- Email: [email protected]
Neuropeptides and capsaicin stimulate the release of inflammatory cytokines in a human bronchial epithelial cell line
Neuropeptides (1999) 33 (6), 447–456 © 1999 Harcourt Publishers Ltd Article no. npep.1999.0761
Neuropeptides and capsaicin stimulate the release of inflammatory cytokines in a human bronchial epithelial cell line B. Veronesi,1 J.D. Carter,1 R.B. Devlin,1 S.A. Simon,2 M. Oortgiesen2 1
National Health and Environmental Effects Research Laboratory, US Environmental Protection Agency, Research Triangle Park, 2Departments of Neurobiology and Anesthesiology, Duke University Medical Center, Durham, NC, USA
Summary The role of neuropeptides in initiating and modulating airway inflammation was examined in a human bronchial epithelial cell line (i.e. BEAS–2B). At a range of concentrations, exposure of BEAS–2B cells to Substance P (SP) or calcitonin gene related protein resulted in immediate increases in intracellular calcium ([Ca2+]i), the synthesis of the transcripts for the inflammatory cytokines, IL-6, IL-8 and TNFα after 2 h exposure, and the release of their proteins after 6 h exposure. Addition of thiorphan (100 nM), an inhibitor of neutral endopeptidase, enhanced the levels of SP-stimulated cytokine release. Stimulation of IL-6 by SP occurred in a conventional receptor-mediated manner as demonstrated by its differential release by fragments SP 4–11 and SP 1–4 and by the blockage of IL-6 release with the non-peptide, NK-1 receptor antagonist, CP-99 994. In addition to the direct stimulation of inflammatory cytokines, SP (0.5 µM), in combination with TNFα (25 units/ml), synergistically stimulated IL-6 release. BEAS–2B cells also responded to the botanical irritant, capsaicin (10 µM) with increases in [Ca2+]i and IL-8 cytokine release after 4 h exposure. The IL-8 release was dependent on the presence of extracellular calcium. Capsaicin-stimulated increases of [Ca2+]i and cytokine release could be reduced to control levels by pre-exposure to capsazepine, an antagonist of capsaicin (i.e. vanilloid) receptor(s) or by deletion of extracellular calcium from the exposure media. The present data indicate that the BEAS–2B human epithelial cell line expresses neuropeptide and capsaicin-sensitive pathways, whose activation results in immediate increases of [Ca2+]i stimulation of inflammatory cytokine transcripts and the release of their cytokine proteins. © 1999 Harcourt Publishers Ltd
INTRODUCTION Airway inflammation underlies various disorders associated with exposure to environmental xenobiotics. The cascade of cellular and subcellular events that compose this sequelae has been described primarily in vascular and immunological terms. However, more recent thinking suggests that neuropeptides initiate and sustain inflammation through the process known as neurogenic
Received 15 August 1999 Accepted 10 October 1999 Correspondence to: Dr Bellina Veronesi, US Environmental Protection Agency, National Health Effects and Environmental Research Laboratory, Neurotoxicology Division MD 74B, Cellular and Molecular Toxicology Branch, MD74B RTP, NC 27711, USA. Tel.: +1 919 541 5780; Fax: +1 919 541 4849; E-mail: [email protected]
inflammation.1–3 This process predominates in tissues (i.e. skin, airways, gut, conjunctiva, mucous membranes, etc.) that are most likely to encounter invading organisms, irritating stimuli or exogenous chemicals. Briefly, this process involves the activation of irritant pathways (e.g. capsaicin, pH sensitive, temperature-sensitive, mechano-receptors) which are found on the sensory C and Aδ fibers that innervate such tissues. In the airways, neuropeptides (e.g. Substance P [SP], calcitonin generelated protein [CGRP] and neurokinin A) that are released upon activation of these receptors, can initiate and modulate symptoms of inflammation by targeting both immune (e.g. lymphocytes, neutrophils, macrophages and eosinophils) and non-immune cells (e.g. smooth muscle and endothelial cells of the vasculature, epithelial cells that line the airway lumen). Neuropeptides act on these target cells directly or 447
448 Veronesi et al.
indirectly as vasodilators, vasoconstrictors, degranulators, chemotactic factors, mitogens, adherents and secretogues. In the airway lumen, sensory fibers, containing both irritant and neuropeptide receptors, terminate between and below the respiratory epithelial cells. Epithelial cells lining the lumen are the first to encounter and respond to irritating airborne stimuli, making their interactions with irritants and neuropeptides a pivotal event in the pathophysiology of neurogenic inflammation in the airways. It has been reported that various types of respiratory epithelial cells contain neuropeptides. For example, neuropeptide immunostaining is found in neuroendocrine cells (i.e. neuroepithelial bodies), Clara cells, Type II alveolar cells, mast cells and serous epithelial cells.4–7 The present study extends these data and demonstrates that SP, CGRP and capsaicin can directly stimulate the release of inflammatory cytokines in a receptor-mediated fashion from a human bronchial epithelial cell line (i.e. BEAS–2B). MATERIALS AND METHODS Cells and maintenance BEAS–2B cells are SV-40/adenovirus-transformed human bronchial epithelial cells8 that have been used in numerous studies to evaluate pulmonary toxicants.9,10 These cells, like most epithelial cells, are highly influenced by their passage and confluency state.11,12 For these studies, cells were obtained commercially (ATTC, Rockville, MD, USA) and used from passage 40–80. Cells were maintained in complete keratinocyte growth media (KGM) consisting of KBM (Clonetics, San Diego, CA, USA) and supplemented with bovine pituitary extract, human epidermal growth factor (5 ng/ml), hydrocortisone (0.5 mg/ml), ethanolamine (0.1 mM), phosphoethanolamine (0.1 mM) and insulin (5 mg/ml). Complete KGM minus calcium and magnesium was purchased from Clonetics (Cat # CC–3004). For cytokine and transcript experiments, cells were grown to 85–95% confluency in 96 or 12 well Costar culture plates. For calcium experiments, cells were grown to 80–90% confluency on fibronectin-coated, 22 mm, # 0 thickness glass coverslips (Carolina Biological Supply, Burlington, NC, USA). Chemicals SP, CGRP, thiorphan, Hanks Balanced Salt Solution (HBSS), M-[2-hydroxyethyl] piperanine-N-{4-butane sulfonic acid (HEPES) and 2-[N-morpholinoethanesulfonic acid] (MES) were purchased from Sigma Chemical Corporation. Capsaicin and its receptor antagonist, capsazepine (CPZ), were purchased from RBI (Natick, MA, Neuropeptides (1999) 33(6), 447–456
USA). The non-peptide, NK-1 (SP-preferring) competitive receptor antagonist, CP–99 994 (13) was a generous gift from Pfizer Company (Grouton, CT, USA). Recombinant TNFα was obtained commercially (R&D Systems, Minneapolis, MN, USA). All chemicals were dissolved at test concentrations in their respective vehicles (i.e. nutrient media, HBSS) immediately before use.
Intracellular calcium measurements Increases in [Ca2+]i were recorded using the fluorescent probe Fluo-3-AM (Molecular Probes, Eugene, OR, USA). In preparation for calcium measurements, cells were incubated for 40 min with Fluo-3-AM (2 µM) at 37°C and then washed for 10 min in HEPES-buffer containing 150 mM NaCl, 5 mM KCI, 2 mM CaCI2, 1 mM MgCl2, 10 mM D-glucose, and 10 mM HEPES. Cells were illuminated with the exciting wavelength of 485 nm and the fluorescent emission was detected at 525 nm using Axon Workbench Imaging software (Axon Instruments, Foster City, CA, USA). After stable baseline recordings were collected, cells were exposed to the test compounds and washed between exposures with HEPES-buffer by rapid superfusion (15 ml/min). Antagonists were tested by preincubating cells for 5–10 min. Then, the test compound was exposed to the cells in the presence of the antagonist. Maximal [Ca2+]i increases were obtained at the end of each experiment by the addition of 2 µM ionomycin (IO). Traces of cells that showed strong IO responses but showed no marked bleaching were collected from 10–20 cells and from three to five different experiments. Experiments were carried out at 25–26°C.
Reverse transcription-polymerase chain reaction (RT–PCR) BEAS–2B (85–90% confluency) were exposed for 2 h to the test compound, washed twice with phosphatebuffered saline (PBS) and harvested in a guanidine thiocyanate-containing lysis buffer (GITC) as previously described.10 RNA was prepared by lysing cells in GITC buffer containing 4 M guanidine isothiocyanate, 25 mM sodium citrate (pH 7.0), 0.5% Sarkosyl, and 10 mM DTT. The lysate was sheared with a 22 gauge needle and layered over an equal volume of 5.7 M cesium chloride and 0.1 M EDTA. Total RNA was pelleted by centrifugal sedimentation at 114 000 g for 2 h. The relative concentrations of mRNAs coding for IL-6, IL-8, and TNFα were estimated by sequential reverse transcription of 200 ng RNA from the cultured cells and subsequent cDNA amplification (RT–PCR). Specific cDNA sequences within the mixture were amplified as previously described.10 Amplifications were performed under oil in 96 well plates using a PTC-100 programmable thermal cycler (MJ © 1999 Harcourt Publishers Ltd
Neuropeptides and human bronchial epithelial cells 449
100
c
90
160 140
70
525 nm intensity
525 nm intensity
80
60 50 40 wash
30 20
SP (10 nM)
wash
SP (10 nM)
IO
10 0
120 100 80 60 40
0
50
100
150
600
650
900
950
time (s)
a
CGRP 10 nM
IO
20
1000
0
c
0
50
100
150
time (s)
160 525 nm intensity
140 120 100 80 60 Ca2+ -free
40 20
Ca2+ -free IO
SP (10 nM)
0 0
50
100
150
200
time (s)
b Research, Watertown, MA, USA). Following an initial 3 min denaturation at 92°C, amplifications were cycled for 1 min at 92°C, 1 min at 56°C, and 2 min at 72°C. The optimum number of cycles varied with the target cDNA and was determined empirically. cDNA was amplified for glyceraldehyde-3-phosphate dehydrogenase (GAPDH), a housekeeping gene and the inflammatory cytokine transcripts for IL-8, IL-6 and TNFα, using gene-specific primers. Relative concentrations of specific DNA sequences were quantified by separating 15 µl of each amplification mixture through 2% agarose gels in 0.5× Trisborate, EDTA buffer. The gel was stained in 1 µg/ml ethidium bromide and photographed under UV illumination with Polaroid type 55 P/N film (Polaroid, Cambridge, MA, USA). The negatives were scanned using a Kodak DC120 Zoom Digital camera. The Kodak Electrophoresis Documentation and Analysis System 120 (Kodak Corp., Rochester, NY, USA) was used to determine the intensities of amplification products that resolved as single bands of the predicted size. The net intensities of IL-6, IL-8 and TNFα DNA bands were divided by the net intensity of the GAPDH-DNA to normalize for any errors in the amount of input RNA, day to day variations in amplification efficiency, gel staining or photography.
