Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The inflammatory and immunomodulatory properties of the complement-derived human anaphylatoxin C5a are induced upon its binding to specific, high-affinity receptors (C5aR, CD88) expressed on the surface of C5a-responsive target cells. The expression and function of C5aR have been well characterized and documented on cells of myeloid origin (1). A hallmark of C5a-mediated responses is the ability to recruit C5aR-bearing myeloid cells and to induce cytokine synthesis and release from peripheral blood mononuclear cells (2, 3). However, there have been a few reports of the presence of C5aR on a variety of nonmyeloid cells including hepatocytes (4), astrocytes (5), and epithelial cells derived from tissues in the lung, gut, and kidney (6). In contrast to what is known about C5aR-bearing myeloid cells, little is understood of the nature of C5aR expression on nonmyeloid cells or the functional role the C5aR/C5a system plays in these cells under normal and aberrant physiologic conditions.

Some insight into a potential functional role for the C5aR/C5a system in cells of nonmyeloid origin was provided by a recent study in which we demonstrated that C5aRs are constitutively expressed on the surface of human bronchial epithelial cells (HBECs). HBECs were shown to respond to C5a by releasing small but significant amounts of interleukin (IL)-8 (7). However, HBECs that were exposed to an inflammatory insult such as cigarette smoke were significantly more responsive to the C5a- mediated release of this cytokine than were untreated HBECs challenged with either C5a alone or cigarette smoke alone (8). This enhanced ability of C5a to induce the release of cytokines from cigarette smoke-treated HBECs was accompanied by the observation that cigarette smoke extract (CSE) did not appear to increase the surface density of C5aRs, but did increase the population of HBECs that expressed C5aR. It is unknown by what mechanism(s) C5a induces the release of cytokines from HBECs that have been previously exposed to cigarette smoke. Because our results suggest that CSE heightens C5aR responsiveness toward C5a, we believe that an understanding of the intracellular signaling events coupled to C5aR, the manner by which these events lead to cytokine release, and how these events are affected by CSE may provide insights into the nature of C5a-mediated inflammatory responses by airway epithelial cells.

An important intracellular signaling component linked to C5aR in human neutrophils is protein kinase C (PKC) (7). Others have shown that CSE activates PKC in monocytes (8), endothelial cells (9), and lung carcinoma cells (10). We have previously identified several isoforms of PKC present in airway epithelial cells (11), and have found that acetaldehyde activates PKC in these cells in both a time- and dose-dependent manner (12). Since acetaldehyde is a known component of the volatile phase of cigarette smoke, we hypothesize that cigarette smoke augments the C5a-mediated release of IL-8 from HBECs by activating PKC. Such a finding may add to the understanding of in vivo mechanisms by which airway epithelial cells increase their response to C5a upon habitual exposure to cigarette smoke, leading to chronic airway inflammation.

In the present study we investigated the activation state of PKC under conditions of CSE and C5a stimulation of HBECs. We correlated the increase in C5a-stimulated IL-8 release with the activation of PKC in response to CSE treatment of the HBECs. In addition, we found that the specific inhibition of PKC results in inhibition of the CSE + C5a-augmented release of IL-8.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Preparation

HBECs were obtained by endobronchial brushing from patients undergoing flexible fiberoptic bronchoscopy (13). After each pass of a bronchial brush, the brush was rinsed in 5 ml of unsupplemented medium L-15 (Biofluids, Rockville, MD). The brush was rinsed twice in sterile saline, and repeat brushings were performed. After completion of the brushings, an antibiotic mixture (streptomycin, penicillin, and amphotericin B) was added to the L-15 medium. Cytocentrifuge preparations were made and cell viability was determined for the subsequent experiments. Cells were then cultured on type I collagen (Vitrogen, Palo Alto, CA)-coated dishes in serum-free medium (LHC-9/ RPMI, 1:1). These cells showed keratin staining but not vimentin staining, as consistent with an epithelial cell origin. The cells were kept frozen in liquid nitrogen, and were successfully recultured and repassaged upon thawing. HBECs were maintained in culture and used in this study for up to six passages. All experiments were performed no less than three separate times, and representative data from these experiments is presented.

