The signaling pathways mediating lysophosphatidic acid (LPA)-stimulated PKD2 activation and the potential contribution of PKD2 in regulating LPA-induced interleukin 8 (IL-8) secretion in nontransformed, human colonic epithelial NCM460 cells were examined. Treatment of serum-deprived NCM460 cells with LPA led to a rapid and striking activation of PKD2, as measured by in vitro kinase assay and phosphorylation at the activation loop (Ser706/710) and autophosphorylation site (Ser876). PKD2 activation induced by LPA was abrogated by preincubation with selective PKC inhibitors GF-I and Ro-31-8220 in a dose-dependent manner. These inhibitors did not have any direct inhibitory effect on PKD2 activity. LPA induced a striking increase in IL-8 production and stimulated NF-κB activation, as measured by NF-κB-DNA binding, NF-κB-driven luciferase reporter activity, and IκBα phosphorylation. PKD2 gene silencing utilizing small interfering RNAs targeting distinct PKD2 sequences dramatically reduced LPA-stimulated NF-κB promoter activity and IL-8 production. PKD2 activation is a novel early event in the biological action of LPA and mediates LPA-stimulated IL-8 secretion in NCM460 cells through a NF-κB-dependent pathway. Our results demonstrate, for the first time, the involvement of a member of the PKD family in the production of IL-8, a potent proinflammatory chemokine, by epithelial cells.
- NCM460 cells
- protein kinase C
- phorbol esters
colonic epithelial cells are uniquely positioned to serve as a direct line of communication between the external environment and the immune system via subepithelial immune cells. These epithelial cells are constantly exposed to luminal contents, including ingested pathogens and commensal microflora that populate the gut. Intestinal epithelial cells have evolved mechanisms for sensing luminal antigens or ligands and can respond to external stimuli by inducing intracellular signaling events, leading to the production and basolateral secretion of inflammatory cytokines (13, 31). In addition, these cells also respond to signals generated in the internal environment. The colonic epithelial cells must then integrate and transduce internal and external signals into biological responses.
PKC, a major target for the tumor-promoting phorbol esters, has been implicated in the signal-transduction pathways that mediate important functions in intestinal epithelial cells, including proliferation (1, 14) and carcinogenesis (35). It is known that intestinal epithelial cells express multiple isoforms of the PKC family, including α, β, δ, ε, and ζ (11). Transgenic overexpression of PKC-βII in murine colonic epithelium (32) or overexpression of certain PKCs in intestinal epithelial cells in culture has been shown to promote growth (26). In addition to its effects on colonic epithelial cell proliferation, multiple PKCs have also been shown to stimulate NF-κB activation (34, 43, 55, 56) in various cell types and thus mediate cell survival as well as proinflammatory signaling (22, 61). Despite the critical role of PKCs in intestinal epithelial cell function and specifically, in initiating the pathway leading to NF-κB activation, downstream signaling targets of PKCs within the NF-κB pathway in intestinal epithelial cells remain poorly understood.
Protein kinase D (PKD; also known initially as PKCμ) has emerged as a major target in the signal-transduction pathways initiated by diacylglycerol and PKC in a variety of cell types (40). PKD is the founding member of a new family of serine/threonine protein kinases, including PKD, PKD2, and PKD3 (reviewed in Ref. 40). These kinases share similarities in overall structure, primary sequence, and enzymological properties (16, 39, 52). PKD has been implicated in the regulation of multiple important cell functions, including signal transduction via MAPK kinase pathways (5, 18, 19), chromatin remodeling via histone deacetylase phosphorylation (10, 20, 27), polarized Golgi function (6, 15, 25), NF-κB activation in response to oxidative stress (29, 47–50), and cell proliferation (46, 64). Although much less is known about the regulation and function of the other PKD isoforms, recent studies revealed interesting differences in the subcellular distribution (38, 39), embryonic expression (33), and regulation (59) of these isoforms. However, studies on the physiological functions of specific PKD isoforms have been hampered by the endogenous expression of more than one isoform in the same cell type and frequently involved the overexpression of wild-type and mutant PKDs in transient transfection systems. The identification of cell types that endogenously express a single member of the PKD family would be of great value to clarify the regulation and physiological function of that isoform.