Oligonucleotide sequences were synthesized using an Applied Biosystems 391 DNA synthesizer (Foster City, CA, USA) based on sequences published in © 1999 Harcourt Publishers Ltd
Fig. 1 (a) Intracellular calcium ([Ca2+]i) levels were measured in BEAS–2B cells using the fluorescent probe Fluo-3 AM. Exposure of BEAS–2B to SP (10 nM) produced a rapid increase in [Ca2+]i, followed by a lower, longer lasting plateau. At both concentrations, approximately 40% of tested cells responded. A second exposure of the cells to SP (10 nM) produced a [Ca2+]i response with a smaller peak amplitude, which suggested receptor desensitization or internalization. Between the two SP exposures, cells were washed for 7 min with control HEPES buffer. IO (2 µM) was applied at the end of each experiment to determine the maximal [Ca2+]i increase. (b) The increase in [Ca2+]i was dependent on extracellular calcium and magnesium since, in media lacking these ions, SP produced only a transient increase in [Ca2+]i which increased when the superfusion medium was changed to one containing SP and 2 mM calcium-magnesium and returned again to baseline when the medium was changed to one lacking calcium-magnesium. IO (2 µM) was applied at the end of each experiment to determine the maximal [Ca2+]i increase. (c) Exposure of BEAS–2B to CGRP (10 nM) produced an increase in cytoplasmic calcium, which declined over 1–5 min. The percentage of responding cells varied and amounted to 20–60%. Application of IO (2 µM) was used to determine the maximal calcium increase. Exposure to CGRP is indicated by horizontal bars.
GenBank human DNA database. The following sense and anti-sense sequences were employed: GAPDH: 5′CCATGGAGAAGGCTGGGG 3′ and 5′CAAAGTTGTCATGGATGACC 3′ IL-8: 5′AACCCTCTGCACCCAGTTTTCCTT3′ and 5′TCCACTCTCAATCACTCTCA 3′ IL-6: 5′CTTCTCCACAAGCGCCTTC 3′ and 5′GGCAAGTCTCCTCATTGAATC 3′ TNFα: 5′CAGGCAGTCAGATCATCTTC 3′ and 5′CTTGATGGCAGAGAGGAGGT 3′ The oligonucleotide primers used for quantitation were designed to amplify segments of the mature mRNA derived from two or more gene exons, so that amplification of genomic sequences would yield products larger than those derived from mature mRNA. Only amplification products of the size predicted by the mature mRNA sequence were seen and quantified. Neuropeptides (1999) 33(6), 447–456
450 Veronesi et al.
Substance P CGRP
20
10
Substance P 1 nM media
6 net intensity
IL-6 pg /ml
30
4
* *
2 0
–9
–8
–7 –6 molarity
a
–5
–4
Substance P CGRP
TNFα pg / ml
40
20
0
b
0
c
60
–9
–8
–7 molarity
–6
*
–5
IL-8 / GAPDH IL-6 / GAPDH TNFα / GAPDH
Fig. 2 (a,b) SP and CGRP stimulated IL-6 and TNFα release in BEAS–2B cells after 6 h exposure. This release followed a poorly-defined dose response curve. Graphs were compiled from data collected from six to eight separate experiments. Media control values (ranging from 2 to 5 pg/ml for IL-6 and 6 to 10 pg/ml for TNFα) were subtracted from all treatment groups. All treatment values were significantly different from controls (P<0.05). (c) PCR data indicated that SP (1 nM) significantly stimulated transcripts for the pro-inflammatory cytokines IL-8 and IL-6 after 2 h exposure. GAPDH, a ‘house-keeping’ gene, is shown as an internal control. Significance (P<0.05) over media controls is denoted by an asterisk (*).BEAS-2B cells were grown to 85–90% confluency on plastic 24 well plates.
Cytokine measurements
Statistics
Exposure media were removed from the exposed BEAS2B cells and held in –86°C before being analysed for the inflammatory cytokines IL-6, IL-8, or TNFα by ELISA (R&D Systems, Minneapolis, MN, USA). Plates were read on a Molecular Device Plate reader (Sunnyvale, CA, USA) and the data analysed, using its SoftMax Pro 1.2 software. Cytokine protein was measured in media taken from 6–8 wells/exposure and was calculated in pg/ml concentrations after being compared to an ‘in plate’ quadratic standard curve. The detectable concentration was typically < 0.7 pg/ml for IL-6, <4.4 pg/ml for TNFα, and <10 pg/ml for IL-8 (R&D Systems). IL-6 and IL-8 were measured interchangeably as markers of inflammation.12,14
Cytokine and transcript data were obtained from multiple6–8 samples from at least three to four separate experiments (i.e. trials). Data analysis was carried out using one-way ANOVA followed by Newman-Keull’s post-comparison test. For analysis of calcium measurements, Student’s t-test was used for pair-wise comparisons. The software package Instat 2 (Prism-3, San Diego, CA, USA) was used for the statistical analysis. Data were expressed as mean +/– SEM and statistically significant differences (P < 0.05) were denoted with an asterisk.
Neutral red cytotoxicity assay Neutral Red (NR) was prepared as a 0.4% aqueous solution. After washing the exposed cells with HBSS, KGM (200 µl), containing 50 µg/ml NR, was added to each well and incubated at 37°C for 3 h. Cells exposed to 0.1% saponin solution for 15 min provided a background reading of 100% lethality. After incubation, the dye media were aspirated and the cells washed with 200 µl of formol-calcium solution for 2–3 min. After removal of this fixative, 200 µl of 1% acetic acid/50% ethanol were added to each well at room temperature for 15 min to extract the extraneous dye. Plates were read at 540 nm absorbance. Neuropeptides (1999) 33(6), 447–456
RESULTS Neuropeptides stimulate intracellular calcium increases and cytokine expression Activation of the NK-1 receptor results in increases in [Ca2+]i (see Discussion). Effects of low (10 nM) and high (10 mM) (data not shown) concentrations of SP on [Ca2+]i were measured in Fluo-3 loaded epithelial cells. Both concentrations produced a rapid increase of [Ca2+]i (~ 10 sec). After a 4–5 min wash, a second exposure to SP (10 nM) produced a smaller response than the initial peak increase of [Ca2+]i (Fig. 1a). The [Ca2+]i response to 10 nM SP was partly dependent on extracellular calcium since in a calcium and magnesium-free medium, SP induced a transient increase in [Ca2+]i which increased to a plateau value when the exposure medium was changed to one containing SP and 2 mM calcium, and © 1999 Harcourt Publishers Ltd
Neuropeptides and human bronchial epithelial cells 451
a
c
SUBSTANCE P SP 4-11 SP 1-4
50
SP 10 µM SP 1 nM SP 10 µM + CP-99,994 1 nM SP 1 nM + CP-99,994 1 CP-99,994 M media
c
IL-6 pg / ml
40 6 30 4
10
* # 2 * # –8
–7
–6
–5
–4 0
molarity
b 600
450
SUBSTANCE P 1 nM SP + CP-99,9994
300 * *
*
–10
–9
150
0
SP
–8
molarity
returned again to baseline values when the medium was changed to a calcium-magnesium free medium (Fig. 1b). Exposure of cells to CGRP (10 nM) also caused a rapid increase of [Ca2+]i (Fig. 1c). Approximately 40% of neurons responded to 10 nM concentrations of SP or CGRP. Exposure (6 h) of BEAS–2B cells (80–85% confluent, Passage 35–50) to SP or CGRP stimulated a low, but statistically significant, release of IL-6 (Fig. 2a) and TNFα (Fig. 2b) over a range of concentrations (1 nM–10 µM). Release of both cytokines followed a poorly defined doseresponse curve. One explanation for this is that BEAS–2B and other epithelial cells contain high concentrations of neutral endopeptidase 24.11 (NEP), an enzyme which rapidly deactivates SP in situ.16,17 To test whether NEP inhibition would increase IL-6 release, we exposed cells (95–100% confluent, Passage 78–80) to several concentrations (1 nM–10 µM) of SP in the presence or absence of thiorphan (100 nM), an inhibitor of NEP. Although a significant 2–3-fold increase in IL-6 release occurred in the thiorphan-supplemented SP exposure media, relative to SP alone (data not shown), a well defined doseresponse curve did not occur, suggesting that additional neuropeptide-deactivating factors are present in respiratory epithelial cells.16,17 Using RT–PCR, BEAS–2B cells, exposed for 2 h to SP (1 nM) in thiorphan-supplemented © 1999 Harcourt Publishers Ltd
* * # #
* #
20
0 –9
IL-8 pg / ml
* #
IL-8 / GAPDH IL-6 / GAPDH TNFα / GAPDH
Fig. 3 (a) Separate groups of BEAS–2B cells were exposed to equimolar concentrations of SP, its amino (SP 1–4) or its carboxyl (SP 4–11) terminal fragments to determine which fragment had potency to stimulate IL-6 release. The IL-6 release in response to SP 1–4 was not significantly different from media controls. In contrast, all concentrations of SP and the more carboxyl end of the SP molecule (SP 4–11) significantly stimulated IL-6 release. Graphs were compiled from data collected from four separate experiments. Media values (ranging from 2 to 4 pg/ml) were subtracted from all treatment groups. (b) BEAS–2B cells were also pretreated with 0.1–10 nM CP-99 994, a non-peptide, competitive antagonist of the NK-1 receptor before exposure to SP (1 nM). IL-8 release was significantly reduced at all concentrations of the antagonist. Graphs were compiled from data collected from three to four separate experiments. Media values (ranging from 20 to 24 pg/ml) are subtracted from all treatment groups. (c) PCR data on inflammatory cytokine transcript levels indicated that transcripts for IL-8, IL-6 and TNFα were significantly depressed in cultures treated with CP-99 994 (1 nM) before a 2 h exposure to SP (10 µM, 1 nM). Data were derived from at least three separate experiments, consisting of three individual exposure wells per treatment group. # indicates significance (P < 0.05) from media controls and asterisks (*) indicate significance (P < 0.05) from SP treated cells. Media control values are shown.
media showed significant synthesis of IL-6, IL-8 and TNFα inflammatory cytokine transcripts (Fig. 2c). Receptor-mediated cytokine release Neuropeptides can stimulate cytokine release in immune and non-immune cells in both a conventional (i.e. NK-1 receptor mediated) and non-conventional fashion (see Discussion). To determine if the release of cytokines by SP was through a receptor mediated event, two approaches using SP-fragments and receptor antagonists were taken. Since, the SP preferring, NK-1 receptor shows affinity for the carboxyl end of the SP molecule,18,19 cells were exposed to serial concentrations (1 pM–100 µM) of either SP, or its terminal fragments SP1–4 and SP 4–11 in regular KGM media. IL-6 was measured after these exposures to determine which fragment had potency to stimulate IL-6 release. The data (Fig. 3a) indicate that the carboxyl end of the SP molecule had significant, cytokine stimulatory properties, that were qualitatively and quantitatively equivalent to SP, whereas, SP 1–4 did not stimulate IL-6 release. The second approach used a Neuropeptides (1999) 33(6), 447–456
452 Veronesi et al.
TNFα TNFα+ 0.5 µM SP TNFα + 0.5 µM CGRP
150
IL-6 pg / ml
exposures produced low levels of IL-6 release, as shown in Figure 2a. After 30 min exposure to each neuropeptide concentration, TNFα was added to the exposure media in combination with the neuropeptide for an additional 5.5 h exposure. IL-6 release in response to these combinations was compared to levels induced by TNFα alone. These data indicated that CGRP stimulated TNFαinduced IL-6 release at all test concentrations, in an additive fashion. However, SP appeared to synergistically stimulate this IL-6 release by ~5-fold when combined with TNFα (25 units/ml).
