The Role of Protein Kinase D in Neurotensin Secretion Mediated by Protein Kinase C-α/-δ and Rho/Rho Kinase*

Neurotensin (NT) is a gut peptide that plays an important role in gastrointestinal (GI) secretion, motility, and growth as well as the proliferation of NT receptor positive cancers. Secretion of NT is regulated by phorbol ester-sensitive protein kinase C (PKC) isoforms-α and -δ and may involve protein kinase D (PKD). The purpose of our present study was: (i) to define the role of PKD in NT release from BON endocrine cells and (ii) to delineate the upstream signaling mechanisms mediating this effect. Here, we demonstrate that small interfering RNA (siRNA) targeted against PKD dramatically inhibited both basal and PMA-stimulated NT secretion; NT release is significantly increased by overexpression of PKD. PKC-α and -δ siRNA attenuated PKD activity, whereas overexpression of PKC-α and -δ enhanced PKD activity. Rho kinase (ROK) siRNA significantly inhibited NT secretion, whereas overexpression of ROKα effectively increased NT release. Rho protein inhibitor C3 dramatically inhibited both NT secretion and PKD activity. In conclusion, our results demonstrate that PKD activation plays a central role in NT peptide secretion; upstream regulators of PKD include PKC-α and -δ and Rho/ROK. Importantly, our results identify novel signaling pathways, which culminate in gut peptide release.

Regulatory hormones, localized to specialized endocrine cells in the small bowel, control numerous physiological functions of the gastrointestinal (GI) tract including secretion, motility, and mucosal growth (1). The gut peptide neurotensin (NT), 1 a tridecapeptide localized to enteroendocrine cells (N cells) of the distal small bowel (2,3), facilitates fatty acid translocation (4), affects gut motility (5), and stimulates growth of normal gut mucosa (6,7). In addition to its trophic effects on normal GI tissues, NT stimulates proliferation of certain pancreatic, colonic, and prostatic cancers bearing NT receptors (NTR) (8). Although the mechanisms for pancreatic hormone release have been well characterized (9), the signal transduction pathways regulating stimuli-induced gut hormone secretion are not entirely understood. One reason for this paucity in our understanding is the relative lack of useful in vitro models that recapitulate in vivo properties of intestinal endocrine cells.
The BON endocrine cell line was established from a human pancreatic carcinoid tumor and characterized in our laboratory (10). These cells have served as an invaluable in vitro model for hormone secretion studies. Similar to the terminally differentiated N cell of the small bowel, BON cells express high levels of NT/neuromedin N mRNA, synthesize and secrete NT peptide, and process the NT/neuromedin N precursor protein in a fashion identical to that of the normal intestine (11). BON cells exhibit morphological and biochemical characteristics consistent with the enteroendocrine cell phenotype, including the presence of numerous dense core granules and the expression and secretion of chromogranin A and other peptides (e.g. pancreastatin) (10,12,13). Thus, the BON cell line provides an excellent model to delineate the mechanisms underlying gut peptide secretion.
We have shown that protein kinase C (PKC), particularly isoforms PKC-␣ and -␦, plays a role in the stimulated release of NT (13)(14)(15). The PKCs comprise a family of intracellular serine/threonine-specific kinases that are, depending on the isoform, typically activated by Ca 2ϩ , lipid second messengers or protein activators and mediate the effects of a wide range of physiological stimuli, including growth factors, hormones, and neurotransmitters (16,17). Protein kinase D (PKD), originally referred to as PKC-, is a serine/threonine protein kinase with unique structural, enzymological, and regulatory properties that are different from those of the PKC family members (18,19). The most distinct characteristics of PKD are the presence of a catalytic domain distantly related to Ca 2ϩ -regulated kinases, a pleckstrin homology (PH) domain within the regulatory region, and a highly hydrophobic stretch of amino acids in its N-terminal region (18,19). Expression of PKD has been demonstrated in some endocrine cells, including insulin-and gastrin-secreting cells (20 -22). PKD activation occurs through several mediators via a PKC-dependent pathway (23)(24)(25)(26).