In this study, we identified PKD2 as the predominant PKD isoform in nontransformed human colonic epithelial NCM460 cells. This provided a unique opportunity to examine the regulation and function of endogenous PKD2 without the potential contribution of the other PKD isoforms. Initially, we showed that stimulation of intact NCM460 cells with the G-protein-coupled receptor (GPCR) agonist lysophosphatidic acid (LPA), a major bioactive lipid of serum (3), induced a rapid and striking PKC-dependent PKD2 activation and promoted the production of the potent proinflammatory chemokine interleukin 8 (IL-8; designated CXCL8 in current chemokine nomenclature). Subsequently, we used small interfering RNA (siRNA)-mediated knockdown of PKD2 to determine the contribution of this PKD isoform in mediating NF-κB activation and IL-8/CXCL8 production in response to LPA in NCM460 cells. Our results demonstrate that endogenous PKD2 mediates a substantial component of LPA-induced IL-8/CXCL8 production in NCM460 cells.
MATERIALS AND METHODS
The nontransformed human colonic epithelial cell line NCM460 was obtained from INCELL (San Antonio, TX). Cultures of these cells were maintained as described in a humidified atmosphere containing 5% CO2 and 95% air at 37°C (30). NCM460 cells were used between passages 35 and 45 in this study. For experimental purposes, cells were plated in medium containing 10% FBS and were allowed to grow to confluency (5–7 days) and then changed to serum-free medium for 18–24 h before the experiment. IEC-18 cells (American Type Culture Collection) were maintained as previously described (9).
Western blot analysis for PKD isoforms and PKD2 phosphorylation.
Serum-starved cultures of NCM460 cells grown on plastic dishes were washed twice with DMEM and then treated as described in the individual experiments. The cells were lysed in 2× SDS-PAGE sample buffer. After SDS-PAGE, proteins were transferred to Immobilon-P membranes (Millipore) and blocked by 3- to 6-h incubation with 5% nonfat milk in PBS (pH 7.2). Membranes were then incubated overnight with the respective primary antibody. Bound primary antibodies to immunoreactive bands were visualized by enhanced chemiluminescence detection with horseradish peroxidase-conjugated anti-mouse or anti-rabbit antibodies. Autoradiograms were scanned with a GS-710 scanner (Bio-Rad), and the labeled bands were quantified with the Quantity One software program (Bio-Rad).
In vitro kinase assay of PKD2.
Cultures of NCM460 cells, treated as described in the individual experiments, were washed and lysed in 50 mM Tris·HCl (pH 7.6), 2 mM EGTA, 2 mM EDTA, 1 mM DTT, 10 μg/ml aprotinin, 10 μg/ml leupeptin, 1 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride hydrochloride, and 1% Triton X-100 (lysis buffer A). Cell lysates were clarified by centrifugation at 15,000 g for 10 min at 4°C. PKD2 was immunoprecipitated at 4°C for 2–4 h with PKD2 antibody (1:200). The immune complexes were recovered by protein A coupled to agarose.
PKD2 autophosphorylation was determined in an in vitro kinase assay by mixing 20 μl of PKD2 immunocomplexes with 10 μl of a phosphorylation mixture containing (final concentration) 100 μM [γ-32P]ATP (specific activity of 400–600 cpm/pmol), 30 mM Tris·HCl (pH 7.4), 10 mM MgCl2, and 1 mM DTT. After 10 min of incubation at 30°C, the reaction was stopped by washing with 200 μl of kinase buffer and then adding an equal volume of 2× SDS-PAGE sample buffer (200 mM Tris·HCl, pH 6.8, 2 mM EDTA, 0.1 M Na3VO4, 6% SDS, 10% glycerol, and 4% 2-mercaptoethanol), followed by SDS-PAGE analysis. The gels were dried, and the 105-kDa radioactive band corresponding to autophosphorylated PKD2 was visualized by autoradiography. Autoradiographs were scanned in a GS-710 calibrated imaging densitometer (Bio-Rad), and the labeled band was quantified by the Quantity One software program.
Exogenous substrate phosphorylation by purified PKD2 was carried out by mixing 20 μl of the washed immunocomplexes with 20 μl of a phosphorylation mixture containing 2.5 mg/ml syntide-2 (PLARTLSVAGLPGKK), a peptide based on phosphorylation site two of glycogen synthase. After 10 min of incubation at 30°C, the reaction was stopped by adding 100 μl of 75 mM H3PO4 and spotting 75 μl of the supernatant on P-81 phosphocellulose paper. Free [γ-32P]ATP was separated from the labeled substrate by washing the P-81 paper four times for 5 min in 75 mM H3PO4. The papers were dried, and the radioactivity incorporated into syntide-2 was determined by Cerenkov counting.