100
50
0
Capsaicin stimulates cytokine release
0
3
6
12.5
25
TNF α units / ml
Fig. 4 SP and CGRP interacted with TNFα to produce IL-6 release. Cells were exposed for 6 h to TNFα (3–25 units/ml) alone to establish a concentration-response curve. Separate plates of cells were pre-exposed for 30 min to SP or CGRP (5 µM). This was followed by exposure to TNFα in combination with SP or CGRP. Significant, but additive increases in IL-6 occurred at all concentrations of TNFα in combination with CGRP. However, IL-6 release was stimulated by 5-fold by SP in combination with the TNFα (25 units/ml). Graphs were compiled from data collected from five to six separate experiments. Media values (ranging from 5 to 8 pg/ml) were subtracted from all treatment groups.
non-peptide, competitive antagonist for the NK-1 receptor, CP-99 994.13 Cells were exposed to CP-99 994 for 30 min (0.1–1–10 nM) before being exposed to SP in combination with the antagonist for an additional 5.5 h. Figure 3b indicates that all tested antagonist concentration significantly reduced IL-8 release in cells exposed to SP (1 nM) at all tested antagonist concentrations. These data were supported by RT–PCR which demonstrated significant reduction of IL-6, TNFα and IL-8 transcripts in cells pretreated with 1 nM of CP-99 994 before exposure to SP (10 µM, 1 nM) (Fig. 3c). Collectively, these data suggested that physiologically relevant concentrations of SP stimulated inflammatory cytokines and their transcripts through activation of the NK-1 receptor. SP and TNFα Experiments were designed to address the possibility that SP and CGRP could modulate inflammation by interacting with other cytokines to stimulate moderate amounts of IL-6 release (Fig. 4). A concentrationresponse (3–25 units/ml) curve, measuring IL-6 release, was established for TNFα. Separate plates of cells were pre-exposed for 20–30 min to SP or CGRP (5 µM). These Neuropeptides (1999) 33(6), 447–456
In addition to direct and synergistic stimulation of cytokine release by neuropeptides, the botanical irritant, capsaicin (10 µM) also stimulated the release of IL-6 (data not shown) and IL-8 from BEAS-2B cells within 4 h exposure (Fig. 5a). This release could be significantly reduced (> 75%) by pre-incubation for 15 min with CPZ, an antagonist of capsaicin receptors.20,21 The capsaicin receptor is associated with an ion channel that mediates calcium entry from external sources.22 IL-8 release was reduced by 90% in cells exposed to capsaicin (10 µM) for 4 h in calcium-magnesium free media. Single cell recordings of BEAS–2B cells exposed to capsaicin showed increases in [Ca2+]i that were reduced in amplitude by repeated exposures, suggesting receptor desensitization (Fig. 5b). Increases in [Ca2+]i in response to capsaicin (10 µM) were totally inhibited by pretreatment with CPZ (15 µM) (Fig. 5c) and absent in cells exposed to capsaicin in calcium-magnesium free media (Fig. 5d). Cell viability, as measured by the neutral red assay, indicated that the above treatments were non-cytotoxic (data not shown). DISCUSSION Data from the present study indicate that human bronchial epithelial cells respond to SP, CGRP and the botanical irritant, capsaicin by inducing increases in [Ca2+]i and subsequent synthesis and release of inflammatory cytokines. These changes can be subserved by various pathways. For example, SP can activate cells by conventional binding to NK-1 receptors (via 7 TM G- protein) or through non-conventional means. In the latter mechanism, the amphiphilic SP molecule can activate G-proteins by inserting its cationic, N-terminal region into the lipid bilayers and directly triggering G-proteins. Although this type of G-protein activation is seen in many of the neuroimmune functions of SP (e.g. mast cell degranulation, stimulation of cytokine release from lymphocytes and monocytes),23–26 our data suggest that SP stimulates inflammatory cytokine release from BEAS–2B cells through a conventional NK-1 receptor © 1999 Harcourt Publishers Ltd
Neuropeptides and human bronchial epithelial cells 453
Capsaicin 10 µM Capsaicin 10 µM / CPZ 30 µM Capsaicin minus calcium and magnesium
a
160 140 525 nm intensity
IL-8 pg / ml
500
c
400 300 200
*
100
120 100 80
wash 10 min
60
*
CPZ 15 µM CAPS 10 µM
40 CAPS 10 µM
0
0
50
700
b
IO
750
800
time (s) 90
d 70 60 50
wash 2 min
wash 7 min
40 CAPS 10 µM 0
100
CAPS 10 µM 600
700
IO 900
time (s)
525 nm intensity
525 nm intensity
80 200 150 100 2+
CA
50 0
free CAPS 10 µM
50
100 time (s)
IO 150
Fig. 5 (a) Capsaicin (10 µM) stimulated IL-8 release after 4 h exposure. This release was significantly (*) reduced (>75%) by pre-exposure to CPZ (30 µM), an antagonist of the vanilloid receptor(s). BEAS–2B cells, exposed to capsaicin (10 µM) in calcium-magnesium free media, failed to release IL-8 which suggested that extracellular calcium sources were required for this release. Graphs were compiled from data collected from three to four separate experiments. Media values (ranging from 5 to 7 pg/ml) were subtracted from all treatment groups. (b) [Ca2+]i levels were measured in response to capsaicin (CAPS). Cells, exposed to 10 µM CAPS produced a rapid increase in [Ca2+]i that declined over time. Washing of the cells with vehicle caused the [Ca2+]i to return to baseline levels. After 7 min washing, a second exposure to CAPS (10 µM) increased [Ca2+]i but with a reduced response amplitude, suggesting receptor desensitization. (c) Effect of CPZ on the capsaicin (CAPS) induced [Ca2+]i response. Increases in [Ca2+]i did not occur in cells that were pre-exposed CPZ (15 µM) for 10 min before stimulation with CAPS (10 µM) in the presence of CPZ. Reapplication of CAPS after 10 min washing with control vehicle caused an immediate increase in [Ca2+]i. (d) Effects of extracellular calcium and magnesium on the capsaicin (CAPS) induced [Ca2+]i response. Increases of [Ca2+]i did not occur when cells were bathed in calcium-magnesium free HEPES buffer. When the wash solution was changed to one containing CAPS (10 µM) in calcium-magnesium media, an immediate increase in [Ca2+]i occurred. Changing the solution from calcium-magnesium free to regular vehicle in the absence of CAPS did not cause significant effects on the [Ca2+]i. Exposures to test solutions are indicated by the horizontal bars. Between exposures, the cells were washed with control HEPES buffer or the test solution, as indicated. IO (2 µM) was applied to determine maximal [Ca2+]i responses.
binding. This is evidenced by the equipotency of SP and its carboxyl fragment SP 4–11 to release IL-6 and the reduction of inflammatory cytokines and transcripts by the specific pharmacological NK-1 receptor antagonist, CP-99 994.13 The NK-1 receptor is a member of the superfamily of 7 TM G-protein coupled receptors whose activation by SP results in phosphoinositol breakdown, increases in IP3-mediated [Ca2+]i from internal stores and an extracellular influx of calcium.27,15 Thus, the increases in [Ca2+]i noted in BEAS–2B cells upon exposure to SP (Fig. 1a,b) are in accordance with NK-1 receptor activation. These increases declined upon a second expo© 1999 Harcourt Publishers Ltd
sure to SP, reflecting possible receptor desensitization. This phenomenon is in agreement with characteristics of the NK-1 receptor, which internalizes after stimulation.28 In addition to direct stimulation of cytokine formation, SP appears to stimulate cytokine release by interacting synergistically with one cytokine, to enhance the production of another. TNF α, an early and crucial player in the inflammatory cascade, acts on various immune and non-immune cell types to signal the release of cytokines.12,14–29 Physiologically speaking, the timing of SP and TNFα release in the inflammatory cascade occurs in tandem, with the initial release of relatively high Neuropeptides (1999) 33(6), 447–456
454 Veronesi et al.
concentrations of SP into the microenvironment, followed by TNFα synthesis and release, shortly thereafter.30 In keeping with SP’s physiological role as a modulator of inflammation, our data (Fig. 4) indicate that physiologically relevant (0.5 µM) concentrations of SP potentiate TNFα to produce 5-fold increases in IL-6. Although the exact manner in which SP and TNFα interact in BEAS–2B cells to amplify IL-6 synthesis and/or release is presently unknown, several reports indicate that SP can directly stimulate TNFα transcript in human nasal mucosa,31 cultured human astrocytes,32 rat uterine mast cells,33 and in both a murine mast cell line (CFTL12) and peritoneal mast cells.29 The effect of SP on TNFα appears to be NK-1 receptormediated, since SP and its carboxyl fragment SP 4–11 (but not SP 1–4) induce significant production of TNFα in the mononuclear cells of whole blood34 and in human monocyte-derived macrophages.19 Our present studies add to these data by showing that SP induces TNFα transcript in a human bronchial epithelial cell line, and that formation of this transcript can be blocked by the NK-1 receptor antagonist, CP-99 994 (Fig. 3c). The existence of bronchial epithelial peptide receptor and their responsiveness to neuropeptides represents a new area of neuroimmune research that impacts on airway inflammation.35,36 Although airway epithelial cells are the first to encounter chemical xenobiotics, their role in expressing signs of airway inflammation is poorly understood. However, several reports show that neuropeptides can stimulate epithelial cells to release cytokines6,7 and eosinophil-chemotactic factors,37 and influence epithelial cell proliferation, growth and migration.38–40 Neuropeptides also exert a protective effect on bronchial epithelial barrier functions41 and increase neutrophil adhesion to bronchial epithelial cells.42 Deleterious interactions can also occur between neuropeptides and their epithelial targets. For example, SP can stimulate eosinophils to secrete cytotoxic agents (e.g. major basic protein, leukotrienes, and platelet activating factor)43 which damage and denude the epithelial cell. This results in loss of the epithelial airway barrier and the direct exposure of nerve fibers and their receptors to airway irritants entering the lumen. Damage and loss of the epithelial lining also results in loss of NEP, the enzyme critical to SP deactivation. Loss of this enzyme prolongs the residency time and effectiveness of SP.17,44 BEAS–2B cells are known to house NEP–24.1116 and, as the present data show, its inhibition by thiorphan results in significant increases of SP-induced cytokine. Although the levels of IL-6 increased, the poorly defined dose-response curve of SP/CGRP induced IL-6 noted in non-supplemented exposure media did not improve in the presence of thiorphan. This might be explained by noting that SP can be hydrolyzed in vitro by various peptidases (e.g. dipeptidyl aminopeptidase B, depiptidyl aminopeptidase Neuropeptides (1999) 33(6), 447–456