Recent studies indicate a close association between PKC isoforms and members of the Rho family of small GTP-binding proteins including Rho (A, B, and C), Rac1, and Cdc42 (27,28). The Rho family members, well-known regulators of the actin cytoskeleton and phosphoinositide metabolism (29), have been implicated in hormone secretion from endocrine cells (30 -32). Rho kinases (ROK) are the first effectors of Rho to be discovered and, to date, two ROK isoforms have been identified: ROK␣ and ROK␤ (33). There are now several examples where Rho GTPases participate in regulated secretory pathways, such as in mast cells (34), PC12 cells (35), and neurons (36). duplicate samples as described previously (42,43).
Real Time Confocal Microscopy-BON cells, transiently expressing GFP-tagged PKD, were cultured in 25-mm round coverslips in 6-well plates and imaged in real time before and after PMA treatment. Cells were placed inside a prewarmed (37°C) chamber on the stage of an LSM 510 META confocal system configured with an Axiovert 200M inverted microscope (Zeiss, Jena, Germany). GFP fluorescence images were acquired using a plan-aprochromat 63ϫ, 1.4 NA oil immersion objectives and the 488 nm line of an argon ion laser for excitation. The image acquisition and processing was carried out using the Zeiss LSM510 workstation (v 3.0) and the Zeiss Image Browser (v3.1) software.
Protein Preparation and Western Blotting-Protein preparation and Western blotting were performed as described previously (14). In brief, equal amounts of protein were resolved on Novex Tris-glycine or Nu-PAGE Bis-Tris gels (Invitrogen) and electrophoretically transferred to polyvinylidene difluoride membranes; the membranes were incubated with primary antibodies overnight at 4°C followed by secondary antibodies conjugated with horseradish peroxidase. Membranes were developed using the ECL detection system.
Immunoprecipitation and in Vitro Kinase Assays-Immunoprecipitation and in vitro kinase assays were performed as described previously (14). In brief, proteins (50 g) were incubated with PKD or GFP antibodies (1:50) on a shaker for 2 h at 4°C followed by another 2 h incubation with 30 l of protein A-Sepharose beads at 4°C. The immunocomplexes were suspended in 20 l of kinase buffer and kinase reaction, with or without 2.5 g of syntide-2 as a substrate, was started by adding 5 Ci of [␥-32 P]ATP and incubated for 10 min at 30°C. Reactions were stopped by the addition of 2ϫ Tris-glycine sample buffer. Samples were denatured by boiling for 5 min and separated by NuPAGE 4 -12% Bis-Tris gels. Gels were incubated in Gel-Dry drying solution (Invitrogen) for 5 min and dried at 60°C for 60 min followed by exposure to x-ray film.
RhoA Activity Assay-RhoA activity was assessed using the Rho binding domain of Rhotekin as described (44,45). In brief, cells (3 ϫ 10 6 ) were transfected with 50 g of GST-C3 or GST protein and plated onto 60-mm dishes in growth medium, allowed to adhere and then serumstarved in secretion medium overnight. Cells were treated with PMA (10 nM) for 10 min and then extracted with radioimmune precipitation assay buffer. After centrifugation at 14,000 ϫ g for 2 min, the extracts were incubated for 30 min at 4°C with glutathione beads (Amersham Biosciences) coupled with bacterially expressed GST-RBD (Rho binding domain of Rhotekin) fusion protein (44), and then washed three times. The RhoA content was determined by immunoblotting samples using rabbit anti-RhoA antibody.
Statistical Analysis-All experiments were repeated at least two times and data are reported as means Ϯ S.E. Data were analyzed using analysis of variance for a two-factor factorial experiment (Figs. 1B, 2A, 6, A-C, and 8C) or a one-factor experiment (Fig. 6D). Fisher's least significant difference procedure was used for multiple comparisons with Bonferroni adjustment for the number of comparisons. A p value Ͻ 0.05 was considered significant.

