Chronic obstructive pulmonary disease (COPD) therapy is complicated by corticosteroid resistance of the interleukin 8 (IL-8)-dependent and granulocyte macrophage-colony stimulating factor (GM-CSF)-dependent chronic airway inflammation, for whose establishment human airway smooth muscle cells (HASMCs) might be crucial. It is unclear whether the release of inflammatory mediators from HASMCs is modulated by cigarette smoking and is refractory to corticosteroids in COPD. Resveratrol, an antiaging drug with protective effects against lung cancer, might be an alternative to corticosteroids in COPD therapy. Vascular endothelial growth factor (VEGF) might offer protection from developing emphysema. We tested the following hypotheses for HASMCs: 1) smoking with or without airway obstruction modulates IL-8, GM-CSF, and VEGF release; and 2) corticosteroids, but not resveratrol, fail to inhibit cytokine release in COPD. Cytokine release from HASMCs exposed to tumor necrosis factor α (TNFα), dexamethasone, and/or resveratrol was measured via enzyme-linked immunosorbent assay and compared between nonsmokers (NS), smokers without COPD (S), and smokers with COPD (all n = 10). In response to TNFα, IL-8 release was increased, but GM-CSF and VEGF release was decreased in S and COPD compared with NS. Dexamethasone and resveratrol inhibited concentration-dependently TNFα-induced IL-8, GM-CSF, and VEGF release. For IL-8 and GM-CSF efficiency of dexamethasone was NS > S > COPD. That of resveratrol was NS = S = COPD for IL-8 and NS = S < COPD for GM-CSF. For VEGF the efficiency of dexamethasone was NS = S = COPD, and that of resveratrol was NS = S > COPD. All resveratrol effects were partially based on p38 mitogen-activated protein kinase blockade. In conclusion, smoking modulates cytokine release from HASMCs. Corticosteroid refractoriness of HASMCs in COPD is cytokine-dependent. Resveratrol might be superior to corticosteroids in COPD therapy, because it more efficiently reduces the release of inflammatory mediators and has limited effects on VEGF in COPD.
Chronic obstructive pulmonary disease (COPD) is a chronic inflammatory airway disease with progressive and irreversible airflow limitation. The major risk factor for the development of COPD is cigarette smoking. Emphysema in COPD is characterized by destruction and enlargement of the alveoli. Because airway inflammation in COPD is refractory to corticosteroids, COPD treatment is hindered and insufficient. The inflammatory cell profile in COPD includes particularly alveolar macrophages (AMs) and neutrophilic granulocytes, but recent literature suggests an involvement of airway smooth muscle cells (ASMCs) (Barnes, 2004; Hogg et al., 2004; Adcock and Barnes, 2008; Barnes and Adcock, 2009). On the cellular level, corticosteroid resistance in COPD has been attributed primarily to AMs (Culpitt et al., 2003b; Adcock and Barnes, 2008), but it is unclear whether the release of inflammatory mediators from other cell types such as human ASMCs (HASMCs) is also insensitive to corticosteroids.
ASMCs orchestrate and perpetuate airway inflammation by promoting the recruitment, activation, and trafficking of inflammatory cells in the airways (Tliba et al., 2008). ASMCs respond to proinflammatory cytokines such as tumor necrosis factor α (TNFα) that is believed to be crucial for the induction and maintenance of inflammation in COPD (Barnes, 2004; Tliba et al., 2008). New evidence suggests ASMCs as an important source of COPD-relevant cytokines because they are capable of releasing macrophage chemotactic and activating factors such as interleukin 8 (IL-8) (Mullan et al., 2008) and granulocyte macrophage-colony stimulating factor (GM-CSF) (Knobloch et al., 2009). Comparing the release of inflammatory cytokines from HASMCs between nonsmokers (NS), smokers with COPD, and smokers without COPD (S) would significantly help to ascertain the importance of HASMCs for smoking-induced inflammation in COPD; however, the literature lacks such studies. Thus, it is currently unclear whether IL-8 and GM-CSF release from HASMCs is modulated by cigarette smoking and/or in COPD.
The molecular pathology leading to the development of emphysema in COPD is currently under discussion. According to one hypothesis, the development of emphysema is believed to contain the progressive loss of capillary endothelial and alveolar epithelial cells by apoptosis (Kanazawa, 2007). The angiogenic cytokine vascular endothelial growth factor (VEGF) is required for the survival of endothelial cells and thus might provide protection from developing emphysema in COPD (Kanazawa, 2007). Indeed, moderate COPD without emphysema has been shown to be associated with an increase of VEGF expression, probably as a protective mechanism to counter-regulate tissue damage. In contrast, VEGF is reduced in COPD with emphysema (Kanazawa, 2007). VEGF is released from several cell types of the lung including HASMCs (Raidl et al., 2007).