IV, post-proline cleaving enzyme, serum acetylcholinesterase and pseudocholinesterase, acid proteinase, SP endopeptidase and NEP–24.11) and at a variety of peptide bonds.17 This suggests that the BEAS–2B cell line may contain additional enzymes that can deactivate SP and CGRP. It should also be noted that the enzyme activity of NEP is located in the brush border of the respiratory epithelial cell, rather than housed in cytoplasmic granules, as is often seen in neuronal cells. This proximity of the enzyme to the site of SP-release facilitates the instantaneous deactivation of neuropeptides both in situ and in culture systems. Baluk and colleagues4 first reported that serous epithelium responded to capsaicin with the release of CGRP granules housed in their cytoplasm. We have extended this observation by demonstrating that BEAS–2B cells respond to capsaicin with immediate increases in [Ca2+]i and subsequent cytokine release through capsaicin-sensitive pathways, since both events were blocked by CPZ and reduced by exposure in calcium-magnesium free media. Capsaicin receptors belong to the vanilloid family and show overlapping characteristics with pH sensitive pathways or pH receptors. Functional capsaicin receptors have been reported on non-neuronal cells (e.g. mast cells and C6 glioma).45,46 The existence of capsaicin and neuropeptide receptors on BEAS–2B cells offers an exciting cell culture model to examine how other chemical irritants stimulate inflammatory changes in respiratory epithelial cells. In this regard, we have used the BEAS–2B cell to demonstrate the relevance of neuropeptide and capsaicin sensitive pathways to inflammatory changes associated with exposure to industrial residual oil fly ash.47 These studies show that expressions of epithelial cell inflammation (i.e. intracellular calcium increases, cytokine transcription and protein release) can be diminished or completely blocked by antagonists to the neuropeptide (i.e. SP, CGRP) and capsaicin receptors. Such studies link the initial cellular and subcellular events of pollutant-induced airway inflammation even closer to neurogenic factors. ACKNOWLEDGMENT This work has been sponsored in part by NIH grant DC01065. Disclaimer This manuscript has been reviewed by the National Health Effects Environmental Research Laboratory, US Environmental Protection Agency and approved for publication. Approval does not signify that the contents necessarily reflect the view and policies of the Agency, endorsement or recommendation for use. © 1999 Harcourt Publishers Ltd
Neuropeptides and human bronchial epithelial cells 455
REFERENCES 1. Baluk P. Neurogenic inflammation in skin and airways. J Investig Dermatol Symp Proc 1997; 2: 76–81. 2. Barnes PJ, Belvisi MG. Sensory Neuropeptides. In: Barnes PJ, Grunstein MM, Leff AR, Woolcock AJ (eds) Asthma, 1st edn. Philadelphia: Lippincott-Raven 1997; 1051–1063. 3. Barnes PJ. Neurogenic inflammation in airways. Int Arch Allergy Appl Immunol 1991; 94: 303–309. 4. Baluk P, Nadel JA, McDonald DM. Calcitonin gene-related peptide in secretory granules of serous cells in the rat tracheal epithelium. Am J Respir Cell Mol Biol 1993; 8: 446–453. 5. Wuenschell CW, Sunday ME, Singh G, Minoo P, Slavkin HC, Warburton D. Embryonic mouse lung epithelial progenitor cells co-express immunohistochemical markers of diverse mature cell lineages. J Histochem Cytochem 1996; 44: 113–123. 6. Mullol J, Baraniuk JN, Logun C, Benfield T, Picado C, Shelhamer JH. Endothelin-1 induces GM-CSF, IL-6 and IL-8 but not G-CSF release from a human bronchial epithelial cell line (BEAS–2B). Neuropeptides 1996; 30: 551–556. 7. Mullol J, Baraniuk JN, Pitale M et al. Vasoactive intestinal peptide (VIP) induces IL-6 and IL-8, but not G-CSF and GM-CSF release from a human bronchial epithelial cell line. Neuropeptides 1997; 31: 119–124. 8. Reddel RR, Ke Y, Gerwin BI et al. Transformation of human bronchial epithelial cells by infection with SV40 or adenovirus12 SV40 hybrid virus, or transfection via strontium phosphate coprecipitation with a plasmid containing SV40 early region genes. Cancer Res 1988; 48: 1904–1909. 9. McKinnon KP, Madden MC, Noah TL, Devlin RB. In vitro ozone exposure increases release of arachidonic acid products from a human bronchial epithelial cell line. Toxicol Appl Pharmacol 1993; 118: 215–223. 10. Carter JD, Ghio AJ, Samet JM, Devlin RB. Cytokine production by human airway epithelial cells after exposure to an air pollution particle is metal-dependent. Toxicol Appl Pharmacol 1997; 146: 180–108. 11. Folkerts G, Nijkamp FP. Airway epithelium: more than just a barrier! Trends Pharmacol Sci 1998; 19: 334–341. 12. Takizawa H. Airway epithelial cells as regulators of airway inflammation. Int J Mol Med 1998; 1: 367–378. 13. McLean S, Ganong A, Seymour PA et al. Pharmacology of CP-99 994, a non-peptide antagonist of the tachykinin neurokinin-1 receptor. J Pharmacol Exp Ther 1993; 267: 472–479. 14. Khair OA, Davies R J, Devalia JL. Bacterial-induced release of inflammatory mediators by bronchial epithelial cells. Eur Respir J 1996; 9: 1913–1922. 15. Burcher E, Mussap CJ, Geraghty DP, McClure-Sharp JM, Watkins DJ. Concepts in characterization of tachykinin receptors. Ann NY Acad Sci 1991; 632: 123–136. 16. Proud D, Subauste MC, Ward PE. Glucocorticoids do not alter peptidase expression on a human bronchial epithelial cell line. Am J Respir Cell Mol Biol 1994; 11: 57–65. 17. Krause JE. On the physiological metabolism of Substance P. In: Jordan CC, Oehme PP (eds) Substance P: Metabolism and Biological Actions. London: Taylor and Francis, 1985; 13–32. 18. Bar-Shavit Z, Goldman R, Stabinsky Y et al. Enhancement of phagocytosis – a newly found activity of Substance P residing in its N-terminal tetrapeptide sequence. Biochem Biophys Res Commun 1980; 94: 1445–1451. 19. Ho WZ, Stavropoulos G, Lai JP et al. Substance P C-terminal octapeptide analogues augment tumor necrosis factor-alpha release by human blood monocytes and macrophages. J Neuroimmunol 1998; 82: 126–132.
© 1999 Harcourt Publishers Ltd
20. Bevan S, Hothi S, Hughes G et al. Capsazepine: a competitive antagonist of the sensory neurone excitant capsaicin. Br J Pharmacol 1992; 107: 544–552. 21. Belvisi MG, Miura M, Stretton D, Barnes PJ. Capsazepine as a selective antagonist of capsaicin-induced activation of C-fibres in guinea-pig bronchi. Eur J Pharmacol 1992; 215: 341–34. 22. Szallasi A, Blumberg PM. Vanilloid (capsaicin) receptors and mechanisms. Pharmacol Rev 1999; 51: 159–212. 23. Kavelaars A, Jeurissen F, Heijnen CJ. Substance P receptors and signal transduction in leukocytes. Immunomethods 1994; 5: 41–48. 24. Kavelaars A, Jeurissen F, von Frijtag D, Herman vR, Rijkers GT, Heijnen CJ. Substance P induces a rise in intracellular calcium concentration in human T lymphocytes in vitro: evidence of a receptor-independent mechanism. J Neuroimmunol 1993; 42: 61–70. 25. Mousli M, Bronner C, Bockaert J, Rouot B, Landry Y. Interaction of Substance P, compound 48/80 and mastoparan with the alpha-subunit C-terminus of G protein. Immunol Lett 1990; 25: 355–357. 26. Mousli M, Bueb JL, Bronner C, Rouot B, Landry Y. G protein activation: a receptor-independent mode of action for cationic amphiphilic neuropeptides and venom peptides. Trends Pharmacol Sci 1990; 11: 358–362. 27. Putney JW. Receptors and the Inositol Phosphate-Calcium Signaling System. In: Buck S H (ed) The Tachykinin Receptors. New Jersey: Humana Press, 1994; 257–284. 28. McConalogue K, Dery O, Lovett M et al. Substance P-induced trafficking of beta-arrestins. J Biol Chem 1999; 274: 16257–16268. 29. Ansel JC, Brown JR, Payan DG, Brown MA. Substance P selectively activates TNF-alpha gene expression in murine mast cells. J Immunol 1993; 150: 4478–4485. 30. Yoshikawa T, Kurimoto I, Streilein JW. Tumour necrosis factor-alpha mediates ultraviolet light B-enhanced expression of contact hypersensitivity. Immunology 1992; 76: 264–271. 31. Okamoto Y, Shirotori K, Kudo K et al. Cytokine expression after the topical administration of Substance P to human nasal mucosa. J Immunol 1993; 151: 4391–4398. 32. Palma C, Minghetti L, Astolfi M et al. Functional characterization of Substance P receptors on cultured human spinal cord astrocytes: synergism of Substance P with cytokines in inducing interleukin-6 and prostaglandin E2 production. Glia 1997; 21: 183–193. 33. Cocchiara R, Bongiovanni A, Albeggiani G, Azzolina A, Geraci D. Substance P selectively activates TNF-alpha mRNA in rat uterine immune cells: a neuroimmune link. Neuroreport 1997; 8: 2961–2964. 34. Nair MP, Schwartz SA. Substance P induces tumor necrosis factor in an ex vivo model system. Cell Immunol 1995; 166: 286–290. 35. Polito AJ, Proud D. Epithelia cells as regulators of airway inflammation. J Allergy Clin Immunol 1998; 102: 714–718. 36. Yu XY, Takahashi N, Croxton TL, Spannhake EW. Modulation of bronchial epithelial cell barrier function by in vitro ozone exposure. Environ Health Perspect 1994; 102: 1068–1072. 37. Koyama S, Sato E, Nomura H, Kubo K, Nagai S, Izumi T. Acetylcholine and Substance P stimulate bronchial epithelial cells to release eosinophil chemotactic activity. J Appl Physiol 1998; 84: 1528–1534. 38. Sanghavi JN, Rabe KF, Kim JS, Magnussen H, Leff A , White SR. Migration of human and guinea pig airway epithelial cells in response to calcitonin gene-related peptide. Am J Respir Cell Mol Biol 1994; 11: 181–187.
Neuropeptides (1999) 33(6), 447–456
456 Veronesi et al.
39. White SR, Garland A, Gitter B et al. Proliferation of guinea pig tracheal epithelial cells in coculture with rat dorsal root ganglion neural cells. Am J Physiol 1995; 268: L957–L965. 40. Reid TW, Murphy CJ, Iwahashi CK, Foster BA, Mannis MJ. Stimulation of epithelial cell growth by the neuropeptide Substance P. J Cell Biochem 1993; 52: 476–485. 41. Yu XY, Undem BJ, Spannhake EW. Protective effect of Substance P on permeability of airway epithelial cells in culture. Am J Physiol 1996; 271: L889–L895. 42. DeRose V, Robbins RA, Snider RM et al. Substance P increases neutrophil adhesion to bronchial epithelial cells. J Immunol 1994; 152: 1339–1346. 43. Barnes PJ. New concepts in the pathogenesis of bronchial hyperresponsiveness and asthma. J Allergy Clin Immunol 1989; 83: 1013–1026.
Neuropeptides (1999) 33(6), 447–456
44. Piedimonte G. Tachykinin peptides, receptors, and peptidases in airway disease. Exp Lung Res 1995; 21: 809–834. 45. Biro T, Maurer M, Modarres S et al. Characterization of functional vanilloid receptors expressed by mast cells. Blood 1998; 91: 1332–1340. 46. Biro T, Brodie C, Modarres S, Lewin NE, Acs P, Blumberg PM. Specific vanilloid responses in C6 rat glioma cells. Brain Res Mol Brain Res 1998; 56: 89–98. 47. Veronesi B, Oortgiesen M, Carter JD, Devlin RB. Particulate matter (PM) initiates inflammatory cytokine release by activation of capsaicin receptors in a human bronchial epithelial cell line. Toxicol Appl Pharmacol 1999; 154: 106–115.