RESULTS
siRNA Directed Against PKD Inhibits NT Secretion and PKD Phosphorylation in BON Cells-Based on findings that both PMA and bryostatin1 induce the activation of endogenous PKD in BON cells (14), we determined whether PKD is involved in PMA-mediated NT secretion. We took advantage of the RNA interference (RNAi) to selectively reduce PKD expression (Fig.  1). BON cells were transfected with the PKD siRNA vector (pSUPER.PKD siRNA) and the control vector (pSUPER). Cells were collected at 48 or 96 h after transfection, and lysates immunoblotted to assess expression of endogenous PKD (Fig.  1A, top). Compared with cells transfected with the control siRNA (pSUPER), transfection of PKD siRNA significantly reduced PKD levels. The blot was stripped and reprobed with ␤-actin demonstrating equal loading. Moreover, the PMA-activated phosphorylation of PKD was reduced significantly by PKD siRNA compared with the PMA-treated control vector (Fig. 1A, bottom).
Next, BON cells were transfected with PKD siRNA or the control vector (pSUPER) and treated with PMA; NT secretion was assayed by RIA (Fig. 1B). Transfection with siRNA directed against PKD decreased both basal and PMA-stimulated NT secretion from BON cells compared with the control vector. Taken together, our results, using complementary techniques, strongly demonstrate an important role for PKD in NT secretion from BON cells.
Overexpression of Wild-type and Constitutively Active PKD Increases PMA-mediated NT Secretion-To further confirm the regulation of PKD on NT secretion, BON cells were transiently transfected with PKD WT , or the constitutively active PKD ⌬PH , both linked to GFP, or the empty vector (pEGFP-N-1), as a control (Fig. 2). 24 h after, the cells were serum-starved, then treated with vehicle (Me 2 SO) or PMA (10 nM) for 30 min, and the medium collected for measurement of NT by RIA ( Fig. 2A).
In the absence of PMA treatment, overexpression of PKD ⌬PH increased NT release compared with the empty vector (pEGFP-N-1) and PKD WT . The PH domain exerts an inhibitory effect on the catalytic domain of PKD (46); therefore, the increase in basal NT release suggests that the PH domain plays an inhibitory role in the regulation of its enzymatic activity. NT secretion was significantly increased by PMA treatment of BON cells transfected with the empty vector. Importantly, PMA treatment of cells transfected with either PKD WT or PKD ⌬PH resulted in a significantly enhanced NT secretion compared with PMA treatment of BON cells transfected with the empty vector, but the increase in secretion from BON cells transfected with PKD ⌬PH was lower than NT secretion noted from PKD WT transfected BON cells.
Overexpression of PKD plasmids was confirmed by immunoblot analysis for GFP tag (Fig. 2B). GFP-tagged PKD was evenly distributed in cells transfected with PKD WT or PKD ⌬PH , either in the presence or absence of PMA, but not in the control vector (upper panel). The membrane was reprobed with an antibody that recognizes phosphorylated serine 744/748 (middle panel). Phosphorylation of exogenous PKD was not detected in BON cells transfected with the PKD WT , in the absence of PMA, but was induced in BON cells transfected with either the PKD WT and PKD ⌬PH treated with PMA. These data indicate that the response of exogenous PKD to PMA is similar to endogenous PKD. A weak signal was detected in BON cells transfected with PKD ⌬PH , in the absence of PMA treatment, demonstrating the constitutively active status of PKD with the PH domain deleted. This observation is in agreement with the increase of NT secretion found in BON cells transfected with PKD ⌬PH without PMA stimulation ( Fig. 2A). Similar levels of endogenous (endo) phosphorylated PKD (ϳ115 kDa) were detected in all BON cells treated with PMA, showing that stimulation of endogenous PKD by PMA was not affected by exogenous PKD. Total endogenous PKD was reprobed as a loading control (Fig. 2B, bottom panel). Taken together, these results further support the findings that activation of overexpressed PKD enhances PMA-mediated NT secretion.