Nuclear extracts were prepared, and EMSA was performed according to the instructions of the manufacturer (Panomics, Redwood City, CA). The EMSA “Gel-Shift” kit (Panomics) was used. The oligonucleotide probes (biotin labeled and cold) encoding the consensus sequence of NF-κB (5′-AGTTGAGGGGACTTTCCCAGGC-3′) transcription factor was purchased from Panomics. EMSAs were performed by incubating biotin-labeled NF-κB probe with nuclear extracts (5 μg) from cells with or without LPA treatment in binding buffer containing nonspecific blocker poly(dI-dC) at 20°C for 30 min. The protein-DNA complexes were analyzed by electrophoresis through a 6% nondenaturing polyacrylamide gel. The protein-DNA complexes were then transferred onto a nylon membrane using a semi-dry method at 300 mA for 30 min. After transfer, the nylon membrane was subjected to a UV cross-linker oven for 3 min. The membrane was blocked, followed by addition of streptavidin-horseradish peroxidase conjugate (1:1,000). The shifted bands corresponding to the protein-DNA complex were visualized relative to the unbound double-strand DNA after exposure to film.
To inhibit PKD2 protein expression, subconfluent cultures of NCM460 cells were transfected with siRNAs. PKD2 siRNAs from Dharmacon (Chicago, IL) and from Ambion (Austin, TX) were used. The sequence for the siRNA oligonucleotides (sense strands) are as follows: UGAGACACCUUCACUUCAUU, CAAGAACAUUGUCCACUGUUU, and GGAAGAUGGGAGAGCGAUAUU. siCONTROL nontargeting siRNA pool (pool of 4 duplexes) was purchased from Dharmacon and used as a control. Transfection of duplex siRNAs was performed with Lipofectamine 2000 according to the manufacturer’s protocol. Forty-eight hours after transfection, cells were used for experiments and subsequent analyses.
Cells (4 × 105/well) were plated onto six-well dishes (Nunc). The next day, a transfection mixture was prepared by adding 1 μg of the NF-κB luciferase construct (pNF-κB-Luc plasmid; Stratagene, La Jolla, CA), 0.25 μg of control luciferase construct phRG-TK (Promega, Madison, WI), and either PKD2 siRNA (as indicated above) or siCONTROL nontargeting siRNA pool to serum-free medium. After a 20-min incubation and addition of Lipofectamine 2000 (Invitrogen, Carlsbad, CA), the transfection mixture was combined with 0.2 ml of complete DMEM and placed on the cells in each well.
Two days after transfection, the cells were treated with agonists. After the indicated times, the cells were washed with PBS and then lysed with passive lysis buffer (150 μl; Promega). The relative light units from firefly and Renilla luciferases (20 μl) were measured with the use of the dual-luciferase activity assay kit (100 μl; Promega) on a Turner TD20/20 luminometer. The ratio of firefly to Renilla luciferase activity was calculated. The results are reported as relative changes in the ratio, with a relative-fold increase of 1.0 signifying no change. The results are reported as means ± SE of three or more sets of transfections done in triplicate.
Serum-starved cultures of NCM460 cells were stimulated with LPA. To determine the amount of IL-8/CXCL8 protein secreted by the cells, conditioned medium was collected and centrifuged at 5,000 g for 5 min to remove cell debris. The collected samples were used for IL-8 ELISA (BD Pharmingen).