© 1999 Harcourt Publishers Ltd
Neuropeptides and capsaicin stimulate the release of inflammatory cytokines in a human bronchial epithelial cell line B. Veronesi,1 J.D. Carter,1 R.B. Devlin,1 S.A. Simon,2 M. Oortgiesen2 1
National Health and Environmental Effects Research Laboratory, US Environmental Protection Agency, Research Triangle Park, 2Departments of Neurobiology and Anesthesiology, Duke University Medical Center, Durham, NC, USA
Summary The role of neuropeptides in initiating and modulating airway inflammation was examined in a human bronchial epithelial cell line (i.e. BEAS–2B). At a range of concentrations, exposure of BEAS–2B cells to Substance P (SP) or calcitonin gene related protein resulted in immediate increases in intracellular calcium ([Ca2+]i), the synthesis of the transcripts for the inflammatory cytokines, IL-6, IL-8 and TNFα after 2 h exposure, and the release of their proteins after 6 h exposure. Addition of thiorphan (100 nM), an inhibitor of neutral endopeptidase, enhanced the levels of SP-stimulated cytokine release. Stimulation of IL-6 by SP occurred in a conventional receptor-mediated manner as demonstrated by its differential release by fragments SP 4–11 and SP 1–4 and by the blockage of IL-6 release with the non-peptide, NK-1 receptor antagonist, CP-99 994. In addition to the direct stimulation of inflammatory cytokines, SP (0.5 µM), in combination with TNFα (25 units/ml), synergistically stimulated IL-6 release. BEAS–2B cells also responded to the botanical irritant, capsaicin (10 µM) with increases in [Ca2+]i and IL-8 cytokine release after 4 h exposure. The IL-8 release was dependent on the presence of extracellular calcium. Capsaicin-stimulated increases of [Ca2+]i and cytokine release could be reduced to control levels by pre-exposure to capsazepine, an antagonist of capsaicin (i.e. vanilloid) receptor(s) or by deletion of extracellular calcium from the exposure media. The present data indicate that the BEAS–2B human epithelial cell line expresses neuropeptide and capsaicin-sensitive pathways, whose activation results in immediate increases of [Ca2+]i stimulation of inflammatory cytokine transcripts and the release of their cytokine proteins. © 1999 Harcourt Publishers Ltd
INTRODUCTION Airway inflammation underlies various disorders associated with exposure to environmental xenobiotics. The cascade of cellular and subcellular events that compose this sequelae has been described primarily in vascular and immunological terms. However, more recent thinking suggests that neuropeptides initiate and sustain inflammation through the process known as neurogenic
Received 15 August 1999 Accepted 10 October 1999 Correspondence to: Dr Bellina Veronesi, US Environmental Protection Agency, National Health Effects and Environmental Research Laboratory, Neurotoxicology Division MD 74B, Cellular and Molecular Toxicology Branch, MD74B RTP, NC 27711, USA. Tel.: +1 919 541 5780; Fax: +1 919 541 4849; E-mail: [email protected]
inflammation.1–3 This process predominates in tissues (i.e. skin, airways, gut, conjunctiva, mucous membranes, etc.) that are most likely to encounter invading organisms, irritating stimuli or exogenous chemicals. Briefly, this process involves the activation of irritant pathways (e.g. capsaicin, pH sensitive, temperature-sensitive, mechano-receptors) which are found on the sensory C and Aδ fibers that innervate such tissues. In the airways, neuropeptides (e.g. Substance P [SP], calcitonin generelated protein [CGRP] and neurokinin A) that are released upon activation of these receptors, can initiate and modulate symptoms of inflammation by targeting both immune (e.g. lymphocytes, neutrophils, macrophages and eosinophils) and non-immune cells (e.g. smooth muscle and endothelial cells of the vasculature, epithelial cells that line the airway lumen). Neuropeptides act on these target cells directly or 447
448 Veronesi et al.
indirectly as vasodilators, vasoconstrictors, degranulators, chemotactic factors, mitogens, adherents and secretogues. In the airway lumen, sensory fibers, containing both irritant and neuropeptide receptors, terminate between and below the respiratory epithelial cells. Epithelial cells lining the lumen are the first to encounter and respond to irritating airborne stimuli, making their interactions with irritants and neuropeptides a pivotal event in the pathophysiology of neurogenic inflammation in the airways. It has been reported that various types of respiratory epithelial cells contain neuropeptides. For example, neuropeptide immunostaining is found in neuroendocrine cells (i.e. neuroepithelial bodies), Clara cells, Type II alveolar cells, mast cells and serous epithelial cells.4–7 The present study extends these data and demonstrates that SP, CGRP and capsaicin can directly stimulate the release of inflammatory cytokines in a receptor-mediated fashion from a human bronchial epithelial cell line (i.e. BEAS–2B). MATERIALS AND METHODS Cells and maintenance BEAS–2B cells are SV-40/adenovirus-transformed human bronchial epithelial cells8 that have been used in numerous studies to evaluate pulmonary toxicants.9,10 These cells, like most epithelial cells, are highly influenced by their passage and confluency state.11,12 For these studies, cells were obtained commercially (ATTC, Rockville, MD, USA) and used from passage 40–80. Cells were maintained in complete keratinocyte growth media (KGM) consisting of KBM (Clonetics, San Diego, CA, USA) and supplemented with bovine pituitary extract, human epidermal growth factor (5 ng/ml), hydrocortisone (0.5 mg/ml), ethanolamine (0.1 mM), phosphoethanolamine (0.1 mM) and insulin (5 mg/ml). Complete KGM minus calcium and magnesium was purchased from Clonetics (Cat # CC–3004). For cytokine and transcript experiments, cells were grown to 85–95% confluency in 96 or 12 well Costar culture plates. For calcium experiments, cells were grown to 80–90% confluency on fibronectin-coated, 22 mm, # 0 thickness glass coverslips (Carolina Biological Supply, Burlington, NC, USA). Chemicals SP, CGRP, thiorphan, Hanks Balanced Salt Solution (HBSS), M-[2-hydroxyethyl] piperanine-N-{4-butane sulfonic acid (HEPES) and 2-[N-morpholinoethanesulfonic acid] (MES) were purchased from Sigma Chemical Corporation. Capsaicin and its receptor antagonist, capsazepine (CPZ), were purchased from RBI (Natick, MA, Neuropeptides (1999) 33(6), 447–456
USA). The non-peptide, NK-1 (SP-preferring) competitive receptor antagonist, CP–99 994 (13) was a generous gift from Pfizer Company (Grouton, CT, USA). Recombinant TNFα was obtained commercially (R&D Systems, Minneapolis, MN, USA). All chemicals were dissolved at test concentrations in their respective vehicles (i.e. nutrient media, HBSS) immediately before use.
Intracellular calcium measurements Increases in [Ca2+]i were recorded using the fluorescent probe Fluo-3-AM (Molecular Probes, Eugene, OR, USA). In preparation for calcium measurements, cells were incubated for 40 min with Fluo-3-AM (2 µM) at 37°C and then washed for 10 min in HEPES-buffer containing 150 mM NaCl, 5 mM KCI, 2 mM CaCI2, 1 mM MgCl2, 10 mM D-glucose, and 10 mM HEPES. Cells were illuminated with the exciting wavelength of 485 nm and the fluorescent emission was detected at 525 nm using Axon Workbench Imaging software (Axon Instruments, Foster City, CA, USA). After stable baseline recordings were collected, cells were exposed to the test compounds and washed between exposures with HEPES-buffer by rapid superfusion (15 ml/min). Antagonists were tested by preincubating cells for 5–10 min. Then, the test compound was exposed to the cells in the presence of the antagonist. Maximal [Ca2+]i increases were obtained at the end of each experiment by the addition of 2 µM ionomycin (IO). Traces of cells that showed strong IO responses but showed no marked bleaching were collected from 10–20 cells and from three to five different experiments. Experiments were carried out at 25–26°C.
Reverse transcription-polymerase chain reaction (RT–PCR) BEAS–2B (85–90% confluency) were exposed for 2 h to the test compound, washed twice with phosphatebuffered saline (PBS) and harvested in a guanidine thiocyanate-containing lysis buffer (GITC) as previously described.10 RNA was prepared by lysing cells in GITC buffer containing 4 M guanidine isothiocyanate, 25 mM sodium citrate (pH 7.0), 0.5% Sarkosyl, and 10 mM DTT. The lysate was sheared with a 22 gauge needle and layered over an equal volume of 5.7 M cesium chloride and 0.1 M EDTA. Total RNA was pelleted by centrifugal sedimentation at 114 000 g for 2 h. The relative concentrations of mRNAs coding for IL-6, IL-8, and TNFα were estimated by sequential reverse transcription of 200 ng RNA from the cultured cells and subsequent cDNA amplification (RT–PCR). Specific cDNA sequences within the mixture were amplified as previously described.10 Amplifications were performed under oil in 96 well plates using a PTC-100 programmable thermal cycler (MJ © 1999 Harcourt Publishers Ltd
Neuropeptides and human bronchial epithelial cells 449
100
c
90
160 140
70
525 nm intensity
525 nm intensity
80
60 50 40 wash
30 20
SP (10 nM)
wash
SP (10 nM)
IO
10 0
120 100 80 60 40
0
50
100
150
600
650
900
950
time (s)
a
CGRP 10 nM
IO
20
1000
0
c
0
50
100
150
time (s)
160 525 nm intensity
140 120 100 80 60 Ca2+ -free
40 20
Ca2+ -free IO
SP (10 nM)
0 0
50
100
150
200
time (s)
b Research, Watertown, MA, USA). Following an initial 3 min denaturation at 92°C, amplifications were cycled for 1 min at 92°C, 1 min at 56°C, and 2 min at 72°C. The optimum number of cycles varied with the target cDNA and was determined empirically. cDNA was amplified for glyceraldehyde-3-phosphate dehydrogenase (GAPDH), a housekeeping gene and the inflammatory cytokine transcripts for IL-8, IL-6 and TNFα, using gene-specific primers. Relative concentrations of specific DNA sequences were quantified by separating 15 µl of each amplification mixture through 2% agarose gels in 0.5× Trisborate, EDTA buffer. The gel was stained in 1 µg/ml ethidium bromide and photographed under UV illumination with Polaroid type 55 P/N film (Polaroid, Cambridge, MA, USA). The negatives were scanned using a Kodak DC120 Zoom Digital camera. The Kodak Electrophoresis Documentation and Analysis System 120 (Kodak Corp., Rochester, NY, USA) was used to determine the intensities of amplification products that resolved as single bands of the predicted size. The net intensities of IL-6, IL-8 and TNFα DNA bands were divided by the net intensity of the GAPDH-DNA to normalize for any errors in the amount of input RNA, day to day variations in amplification efficiency, gel staining or photography.