To confirm the above results of PKD phosphorylation, we directly assayed the activity of PKD using the same cell lysates as above by in vitro kinase assays of GFP immunoprecipitates (Fig. 2C). PKD activity in immunocomplexes was determined by phosphorylation of syntide-2, a synthetic peptide reported to be an excellent substrate for PKD (47) (lower bands), as well as by autophosphorylation (upper bands). Extracts from BON cells transfected with the empty vector showed a basal level of syntide-2 (with or without stimulation); PKD WT activity was increased in the absence of PMA compared with the basal level of syntide-2, suggesting the existence of weak kinase activity in the wild-type PKD even without PMA stimulation. The phosphorylation of syntide-2 was enhanced by PMA in BON cells transfected with PKD WT , further demonstrating the stimulation of PKD by PMA. BON cells expressing constitutively active PKD exhibited a similar level of syntide-2 phosphorylation as FIG. 3. PMA induces translocation of wild-type PKD and constitutively active PKD in BON cells. BON cells were transiently transfected with GFP-tagged wild-type PKD, constitutively active PKD or the empty vector and allowed to recover for 24 h. Cells were then serum-starved in secretion medium overnight. BON cells were incubated in PBS, pH 7.4, and placed inside a prewarmed (37°C) chamber and imaged in real time as described under "Experimental Procedures." Top, PKD WT in quiescent BON cells was distributed evenly throughout the cytosol of these cells with no apparent association with specific intracellular compartments. PKD WT rapidly translocated to the plasma membrane when PMA was added. Middle, the constitutively active PKD ⌬PH localized to the cytosol as well as the nucleus. Moreover, PKD ⌬PH was noted to translocate from the cytosol to the membrane by PMA stimulation. The PKD ⌬PH in the nucleus failed to translocate to the membrane. Bottom, compared with wild-type PKD, the GFP control vector was noted throughout the cytosol and the nucleus. Translocation was not found in the cells transfected with the empty vector after PMA stimulation. A representative result from three experiments is shown. noted with PKD WT . With PMA stimulation, the activity of both PKD WT and PKD ⌬PH was increased with PKD WT being more pronounced. These results are consistent with the findings of NT release following transfection with these constructs (Fig.  2A). Increased phosphorylation of syntide-2 was not noted in BON cells transfected with PKD ⌬PH compared with BON cells transfected with PKD WT in the presence of PMA. This is consistent with the result in Fig. 2A in which NT secretion is lower in BON cells transfected with PKD ⌬PH than BON cells transfected with PKD WT in the presence of PMA but inconsistent with the result in Fig. 2B in which PKD phosphorylation is much higher in BON cells transfected with PKD ⌬PH than BON cells transfected with PKD WT . These results may be explained by different reaction conditions in the kinase assay in vitro compared with the in vivo conditions. PMA Induces Translocation of PKD WT and PKD ⌬PH in BON Cells-Translocation is considered to be a marker of PKC activation (48). To further confirm the role of PKD on NT secretion, translocation of PKD tagged with GFP was visualized in BON cells using time-lapse confocal microscopy (Fig. 3). In-tense fluorescence was observed in BON cells transfected with PKD WT (top panel). The addition of PMA (100 nM) induced a rapid translocation of PKD WT from the cytosol to the plasma membrane within 1 min after stimulation. These results are consistent with our report that endogenous PKD translocated to the membrane from the cytosol after PMA stimulation in fixed cells (14). In quiescent BON cells transfected with PKD ⌬PH , the fluorescence was present in the cytosol with intense fluorescence in the nuclei (middle panel). The addition of PMA induced a rapid translocation of PKD ⌬PH from the cytosol to the plasma membrane; the intense fluorescence noted in the nuclei persisted after PMA treatment. In addition, vesicle-like structures were observed close to the membrane in BON cells transfected with either PKD WT or PKD ⌬PH (Supplementary video of Fig. 3, middle panel) after addition of PMA. As a control, fluorescence in BON cells transfected with empty vector was observed before and after the addition of PMA (bottom panel); no changes in GFP localization were noted after PMA treatment. These results demonstrate the activation of PKD by PMA and further support the role of PKD in NT secretion.

