stephen A. Ernst, John A. Williams

Abstract

Cholecystokinin (CCK) has been shown to activate RhoA and Rac1, as well as reorganize the actin cytoskeleton and, thereby, modify acinar morphology and amylase secretion in mouse pancreatic acini. The aim of the present study was to determine which heterotrimeric G proteins activate RhoA and Rac1 upon CCK stimulation. Gα13, but not Gα12, was identified in mouse pancreatic acini by RT-PCR and Western blotting. Using specific assays for RhoA and Rac1 activation, we showed that only active Gα13 activated RhoA. By contrast, active Gα13 and Gαq, but not Gαs, slightly increased GTP-bound Rac1 levels. A greater increase in Rac1 activation was observed when active Gα13 and active Gαq were coexpressed. Gαi was not required for CCK-induced RhoA or Rac1 activation. The regulator of G protein signaling (RGS) domain of p115-Rho guanine nucleotide exchange factor (p115-RGS), a specific inhibitor of Gα12/13-mediated signaling, abolished CCK-stimulated RhoA activation. By contrast, both RGS-2, an inhibitor of Gαq, and p115-RGS abolished CCK-induced Rac1 activation, which was PLC pathway-independent. Active Gαq and Gα13, but not Gαs, induced morphological changes and actin redistribution similar to 1 nM CCK. CCK-induced actin cytoskeletal reorganization was inhibited by RGS-2, but not by p115-RGS, whereas CCK-induced amylase secretion was blocked by both inhibitors. Together, these findings indicate that, in mouse pancreatic acini, Gα13 links CCK stimulation to the activation of RhoA, whereas both Gα13 and Gαq link CCK stimulation to the activation of Rac1. CCK-induced actin cytoskeletal reorganization is mainly mediated by Gαq. By contrast, Gα13 and Gαq signaling are required for CCK-induced amylase secretion.

  • actin cytoskeleton
  • amylase secretion
  • bleb formation

cholecystokinin (CCK) activates multiple signaling pathways in pancreatic acinar cells through a specific G protein-coupled receptor, CCK1, which is coupled to heterotrimeric G proteins (38, 55). Activated G protein-coupled receptors interact with and produce a conformational change in G protein α-subunits (Gα), which promotes exchange of GDP for GTP. GTP-bound Gα dissociates from Gβγ and activates downstream effectors. Activation of Gαq by CCK results in stimulation of PLCβ, leading to formation of inositol-1,4,5-trisphosphate, which increases intracellular Ca2+ levels, and diacylglycerol, which activates PKC (49). Activation of Gαs stimulates adenylate cyclase (14), which in pancreatic acinar cells is associated with an increase in cAMP-dependent protein kinase activity (26). CCK activates several forms of Gαi in rat pancreatic acinar cells (38), which in various cells inhibits adenylate cyclase. In pancreatic acini, pertussis toxin blocks Gαi and, thereby, increases cAMP formation (46) but has no effect on Ca2+ mobilization (28, 46). G12/13 is another well-characterized heterotrimeric G protein family that can initiate intracellular signaling (51). A previous study showed that, in NIH 3T3 cells expressing the CCK1 receptor, CCK activates Gα12/13 and, thereby, the small GTP-binding protein RhoA, inducing actin cytoskeletal reorganization (23). In intestinal smooth muscle, CCK stimulates phospholipase D activity via RhoA through Gα13 activation (30).

Rho proteins, members of the Ras superfamily of small GTP-binding proteins, participate in several cellular processes, including cytoskeletal rearrangement, cell cycle progression, gene transcription, and cytokinesis (5, 17). The Rho family includes at least three well-studied subfamilies: Rho, Rac, and Cdc42. Like all small GTP-binding proteins, members of the Rho family cycle between two forms: an active GTP-bound form and an inactive GDP-bound form. This cycle is regulated by three groups of regulatory proteins: guanine nucleotide exchange factors (GEFs), which induce the binding of GTP to the small GTP-binding protein; GTPase-activating proteins, which induce inactivation by hydrolysis of GTP; and guanine nucleotide-dissociation inhibitors, which, in the cytosol, bind the GDP-bound form of the small GTP-binding proteins, thereby preventing exchange to the GTP-bound form (17). One group of RhoGEFs, including p115-RhoGEF, contains a regulator of G protein signaling (RGS) domain, which interacts with the α-subunit of the heterotrimeric G protein G12/13 and, when expressed as an isolated domain, inhibits G12/13 signaling (17). A number of RacGEFs are known, and some are activated by second messengers downstream of Gαq (5).

In pancreatic acinar cells, RhoA and Rac1 have been implicated in the regulation of CCK-induced amylase secretion through an actin cytoskeleton-dependent cellular process (3, 4). CCK not only increases the amount of GTP-bound RhoA and GTP-bound Rac1, but it also induces translocation of both from the cytosol to the membrane (4). Constitutively active RhoA and Rac1 expression induces morphological changes, actin cytoskeletal reorganization, and amylase secretion (3), whereas dominant-negative RhoA and Rac1 expression reduces CCK-induced acinar morphological changes, actin reorganization, and amylase secretion (3, 4).