Current treatment for COPD is symptomatic but does not inhibit disease progression. The use of corticosteroids in COPD is disputed, because their anti-inflammatory properties are impaired in smoking-related COPD compared with other chronic inflammatory lung diseases such as asthma (Adcock and Barnes, 2008; Barnes, 2008a). Congruously, there is an urgent need to establish alternative therapeutic approaches. Resveratrol (trans-3,5,4′-trihydroxystilbene), a polyphenolic molecule and constituent of red wine extract, exhibits antioxidative, antiaging, and anti-inflammatory properties and has protective effects against lung cancer (Whyte et al., 2007; Rahman, 2008). Thus, resveratrol is believed to contribute to the health benefits ascribed to the consumption of red wine. First evidence for its possible utility in COPD was provided by a study showing that resveratrol reduces basal and IL-1β- and cigarette smoke-induced release of IL-8 and GM-CSF from AMs in smokers with and without COPD (Culpitt et al., 2003a). However, it is currently unknown whether resveratrol can modulate cytokine release from other cell types, such as HASMCs, in the airways.
Here, we investigated whether cigarette smoking with or without airway obstruction modulates IL-8, GM-CSF, and VEGF release from HASMCs in response to TNFα. To address the question of corticosteroid refractoriness of HASMCs in COPD, we studied whether the release of these cytokines is sensitive to dexamethasone comparative between nonsmokers and smokers with and without COPD. Finally, we studied for the first time the effect of resveratrol on cytokine release from HASMCs (including the underlying molecular mechanisms) and compared it with that of dexamethasone to explore whether resveratrol might be superior to corticosteroids in COPD.
Materials and Methods
The study population for HASMCs consisted of 10 nonsmokers without smoking history and without respiratory symptoms or airflow limitation, 10 current smokers (≥ 10 pack years) without respiratory symptoms or airflow limitation, and 10 current smokers with respiratory symptoms and mild to severe airflow limitation without emphysema (GOLD stages I-III) (Table 1). None of the subjects were using oral or inhaled corticosteroids or receiving immunosuppressive treatment, and none reported any other serious illness (with the exception of lung carcinoma, see below) or acute viral disease during the 2 months preceding the test or had tuberculosis, parasite infections, or histories of allergies or asthma. COPD was diagnosed according to the criteria recommended by the National Institutes of Health (Fabbri et al., 2004). The smokers with COPD (≥ 10 pack years, GOLD stages I-III) had a history of cough with sputum production and/or dyspnea on most of the days of the month for at least 3 months a year during >2 years before the study and airflow limitation in spirometry (FEV1/FVC <70%) as defined by the GOLD initiative (Pauwels et al., 2001; Fabbri et al., 2004). The airflow limitation in these patients was irreversible as shown by negative immediate response to inhalation of 200 μg of albuterol (≤ 12% reversibility). Their pulmonary function had been stable for several months under observation. The study was approved by the Ethics Committee of the University of Cologne, Germany.
Isolation and Cultivation of Human Airway Smooth Muscle Cells.
HASMCs were dissected out from lobar or main bronchus tissue obtained from patients undergoing lung resection for carcinoma of the bronchus as described previously (Raidl et al., 2007). For HASMC isolation, tissue without tumor infiltration was used. Characterization and cultivation of HASMCs was performed as described elsewhere (Raidl et al., 2007). Before stimulation, subconfluent cell monolayers (approximately 80% confluence) in six-well cell culture plates were deprived of serum for 24 h as described previously (Raidl et al., 2007). HASMCs at passages 3 to 7 were stimulated with human TNFα (R&D Systems, Minneapolis, MN) at 20 ng/ml for the indicated times. Resveratrol (Sigma, Munich, Germany), dexamethasone [(8S,9R,10S,11S,13S,14S,16R,17R)-9-fluoro-11,17-dihydroxy-17-(2-hydroxyacetyl)-10,13,16-trimethyl-6,7,8,9,10,11,12,13,14,15, 16,17-dodecahydro-3H-cyclopenta[a]phenanthren-3-one; Sigma], and/or anisomycin [(2S,3R,4R)-4-hydroxy-2-(4-methoxybenzyl)pyrrolidin-3-yl acetate; Sigma] was added 60 min before TNFα stimulation. SB203580 [4-[4-(4-fluorophenyl)-2-(4-methylsulfinylphenyl)-1H-imidazol-5-yl]pyridine; Calbiochem/VWR, Darmstadt, Germany] was added at the times indicated. Cell viability was determined by trypan blue staining as described previously (Knobloch et al., 2007).
Cytokine and p38 Mitogen-Activated Protein Kinase Activity Measurements.