[γ-32P]ATP (370 MBq/ml) was from Amersham Pharmacia Biotech (Piscataway, NJ). Bisindolylmaleimide I (GF-109203X), bisindolylmaleimide V, AG-1478, LY-294002, PD-98059, U-0126, rapamycin, and Ro-31-8220 were purchased from Calbiochem. LPA, phorbol 12,13-dibutyrate (PDBu), and trypsin-EDTA solution (1×) were obtained from Sigma (St. Louis, MO). Phosphoserine 744/748 PKD antibody, anti-phospho-ERK1/2 (Thr202/Tyr204) MAb, and anti-phospho-p38MAPK (Thr180/Tyr182) MAb were obtained from Cell Signaling Technology (Beverly, MA). Purified PKD2 protein was obtained from Stressgen (Santa Diego, CA). TRIzol reagent, Lipofectamine 2000, and the SuperScript III one-step RT-PCR system with Platinum Taq DNA polymerase were obtained from Invitrogen. Anti-phospho-PKD2 (Ser876) was obtained from Upstate (Lake Placid, NY). PKD2 siRNA oligonucleotides and siCONTROL nontargeting siRNA pool were obtained from Dharmancon and Ambion. The PKD2 polyclonal antibody was used for immunoprecipitation, and Western blotting was generated against the synthetic peptide GLPTDRDLGGACPPQD, corresponding to amino acids 850–865 of PKD2 (by the Antibody Core of CURE: Digestive Diseases Research Center) or obtained from Upstate. Similar results were obtained with the PKD2 antibodies from both sources. The polyclonal anti-PKD3 N17 antibody was raised by immunizing rabbits with a peptide corresponding to the human PKD3 17 amino-terminal residues SANNSPPSAQKSVLPTA using standard procedures (39). Antibodies (PKD C-20, PKC-ζ C-20, PKC-ε C-15, PKC-δ, and PKC-α C-15; anti-ERK2 polyclonal antibody) used in Western blot analysis were obtained from Santa Cruz Biotechnologies (Santa Cruz, California). Caco-2, HT-29, and T84 cells were obtained from American Type Culture Collection. Other items were from standard suppliers or as indicated in the text.
Paired t-test was applied. All data are expressed as means ± SE. Significance at P < 0.05 and P < 0.01 was as indicated.
PKD2 is the only detectable member of the PKD family in NCM460 cells.
To determine which PKD isoform(s) is expressed in the nontransformed human colonic epithelial cell line NCM460, cultures of these cells were lysed and analyzed by SDS-PAGE, followed by Western blot analysis with a polyclonal antibody that recognizes the COOH-terminal region of both PKD and PKD2 (C-20; Santa Cruz Biotechnology) (52). Lysates from rat IEC-18 cells, known to express PKD and PKD2 (9, 38), were analyzed in parallel. Interestingly, NCM460 cells yielded only the faster migrating band compared with IEC-18 cells (Fig. 1A). Similarly, Western blot analysis of NCM460 cell lysates immunoprecipitated with the antibody that recognizes the COOH-terminal region of both PKD and PKD2 yielded only the 105-kDa band (results not shown). In addition to rat IEC-18 and NCM460 cells, we also tested three human colon adenocarcinoma cell lines (Caco-2, HT-29, and T84) and mucosal scrapings of mouse ileum and colon to determine whether PKD2 is also the predominant PKD family isoform in these colon epithelial cells. As shown in Fig. 1A, left, three of four human colonic epithelial cell lines tested, including NCM460 cells, and mouse intestinal lining express predominantly the faster migrating band, corresponding to the apparent molecular mass of PKD2. To further substantiate these results, Western blot analysis of IEC-18 and all four human colon cell line lysates was performed with the use of a different PKD2 antibody that was generated against amino acids 850–865 (GLPTDRDLGGACPPQD) of human PKD2 (either produced in our laboratory or available from Upstate; see materials and methods). This antibody is highly specific for human PKD2 and has weak reactivity against murine PKD2. As shown in Fig. 1A, right, the PKD2-specific antibody detected the faster migrating band of the doublet recognized by C-20 antibody in the cell lines. Together, these results indicate that human colonic epithelial NCM460 cells express predominantly the PKD2 isoform of the PKD family.
To substantiate further this conclusion, we immunoprecipitated cell lysates with the PKD2-specific polyclonal antibody and the immunocomplexes were then analyzed by SDS-PAGE, followed by Western blot analysis with the same PKD2 antibody. As shown in Fig. 1B, left lane, we found a well-defined band migrating with an apparent molecular mass of 105 kDa, corresponding to PKD2. The detection of this band was completely blocked by inclusion of the immunizing peptide during immunoprecipitation, confirming the specificity of the PKD2 antibody (Fig. 1B, right lane).
We next determined whether NCM460 cells endogenously express the PKD3 isoform of the PKD family. As shown in Fig. 1C, we did not detect any immunoreactive signal in lysates of these cells by Western blot analysis using a specific antibody directed against PKD3 (39). As positive controls, the same antibody detected ectopic human PKD3 protein in COS-7 cells transfected with green fluorescent protein-PKD3 and endogenous PKD3 protein in rat IEC-18 cells (Fig. 1C). These results indicate that PKD2 protein is the predominate isoform of the PKD family endogenously expressed by NCM460 cells. Thus these human epithelial cells provide a model system to study the regulation and function of PKD2 without significant contribution of the other PKD isoforms.