Because little is known of the upstream regulators of RhoA and Rac1 in mouse pancreatic acini, the aim of the present study was to determine which heterotrimeric G proteins link CCK stimulation to activation of RhoA and Rac1. Four families of heterotrimeric G proteins were studied, Gq, Gi, Gs, and G12/13, which have been implicated in the activation of RhoA and Rac1 in several cell types (9, 12, 13, 23, 27, 32, 44, 45). We used pull-down assays to study RhoA and Rac1 activation and isolated pancreatic acini infected with adenoviruses encoding the constitutively active α-subunits, Gαq, Gα13, and Gαs, as well as RGS domains of the p115-RhoGEF and RGS-2, which inhibit Gα12/13 and Gαq, respectively. Pretreatment with pertussis toxin was used to study the participation of Gαi. The present findings show that Gα13 is the only member of the G12/13 family expressed in mouse pancreatic acini. We also demonstrate that Gα13 activation mediates the response to CCK by RhoA, whereas Gαq and Gα13 activation mediates the response to CCK by Rac1 in a PLC pathway-independent manner. Finally, we found that although Gα13 is able to induce actin reorganization and bleb formation via the RhoA/Rho kinase pathway, Gαq is implicated in CCK-induced actin cytoskeletal reorganization and acinar morphological changes, whereas Gαq and Gα13 are required for CCK-induced amylase secretion.

MATERIALS AND METHODS

Materials.

Collagenase was purchased from Crescent Chemical (Islandia, NY); BSA and soybean trypsin inhibitor from Sigma Chemical (St. Louis, MO), and DMEM and Alexa 594-conjugated phalloidin from Invitrogen (Carlsbad, CA). The following stimuli and inhibitors were used: sulfated CCK octapeptide (Research Plus, Bayonne, NJ); vasoactive intestinal polypeptide (VIP; American Peptide, Sunnyvale, CA); A-23187 and GF-109203X (Calbiochem, La Jolla, CA); and PMA, Y-27632, pertussis toxin, and U-73122 (Sigma Chemical). All other chemicals were of reagent grade.

Antibodies against the following proteins were used: rabbit polyclonal antibodies to Rap1, RhoA, Gαq, Gα12, and Gα13 (Santa Cruz Biotechnology, Santa Cruz, CA); rabbit polyclonal antibody to Gα12 (Abcam, Cambridge, MA); mouse monoclonal antibody to hemagglutinin (HA)- and Myc-tagged proteins (Cell Signaling Technology, Beverly, MA); and mouse monoclonal antibody to Rac1 (Pierce Biotechnology, Rockford, IL).

Fed male ICR mice (22–27 g body wt) were used in the experiments. All the experimental protocols were approved by The University of Michigan University Committee on the Use and Care of Animals.

Construction of recombinant adenoviruses.

Plasmids of constitutively active G protein α-subunits, Gαq(Q209L) and Gα13(Q226L), denoted active Gαq and Gα13, respectively, were obtained from J. Silvio Gutkind (National Institutes of Health, Bethesda, MD) and Diana Barber (University of California, San Francisco, CA). The plasmid of Gαs(Q227L), denoted Gαs, was purchased from American Type Culture Collection (Manassas, VA). Recombinant adenoviruses were prepared using the pAdTrack System (Addgene, Cambridge, MA) according to the method of He et al. (16), as described previously (3), and amplified and purified using the ViraBind adenovirus purification kit (Cell Biolabs, San Diego, CA). Adenoviruses expressing the RGS domain of p115-RhoGEF, Myc-tagged p115-RGS, the mutant of the RGS domain of p115-RhoGEF, Myc-tagged p115-RGS(E29K), and HA-tagged RGS-2 were obtained from Patrick J. Casey (Duke University, Durham, NC). Adenovirus expressing β-galactosidase (β-Gal) was used as a control.

Preparation, short-term culture, and viral infection of pancreatic acini.

Mouse pancreatic acini from ICR mice were prepared by enzymatic digestion with collagenase followed by mechanical shearing, as previously described (4). Acini were cultured in suspension without shaking at low density in 10-cm petri dishes in DMEM enriched with 0.1% BSA, 0.01% soybean trypsin inhibitor, and antibiotics and incubated overnight at 37°C with 5% CO2. For the viral infection experiments, 106 plaque-forming units/ml of constitutively active Gα13, Gαq, and Gαs or 107 plaque-forming units/ml of p115-RGS, p115-RGS(E29K), and RGS-2 were added to the culture medium at the beginning of the overnight incubation, unless otherwise indicated. Under such conditions, >95% of acinar cells express adenoviral-driven proteins (4). In another experiment, acutely dissociated acini were incubated with or without pertussis toxin (2 μg/ml) for 2 h at 37°C, as previously described (36).

Detection of Gα12 and Gα13 expression in mouse pancreatic acini.