IL-8, GM-CSF, and VEGF were measured in HASMC culture supernatants, using paired antibody quantitative ELISAs (R&D Systems) as described previously (Koch et al., 2004a,b; Raidl et al., 2007). Resveratrol has previously been reported not to affect cytokine ELISAs (Culpitt et al., 2003a). p38 mitogen-activated protein kinase (p38MAPK) activity was measured as described previously (Knobloch et al., 2009) with an intracellular ELISA specific for phosphorylated p38MAPK with total p38MAPK as reference.
RNA Isolation and Quantitative RT-PCR.
DNA-free total RNA was extracted from 80% subconfluent HASMCs with the chromatography-based RNeasy technique (QIAGEN GmbH, Hilden, Germany) (Knobloch et al., 2009). Quantitative RT-PCR (with PCR cycles in the linear range) with IL-8-, GM-CSF-, VEGF-, and endothelin 1 (ET-1)-specific primers and GAPDH as a reference and subsequent evaluation of signals by densitometry using Alpha Innotech (San Leandro, CA) software, version 184.108.40.206 was performed as described previously (Knobloch et al., 2009). Primer sequences for GM-CSF, ET-1, and GAPDH have been published previously (Knobloch et al., 2009): IL-8, forward 5′-ACC ACC GGA AGG AAC CAT-3′, reverse 5′-TCT GTG TTG GCG CAG TGT-3′; VEGF, forward 5′-CAA GAA AAT CCC TGT GGG CC-3′, reverse 5′-CGT TTA ACT CAA GCT GCC TCG-3′.
Statistical analyses were performed to examine the effects of TNFα alone or in combination with resveratrol, dexamethasone, SB203580, and/or anisomycin on cytokine release from cells comparative between NS, S, and COPD. Histogram analyses and the Kolmogorov-Smirnov test were used to test for Gaussian distribution. A Gaussian distribution was confirmed for all datasets, and the results are expressed as mean ± S.E.M. Comparisons across different cohorts were analyzed by one-way ANOVA with 95% confidence intervals. Comparisons in time response experiments and across different stimulations were analyzed by one-way repeated measures ANOVA with 95% confidence intervals. For separate comparisons of each stimulation or cohort post hoc Bonferroni-Holm tests were performed. Comparisons between two stimulations within one cohort (e.g., TNFα induction versus unstimulated control) were analyzed with two-tailed, paired Student's t test. A p value <0.05 was considered statistically significant. Concentration-response curves, EC50, and Emax values were determined by performing nonlinear regression curve fit. All tests were done with Prism (version 3.00) (GraphPad Software Inc., San Diego, CA).
Effect of TNFα on IL-8, GM-CSF, and VEGF Release from HASMCs Comparative between Nonsmokers, Smokers without COPD, and Smokers with COPD.
Time-response experiments (4–72 h) revealed that the inducing effect of TNFα on IL-8 release in this period is maximal at 72 h (Fig. 1A), and TNFα induces a significant increase of GM-CSF and VEGF release from HASMCs earliest after 72 h of stimulation (Fig. 1B; Raidl et al., 2007). Thus, we used 72 h of TNFα treatment for cytokine measurements.
Baseline IL-8, GM-CSF, and VEGF release of HASMCs after 72 h of cultivation was in each case without significant differences between NS, S, and COPD (Fig. 1, C–E).
TNFα-induced IL-8 release from HASMCs was increased in S and COPD compared with NS (Fig. 1C). TNFα-induced GM-CSF release from HASMCs was reduced in COPD compared with S as well as in S compared with NS (Fig. 1D). TNFα-induced VEGF release was reduced in S and COPD compared with NS (Fig. 1E).
Effect of Dexamethasone and Resveratrol on TNFα-Induced IL-8, GM-CSF, and VEGF Release.
In cell culture, dexamethasone concentrations of 10−12 or 10−4 M represent a low or a high concentration, respectively, in terms of clinical relevance. Thus, we performed concentration-response experiments with concentrations from 10−12 to 10−4 M. Dexamethasone partially reduced IL-8 release from TNFα-stimulated HASMCs, and this inhibitory effect was significantly stronger in NS compared with S (Fig. 2A; Table 2). However, in all concentrations used dexamethasone did not significantly modulate IL-8 release from TNFα-stimulated HASMCs of COPD (Fig. 2A; Table 2). Resveratrol use in human requires relatively high dosages (Elliott and Jirousek, 2008), and up to 10−3 M the drug did not significantly affect viability of cultured HASMCs (data not shown). Thus we used concentrations from 10−7 to 10−3 M for concentration-response experiments. In contrast to dexamethasone, resveratrol reduced IL-8 release from TNFα-stimulated HASMCs down to baseline level in all three cohorts. This inhibitory effect was observed at high concentrations with almost equal values for EC50 and Emax in the three cohorts (Fig. 2B; Table 2).