PKC Activity Assay

PKC activity was determined in crude whole-cell fractions of HBECs. The assay employed was a modification of procedures previously described (14), and is done with 900 µM PKC substrate peptide (Peninsula, Belmont, CA), 12 mM Ca(C₂H₃O₂)₂, 8 µM phosphatidyl-L-serine, 24 µg/ml phorbol 12-myristate 13-acetate (PMA), 30 mM dithiothreitol, 150 µM adenosine triphosphate (ATP), 45 mM Mg(C₂H₃O₂)₂, and 10 µCi/ml [-³²P]ATP in a Tris-HCl buffer (pH 7.5). Samples (20 µl) were added to 40 µl of this reaction mixture and were incubated for 15 min at 30°C. Incubations were halted by spotting 50 µl of each sample onto P-81 phosphocellulose papers (Whatman, Hillsboro, OR). Papers were then washed five times for 5 min each in phosphoric acid (75 mM), washed once in ethanol, dried, and counted in nonaqueous scintillant as previously described (15). Kinase activity was expressed in relationship to total cellular protein assayed, and was calculated in pmol/min/mg. As previously reported with regard to variability in IL-8 release (16), baseline unstimulated values for PKC activity in HBECs vary among cell-lines and among passages. The trends in PKC activation are consistent, whereas the empirical magnitude of the kinase activity may differ among cells from different donors and passages. We assayed all samples in triplicate, and performed no less than three separate experiments per unique parameter. Data were analyzed for statistical significance with Student's paired t test.

Acetaldehyde Concentration Assay

We measured concentrations of acetaldehyde in the extracellular milieu by transferring the medium from exposed cells to 1.8-ml microcentrifuge tubes, which were filled to the top, capped, sealed with Parafilm, and stored at 80°C until analysis. We determined acetaldehyde and ethanol concentrations by headspace gas chromatography, using the method of Eriksson and colleagues (17).

Cytokine Release Assay

IL-8 levels in culture supernatants were quantified with a sandwich-type enzyme-linked immunosorbent assay (ELISA). Ninety-six well flat bottomed polystyrene microtiter plates (Immulon, Chantilly, VA) were coated with 200 µl/well of purified (goat) antihuman IL-8 antibody (R&D Systems, Minneapolis, MN) or IL-6 antibody (ICN Biomedical, Costa Mesa, CA) diluted 1:2,000 in Voller's buffer (pH 9.6) for 24 h at 4°C. After washing the plates thrice in phosphate-buffered saline (PBS)-Tween, we applied undiluted culture supernatants and human recombinant IL-8 (rIL-8) standards (Sigma, St. Louis, MO) to the plates and incubated them at room temperature for 90 min. Plates were washed thrice with PBS-Tween, followed by the addition of (rabbit) antihuman IL-8 antibody (UBI, Lake Placid, NY) or IL-6 antibody (Sigma) diluted 1:4,000 in PBS-Tween/BLOTTO (0.2% instant nonfat milk in PBS-Tween) for 60 min. After three washes, human- serum-absorbed, peroxidase conjugated (goat) antirabbit immunoglobulin (Ig)G (ICN) was added at 1:2,000 in PBS-Tween/BLOTTO for a 45-min incubation. The plates were again washed thrice and 200 µl/well of peroxidase substrate (10 ng/ml orthophenylenediamine [Sigma] and 0.003% H₂O₂ in distilled H₂O [dH₂O]) was added. The reaction was terminated with 27.5 µl/well of 8 M sulfuric acid, and plates were read at 492 nm in an automated ELISA reader (BioRad, Hercules, CA).

CSE Preparation

CSE was made freshly immediately before all experimental procedures. The cigarettes used (unfiltered, Code 2R1) were obtained from the University of Kentucky Tobacco-Health Research Division. A cigarette was connected to a peristaltic pump apparatus and lit, and the smoke was bubbled through 25 ml of sterile RPMI medium. One cigarette per 25 ml volume was used. The peristaltic pump was equilibrated at a rate of one cigarette per 10 min, or 160 cm³ of tobacco smoke produced per minute. The cigarettes were 85 mm long, and approximately 60 mm was consumed. The CSE medium was then sterile-filtered and diluted into LHC9-RPMI medium as a percentage of the total volume. CSE was diluted from 5-20% for use in these studies. No significant cell death was observed at less than 20% CSE.