Expression of Gα12 and Gα13 in mouse pancreatic acini was assessed by RT-PCR. The following primers, obtained from Invitrogen, were used: mouse Gα12 [GenBank accession no. NM_010302; 5′-ACAAGATGGACCTCCTGGTG-3′ (sense), 5′-GTCTATGGCGGTGGTGAAGT-3′ (antisense)] and mouse Gα13 [GenBank accession no. NM_010303; 5′-CGTGATGCCCGTTTTCTTAT-3′ (sense), 5′-TGGTCAGACCTGACTGCTTG-3′ (antisense)]. The primers were designed with Invitrogen Oligoperfect Designer based on gene sequences obtained from the GenBank National Center for Biotechnology Information Sequence Viewer (http//www.ncbi.nlm.nih.gov). The presence of Gα12 and Gα13 protein was evaluated by Western blotting.

Determination of RhoA, Rac1, and Rap1 activation.

RhoA and Rac1 activation were determined by pull-down assay according to methods modified from Ren et al. (33) and Sander et al. (37), respectively, as described previously (4). Acini were prepared for the assay as described above and washed once with ice-cold Tris-buffered saline and cold lysis buffer containing 50 mM Tris · HCl (pH 7.5), 1% Triton X-100, 500 mM NaCl, 10 mM MgCl2, 10% (vol/vol) glycerol, 10 μg/ml aprotinin, 10 μg/ml leupeptin, and 1 mM PMSF, and 1 mM sodium orthovanadate and 10 mM NaF were added. Acini were rapidly sonicated and cleared at 13,000 g for 10 min at 4°C. The resulting supernatants (2,000–2,500 μg of protein) were quickly transferred to tubes with GST-rhotekin-RhoA binding domain (RBD, 30 μg; Cell Biolabs, San Diego, CA) or GST-p21-activated kinase protein-binding domain (PAK1-PBD, 20 μg; Cell Biolabs and Pierce, Rockford, IL), followed by a 1 h of rotation at 4°C. The supernatants were subjected to extensive washing, and the levels of GTP-Rac1 and GTP-RhoA were quantified by Western blot analysis. Peroxidase activity was visualized using the SuperSignal West Femto sensitivity substrate kit (Pierce). Rap1 activation was determined as previously described (35).

Determination of Gα13 translocation.

Isolated pancreatic acini were resuspended in lysis buffer containing 50 mM HEPES (pH 7.4), 50 mM NaCl, 1 mM MgCl2, 2 mM EDTA, 1 mM PMSF, 10 μg/ml leupeptin, 1 mM sodium orthovanadate, 5 mM NaF, and 1 mM dithiothreitol and lysed by five passages through a 27-gauge needle; then debris and nuclei were removed by centrifugation for 10 min at 500 g at 4°C. The supernatants were centrifuged at 120,000 g for 45 min in a Beckman Optima TLX ultracentrifuge. After the supernatant was collected as the cytosol fraction, the pellet was resuspended in the lysis buffer containing 1% Triton X-100 and centrifuged at 10,000 g for 10 min; the supernatant was collected as the membrane fraction.

Fluorescence staining of actin.

After overnight incubation, isolated mouse pancreatic acini were allowed to settle in test tubes and then fixed for 1 h at room temperature with 4% paraformaldehyde in PBS. Acinar preparations were rinsed with PBS, cryoprotected, and frozen as previously described (35). Cryostat sections were incubated with Alexa 594-conjugated phalloidin (1:50 dilution), which stains filamentous actin (F-actin). Digitized images of Z series of 10–17 optical sections (0.5 μm thick) extending through the thickness of the cryosections were collected with an Olympus 500 Fluoview confocal microscope. Z-series stacks were collapsed with the Olympus software and then processed using Photoshop CS software (Adobe System, Mountain View, CA).

Measurement of 45Ca2+ efflux.

Cellular 45Ca2+ mobilization in isolated pancreatic acini was measured using a previously described procedure (48). Acini were suspended in HEPES-buffered Ringer solution and preincubated with 2 μCi/ml 45CaCl2 (New England Nuclear) for 1 h at 37°C. Then labeled acini were washed twice with nonradioactive medium, suspended in fresh medium, and incubated with or without CCK for a specified time at 37°C. 45Ca2+ remaining in acini was calculated by difference between total 45Ca2+ and 45Ca2+ released from the acini into the extracellular medium.

Quantification of intracellular cAMP content.

cAMP generation was determined as previously described (35). cAMP was extracted in ethanol and measured using a cAMP colorimetric enzyme immunoassay kit (Cayman Chemical, Ann Arbor, MI). Results are expressed as fold increase vs. control.

Measurement of amylase secretion.

Overnight cultured acini were stimulated with CCK in 1-ml aliquots in plastic blood dilution vials for 30 min. Samples were then centrifuged for 30 s in a microcentrifuge, and the supernatant was assayed for amylase activity with Phadebas reagents (Magle Life Sciences, Lund, Sweden). Results are expressed as percentage of initial acinar amylase content.

Statistical analysis.

Values are means ± SE of three to six separate experiments. Statistical analysis was performed by ANOVA following by the Student-Newman-Keuls test. P ≤ 0.05 was considered statistically significant.