Oligonucleotide sequences were synthesized using an Applied Biosystems 391 DNA synthesizer (Foster City, CA, USA) based on sequences published in © 1999 Harcourt Publishers Ltd
Fig. 1 (a) Intracellular calcium ([Ca2+]i) levels were measured in BEAS–2B cells using the fluorescent probe Fluo-3 AM. Exposure of BEAS–2B to SP (10 nM) produced a rapid increase in [Ca2+]i, followed by a lower, longer lasting plateau. At both concentrations, approximately 40% of tested cells responded. A second exposure of the cells to SP (10 nM) produced a [Ca2+]i response with a smaller peak amplitude, which suggested receptor desensitization or internalization. Between the two SP exposures, cells were washed for 7 min with control HEPES buffer. IO (2 µM) was applied at the end of each experiment to determine the maximal [Ca2+]i increase. (b) The increase in [Ca2+]i was dependent on extracellular calcium and magnesium since, in media lacking these ions, SP produced only a transient increase in [Ca2+]i which increased when the superfusion medium was changed to one containing SP and 2 mM calcium-magnesium and returned again to baseline when the medium was changed to one lacking calcium-magnesium. IO (2 µM) was applied at the end of each experiment to determine the maximal [Ca2+]i increase. (c) Exposure of BEAS–2B to CGRP (10 nM) produced an increase in cytoplasmic calcium, which declined over 1–5 min. The percentage of responding cells varied and amounted to 20–60%. Application of IO (2 µM) was used to determine the maximal calcium increase. Exposure to CGRP is indicated by horizontal bars.
GenBank human DNA database. The following sense and anti-sense sequences were employed: GAPDH: 5′CCATGGAGAAGGCTGGGG 3′ and 5′CAAAGTTGTCATGGATGACC 3′ IL-8: 5′AACCCTCTGCACCCAGTTTTCCTT3′ and 5′TCCACTCTCAATCACTCTCA 3′ IL-6: 5′CTTCTCCACAAGCGCCTTC 3′ and 5′GGCAAGTCTCCTCATTGAATC 3′ TNFα: 5′CAGGCAGTCAGATCATCTTC 3′ and 5′CTTGATGGCAGAGAGGAGGT 3′ The oligonucleotide primers used for quantitation were designed to amplify segments of the mature mRNA derived from two or more gene exons, so that amplification of genomic sequences would yield products larger than those derived from mature mRNA. Only amplification products of the size predicted by the mature mRNA sequence were seen and quantified. Neuropeptides (1999) 33(6), 447–456
450 Veronesi et al.
Substance P CGRP
20
10
Substance P 1 nM media
6 net intensity
IL-6 pg /ml
30
4
* *
2 0
–9
–8
–7 –6 molarity
a
–5
–4
Substance P CGRP
TNFα pg / ml
40
20
0
b
0
c
60
–9
–8
–7 molarity
–6
*
–5
IL-8 / GAPDH IL-6 / GAPDH TNFα / GAPDH
Fig. 2 (a,b) SP and CGRP stimulated IL-6 and TNFα release in BEAS–2B cells after 6 h exposure. This release followed a poorly-defined dose response curve. Graphs were compiled from data collected from six to eight separate experiments. Media control values (ranging from 2 to 5 pg/ml for IL-6 and 6 to 10 pg/ml for TNFα) were subtracted from all treatment groups. All treatment values were significantly different from controls (P<0.05). (c) PCR data indicated that SP (1 nM) significantly stimulated transcripts for the pro-inflammatory cytokines IL-8 and IL-6 after 2 h exposure. GAPDH, a ‘house-keeping’ gene, is shown as an internal control. Significance (P<0.05) over media controls is denoted by an asterisk (*).BEAS-2B cells were grown to 85–90% confluency on plastic 24 well plates.
Cytokine measurements
Statistics
Exposure media were removed from the exposed BEAS2B cells and held in –86°C before being analysed for the inflammatory cytokines IL-6, IL-8, or TNFα by ELISA (R&D Systems, Minneapolis, MN, USA). Plates were read on a Molecular Device Plate reader (Sunnyvale, CA, USA) and the data analysed, using its SoftMax Pro 1.2 software. Cytokine protein was measured in media taken from 6–8 wells/exposure and was calculated in pg/ml concentrations after being compared to an ‘in plate’ quadratic standard curve. The detectable concentration was typically < 0.7 pg/ml for IL-6, <4.4 pg/ml for TNFα, and <10 pg/ml for IL-8 (R&D Systems). IL-6 and IL-8 were measured interchangeably as markers of inflammation.12,14
Cytokine and transcript data were obtained from multiple6–8 samples from at least three to four separate experiments (i.e. trials). Data analysis was carried out using one-way ANOVA followed by Newman-Keull’s post-comparison test. For analysis of calcium measurements, Student’s t-test was used for pair-wise comparisons. The software package Instat 2 (Prism-3, San Diego, CA, USA) was used for the statistical analysis. Data were expressed as mean +/– SEM and statistically significant differences (P < 0.05) were denoted with an asterisk.
Neutral red cytotoxicity assay Neutral Red (NR) was prepared as a 0.4% aqueous solution. After washing the exposed cells with HBSS, KGM (200 µl), containing 50 µg/ml NR, was added to each well and incubated at 37°C for 3 h. Cells exposed to 0.1% saponin solution for 15 min provided a background reading of 100% lethality. After incubation, the dye media were aspirated and the cells washed with 200 µl of formol-calcium solution for 2–3 min. After removal of this fixative, 200 µl of 1% acetic acid/50% ethanol were added to each well at room temperature for 15 min to extract the extraneous dye. Plates were read at 540 nm absorbance. Neuropeptides (1999) 33(6), 447–456
RESULTS Neuropeptides stimulate intracellular calcium increases and cytokine expression Activation of the NK-1 receptor results in increases in [Ca2+]i (see Discussion). Effects of low (10 nM) and high (10 mM) (data not shown) concentrations of SP on [Ca2+]i were measured in Fluo-3 loaded epithelial cells. Both concentrations produced a rapid increase of [Ca2+]i (~ 10 sec). After a 4–5 min wash, a second exposure to SP (10 nM) produced a smaller response than the initial peak increase of [Ca2+]i (Fig. 1a). The [Ca2+]i response to 10 nM SP was partly dependent on extracellular calcium since in a calcium and magnesium-free medium, SP induced a transient increase in [Ca2+]i which increased to a plateau value when the exposure medium was changed to one containing SP and 2 mM calcium, and © 1999 Harcourt Publishers Ltd
Neuropeptides and human bronchial epithelial cells 451
a
c
SUBSTANCE P SP 4-11 SP 1-4
50
SP 10 µM SP 1 nM SP 10 µM + CP-99,994 1 nM SP 1 nM + CP-99,994 1 CP-99,994 M media
c
IL-6 pg / ml
40 6 30 4
10
* # 2 * # –8
–7
–6
–5
–4 0
molarity
b 600
450
SUBSTANCE P 1 nM SP + CP-99,9994
300 * *
*
–10
–9
150
0
SP
–8
molarity
returned again to baseline values when the medium was changed to a calcium-magnesium free medium (Fig. 1b). Exposure of cells to CGRP (10 nM) also caused a rapid increase of [Ca2+]i (Fig. 1c). Approximately 40% of neurons responded to 10 nM concentrations of SP or CGRP. Exposure (6 h) of BEAS–2B cells (80–85% confluent, Passage 35–50) to SP or CGRP stimulated a low, but statistically significant, release of IL-6 (Fig. 2a) and TNFα (Fig. 2b) over a range of concentrations (1 nM–10 µM). Release of both cytokines followed a poorly defined doseresponse curve. One explanation for this is that BEAS–2B and other epithelial cells contain high concentrations of neutral endopeptidase 24.11 (NEP), an enzyme which rapidly deactivates SP in situ.16,17 To test whether NEP inhibition would increase IL-6 release, we exposed cells (95–100% confluent, Passage 78–80) to several concentrations (1 nM–10 µM) of SP in the presence or absence of thiorphan (100 nM), an inhibitor of NEP. Although a significant 2–3-fold increase in IL-6 release occurred in the thiorphan-supplemented SP exposure media, relative to SP alone (data not shown), a well defined doseresponse curve did not occur, suggesting that additional neuropeptide-deactivating factors are present in respiratory epithelial cells.16,17 Using RT–PCR, BEAS–2B cells, exposed for 2 h to SP (1 nM) in thiorphan-supplemented © 1999 Harcourt Publishers Ltd
* * # #
* #
20
0 –9
IL-8 pg / ml
* #
IL-8 / GAPDH IL-6 / GAPDH TNFα / GAPDH
Fig. 3 (a) Separate groups of BEAS–2B cells were exposed to equimolar concentrations of SP, its amino (SP 1–4) or its carboxyl (SP 4–11) terminal fragments to determine which fragment had potency to stimulate IL-6 release. The IL-6 release in response to SP 1–4 was not significantly different from media controls. In contrast, all concentrations of SP and the more carboxyl end of the SP molecule (SP 4–11) significantly stimulated IL-6 release. Graphs were compiled from data collected from four separate experiments. Media values (ranging from 2 to 4 pg/ml) were subtracted from all treatment groups. (b) BEAS–2B cells were also pretreated with 0.1–10 nM CP-99 994, a non-peptide, competitive antagonist of the NK-1 receptor before exposure to SP (1 nM). IL-8 release was significantly reduced at all concentrations of the antagonist. Graphs were compiled from data collected from three to four separate experiments. Media values (ranging from 20 to 24 pg/ml) are subtracted from all treatment groups. (c) PCR data on inflammatory cytokine transcript levels indicated that transcripts for IL-8, IL-6 and TNFα were significantly depressed in cultures treated with CP-99 994 (1 nM) before a 2 h exposure to SP (10 µM, 1 nM). Data were derived from at least three separate experiments, consisting of three individual exposure wells per treatment group. # indicates significance (P < 0.05) from media controls and asterisks (*) indicate significance (P < 0.05) from SP treated cells. Media control values are shown.
media showed significant synthesis of IL-6, IL-8 and TNFα inflammatory cytokine transcripts (Fig. 2c). Receptor-mediated cytokine release Neuropeptides can stimulate cytokine release in immune and non-immune cells in both a conventional (i.e. NK-1 receptor mediated) and non-conventional fashion (see Discussion). To determine if the release of cytokines by SP was through a receptor mediated event, two approaches using SP-fragments and receptor antagonists were taken. Since, the SP preferring, NK-1 receptor shows affinity for the carboxyl end of the SP molecule,18,19 cells were exposed to serial concentrations (1 pM–100 µM) of either SP, or its terminal fragments SP1–4 and SP 4–11 in regular KGM media. IL-6 was measured after these exposures to determine which fragment had potency to stimulate IL-6 release. The data (Fig. 3a) indicate that the carboxyl end of the SP molecule had significant, cytokine stimulatory properties, that were qualitatively and quantitatively equivalent to SP, whereas, SP 1–4 did not stimulate IL-6 release. The second approach used a Neuropeptides (1999) 33(6), 447–456
452 Veronesi et al.
TNFα TNFα+ 0.5 µM SP TNFα + 0.5 µM CGRP
150
IL-6 pg / ml
exposures produced low levels of IL-6 release, as shown in Figure 2a. After 30 min exposure to each neuropeptide concentration, TNFα was added to the exposure media in combination with the neuropeptide for an additional 5.5 h exposure. IL-6 release in response to these combinations was compared to levels induced by TNFα alone. These data indicated that CGRP stimulated TNFαinduced IL-6 release at all test concentrations, in an additive fashion. However, SP appeared to synergistically stimulate this IL-6 release by ~5-fold when combined with TNFα (25 units/ml).