Materials

LHC basal medium was purchased from Biofluids. RPMI 1640, Dulbecco's modified Eagle's medium, modified Eagle's medium, streptomycin-penicillin, and Fungizone were purchased from GIBCO-BRL (Chagrin Falls, OH). Extraction of frozen bovine pituitaries, obtained from Pel Freez (Rogers, AR), was performed as previously described, and yielded an extract containing 10 mg/ml protein (18). Purified PKC was obtained from Calbiochem (San Diego, CA). [³²P]ATP was obtained from Amersham (Arlington Heights, IL), 2-diethylaminoethanol and phosphocellulose P-81 paper from Whatman, and heptapeptide substrates for PKC from Peninsula. All other reagents not specified were purchased from Sigma.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Activation of PKC by CSE

In an attempt to identify PKC activity in crude cell fractions, we assayed total PKC activity in combined cytosolic and particulate fractions of HBECs. Cells were either treated with CSE (5-20%) for 1 h or PMA (100 ng/ml) for 15 min to activate PKC. PMA stimulation of these cells was used as a positive control for PKC activity in crude HBEC assays for PKC involving other possible PKC- mediating agents such as CSE. We observed a significant increase in PKC activity in the particulate fraction of HBECs treated with PMA as compared with the cytosolic fraction (data not shown).

Primary cultures of HBECs were treated with various concentrations of CSE (5-20%) for various periods (0.5- 48 h) in multiple 60-mm tissue culture dishes. Concentrations of CSE in excess of 20% were determined to be cytotoxic, whereas 5-10% CSE had no apparent toxicity to monolayer cells as determined with the trypan blue exclusion assay. Subcellular fractions were assayed for PKC kinase activity. Changes in PKC activity for cells stimulated with CSE were compared with changes in unstimulated control cells as well as in cells whose PKC had been maximally activated by PMA. We found that 5% CSE significantly stimulated PKC (Figure 1). CSE-stimulated PKC activity reached its maximal levels at 1-3 h after exposure, and the activity subsided to nearly unstimulated levels thereafter. This loss of sustained PKC activation by CSE correlated with the loss of acetaldehyde in the stimulation medium (Figure 1, insert). The specific PKC inhibitor calphostin C, and the nonspecific PKC inhibitor H-7, abrogated the CSE-mediated PKC activation (Figure 2). As expected, calphostin C (1 µM) inhibited the activity of PKC in whole-cell extracts, whereas H-7 (10 µM) reduced PKC activity to that of unstimulated cells.

View larger version (20K): [in this window] [in a new window]   Figure 1.   PKC activation at 0-24 h in HBECs treated with 5% CSE. Insert: Acetaldehyde concentration in the supernatant medium of these same cell samples from 0-6 h. Each data point represents the average of triplicate measurements of three or more samples within an experiment (P  0.01 for 1- to 3-h CSE-stimulated PKC activity and acetaldehyde concentration as compared with cells in medium only).

View larger version (19K): [in this window] [in a new window]   Figure 2.   PKC activation in HBECs treated with 0-20% CSE for 1 h in the presence or absence of the PKC inhibitors calphostin C (CC) or H-7. Each data point represents the average of triplicate measurements of three or more samples within an experiment (P  0.05 for 1 µM calphostin C-pretreated cells and P  0.01 for 10 µM H-7-pretreated cells, as compared with cells in medium only at 1 h).

Activation of PKC by Acetaldehyde

Acetaldehyde, a component of ethanol metabolism and cigarette smoke, was investigated for its ability to alter the PKC response in HBECs. PKC activity was maximally increased in HBECs treated with 10 mM acetaldehyde at 15-90 min after treatment (Figure 3). This activity decreased at 2-6 h, with a return to baseline unstimulated kinase activity levels by 24 h. When added directly to the incubation medium, acetaldehyde was quickly volatilized, leaving the resulting concentration in the medium at 1 h in the range of 100 µM, and no detectable acetaldehyde by 24 h (data not shown).

View larger version (14K): [in this window] [in a new window]   Figure 3.   PKC activation at 0-72 h in HBECs treated with 10 mM acetaldehyde. Each time point represents the average of triplicate measurements of three or more samples within an experiment (P  0.01 at 15 min to 2 h for acetaldehyde-stimulated PKC activity, as compared with cells in medium only).