RESULTS

13, but not Gα12, is present in mouse pancreatic acini and associated with cellular membranes.

Because Gα12/13 has not been studied in detail in mouse pancreatic acinar cells, we first established the forms present in acini. Gα13 mRNA expression was observed in whole pancreas and pancreatic acini with a PCR product of the expected size, whereas Gα12 mRNA expression was detected only in brain and lung (see supplemental Fig. 1A in the online version of this article). At the protein level, Gα13 was found not only in whole pancreas but also in pancreatic acini. Unlike Gα13, Gα12 protein could not be detected in pancreas with use of two different antibodies (see supplemental Fig. 1B). Although Gα13 is often tightly associated with cellular membranes (2, 41), in some cells Gα13 is found in the cytosol and, upon stimulation, translocates to the membrane (53). In pancreatic acini in nonstimulated and stimulated conditions, Gα13 was predominantly present in the membrane fraction (see supplemental Fig. 1C). These results indicate that G13 is the major G protein of the G12/13 family present in mouse pancreatic acini and associated with the membranes.

13 signaling induces RhoA activation, whereas Gαq and Gα13 signaling induce Rac1 activation.

As previously reviewed, CCK induces RhoA and Rac1 activation in pancreatic acini (4). The heterotrimeric G proteins G12, G13, Gi, and Gq have been implicated in the activation of RhoA and Rac1 through specific GEFs in several cell types (9, 21, 45). Since Gα13 was found in mouse pancreatic acini, we tested the ability of constitutively active forms of Gα13, Gαq, and Gαs to promote RhoA and Rac1 activation. RhoA activation was induced by active Gα13, but not by active Gαq or Gαs (Fig. 1A). By contrast, Rac1 activation was weakly induced by active Gαq or Gα13 (Fig. 1B). When active Gαq and Gα13 were coexpressed, an additive effect on Rac1 activation was observed (Fig. 1C). The effectiveness of Gαs expression was shown by an increase in cAMP levels (data not shown). The activation of Gαi is not required in the actions of CCK, since pretreatment with pertussis toxin did not modify the activation of RhoA or Rac1 induced by CCK (see supplemental Fig. 2, A and B). These findings indicate that whereas Gα13 activates RhoA, Gα13 and Gαq activate Rac1.

Fig. 1.

Expression of active Gα13 activates RhoA, whereas expression of active Gαq and Gα13 additively activates Rac1. β-Galactosidase [β-Gal (vector control)], Gα13, Gαq, and Gαs were expressed in isolated mouse pancreatic acini by means of recombinant adenoviruses with overnight incubation. In another group of acini, active Gα13 and Gαq were coexpressed. β-Gal-expressing pancreatic acini were treated with or without 1 nM cholecystokinin (CCK) for 10 min. Acini were lysed and assayed for activation of RhoA (A) and Rac1 (B and C) using rhotekin Rho-binding domain (RBD) and p21-activated kinase binding domain (PAK1-PBD), respectively, as activation-specific probes. Expression of active Gα13 increased GTP-RhoA and GTP-Rac1 levels, whereas expression of active Gαq increased GTP-Rac1 levels only. Coexpression of active Gα13 and Gαq induces an additive effect on Rac1 activation (C). Top: representative immunoblots for GTP-RhoA, GTP-Rac1, total RhoA, total Rac1, Gα13, and Gαq. Bottom: quantitative analysis of RhoA and Rac1 activation. Values are means ± SE (n = 4–5 experiments). *P < 0.05; **P < 0.01 vs. control. CTL, control.

13 and Gαq, but not Gαs, evoke actin cytoskeletal reorganization and acinar morphological changes.

Supramaximal stimulation by CCK modifies actin cytoskeletal assembly, with shifting of F-actin from subapical to basolateral membranes and induction of cytoplasmic blebs (7, 43). In an earlier study, we showed that CCK-induced actin reorganization involves Rac1 and RhoA activation (3). Since Gα13 is involved in RhoA activation, whereas Gα13 and Gαq are involved in Rac1 activation, the participation of Gα13 and Gαq in cytoskeletal reorganization and bleb formation was studied. Fluorescence staining for F-actin in pancreatic acini expressing β-Gal or active Gαs was pronounced along apical membranes (arrows in Fig. 2, A and B); weaker staining was present at basolateral membranes, and cytoplasmic actin was diffuse and generally of low intensity. Only a few or no cytoplasmic blebs were observed. Consistent with the lack of bleb formation in Gαs-expressing acini, secretagogues that act through Gαs, such as VIP, also do not induce acinar morphological changes (1). By contrast, acini expressing active Gα13 and Gαq exhibited enhanced fluorescence intensity of basolateral membrane-associated actin (Fig. 2, C and D). These changes were accompanied by the presence of blebs (arrowheads in Fig. 2, C and D; also see supplemental Fig. 3). These results are consistent with previous results obtained with 1 nM CCK and active RhoA and Rac1 (3).

Fig. 2.