LPA stimulates PKD2 activation through a PKC-dependent pathway.
LPA is a physiological ligand that mediates its biological effects predominantly via members of the endothelial differentiation gene subfamily of GPCRs, namely LPA1, LPA2, and LPA3 (3). All three LPA receptor subtypes led to the generation of the second messengers inositol 1,4,5-trisphosphate, which stimulates Ca2+ mobilization from intracellular stores, and diacylglycerol, which activates classic and novel PKCs (3). To determine whether PKD2 is regulated through GPCR-mediated pathways, serum-starved NCM460 cells were treated with LPA (5 μM) for various times, lysed, and immunoprecipitated with PKD2 antibody. The immunocomplexes recovered with protein A agarose were then incubated with [γ-32P]ATP, and the incorporation of 32P into PKD2 was analyzed by SDS-PAGE and autoradiography to examine the level of autophosphorylation.
As shown in Fig. 2A, PKD2 isolated from unstimulated NCM460 cells had low catalytic activity. Treatment of NCM460 cells with LPA induced a rapid and striking increase in PKD2 kinase activity that was maintained during cell lysis and immunoprecipitation. PKD2 activation was detectable in 30 s and reached a maximum (∼10-fold) within 1 min of LPA stimulation (Fig. 2A, top).
PKD2 has two serine residues at 706/710 located within the activation loop of its catalytic domain. The transphosphorylation of these residues by PKCs is thought to stabilize the activation loop in an active conformation (40). Stimulation of NCM460 cells with LPA strikingly stimulated PKD2 phosphorylation at Ser706/710 and also at Ser876, an autophosphorylation site (51), in a time-dependent manner (Fig. 2A). These results indicate that stimulation of NCM460 cells with LPA induces rapid and striking catalytic activation and multisite phosphorylation of endogenous PKD2.
In many cases, PKD is activated through PKC-dependent activation loop phosphorylation, but PKC-independent pathways, including tyrosine phosphorylation of residues in the pleckstrin homology (PH) domain and other less defined pathways, have also been described (40). Much less is known about the mechanisms leading to the activation of PKD2 in normal epithelial cells. To determine whether PKCs mediate PKD2 activation induced by LPA in NCM460 cells, cultures of these cells were treated with various concentrations of the selective PKC inhibitors Ro-31-8220 (58) or GF-I (54) before LPA stimulation. Control cells received either an equivalent amount of solvent or GF-V, a biologically inactive analog of GF-I, before addition of LPA. As shown in Fig. 2B, treatment with either Ro-31-8220 or GF-I potently blocked PKD2 activation induced by subsequent addition of LPA. Because a previous report suggested that GF-I and Ro-31-8220 also inhibit PKD2 activity (51), we tested the effect of these inhibitors on the activity of purified PKD2 and PKC-ε using the exogenous substrate syntide-2. As shown in Fig. 2C, syntide-2 phosphorylation catalyzed by PKD2 was not inhibited by the addition of either Ro-31-8220 or GF-I to the incubation mixture, whereas syntide-2 phosphorylation by PKC-ε was completely blocked. These results indicate that Ro-31-8220 and GF-I interfere with LPA-induced PKD2 activation in intact NCM460 cells by inhibiting PKCs rather than PKD2.
A role of PKCs in PKD2 activation is further substantiated by experiments in which NCM460 cells were exposed to biologically active phorbol esters that directly activate classic and novel PKCs and subsequently, with long-term exposure, promote their downregulation. As shown in Fig. 2D, treatment with 100 nM PDBu induced a rapid and striking PKD2 phosphorylation that remains robust at 30 min. Conversely, downregulation of PKCs by exposure to PDBu for 24 h inhibited PKD2 phosphorylation at Ser876, elicited by subsequent stimulation with LPA (Fig. 2E). Collectively, these results indicate that LPA stimulates PKD2 activation through a PKC-dependent pathway.
LPA stimulates IL-8 secretion and NF-κB activity.