CSE-Stimulated IL-8 Release

In order to examine the role of PKC in CSE-stimulated IL-8 release, HBECs were treated with CSE for various periods and the amount of IL-8 released into the supernatant medium was measured. HBECs treated with CSE for 1 h before assay showed small but significant increases in IL-8 release as compared with control cells (Figure 4). Exposure of the cells to CSE for 1 h followed by a 2-h treatment with 50 nM C5a resulted in enhanced release of IL-8. However, stimulation of the cells with only C5a produced increases in IL-8 release comparable to those with CSE treatment alone. These findings were similar to prior results obtained with C5a and CSE alone or in combination (6). As with the inhibition of PKC activity, HBECs pretreated with the PKC inhibitor calphostin C (20 nM-1 µM) for 30 min before CSE stimulation showed a dose-dependent reduction in IL-8 release (Figure 4). Incubation of HBECs with these inhibitors alone produced no significant change in IL-8 from that in unstimulated control cells.

View larger version (20K): [in this window] [in a new window]   Figure 4.   IL-8 release in HBECs treated with 5% CSE for 2 h, followed by 50 nM C5a for 45 min, and inhibition by 20 nM-1 µM calphostin C of CSE + C5a-stimulated IL-8 release. Cells were exposed to control medium (CON) alone or treated with CSE, C5a, or both (C + C). Each bar represents the average of triplicate measurements of three or more samples within an experiment (P  0.001 for cells treated with C + C as compared with control cells).

CSE-Stimulated IL-6 Release

In order to determine whether cytokines other than IL-8 were released in a CSE + C5a-dependent manner, HBECs were treated with CSE for various periods and the amount of IL-6 released into the supernatant medium was measured. HBECs treated with CSE for 1 h prior to assay showed small but significant increases in IL-6 release as compared with control cells (Figure 5). Stimulation of the cells with only C5a produced increases in IL-6 release comparable to those with CSE treatment alone. However, exposure of the cells to CSE for 1 h, followed by a 2-h treatment with 50 nM C5a, resulted in an increased release of IL-6. As with the inhibition of PKC activity, HBECs pretreated with the PKC inhibitor calphostin C (20 nM-1 µM) for 30 min before CSE stimulation exhibited a dose-dependent reduction in IL-8 release (Figure 5). Incubation of HBECs with the inhibitors alone resulted in no significant change in IL-6 release from unstimulated control cells.

View larger version (42K): [in this window] [in a new window]   Figure 5.   IL-6 release in HBECs treated with 5% CSE for 2 h, followed by 50 nM C5a for 45 min, and inhibition by 20 nM-1 µM calphostin C of CSE + C5a-stimulated IL-6 release. Cells were exposed to control medium (CON) alone or treated with CSE, C5a, or both (C + C). Each bar represents the average of triplicate measurements of three or more samples within an experiment (P  0.001 for cells treated with CSE plus C5a compared to control cells).

Correlation of PKC Activation and IL-8 Release

In order to determine whether PKC activation is necessary for the release of IL-8, both PKC activity and IL-8 release were assayed in the same cultures of HBECs. As shown in Figure 6, PKC activation occurred only in cells treated with 5% CSE for 1 h. However, maximum IL-8 release did not occur in cells treated with CSE alone. Conversely, increases in IL-8 release were demonstrated only in cells treated with C5a, but PKC activation was not observed in cells treated only with C5a. The maximal increase in IL-8 release occurred only under conditions of C5a stimulation in cells in which PKC was activated by CSE (Figure 6). In the presence of calphostin C (1 µM), both IL-8 release and PKC activity were inhibited. Similar results were observed for IL-6 (data not shown). This suggests that the activation of PKC is a necessary precedent to the combination of cigarette smoke- and C5a-mediated IL-8 and IL-6 release.

View larger version (21K): [in this window] [in a new window]   Figure 6.   Inhibition of both PKC activity and IL-8 release by calphostin C in CSE + C5a-stimulated HBECs. Cells were exposed to control medium alone or treated with CSE, C5a, or both (C + C) in the presence or absence of calphostin C (CC). Each data point represents the average of triplicate measurements of three or more samples within an experiment (P  0.05 for CSE or CSE + C5a-stimulated PKC activity and IL-8 concentration in the presence of CSE + C5a as compared with control cells in medium only).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

HBECs produce and release the proinflammatory cytokine IL-8 in response to various exogenous stimuli (16). CSE can also stimulate small increases in IL-8 released by HBECs (16). The release of IL-8 is mediated by a volatile substance in CSE, although the exact component responsible for this effect remains unknown. We found that C5a also stimulates small increases in IL-8 release, and that pretreatment of HBECs with CSE prior to C5a stimulation evokes a large release of IL-8, which is significantly greater than the additive effect of either CSE or C5a (6). CSE increased the number of HBECs expressing C5aRs, but not the density of C5aRs, suggesting that an increase in the total number of cells expressing C5aR might lead to increased CSE-mediated (via PKC) C5a-stimulated IL-8 release. However, we have previously found that only the distribution of C5aRs, and not the total number of C5aRs in this cell population, appears to increase (6).