Expression of active Gα13 or Gαq, but not β-Gal and active Gαs, induces actin cytoskeletal disruption and acinar morphological changes. β-Gal and active Gα13, Gαq, and Gαs were expressed in isolated mouse pancreatic acini. Representative confocal microscopic images of Alexa 594-conjugated phalloidin staining of filamentous actin (F-actin) are shown. In acini expressing β-Gal (A) or active Gαs (B), F-actin stained intensely at apical membranes (arrows) and weakly at basolateral membranes. In acini expressing active Gα13 (C) or active Gαq (D), F-actin strongly localized to basolateral membranes, and cytoplasmic blebs (arrowheads) were observed.

ROCK mediates active Gα13-induced acinar morphological changes.

In several cell types, ROCK, an effector of RhoA, is responsible for myosin light chain phosphorylation and subsequent actin cytoskeletal reorganization and bleb formation (29, 42). In rat pancreatic acini, upon CCK stimulation, ROCK II has been associated with RhoA (22) and implicated in basolateral bleb formation (6, 43). We found that active Gα13 activates RhoA and that blebbing induced by Gα13, but not Gαq, was prevented by the ROCK inhibitor Y-27632 (10 μM), whereas the effect of CCK was partially inhibited by Y-27632 (arrows in supplemental Fig. 3). Thus acinar morphological changes evoked by Gα13 seem to be RhoA/ROCK pathway-dependent.

13 is required for CCK-induced RhoA activation.

To test the importance of Gα13 signaling in CCK-induced RhoA activation, acini were infected with adenoviruses encoding p115-RGS domain, which selectively binds Gα12 and Gα13, thereby preventing them from interacting with their downstream effectors (8, 15, 19, 40). The expression of p115-RGS abolished RhoA activation stimulated by CCK (Fig. 3A). To confirm the specificity of this protein domain, we expressed in pancreatic acini a mutant form of p115-RGS, p115-RGS(E29K), which is unable to bind to and inactivate Gα12/13 proteins (8). The mutant p115-RGS(E29K) expression did not modify CCK-induced RhoA activation (Fig. 3B). Since the CCK receptor also couples to Gαq, we inhibited Gαq signaling using acini infected with adenoviruses encoding RGS-2, which inhibits CCK-evoked Ca2+ signaling mediated by Gαq in mouse pancreatic acini (52). The expression of RGS-2 did not affect the CCK-induced RhoA activation (Fig. 3C). RGS-2 specifically inhibited the activity of Gαq, since the expression of RGS-2 abolished CCK-induced Ca2+ mobilization but did not modify VIP-induced cAMP formation (see supplemental Fig. 4, A and B). These results indicate that G13 is the major heterotrimeric protein involved in CCK-induced RhoA activation.

Fig. 3.

Expression of p115 regulator of G protein signaling (RGS) inhibits CCK-induced RhoA activation. p115-RGS (A), which inhibits Gα12/13 activation, the mutant p115-RGS(E29K) (B), which is unable to inactivate Gα12/13, and RGS-2 (C), which inhibits Gαq activation, were expressed in isolated pancreatic acini for 16 h at 37°C. β-Gal-expressing acini were used as a control. Acini were then incubated with or without 1 nM CCK for 10 min and lysed, and RhoA activation was studied (A–C). Representative immunoblots show that expression of p115-RGS, but not p115-RGS(E29K) and RGS-2, inhibits RhoA activation. Top: representative immunoblots for GTP-RhoA, total RhoA, Myc tag, and hemagglutinin (HA) tag. Bottom: quantitative analysis of RhoA activation. Values are means ± SE (n = 4–5 experiments). *P < 0.05; **P < 0.01 vs. control.

13 and Gαq are required for CCK-induced Rac1 activation.

We also studied which heterotrimeric G proteins mediate CCK-induced Rac1 activation. We found that expression of RGS-2 and p115-RGS abolished the increase in GTP-Rac1 levels induced by CCK, whereas expression of the mutant of p115-RGS was unable to affect CCK-induced activation of GTP-Rac1 (Fig. 4, A and B). Both p115-RGS and RGS-2 also had similar effects on RhoA and Rac1 activation when a lower concentration of CCK (100 pM) was used (data not shown). The specificity of p115-RGS on RhoA and Rac1 was further demonstrated by the observation that CCK-stimulated Rap1 activation (35) was not affected by p115-RGS expression (Fig. 5A). p115-RGS does not affect other signaling pathways in pancreatic acini, since the expression of p115-RGS did not inhibit CCK-induced Ca2+ mobilization or VIP-evoked cAMP generation (Fig. 5, B and C). Thus Gαq and Gα13 are necessary for CCK-induced Rac1 activation in mouse pancreatic acini.

Fig. 4.

Expression of p115-RGS or RGS-2 inhibits CCK-induced Rac1 activation. Isolated pancreatic acini expressing RGS-2 (A) and p115-RGS and the mutant p115-RGS(E29K) (B) were stimulated with 1 nM CCK for 10 min and lysed, and Rac1 activation was determined. β-Gal-expressing acini were used as a control. A representative immunoblot shows that expression of p115-RGS and RGS-2 inhibits CCK-induced Rac1 activation (top). Expression of RGS-2 was confirmed by Western blotting of the HA tag. Bottom: quantitative analysis of Rac1 activation. Values are means ± SE (n = 3–4 experiments). *P < 0.05; **P < 0.01; ***P < 0.001 vs. control.