Classic or Novel PKC and Rho Kinase Inhibitors Attenuate
PMA-mediated PKD Activation-PKD is activated in a PKCdependent fashion in some cell types (23)(24)(25)(26). To determine whether PKC isoforms are involved in PMA-induced PKD activation in BON cells, we examined the effect of three PKC inhibitors, Gö6983, GFX, and Ro31-8220, which inhibit activity of classic and novel PKC isoforms but not PKD (49,50) (Fig.  4). BON cells were pretreated with these inhibitors (1 M for each) for 30 min prior to a 10-min exposure to PMA (10 nM). As shown in Fig. 4A, Gö6983, GFX, and Ro31-8220 completely blocked PKD transphosphorylation at Ser 744/748 (upper panel) and significantly attenuated PKD autophosphorylation at Ser 916 (middle panel), suggesting the requirement of upstream PKC isoforms in the activation of PKD. We examined whether ROK inhibitors, Y27632 and HA1077 (both at a concentration of 15 M), affect the activation of PKD. Both inhibitors blocked significantly attenuated PKD phosphorylation with the most pronounced inhibition noted with the Y27632 compound. These findings demonstrate the involvement of upstream ROK in the activation of PKD. Total PKD was probed to assess loading equality (bottom panel).
To further confirm these results, in vitro kinase assays were performed (Fig. 4B). The same cell extracts above were immunoprecipitated with PKD antibody, and the catalytic activity was then assayed with syntide-2 as an exogenous substrate (47). PKD activity was increased by PMA stimulation compared with the activity without PMA treatment. Activation of PKD was attenuated by all three PKC inhibitors as well as the two ROK inhibitors, particularly the Y27632 compound, which is consistent with the results shown in Fig. 4A. Taken together, these results suggest that PKC and Rho kinase are required for PMA-induced PKD activation in BON cells.
PKD Activation Is PKC-␣-and -␦-dependent-We have reported that PMA-mediated translocation of PKC-␣ and -␦ from the cytosol to the membrane associated with PMA-mediated NT secretion (14), suggesting that PKC-␣ and -␦ might act as upstream kinases for PKD. We first examined whether PKC-␣ contributed to PMA-induced PKD activation by overexpression of wild-type PKC-␣ (Fig. 5A). Phosphorylation of PKD was up-regulated in BON cells overexpressing wild-type PKC-␣ compared with the control vector in the presence of PMA (top). The membrane was stripped and reprobed with PKD antibody to monitor the loading equality (middle). The membrane was reprobed with anti-HA antibody to show the overexpression of PKC-␣ (bottom). Regulation of PKC-␣-mediated PKD phosphorylation was further confirmed using PKC-␣ siRNA (Fig. 5B). BON cells were transfected with the PKC-␣ siRNA and the control siRNA; cells were lysed at 48 h after transfection. Western blotting using anti-PKC-␣ antibody was performed to confirm PKC-␣ inhibition (top). The blot was reprobed with anti-PKD (Ser 744/748 ) antibody; PKD phosphorylation was markedly reduced by PKC-␣ siRNA (middle). The membrane was stripped and reprobed with PKD antibody to monitor the loading equality (bottom).