100
50
0
Capsaicin stimulates cytokine release
0
3
6
12.5
25
TNF α units / ml
Fig. 4 SP and CGRP interacted with TNFα to produce IL-6 release. Cells were exposed for 6 h to TNFα (3–25 units/ml) alone to establish a concentration-response curve. Separate plates of cells were pre-exposed for 30 min to SP or CGRP (5 µM). This was followed by exposure to TNFα in combination with SP or CGRP. Significant, but additive increases in IL-6 occurred at all concentrations of TNFα in combination with CGRP. However, IL-6 release was stimulated by 5-fold by SP in combination with the TNFα (25 units/ml). Graphs were compiled from data collected from five to six separate experiments. Media values (ranging from 5 to 8 pg/ml) were subtracted from all treatment groups.
non-peptide, competitive antagonist for the NK-1 receptor, CP-99 994.13 Cells were exposed to CP-99 994 for 30 min (0.1–1–10 nM) before being exposed to SP in combination with the antagonist for an additional 5.5 h. Figure 3b indicates that all tested antagonist concentration significantly reduced IL-8 release in cells exposed to SP (1 nM) at all tested antagonist concentrations. These data were supported by RT–PCR which demonstrated significant reduction of IL-6, TNFα and IL-8 transcripts in cells pretreated with 1 nM of CP-99 994 before exposure to SP (10 µM, 1 nM) (Fig. 3c). Collectively, these data suggested that physiologically relevant concentrations of SP stimulated inflammatory cytokines and their transcripts through activation of the NK-1 receptor. SP and TNFα Experiments were designed to address the possibility that SP and CGRP could modulate inflammation by interacting with other cytokines to stimulate moderate amounts of IL-6 release (Fig. 4). A concentrationresponse (3–25 units/ml) curve, measuring IL-6 release, was established for TNFα. Separate plates of cells were pre-exposed for 20–30 min to SP or CGRP (5 µM). These Neuropeptides (1999) 33(6), 447–456
In addition to direct and synergistic stimulation of cytokine release by neuropeptides, the botanical irritant, capsaicin (10 µM) also stimulated the release of IL-6 (data not shown) and IL-8 from BEAS-2B cells within 4 h exposure (Fig. 5a). This release could be significantly reduced (> 75%) by pre-incubation for 15 min with CPZ, an antagonist of capsaicin receptors.20,21 The capsaicin receptor is associated with an ion channel that mediates calcium entry from external sources.22 IL-8 release was reduced by 90% in cells exposed to capsaicin (10 µM) for 4 h in calcium-magnesium free media. Single cell recordings of BEAS–2B cells exposed to capsaicin showed increases in [Ca2+]i that were reduced in amplitude by repeated exposures, suggesting receptor desensitization (Fig. 5b). Increases in [Ca2+]i in response to capsaicin (10 µM) were totally inhibited by pretreatment with CPZ (15 µM) (Fig. 5c) and absent in cells exposed to capsaicin in calcium-magnesium free media (Fig. 5d). Cell viability, as measured by the neutral red assay, indicated that the above treatments were non-cytotoxic (data not shown). DISCUSSION Data from the present study indicate that human bronchial epithelial cells respond to SP, CGRP and the botanical irritant, capsaicin by inducing increases in [Ca2+]i and subsequent synthesis and release of inflammatory cytokines. These changes can be subserved by various pathways. For example, SP can activate cells by conventional binding to NK-1 receptors (via 7 TM G- protein) or through non-conventional means. In the latter mechanism, the amphiphilic SP molecule can activate G-proteins by inserting its cationic, N-terminal region into the lipid bilayers and directly triggering G-proteins. Although this type of G-protein activation is seen in many of the neuroimmune functions of SP (e.g. mast cell degranulation, stimulation of cytokine release from lymphocytes and monocytes),23–26 our data suggest that SP stimulates inflammatory cytokine release from BEAS–2B cells through a conventional NK-1 receptor © 1999 Harcourt Publishers Ltd
Neuropeptides and human bronchial epithelial cells 453
Capsaicin 10 µM Capsaicin 10 µM / CPZ 30 µM Capsaicin minus calcium and magnesium
a
160 140 525 nm intensity
IL-8 pg / ml
500
c
400 300 200
*
100
120 100 80
wash 10 min
60
*
CPZ 15 µM CAPS 10 µM
40 CAPS 10 µM
0
0
50
700
b
IO
750
800
time (s) 90
d 70 60 50
wash 2 min
wash 7 min
40 CAPS 10 µM 0
100
CAPS 10 µM 600
700
IO 900
time (s)
525 nm intensity
525 nm intensity
80 200 150 100 2+
CA
50 0
free CAPS 10 µM
50
100 time (s)
IO 150
Fig. 5 (a) Capsaicin (10 µM) stimulated IL-8 release after 4 h exposure. This release was significantly (*) reduced (>75%) by pre-exposure to CPZ (30 µM), an antagonist of the vanilloid receptor(s). BEAS–2B cells, exposed to capsaicin (10 µM) in calcium-magnesium free media, failed to release IL-8 which suggested that extracellular calcium sources were required for this release. Graphs were compiled from data collected from three to four separate experiments. Media values (ranging from 5 to 7 pg/ml) were subtracted from all treatment groups. (b) [Ca2+]i levels were measured in response to capsaicin (CAPS). Cells, exposed to 10 µM CAPS produced a rapid increase in [Ca2+]i that declined over time. Washing of the cells with vehicle caused the [Ca2+]i to return to baseline levels. After 7 min washing, a second exposure to CAPS (10 µM) increased [Ca2+]i but with a reduced response amplitude, suggesting receptor desensitization. (c) Effect of CPZ on the capsaicin (CAPS) induced [Ca2+]i response. Increases in [Ca2+]i did not occur in cells that were pre-exposed CPZ (15 µM) for 10 min before stimulation with CAPS (10 µM) in the presence of CPZ. Reapplication of CAPS after 10 min washing with control vehicle caused an immediate increase in [Ca2+]i. (d) Effects of extracellular calcium and magnesium on the capsaicin (CAPS) induced [Ca2+]i response. Increases of [Ca2+]i did not occur when cells were bathed in calcium-magnesium free HEPES buffer. When the wash solution was changed to one containing CAPS (10 µM) in calcium-magnesium media, an immediate increase in [Ca2+]i occurred. Changing the solution from calcium-magnesium free to regular vehicle in the absence of CAPS did not cause significant effects on the [Ca2+]i. Exposures to test solutions are indicated by the horizontal bars. Between exposures, the cells were washed with control HEPES buffer or the test solution, as indicated. IO (2 µM) was applied to determine maximal [Ca2+]i responses.
binding. This is evidenced by the equipotency of SP and its carboxyl fragment SP 4–11 to release IL-6 and the reduction of inflammatory cytokines and transcripts by the specific pharmacological NK-1 receptor antagonist, CP-99 994.13 The NK-1 receptor is a member of the superfamily of 7 TM G-protein coupled receptors whose activation by SP results in phosphoinositol breakdown, increases in IP3-mediated [Ca2+]i from internal stores and an extracellular influx of calcium.27,15 Thus, the increases in [Ca2+]i noted in BEAS–2B cells upon exposure to SP (Fig. 1a,b) are in accordance with NK-1 receptor activation. These increases declined upon a second expo© 1999 Harcourt Publishers Ltd
sure to SP, reflecting possible receptor desensitization. This phenomenon is in agreement with characteristics of the NK-1 receptor, which internalizes after stimulation.28 In addition to direct stimulation of cytokine formation, SP appears to stimulate cytokine release by interacting synergistically with one cytokine, to enhance the production of another. TNF α, an early and crucial player in the inflammatory cascade, acts on various immune and non-immune cell types to signal the release of cytokines.12,14–29 Physiologically speaking, the timing of SP and TNFα release in the inflammatory cascade occurs in tandem, with the initial release of relatively high Neuropeptides (1999) 33(6), 447–456
454 Veronesi et al.
concentrations of SP into the microenvironment, followed by TNFα synthesis and release, shortly thereafter.30 In keeping with SP’s physiological role as a modulator of inflammation, our data (Fig. 4) indicate that physiologically relevant (0.5 µM) concentrations of SP potentiate TNFα to produce 5-fold increases in IL-6. Although the exact manner in which SP and TNFα interact in BEAS–2B cells to amplify IL-6 synthesis and/or release is presently unknown, several reports indicate that SP can directly stimulate TNFα transcript in human nasal mucosa,31 cultured human astrocytes,32 rat uterine mast cells,33 and in both a murine mast cell line (CFTL12) and peritoneal mast cells.29 The effect of SP on TNFα appears to be NK-1 receptormediated, since SP and its carboxyl fragment SP 4–11 (but not SP 1–4) induce significant production of TNFα in the mononuclear cells of whole blood34 and in human monocyte-derived macrophages.19 Our present studies add to these data by showing that SP induces TNFα transcript in a human bronchial epithelial cell line, and that formation of this transcript can be blocked by the NK-1 receptor antagonist, CP-99 994 (Fig. 3c). The existence of bronchial epithelial peptide receptor and their responsiveness to neuropeptides represents a new area of neuroimmune research that impacts on airway inflammation.35,36 Although airway epithelial cells are the first to encounter chemical xenobiotics, their role in expressing signs of airway inflammation is poorly understood. However, several reports show that neuropeptides can stimulate epithelial cells to release cytokines6,7 and eosinophil-chemotactic factors,37 and influence epithelial cell proliferation, growth and migration.38–40 Neuropeptides also exert a protective effect on bronchial epithelial barrier functions41 and increase neutrophil adhesion to bronchial epithelial cells.42 Deleterious interactions can also occur between neuropeptides and their epithelial targets. For example, SP can stimulate eosinophils to secrete cytotoxic agents (e.g. major basic protein, leukotrienes, and platelet activating factor)43 which damage and denude the epithelial cell. This results in loss of the epithelial airway barrier and the direct exposure of nerve fibers and their receptors to airway irritants entering the lumen. Damage and loss of the epithelial lining also results in loss of NEP, the enzyme critical to SP deactivation. Loss of this enzyme prolongs the residency time and effectiveness of SP.17,44 BEAS–2B cells are known to house NEP–24.1116 and, as the present data show, its inhibition by thiorphan results in significant increases of SP-induced cytokine. Although the levels of IL-6 increased, the poorly defined dose-response curve of SP/CGRP induced IL-6 noted in non-supplemented exposure media did not improve in the presence of thiorphan. This might be explained by noting that SP can be hydrolyzed in vitro by various peptidases (e.g. dipeptidyl aminopeptidase B, depiptidyl aminopeptidase Neuropeptides (1999) 33(6), 447–456