Fig. 5.

Expression of p115-RGS does not modify other intracellular signaling. β-Gal- and p115-RGS-expressing acini were stimulated with CCK or 10 nM vasoactive intestinal polypeptide (VIP), and Rap1 activation (A), 45Ca2+ mobilization (B), and cAMP formation (C) were determined. Expression of p115-RGS did not affect any of these CCK-induced parameters. A: acini were stimulated with 1 nM CCK for 10 min, then a pull-down assay was used to detect GTP-Rap1. Top: immunoblot for GTP-Rap1 and total Rap1. Bottom: quantitative analysis of Rap1 activation. B: acini were stimulated with 300 pM CCK for 30 min, then 45Ca2+ efflux was determined. Values are means ± SE (n = 3 experiments) of 45Ca2+ remaining expressed as percentage of initial. C: cAMP expressed as fold increase vs. control. *P < 0.05; **P < 0.01 vs. control. †P < 0.05 vs. control p115-RGS.

The PLC pathway is not involved in CCK-induced Rac1 activation.

Upon CCK stimulation, the PLC pathway is activated by Gαq in pancreatic acinar cells (49). Although PLC and PKC have been implicated in the activation of Rac1 by Gαq for HIV-1-induced membrane fusion in U87.CD4.CCR5 cells (18) and for cell death in CHO-K1 cells expressing M3 receptors (39), in mouse pancreatic acini CCK-induced activation of Rac1 was not mediated by PLC, since blockade of PLC with U-73122 (56) did not affect CCK-induced activation of Rac1 (Fig. 6A). A similar lack of effect of U-73122 was seen when the threshold concentration of CCK (10 pM) was used (data not shown). Reduction of CCK-induced Ca2+ mobilization by U-73122 (data not shown) confirmed the ability of U-73122 to inhibit PLC. Ca2+ and PKC were not involved, since neither the Ca2+ ionophore A-23187 nor the PKC stimulator PMA induced Rac1 activation by themselves and the PKC inhibitor GF-109203X (11, 25) did not affect CCK-induced Rac1 activation (Fig. 6B). These results indicate that CCK induces Rac1 activation through a PLC pathway-independent mechanism.

Fig. 6.

CCK induces Rac1 activation through a PLC pathway-independent mechanism. Isolated pancreatic acini were pretreated with vehicle or the PLC inhibitor U-73122 (U, 5 μM; A) or the PKC inhibitor GF-109203X (GF, 1 μM; B) for 30 min and then stimulated with 1 nM CCK. Acini were also stimulated with the Ca2+ ionophore A-23187 (1 μM) or the PKC stimulator PMA (500 nM; B). Top: representative immunoblots for GTP-Rac1 and total Rac1. Bottom: quantitative analysis of Rac1 activation. Values are means ± SE (n = 3–4 experiments). *P < 0.05; **P < 0.01 vs. control.

q is required for CCK-induced actin cytoskeletal reorganization and acinar morphological changes.

Since RhoA and Rac1 are involved in CCK-induced actin cytoskeletal reorganization (3) and we have shown that, upon CCK stimulation, Gαq and Gα13 activated Rac1 and RhoA, we studied whether the heterotrimeric G proteins are required for CCK-induced actin cytoskeletal rearrangement and acinar morphological changes. Acini expressing β-Gal, p115-RGS, or RGS-2 showed normal actin distribution and strong staining intensity of actin at apical membranes (arrows in Fig. 7, A, C, and E). Treatment of β-Gal-expressing acini with 1 nM CCK caused pronounced disorganization of actin, with strong staining at basolateral membranes and abundant blebs (arrowheads in Fig. 7B), as previously shown (3). The expression of RGS-2 inhibited CCK-induced apical actin disruption (arrows in Fig. 7F) and bleb formation, whereas the expression of p115-RGS resulted in little, if any, inhibition of the morphological effects induced by supramaximal CCK stimulation (Fig. 7D). The specificity of each inhibitor was confirmed, since the expression of RGS-2 specifically inhibited blebs induced by constitutively active Gαq without affecting those induced by constitutively active Gα13, which were, rather, inhibited specifically by p115-RGS (data not shown). These findings indicate that CCK-induced actin reorganization is mainly Gαq-dependent.

Fig. 7.

Expression of RGS-2 inhibits actin cytoskeletal reorganization and acinar morphological changes induced by CCK. β-Gal, RGS-2, and p115-RGS were expressed in pancreatic acini during overnight (16 h) incubation, then acini were incubated with or without 1 nM CCK for 30 min. Representative confocal microscopic images show Alexa 594-conjugated phalloidin staining of F-actin. Acini expressing β-Gal (A), p115-RGS (C), or RGS-2 (E) showed normal distribution and staining of actin at apical membranes (arrows). Treatment of β-Gal- or p115-RGS-expressing acini with 1 nM CCK (B and D, respectively) caused pronounced disorganization of actin, with strong staining of actin at basolateral membranes and appearance of numerous cytoplasmic blebs (arrowheads). In RGS-2-expressing acini stimulated with 1 nM CCK (F), little alteration in actin staining or induction of blebs is shown.