Similar experiments were performed to determine the regulation of PKC-␦ on PKD phosphorylation (Fig. 5C). Overexpression of wild-type PKC-␦ dramatically up-regulated PKD phosphorylation compared with the control vector in the presence of PMA (top). The loading equality was controlled by probing for total PKD expression using the same membrane (middle). The membrane was reblotted with anti-HA antibody to confirm overexpression of PKC-␦ (bottom). Treatment with PKC-␦ siRNA, as shown by Fig. 5D, blocked PKC-␦ expression (top) as well as PKD phosphorylation (middle). Total PKD expression was assessed as the loading control (bottom). This result further confirmed the upstream regulation of PKC-␦ on PKD activity.

Rho/ROK Pathway Contributes to PMA-mediated NT Secretion-Based on the finding that PKD activation is decreased by
ROK-specific inhibitors, we examined whether Rho/ROK pathway is involved in PMA-mediated NT secretion. BON cells were transfected with Clostridium botulinum C3 toxin, which specifically ADP-ribosylates Rho and impairs its function (51), and NT secretion was measured by RIA. C3 toxin significantly decreased PMA-stimulated NT secretion compared with control GST in the presence of PMA (Fig. 6A). To further confirm the involvement of Rho/ROK pathway, the role of ROK␣ (downstream effector of Rho proteins) on NT secretion was examined using ROK␣ siRNA (Fig. 6B). PMA-mediated NT secretion was decreased in BON cells transfected with ROK␣ siRNA compared with the control siRNA transfected BON cells. To extend these findings on the role of ROK␣ in PMA-mediated NT secretion, HA-tagged ROK␣, including full-length (ROK␣ F/L ), constitutively active (ROK␣ 1-543 ), and control vector (pXJ40) were transfected into BON cells (Fig. 6C). The expression level was first examined by Western blotting using anti-HA antibody (top). Expression of ROK␣ F/L (ϳ180 kDa) and ROK␣ 1-543 (ϳ60 kDa) was detected in BON cells. As expected, no signal was detected in BON cells transfected with the control vector. NT secretion was measured in BON cells transfected with ROK␣ F/L , ROK␣ 1-543 , and the control vector pXJ40 (bottom). Overexpression of ROK␣ F/L , but not ROK␣ 1-543, significantly increased NT secretion compared with the control vector in the presence of 10 nM of PMA. The role of ROK mediating PMAstimulated NT secretion was further supported using ROK inhibitors (Fig. 6D). BON cells were pretreated with varying dosages of Y27632 and HA1077 for 30 min and then treated with PMA (10 nM). Treatment with Y27632 decreased PMAmediated NT secretion in a dose-dependent fashion. Compared with Y27632, HA1077 was less effective in blocking PMAmediated NT secretion; significant inhibition was noted only at a concentration of 15 M which correlates with the effects on PKD phosphorylation as shown in Fig. 4A. These results strongly support a role for Rho/ROK in PMA-mediated NT secretion in BON cells.
Rho Proteins Regulate PKD Phosphorylation-To determine whether PKD activation is induced by signaling pathway(s) initiated by Rho proteins, BON cells were transfected with C3 PKD siRNA Decreases BBS-stimulated NT Secretion in BON/ GRPR Cells-GRP or its amphibian equivalent BBS stimulate the release of all intestinal hormones, including NT (52). To determine whether the effects noted with PMA treatment also occurred with physiologic agents known to stimulate NT release, BON cells stably transfected with the human GRPR tagged with GFP were established. Increased expression of GRPR was noted on membranes in BON cells transfected with GRPR-GFP (Fig. 8A). To determine the role of PKD in BBSstimulated NT secretion, PKD siRNA was transfected into BON/GRPR-GFP cells (Fig. 8B). 48 h after transfection, cells were treated with either vehicle control (Me 2 SO) or various concentrations of BBS. Inhibition of PKD expression was noted in BON/GRPR-GFP cells transfected with PKD siRNA compared with PKD expression in BON/GRPR-GFP cells transfected with control siRNA (Fig. 8B, upper panel). Treatment with BBS induced a dose-dependent increase in PKD Ser 744/748 phosphorylation in BON/GRPR-GFP cells transfected with control siRNA; PKD Ser 744/748 phosphorylation was significantly attenuated in BON/GRPR-GFP cells transfected with PKD siRNA compared with the control siRNA transfected BON/ GRPR-GFP cells in the presence of BBS (Fig. 8B, middle  panel). PKD Ser 744/748 