We next investigated whether LPA, which is increasingly implicated in immunoregulation (23, 41), could regulate a known and important biological function of colonic epithelial cells, namely, production of cytokines/chemokines. Initially, using a human cytokine microarray assay, we screened the expression profile of cytokines from serum-starved NCM460 cells. The membranes were hybridized with supernatant from LPA-treated (10 μM) for 24 h or vehicle-treated (control) NCM460 cells. The corresponding proteins were detected by a mixture of antibodies and visualized by an enhanced chemiluminescence system. Interestingly, of the 40 anti-cytokine antibodies on the array, only IL-8/CXCL8 was markedly increased after incubation with LPA for 24 h (Fig. 3A).
To further characterize the kinetics and magnitude of GPCR agonist-induced IL-8/CXCL8 production, we stimulated NCM460 cells with LPA (10 μM) for various times, and the levels of IL-8 in the medium were measured by ELISA. As shown in Fig. 3B, stimulation of NCM460 cells with 10 μM LPA induced a striking increase in IL-8/CXCL8 production in a time-dependent manner. The level of IL-8/CXCL8 increased by over 20-fold (1,444 ± 113 pg/ml) after 6 h of exposure to LPA.
The transcription factor NF-κB plays an important role in the regulation of cytokine/chemokine expression in a variety of cell types (17). Here, we tested whether treatment of NCM460 cells with LPA stimulates NF-κB activation by measuring NF-κB-DNA binding, NF-κB-driven luciferase reporter activity, and IκBα phosphorylation. As shown in Fig. 4A, stimulation of NCM460 cells with 10 μM LPA for 30 min induced NF-κB binding to oligonucleotide probes encoding the consensus sequence of NF-κB. Similarly, LPA increased NF-κB-driven luciferase reporter activity up to approximately fivefold at 4 h (Fig. 4B). Correspondingly, LPA induced IκBα phosphorylation, the prelude to NF-κB nuclear translocation and subsequent activation (Fig. 4C).
PKD2 knockdown decreases LPA-stimulated NF-κB activation and IL-8 secretion.
Given that LPA elicited a striking increase in IL-8/CXCL8 production, NF-κB promoter activity, and PKC-dependent PKD2 activation, we hypothesized that these events lie in a signal-transduction pathway, namely, PKC/PKD2/NF-κB/IL-8 production. To determine directly whether PKD2 mediates NF-κB activation in response to LPA, we decreased the endogenous levels of PKD2 using gene silencing by RNA interference. Subconfluent cultures of NCM460 cells were transiently transfected with PKD2 siRNA or nontargeted negative control duplex. To minimize the possibility that the siRNA oligonucleotide may be affecting the expression of a gene other than PKD2, NCM460 cells were transfected with siRNA targeting distinct regions of PKD2. As shown in Figs. 5A and 6B, the PKD2 protein level in NCM460 cells transfected with PKD2 siRNA was dramatically reduced (∼80%) compared with cells transfected with nontargeted negative control duplex. In contrast, PKC-α protein levels, determined as a control, were not affected, thus further demonstrating the selectivity of PKD2 gene silencing (Figs. 5A and 6B). PKD2 knockdown dramatically reduced LPA-induced IκBα phosphorylation (Fig. 5A), but it did not interfere with the phosphorylation of either ERK1/2 or p38 MAPK stimulated by exposure to LPA for 30 or 240 min (Fig. 5B). Furthermore, PKD2 siRNA significantly reduced NF-κB-driven luciferase reporter activity (Fig. 5C). These results indicate that PKD2 plays a major role in mediating LPA-induced NF-κB activation.
Next, we determined whether PKC/PKD2 mediates LPA-stimulated IL-8/CXCL8 production. As a first line of evidence supporting this hypothesis, we demonstrated that treatment of NCM460 cells with the selective PKC inhibitor GF-I, a potent inhibitor of PKD2 activation (see Fig. 2), dramatically (∼80%) inhibited the production of IL-8/CXCL8 induced by LPA (Fig. 6A).
Very recently, Zhao et al. (62, 63) demonstrated that in normal human bronchial epithelial cells IL-8 production in response to LPA is dependent on EGF receptor transactivation stimulated by PKC-δ, Lyn kinase, matrix metalloproteinases, and human bronchial EGF. Therefore, if LPA stimulates IL-8 production in NCM460 cells through a similar pathway, inhibition of EGF receptor tyrosine kinase should prevent LPA-stimulated IL-8 production in these cells. As shown in Fig. 6A, treatment of NCM460 cells with the selective EGF receptor tyrosine kinase inhibitor AG-1478 attenuated LPA-stimulated IL-8 production only slightly. These results are consistent with the hypothesis that LPA induces IL-8 production through a PKC-dependent but EGF receptor-independent pathway in human intestinal NCM460 cells and imply that IL-8/CXCL8 production is regulated through different pathways in different epithelial cell types.