Acinar morphological changes induced by CCK are independent of the PLC pathway.

Although CCK-induced morphological changes were dependent on Gαq, the response to CCK was independent of PLC and PKC activation, since U-73122 and GF-109203X were not able to modify CCK-induced bleb formation. Moreover, the PKC stimulator PMA did not stimulate bleb formation by itself (arrows in supplemental Fig. 5).

13 and Gαq participate in the amylase secretion biphasic response to CCK.

Since Gα13 and Gαq activated RhoA and Rac1, which are required for CCK-induced amylase secretion (3), we studied the participation of both heterotrimeric G proteins in the CCK-induced secretory response. The results showed that the expression of p115-RGS, but not RGS-2, inhibited basal amylase release by 36%. Moreover, the expression of p115-RGS or RGS-2 converted the CCK dose-response curve for amylase release from a biphasic to a monophasic pattern, and the expression of RGS-2 tended to increase amylase release at high CCK concentrations (Fig. 8, A and B). These findings indicate that only Gα13 is required for the basal secretory response, whereas Gαq and Gα13 are required for the CCK-stimulated secretory response. Although ROCK is required for the response to active Gα13 on bleb formation, ROCK did not participate in the effect of CCK via Gα13 activation on amylase secretion, since the ROCK inhibitor Y-27632 did not modify the biphasic concentration-response curve induced by CCK (see supplemental Fig. 6). This result indicates that effectors of RhoA other than ROCK are involved in the stimulatory effect of CCK on amylase secretion following Gα13/RhoA activation.

Fig. 8.

Expression of p115-RGS or RGS-2 abolishes the biphasic response induced by CCK on amylase secretion. Isolated pancreatic acini expressing p115-RGS (A) and RGS-2 (B) were stimulated with different concentrations of CCK for 30 min, and amylase release was measured. Expression of p115-RGS inhibited basal amylase release, and expression of p115-RGS or RGS-2 abolished the biphasic response to CCK on amylase release. Values are means ± SE (n = 4–5 experiments). *P < 0.05; **P < 0.01 vs. β-Gal.

DISCUSSION

The identity of the heterotrimeric G proteins mediating CCK-induced RhoA and Rac1 activity, and thereby cytoskeletal reorganization and amylase secretion in mouse pancreatic acini, was determined in the present study. First, we show that only G13 of the G12/13 family is found in mouse pancreatic acini and is associated with the cellular membranes. We further show that expression of the active form of Gα13 induced an increase in the levels of GTP-bound RhoA and, to a lesser extent, GTP-bound Rac1. When Gα13 signaling was inhibited using the RGS domain of p115-RhoGEF, CCK-induced RhoA and Rac1 activation were inhibited. On the other hand, expression of the active form of Gαq induced a modest increase in the levels of GTP-bound Rac1. An additive effect on Rac1 activation was observed when active Gαq and Gα13 were coexpressed. When Gαq signaling was inhibited using RGS-2, only Rac1 activation induced by CCK was inhibited. These findings indicate that Gαq and Gα13 are activated by CCK in mouse pancreatic acini. Gα13 signaling is able to activate RhoA and Rac1, whereas Gαq signaling activates Rac1. We also demonstrate that, although Gαq and Gα13 are able to induce actin cytoskeletal reorganization and acinar morphological changes, only Gαq is required in CCK-induced actin cytoskeletal reorganization and bleb formation. Our findings further demonstrate that the mechanisms of Gα13- and Gαq-induced cytoskeletal reorganization are different: whereas Gα13 induces actin disruption mediated by the RhoA/Rho kinase pathway, Gαq induces actin disruption independent of the PLC pathway. We also demonstrate that Gαq and Gα13 are implicated in CCK-induced amylase secretion, indicating that both heterotrimeric G proteins have a functional relevance in the acinar secretory process. Finally, we show that all CCK-induced intracellular events through Gα13 activation are Ca2+-independent, since in p115-RGS-expressing acini, CCK still induced Ca2+ mobilization.

The participation of Gαi in RhoA and Rac1 activation is seen in some, but not all, systems. For example, in rabbit gastric smooth muscle, Gαi-coupled receptors do not activate RhoA (31); in alveolar epithelial cells, however, Gαi-coupled receptors activate RhoA (24). In other cells, such as human embryonic kidney cells and neonatal rat cardiomyocyte-derived H10 cells, Gαi-coupled receptors are able to activate RhoA and Rac1 (44). In pancreatic acini, although Gαi is activated by CCK (38), Gαi is not required for CCK-induced RhoA or Rac1 activation, since pretreatment with pertussis toxin did not prevent the response to CCK. The lack of participation of Gαi is expected, because RhoA and Rac1 have been involved in a number of CCK-elicited events, such as amylase secretion and cytoskeletal reorganization, which are not mediated by Gαi activation (1, 46).