phosphorylation was not detected in the parental BON cells treated with BBS. Blots were reprobed for ␤-actin as a loading control (Fig. 8B, lower panel). NT secretion was increased with BBS stimulation in a dose-dependent fashion in BON/GRPR-GFP cells transfected with control siRNA. BBS-stimulated NT secretion was significantly reduced in BON/GRPR-GFP cells transfected with PKD siRNA compared with that in BON/GRPR-GFP cells transfected with control siRNA either in the presence or absence of BBS (Fig. 8C). Similar results were obtained using another BON/GRPR-GFP cell clone (data not shown). Collectively, these findings demonstrate that PKD plays a role in NT secretion induced by BBS, a physiologic stimulant of NT release, suggesting that PKD may be a common regulator of stimulus-induced NT secretion. DISCUSSION The novel PKD protein is expressed in certain endocrine cells (14, 20 -22); however, to date, there has been little evidence to suggest a role for PKD in hormone secretion. The findings in our present study are the first to identify a role for PKD in stimulated gut peptide release. Either PMA or bryostatin1, which are potent activators of PKC, resulted in the activation and cellular redistribution of PKD. Silencing PKD expression by RNA interference resulted in decreased basal and PMA-stimulated NT secretion. Overexpression of both wild-type and constitutively active PKD increased PMA-mediated NT secretion from BON cells. In addition, GRPR-expressing BON cells demonstrated increased PKD activation and NT secretion in response to the ligand BBS, thus further supporting a role for PKD in the physiological release of NT. Therefore, our current study, utilizing a combinatorial approach (i.e. specific inhibition and overexpression of PKD) demonstrates a critical role for PKD in the signal transduction pathway leading to NT peptide release.
PKD has been implicated in the organization of the Golgi apparatus, regulating the fission of vesicles from the trans-Golgi network (53). PKD is a resident protein of the Golgi compartment in certain cell types (22,53,54) where it is involved in constitutive transport processes. In our present study, we found that PKD WT -GFP was evenly distributed in the cytosol of untreated BON cells with no discernable Golgi localization. In the presence of PMA, translocation of PKD WT -GFP was noted to occur from the cytosol to the membrane. Consistent with our findings, Matthews et al. (55) noted that PKD was localized to membranes surrounding cytoplasmic granules and not in the Golgi of RBL 2H3 mast cells after treatment with phorbol ester. This report, in combination with our present study, suggests a role for PKD in stimulated peptide secretion that is distinct from its role in Golgi transport processes.
PKD activation can be mediated by a PKC-dependent signal transduction pathway that involves the phosphorylation of Ser 744/748 of PKD within the activation loop of the catalytic domain of PKD and occurs by diverse stimuli including neuropeptides and growth factors (56) in non-endocrine cells. For example, activation of PKD can occur through activation of PKC-⑀ in HEK393 cells (22), PKC-in COS-7 cells (26), and PKC-in T cells (25). We demonstrate PKD phosphorylation was inhibited in the novel human endocrine cell line BON by pretreatment with PKC inhibitors and by PKC-␣ and -␦ siRNA and enhanced by overexpression of wild-type PKC-␣ and -␦. Therefore, our findings indicate that downstream activation of PKD is PKC-dependent, particularly PKC-␣ and -␦, in the BON endocrine cell line. Similar to our results, Tan et al. (23) demonstrated PKC-␦-mediated PKD activation was stimulated by thrombin in aortic smooth muscle cells. Recently, Storz et al. (57) reported that PKC-␦ participates in the activation of PKD in HeLa cells exposed to oxidative stress. To date, although it has been shown that PKD2 is activated via PKC-␣ through the CCK B /gastrin receptor in human gastric cancer cells (58), no direct interaction between PKC-␣ and PKD has been reported. PKD appears to be a scaffold protein and is an important regulator of diverse intracellular signaling pathways (19); therefore, it is possible that PKD forms a complex with PKC-␣ or PKC-␦ or other proteins mediating NT secretion in BON cells. However, the exact relationship between PKC-␣ and PKC-␦ with PKD will require further investigation.