We next determined the contribution of PKD2 to LPA-induced IL-8/CXCL8 production in NCM460 cells. As shown in Fig. 6B, transfection of NCM460 cells with siRNAs targeting PKD2 resulted in selective knockdown of PKD2, in agreement with results in Fig. 5A. The salient feature of the results shown in Fig. 6 is that PKD2 knockdown markedly reduced LPA-stimulated IL-8/CXCL8 production by 60% (Fig. 6C). Collectively, the results presented here indicate that a substantial component of LPA-stimulated IL-8/CXCL8 production is mediated by endogenous PKD2 through a NF-κB-dependent pathway in normal human intestinal NCM460 cells.
Mucosal surfaces, such as the lung and the intestine, are lined by a single layer of epithelial cells. The human epithelial lining is positioned at the interface between the luminal/external environment and the organism. Not only does the intestinal epithelial lining serve as a physical barrier, but it also serves as a sensor of nutrients, toxins, and microorganisms. Thus colonic epithelial cells can respond to luminal stimuli by inducing intracellular signaling events, leading to the production and release of chemokines through the basolateral surface (13, 31).
LPA has been shown to play a role in immunoregulation, affecting various cell types. In human umbilical endothelial cells, LPA stimulates expression of intercellular adhesion molecule 1, which interacts with activated lymphocytes (23). More recently, LPA has been shown to enhance IL-13 expression in T cells (41). Furthermore, LPA has been shown to stimulate IL-8/CXCL8 production in human bronchial epithelial cells (42) and in human colon cancer cell lines (45, 60). Zhao et al. (62, 63) recently demonstrated that LPA-induced EGF receptor transactivation is a major mediator of IL-8 production, using human bronchial epithelial cells as their model system. In contrast, our study demonstrates that exposure of human intestinal NCM460 cells to the specific EGF receptor tyrosine kinase inhibitor AG-1478 inhibited the production of IL-8/CXCL8 in response to LPA only modestly (∼20%). Thus, although in bronchial epithelial cells the majority of the IL-8/CXCL8 production induced by LPA is mediated by EGF receptor transactivation, this pathway has only a modest contribution to the response in intestinal NCM460 cells, the experimental system used in this study.
There is increasing evidence that the PKD family of protein kinases plays fundamental cellular functions in multiple cell types. However, the preponderance of this data has been generated from studies of PKD, the founding member of this protein kinase family. Although the isoforms of the PKD family share extensive homology in their catalytic domains, recent studies demonstrated that these enzymes localize to different organelles in the cell. Furthermore, many aspects of PKD regulation and physiological functions are based on overexpression of wild-type and mutant PKDs in transient transfection systems, which have inherent limitations. Previous studies implicated PKD in a pathway that promotes NF-κB activation in response to oxidative stress (47–50). However, most of these studies utilized transfection of PKD mutants or tagged PKD in cells, which express both PKD and PKD2 (29, 47–50). None of the previous studies demonstrated a role of endogenous PKDs in mediating GPCR-mediated pathways leading to NF-κB activation (as opposed to oxidative stress), and none identified a role of any of the PKDs in chemokine production.
In this study, we demonstrated that PKD2 is the only PKD family protein detected in nontransformed, human colonic epithelial NCM460 cells. LPA stimulated PKD2 activation in these cells, as judged by measurements of catalytic activity in immunoprecipitates and by phosphorylation of activation loop residues, Ser706/710 and Ser876, the autophosphorylation site in the COOH terminus of PKD2. These results indicate that LPA induces striking endogenous PKD2 activation and multisite phosphorylation in human epithelial cells.
Several lines of evidence indicate that LPA-induced PKD2 activation proceeds via a PKC-dependent pathway. Either pharmacological PKC inhibitors or downregulation of classic and novel isoforms of PKC by long-term exposure to phorbol esters dramatically reduced LPA-stimulated PKD2 phosphorylation at Ser706/710 and Ser876. A prior report suggested that the PKC inhibitors GF-1 and Ro-31-8220 have direct inhibitory effects on immunoprecipitated PKD2 in vitro (51). To clarify the mechanism by which the PKC inhibitors interfere with LPA-induced PKD2 activation in intact cells, we tested the activity of purified PKD2 in the absence or presence of Ro-31-8220 or GF-I. Our results indicate that neither Ro-31-8220 nor GF-I had a direct effect on PKD2 catalytic activity. We verified that Ro-31-8220 or GF-I used in our experiments completely inhibited PKC-ε-catalyzed substrate phosphorylation at concentrations that did not exert any inhibitory effect on PKD2 activity. These results are consistent with the notion that LPA stimulates PKD2 activation through a PKC-dependent pathway.