In pancreatic acini from the rat, unlike the mouse, both members of the G12/13 family have been identified, but only Gα13 was shown to form a complex with the RhoGEF Vav-2 and RhoA (20), supporting the concept that, in pancreatic acini, Gα13 activates RhoA signaling.

Interestingly, CCK-stimulated Gα13 signaling is able to activate not only RhoA but also Rac1. These findings suggest the existence of a cross talk between Rac1 and RhoA in pancreatic acini. Several reports show that, in some tissues, Rac1 acts upstream of RhoA during actin cytoskeletal reorganization (10), whereas in other tissues, RhoA acts upstream of Rac1 (50). Alternatively, RhoA and Rac1 may be acting in parallel, as previously shown (54). However, the exact mechanism by which Gα13 activates RhoA and Rac1 remains unknown.

Several reports indicate that the α-subunits of G12, G13, and Gq are able to activate Rho proteins in different cell types and that the participation of one or another Gα protein depends on the type of RhoGEF found in the cell (34). Although there are a number of RhoGEFs, only p115-RhoGEF is directly activated through interaction with Gα13, but not Gα12 (15, 21). In mouse pancreatic acini, mRNA coding for p115-RhoGEF has been found by RT-PCR (Y. Bi and J. A. Williams, unpublished data). Our finding does not mean that RhoA in mouse pancreatic acini is activated through p115-RhoGEF only, since other RhoGEFs, including leukemia-associated RhoGEF and PDZ-RhoGEF, can also be activated by Gα13 (47); only leukemia-associated RhoGEF has also been identified in mouse pancreatic acini by RT-PCR (Y. Bi and J. A. Williams, unpublished data).

The actin cytoskeletal reorganization induced by CCK was mimicked by transfection of the constitutively active Gα13 and Gαq, which induced disruption of the actin filaments under the apical membrane and enhanced actin staining at the basolateral membrane, as well as formation of basolateral membrane protrusions (blebbing). However, only expression of the Gαq inhibitor RGS-2 inhibited CCK-induced actin reorganization and bleb formation; p115-RGS had little effect on either of these CCK-mediated events. These results indicate that although Gαq and Gα13 can contribute to regulation of actin cytoskeletal reorganization, Gαq is the heterotrimeric G protein mainly involved in CCK-induced actin cytoskeletal disruption and acinar morphological changes. These differences may be related to the differences in affinity between the CCK receptor and the heterotrimeric G protein. Furthermore, these findings could explain why bleb formation, actin reorganization, and a supramaximal inhibition of secretion are induced by secretagogues that activate Gαq, at high concentrations, and not by those that activate Gαs activation, such as VIP (1). In support of these findings, in the present study, we showed that RhoA and Rac1 activation, which are implicated in actin cytoskeletal reorganization in pancreatic acini (3), are independent of signaling pathways mediated by Gαs.

Rearrangement of the cytoskeleton is believed to be involved in CCK-induced submaximal secretory response at high CCK concentrations (49). In particular, cytochalasin B, which disrupts the cytoplasmic network of actin microfilaments and inhibits blebbing, converts the CCK concentration-response curve for amylase secretion from its biphasic to monophasic shape and, at high CCK concentrations, tends to increase, rather than decrease, amylase secretion (7). Like cytochalasin B, the expression of RGS-2 or p115-RGS led to a monophasic shape of CCK-induced amylase secretion-concentration response curve and tended to enhance amylase secretion at high CCK concentrations. Although the expression of p115-RGS did not affect CCK-induced actin cytoskeletal reorganization and bleb formation as much as the expression of RGS-2, the expression of p115-RGS decreased basal amylase release. This is similar to the simultaneous expression of dominant-negative RhoA and Rac1 (4), supporting the notion that Gα13 activates RhoA and Rac1. Moreover, the effects of Gα13 on amylase secretion were Ca2+-independent, since expression of p115-RGS did not modify CCK-induced Ca2+ mobilization, although it inhibited CCK-induced amylase secretion.

In summary, these findings indicate that the heterotrimeric G proteins Gq and G13 participate in activation of the small GTP-binding proteins RhoA and Rac1: Gαq activates Rac1 in a PLC pathway-independent manner, and Gα13 activates RhoA and Rac1. Moreover, Gαq and Gα13 are required for CCK-induced actin cytoskeletal reorganization, bleb formation, and amylase secretion: Gαq is implicated in actin cytoskeletal reorganization, and both Gαq and Gα13 are involved in pancreatic amylase secretion.

GRANTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-41122 (to J. A. Williams) and Cell Biology and Cell Imaging Core of the Michigan Gastrointestinal Peptide Center (Grant P30 DK-34933) and the Michigan Diabetes Research and Training Center supported by Grant P60 DK-20572.

DISCLOSURES

No conflicts of interest are declared by the author(s).

ACKNOWLEDGMENTS

We thank N. Vogel and B. Nelson for excellent technical assistance.

Footnotes

  • * Y. Bi should be considered as co-first author.

REFERENCES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
View Abstract