IV, post-proline cleaving enzyme, serum acetylcholinesterase and pseudocholinesterase, acid proteinase, SP endopeptidase and NEP–24.11) and at a variety of peptide bonds.17 This suggests that the BEAS–2B cell line may contain additional enzymes that can deactivate SP and CGRP. It should also be noted that the enzyme activity of NEP is located in the brush border of the respiratory epithelial cell, rather than housed in cytoplasmic granules, as is often seen in neuronal cells. This proximity of the enzyme to the site of SP-release facilitates the instantaneous deactivation of neuropeptides both in situ and in culture systems. Baluk and colleagues4 first reported that serous epithelium responded to capsaicin with the release of CGRP granules housed in their cytoplasm. We have extended this observation by demonstrating that BEAS–2B cells respond to capsaicin with immediate increases in [Ca2+]i and subsequent cytokine release through capsaicin-sensitive pathways, since both events were blocked by CPZ and reduced by exposure in calcium-magnesium free media. Capsaicin receptors belong to the vanilloid family and show overlapping characteristics with pH sensitive pathways or pH receptors. Functional capsaicin receptors have been reported on non-neuronal cells (e.g. mast cells and C6 glioma).45,46 The existence of capsaicin and neuropeptide receptors on BEAS–2B cells offers an exciting cell culture model to examine how other chemical irritants stimulate inflammatory changes in respiratory epithelial cells. In this regard, we have used the BEAS–2B cell to demonstrate the relevance of neuropeptide and capsaicin sensitive pathways to inflammatory changes associated with exposure to industrial residual oil fly ash.47 These studies show that expressions of epithelial cell inflammation (i.e. intracellular calcium increases, cytokine transcription and protein release) can be diminished or completely blocked by antagonists to the neuropeptide (i.e. SP, CGRP) and capsaicin receptors. Such studies link the initial cellular and subcellular events of pollutant-induced airway inflammation even closer to neurogenic factors. ACKNOWLEDGMENT This work has been sponsored in part by NIH grant DC01065. Disclaimer This manuscript has been reviewed by the National Health Effects Environmental Research Laboratory, US Environmental Protection Agency and approved for publication. Approval does not signify that the contents necessarily reflect the view and policies of the Agency, endorsement or recommendation for use. © 1999 Harcourt Publishers Ltd
Neuropeptides and human bronchial epithelial cells 455
REFERENCES 1. Baluk P. Neurogenic inflammation in skin and airways. J Investig Dermatol Symp Proc 1997; 2: 76–81. 2. Barnes PJ, Belvisi MG. Sensory Neuropeptides. In: Barnes PJ, Grunstein MM, Leff AR, Woolcock AJ (eds) Asthma, 1st edn. Philadelphia: Lippincott-Raven 1997; 1051–1063. 3. Barnes PJ. Neurogenic inflammation in airways. Int Arch Allergy Appl Immunol 1991; 94: 303–309. 4. Baluk P, Nadel JA, McDonald DM. Calcitonin gene-related peptide in secretory granules of serous cells in the rat tracheal epithelium. Am J Respir Cell Mol Biol 1993; 8: 446–453. 5. Wuenschell CW, Sunday ME, Singh G, Minoo P, Slavkin HC, Warburton D. Embryonic mouse lung epithelial progenitor cells co-express immunohistochemical markers of diverse mature cell lineages. J Histochem Cytochem 1996; 44: 113–123. 6. Mullol J, Baraniuk JN, Logun C, Benfield T, Picado C, Shelhamer JH. Endothelin-1 induces GM-CSF, IL-6 and IL-8 but not G-CSF release from a human bronchial epithelial cell line (BEAS–2B). Neuropeptides 1996; 30: 551–556. 7. Mullol J, Baraniuk JN, Pitale M et al. Vasoactive intestinal peptide (VIP) induces IL-6 and IL-8, but not G-CSF and GM-CSF release from a human bronchial epithelial cell line. Neuropeptides 1997; 31: 119–124. 8. Reddel RR, Ke Y, Gerwin BI et al. Transformation of human bronchial epithelial cells by infection with SV40 or adenovirus12 SV40 hybrid virus, or transfection via strontium phosphate coprecipitation with a plasmid containing SV40 early region genes. Cancer Res 1988; 48: 1904–1909. 9. McKinnon KP, Madden MC, Noah TL, Devlin RB. In vitro ozone exposure increases release of arachidonic acid products from a human bronchial epithelial cell line. Toxicol Appl Pharmacol 1993; 118: 215–223. 10. Carter JD, Ghio AJ, Samet JM, Devlin RB. Cytokine production by human airway epithelial cells after exposure to an air pollution particle is metal-dependent. Toxicol Appl Pharmacol 1997; 146: 180–108. 11. Folkerts G, Nijkamp FP. Airway epithelium: more than just a barrier! Trends Pharmacol Sci 1998; 19: 334–341. 12. Takizawa H. Airway epithelial cells as regulators of airway inflammation. Int J Mol Med 1998; 1: 367–378. 13. McLean S, Ganong A, Seymour PA et al. Pharmacology of CP-99 994, a non-peptide antagonist of the tachykinin neurokinin-1 receptor. J Pharmacol Exp Ther 1993; 267: 472–479. 14. Khair OA, Davies R J, Devalia JL. Bacterial-induced release of inflammatory mediators by bronchial epithelial cells. Eur Respir J 1996; 9: 1913–1922. 15. Burcher E, Mussap CJ, Geraghty DP, McClure-Sharp JM, Watkins DJ. Concepts in characterization of tachykinin receptors. Ann NY Acad Sci 1991; 632: 123–136. 16. Proud D, Subauste MC, Ward PE. Glucocorticoids do not alter peptidase expression on a human bronchial epithelial cell line. Am J Respir Cell Mol Biol 1994; 11: 57–65. 17. Krause JE. On the physiological metabolism of Substance P. In: Jordan CC, Oehme PP (eds) Substance P: Metabolism and Biological Actions. London: Taylor and Francis, 1985; 13–32. 18. Bar-Shavit Z, Goldman R, Stabinsky Y et al. Enhancement of phagocytosis – a newly found activity of Substance P residing in its N-terminal tetrapeptide sequence. Biochem Biophys Res Commun 1980; 94: 1445–1451. 19. Ho WZ, Stavropoulos G, Lai JP et al. Substance P C-terminal octapeptide analogues augment tumor necrosis factor-alpha release by human blood monocytes and macrophages. J Neuroimmunol 1998; 82: 126–132.
© 1999 Harcourt Publishers Ltd
20. Bevan S, Hothi S, Hughes G et al. Capsazepine: a competitive antagonist of the sensory neurone excitant capsaicin. Br J Pharmacol 1992; 107: 544–552. 21. Belvisi MG, Miura M, Stretton D, Barnes PJ. Capsazepine as a selective antagonist of capsaicin-induced activation of C-fibres in guinea-pig bronchi. Eur J Pharmacol 1992; 215: 341–34. 22. Szallasi A, Blumberg PM. Vanilloid (capsaicin) receptors and mechanisms. Pharmacol Rev 1999; 51: 159–212. 23. Kavelaars A, Jeurissen F, Heijnen CJ. Substance P receptors and signal transduction in leukocytes. Immunomethods 1994; 5: 41–48. 24. Kavelaars A, Jeurissen F, von Frijtag D, Herman vR, Rijkers GT, Heijnen CJ. Substance P induces a rise in intracellular calcium concentration in human T lymphocytes in vitro: evidence of a receptor-independent mechanism. J Neuroimmunol 1993; 42: 61–70. 25. Mousli M, Bronner C, Bockaert J, Rouot B, Landry Y. Interaction of Substance P, compound 48/80 and mastoparan with the alpha-subunit C-terminus of G protein. Immunol Lett 1990; 25: 355–357. 26. Mousli M, Bueb JL, Bronner C, Rouot B, Landry Y. G protein activation: a receptor-independent mode of action for cationic amphiphilic neuropeptides and venom peptides. Trends Pharmacol Sci 1990; 11: 358–362. 27. Putney JW. Receptors and the Inositol Phosphate-Calcium Signaling System. In: Buck S H (ed) The Tachykinin Receptors. New Jersey: Humana Press, 1994; 257–284. 28. McConalogue K, Dery O, Lovett M et al. Substance P-induced trafficking of beta-arrestins. J Biol Chem 1999; 274: 16257–16268. 29. Ansel JC, Brown JR, Payan DG, Brown MA. Substance P selectively activates TNF-alpha gene expression in murine mast cells. J Immunol 1993; 150: 4478–4485. 30. Yoshikawa T, Kurimoto I, Streilein JW. Tumour necrosis factor-alpha mediates ultraviolet light B-enhanced expression of contact hypersensitivity. Immunology 1992; 76: 264–271. 31. Okamoto Y, Shirotori K, Kudo K et al. Cytokine expression after the topical administration of Substance P to human nasal mucosa. J Immunol 1993; 151: 4391–4398. 32. Palma C, Minghetti L, Astolfi M et al. Functional characterization of Substance P receptors on cultured human spinal cord astrocytes: synergism of Substance P with cytokines in inducing interleukin-6 and prostaglandin E2 production. Glia 1997; 21: 183–193. 33. Cocchiara R, Bongiovanni A, Albeggiani G, Azzolina A, Geraci D. Substance P selectively activates TNF-alpha mRNA in rat uterine immune cells: a neuroimmune link. Neuroreport 1997; 8: 2961–2964. 34. Nair MP, Schwartz SA. Substance P induces tumor necrosis factor in an ex vivo model system. Cell Immunol 1995; 166: 286–290. 35. Polito AJ, Proud D. Epithelia cells as regulators of airway inflammation. J Allergy Clin Immunol 1998; 102: 714–718. 36. Yu XY, Takahashi N, Croxton TL, Spannhake EW. Modulation of bronchial epithelial cell barrier function by in vitro ozone exposure. Environ Health Perspect 1994; 102: 1068–1072. 37. Koyama S, Sato E, Nomura H, Kubo K, Nagai S, Izumi T. Acetylcholine and Substance P stimulate bronchial epithelial cells to release eosinophil chemotactic activity. J Appl Physiol 1998; 84: 1528–1534. 38. Sanghavi JN, Rabe KF, Kim JS, Magnussen H, Leff A , White SR. Migration of human and guinea pig airway epithelial cells in response to calcitonin gene-related peptide. Am J Respir Cell Mol Biol 1994; 11: 181–187.
Neuropeptides (1999) 33(6), 447–456
456 Veronesi et al.
39. White SR, Garland A, Gitter B et al. Proliferation of guinea pig tracheal epithelial cells in coculture with rat dorsal root ganglion neural cells. Am J Physiol 1995; 268: L957–L965. 40. Reid TW, Murphy CJ, Iwahashi CK, Foster BA, Mannis MJ. Stimulation of epithelial cell growth by the neuropeptide Substance P. J Cell Biochem 1993; 52: 476–485. 41. Yu XY, Undem BJ, Spannhake EW. Protective effect of Substance P on permeability of airway epithelial cells in culture. Am J Physiol 1996; 271: L889–L895. 42. DeRose V, Robbins RA, Snider RM et al. Substance P increases neutrophil adhesion to bronchial epithelial cells. J Immunol 1994; 152: 1339–1346. 43. Barnes PJ. New concepts in the pathogenesis of bronchial hyperresponsiveness and asthma. J Allergy Clin Immunol 1989; 83: 1013–1026.
Neuropeptides (1999) 33(6), 447–456
44. Piedimonte G. Tachykinin peptides, receptors, and peptidases in airway disease. Exp Lung Res 1995; 21: 809–834. 45. Biro T, Maurer M, Modarres S et al. Characterization of functional vanilloid receptors expressed by mast cells. Blood 1998; 91: 1332–1340. 46. Biro T, Brodie C, Modarres S, Lewin NE, Acs P, Blumberg PM. Specific vanilloid responses in C6 rat glioma cells. Brain Res Mol Brain Res 1998; 56: 89–98. 47. Veronesi B, Oortgiesen M, Carter JD, Devlin RB. Particulate matter (PM) initiates inflammatory cytokine release by activation of capsaicin receptors in a human bronchial epithelial cell line. Toxicol Appl Pharmacol 1999; 154: 106–115.
© 1999 Harcourt Publishers Ltd