Human intestinal epithelial cells in vivo are thought to represent a major local source of proinflammatory cytokines and, in particular, for the IL-8/CXCL8 chemokine. Indeed, epithelial-derived IL-8/CXCL8 has been shown to support neutrophil infiltration in human mucosal inflammatory diseases of the colon, including ulcerative colitis (21, 28). In the present study, we demonstrate that LPA potently stimulates IL-8/CXCL8 production in nontransformed human colonic epithelial cells NCM460. The levels of IL-8/CXCL8 production in response to LPA were comparable to those induced by pathogenic bacteria (37, 44). Either inhibition of PKC or PKD2 gene silencing utilizing siRNAs targeting distinct PKD2 sequences markedly reduced LPA-stimulated IL-8/CXCL8 production and NF-κB promoter activity. Our results indicate that PKD2 is a novel signal-transduction element in the pathway that mediates IL-8/CXCL8 production in response to LPA via NF-κB.
In addition to NF-κB, the IL-8 promoter contains binding sites for the transcription factors activator protein-1 (AP-1), C-EBP/NF-IL-6, and Tcf/Lef (17, 24). NF-κB is considered to be essential, whereas the others are dispensable for transcriptional activation in some cells but contribute to activation in others (24). Therefore, in some cell types, optimal IL-8/CXCL8 transcription and subsequent expression require contribution of other transcription factors, such as AP-1, C-EBP/NF-IL-6, and/or Tcf/Lef. Although our results with NCM460 cells demonstrate a prominent role of PKD2 in LPA-stimulated IL-8/CXCL8 production, the AP-1/ERK pathway could be required for promoting a maximal IL-8/CXCL8 response to LPA stimulation.
In different intestinal epithelial cell lines, LPA stimulates intestinal epithelial cell migration (53), proliferation (57), and IL-8/CXCL8 production (45, 60). The serum concentrations of this bioactive lipid range from 2 to 20 μM (3). Major sources of LPA in the vicinity of injured epithelial surfaces and inflammatory processes are activated platelets and stimulated fibroblasts (3). We envisage that, after intestinal epithelial injury, bleeding and blood clots ensue with subsequent activation of platelets, resulting in high concentrations of LPA in the local microenvironment. LPA receptor activation (at concentrations such as those used in this study) could then lead to PKD2-mediated NF-κB activation that results in IL-8/CXCL8 production. This in turn may lead to neutrophil recruitment to counter potential bacterial invasion and stimulation of CXCR1/CXCR2-expressing cells, such as myofibroblasts (36), in the lamina propria. In addition to its neutrophil chemoattractant ability, IL-8/CXCL8 has recently been shown to mediate spreading and wound closure by fibroblasts (12). These are important events in the repair of intestinal injury that compromise the integrity of the basement membrane.
It is conceivable that the PKC/PKD2/NF-κB/IL-8 pathway delineated in this study could also operate in intestinal epithelial cells challenged by other stimuli. Toll-like receptors (TLRs), which recognize microbial products known as pathogen-associated molecular patterns, are expressed in colonic epithelial cells (2). Recently, PKCs have been implicated as a downstream targets of TLR signaling (4, 7, 8). Thus PKD2 may serve as a point of convergence that integrates signals from both TLRs and GPCRs.
In summary, our results indicate that a substantial component of the production of the potent proinflammatory chemokine IL-8/CXCL8 is mediated by PKD2 through a NF-κB-dependent pathway in NCM460 cells. This is the first time that endogenous PKD2, and indeed any member of the PKD family, is shown to play an important role in mediating the production of a chemokine in epithelial cells or in any cell type.
This work was supported by a Howard Hughes Medical Institute Research Resources Faculty Development Award and by National Institute of Diabetes and Digestive and Kidney Diseases Grants K08 DK-063983 to T. T. Chiu and R0–1 DK-55003 and R0–1 DK-56930 to E. Rozengurt.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society