Previously, we showed that four distinct PKC isoenzymes are present in the bronchial epithelial cells, and that tumor necrosis factor (TNF)- stimulates PKC-dependent bronchial epithelial cell migration (11). In the present study, we investigated the effects of acetaldehyde, cigarette smoke, and C5a on PKC activity in HBECs. CSE stimulates a rapid (within 1 h) activation of PKC in HBECs. C5a alone does not activate PKC in HBECs. Pretreatment of HBECs with CSE, followed by C5a stimulation, results in a large release of IL-8 and IL-6. Neither CSE nor C5a alone stimulates this large interleukin release. Inhibition of PKC by calphostin C results in the inhibition of CSE + C5a-mediated interleukin release. These results suggest that CSE facilitates C5a-stimulated IL-6 and IL-8 release through a PKC-mediated pathway. The mechanism(s) regulating IL-6 and IL-8 production and release as a result of PKC activation remain to be determined. One possible mediator of PKC found in cigarette smoke is acetaldehyde.

Tobacco smoking and alcohol consumption promote upper respiratory tract infections, pneumonia, and cancer. The synergistic effects of these products may be traced to acetaldehyde, which is a product of ethanol oxidation as well as a major component of tobacco smoke. Acetaldehyde production is increased in cells from bronchopulmonary washings of individuals who actively consume alcohol and tobacco (19). Acetaldehyde exposure may play a key role in reducing the host defense among chronic drinkers and smokers (20, 21). It has been suggested that airway acetaldehyde derived from ethanol metabolism or cigarette smoke damages cilia and interferes with airway clearance, although the mechanism for this inhibition is unknown (22, 23). One mechanism by which acetaldehyde is thought to exert its bioactive effects is through the formation of covalent adducts with nucleophilic amino groups of target proteins (24). It has specifically been shown in bovine bronchial epithelial cells that acetaldehyde impairs ciliary beating in cultured cells, presumably due to the binding of acetaldehyde to proteins contained in cilia (23). Physiologic concentrations of ethanol-derived, vapor-phase acetaldehyde were shown to be sufficient to slow ciliary beat frequency in bovine bronchial epithelial cells (22). This effect was greatly augmented by cyanamide, which blocks acetaldehyde dehydrogenase (27) and sensitizes the cells to even greater cilistasis.

Acetaldehyde alone (10 mM) is sufficient to activate PKC in HBECs for a short and transient period. This exceedingly high concentration of acetaldehyde volatilized rapidly under our culture conditions, and may have been completely absent from the cell medium after 3-4 h. At the time of PKC activation, approximately 0.1-0.5 mM acetaldehyde remained in the culture medium. This rapid loss of acetaldehyde corresponded with the time course of loss of PKC activity in the HBECs. We routinely measured acetaldehyde concentrations of 50-100 µM in CSE preparations. The higher concentrations of acetaldehyde found in cigarette smoke may be of far more importance to airway pathophysiology than the lower concentrations of lung acetaldehyde produced through the metabolism of alcohol by alcohol dehydrogenase. Thus, tobacco-derived acetaldehyde could mediate the activation of PKC and subsequent release of IL-8 under conditions of increased C5a in the airways.

When applied to our model of airway epithelial cell signaling, some component of cigarette smoke, possibly acetaldehyde, would trigger the release of proinflammatory cytokines in response to C5a to a much greater degree than would be expected in nonsmokers. Thus, the inflammation occurring in response to a bacterial infectious challenge would be increased, less controlled, and more disposed to causing tissue damage than would be observed in nonsmokers. Given this, and the observations made by others that CSE and acetaldehyde cause ciliastasis, the clearance of such an infectious agent would be minimized, further impeding the repair process at the level of the airway epithelium.