The Rho family and their upstream or downstream factors have been implicated in various cellular functions such as regulation of actin filament reorganization and exocytosis (27,28). In mast and chromaffin cells, one or more members of the Rho family play an active role in the exocytotic process (27,28,59,60). In addition, Cdc42 and Rac control regulated secretion in pancreatic ␤-cells (61). ROK, an effector of Rho proteins but not Rac or Cdc42, is also involved in exocytosis (35,62). But the mechanisms how Rho family regulates the exocytotic process remains unknown. To our knowledge, our findings demonstrate, for the first time, that Rho/ROK pathway promoted PMA-mediated NT secretion as demonstrated by inhibition of PMA-stimulated NT secretion using the Rho inhibitor C. botulinum C3, ROK␣ siRNA, and ROK inhibitors Y27632 and HA1077 or, conversely, by the augmentation of NT secretion with the full-length ROK␣. We also demonstrate for the first time a reduction of PKD activity in intact BON endocrine cells by C3 toxin, dominant negative RhoA and the ROK inhibitors. BON cells express endogenous RhoA and RhoC but not RhoB, 2 therefore, our results do not exclude the possible involvement of RhoC since C3 toxin inhibits all three Rho proteins. However, taken together, our results provide strong evidence supporting the activation of Rho/ROK as a pathway leading to PKD activation, which is in agreement with studies demonstrating that activation of PKD occurs through upstream Rho stimulation in response to agonist activation of BBS receptors in COS-7 cells (63).
Increasingly, the interaction of Rho/ROK and PKC signaling pathways has been reported in a variety of cell types. For example, the interaction between PKC and Rho regulates enzyme secretion in pancreatic acini (64). In epithelial and endothelial cells, treatment with C3 toxin prevented PKC translocation and activation (65). Recently, Slater et al. (27,28) demonstrated that Rho-GTP potently stimulates PKC-␣ activity in vitro. In other studies, PKC has been shown to activate Rho (66,67). We found that treatment of BON cells with PMA stimulated PKC and Rho/ROK pathways, leading to PKD-regulated NT secretion. Rho has been predominantly described in the regulation of actin filament reorganization (27,28), and PKCs are also known to be important regulators of the cy-toskeleton (68). Therefore, it is possible that PKC-␣ and PKC-␦ and Rho/ROK-mediated PKD activation leads to NT secretion through the reorganization of actin filaments. The precise sequence of events culminating in stimulated NT release remains to be fully elucidated.
In conclusion, the results of this study identify the novel PKD protein as a critical regulator of PMA-or BBS-mediated NT secretion from the BON or BON/GRPR cell lines, respectively. We demonstrate that PKD phosphorylation is regulated by upstream PKC-␣ and -␦, and the Rho/ROK pathway ( Fig. 9 summarizes our proposed model for PKD-mediated NT secretion). Our findings identify novel signaling pathways contributing to stimulated peptide secretion from specialized endocrine cells of the GI tract. The coordinated release of intestinal peptides in response to extracellular mediators is essential for the regulation of intestinal digestion, secretion and motility; therefore, our results delineate critical signaling molecules that contribute to this important physiologic process. Interestingly, prominent regulators of PKD and NT secretion, namely PMA-sensitive PKCs (69) and RhoA (70) contribute to tumor progression. Since NT can stimulate proliferation of NTR positive tumors, it is interesting to speculate that pathways, such as RhoA and PKCs, may promote tumorigenesis through the dysregulated secretion of trophic factors such as NT.