Dexamethasone reduced GM-CSF release from TNFα-stimulated HASMCs in all three cohorts concentration-dependently down to baseline level (Fig. 2C; Table 2). This inhibitory effect of dexamethasone was significantly stronger in HASMCs of NS compared with S and also significantly stronger in S compared with COPD (Fig. 2C; Table 2). Resveratrol also reduced GM-CSF release from TNFα-stimulated HASMCs in all three cohorts concentration-dependently down to baseline level. In contrast to dexamethasone, the inhibitory effect of resveratrol was significantly stronger in HASMCs of COPD compared with S and NS (Fig. 2D; Table 2).
Dexamethasone clearly reduced VEGF release from TNFα-stimulated HASMCs in all three cohorts in a concentration-dependent manner. This inhibitory effect of dexamethasone was without significant differences between NS, S, and COPD (Fig. 2E; Table 2). Resveratrol reduced VEGF release from TNFα-stimulated HASMCs in all three cohorts concentration-dependently down to baseline level. In contrast to dexamethasone, the inhibitory effect of resveratrol was significantly stronger in HASMCs of NS and S compared with COPD (Fig. 2F; Table 2).
Effect of Resveratrol on TNFα-Induced IL-8, GM-CSF, and VEGF Transcription.
To get insights into the molecular mechanisms underlying resveratrol effects in HASMCs we first investigated the effects of the drug on cytokine gene transcription. There were two maxima of GM-CSF mRNA levels, at 2 and 8 h, in response to TNFα (Knobloch et al., 2009; Fig. 3A). For IL-8 and VEGF, mRNA induction was weak after 2 h, but maxima were reached after 4 h and got maintained until 8 h of stimulation (Fig. 3A). Thus, we used the 8-h time point for our analyses.
The differences between the cohorts seen for the release of all three cytokines from TNFα-stimulated HASMCs (Fig. 1, C–E) were, at least by trend, also found on the mRNA levels (Fig. 3, B–D). This indicates that cigarette smoking and COPD pathology modulates IL-8, GM-CSF, and VEGF gene transcription. It is noteworthy that the reducing effects of resveratrol on IL-8 and GM-CSF were also reflected on mRNA levels (Fig. 3, B and C), and notably this included the differences between COPD versus NS and S at 10−5 M for GM-CSF (compare Fig. 3C with Fig. 2D). Although resveratrol also reduced VEGF transcription in response to TNFα, the differences between the cohorts seen for resveratrol effects on VEGF releases were not reflected at the mRNA level (compare Fig. 3D with Fig. 2F).
Resveratrol Reduces p38MAPK Activity in HASMCs.
Resveratrol cell type- and concentration-dependently activates or blocks p38MAPK activity (Pervaiz and Holme, 2009), and transcription of inflammatory cytokine genes is often linked to p38MAPK signaling (Barnes, 2008b). Thus we asked whether modulation of p38MAPK activity could explain some of the resveratrol effects on HASMCs. As shown previously (Knobloch et al., 2009), TNFα induces two peak levels of p38MAPK activation after 15 min (first peak) and 5 h (second peak) of stimulation (Fig. 4A). Although the differences were not statistically significant, there is a trend for the first peak toward an increased p38MAPK activity in COPD versus S and in S versus NS. For the second peak the trend is opposite, and the difference between NS and COPD is statistically significant (Fig. 4A). Resveratrol concentration-dependently blocks the first p38MAPK activity peak without differences between the three cohorts (Fig. 4B). Resveratrol also reduces the second peak, but, notably, at 10−5 M this effect was stronger in COPD compared with NS and S (Fig. 4C). We have previously shown that this second peak is a consequence of TNFα-induced ET-1 up-regulation, which autocrinally reactivates p38MAPK via a positive feedback mechanism (Knobloch et al., 2009). In agreement, resveratrol blocks ET-1 transcription, and this effect was stronger in COPD compared with NS and S (Fig. 4D).
Resveratrol Reduces IL-8, GM-CSF, and VEGF Transcription by Interfering with p38MAPK Activity.
The data of Figs. 3 and 4 suggest that differences in the dependence of gene transcription on the second p38MAPK activity peak could explain the variable influence of COPD on resveratrol effects on IL-8, GM-CSF, and VEGF. In agreement with previous reports (Raidl et al., 2007; Knobloch et al., 2009, Munoz et al., 2010), preincubation with a specific p38MAPK inhibitor, which targets both p38MAPK activity peaks, significantly reduced IL-8, GM-CSF, and VEGF mRNA levels (Fig. 5, A–C). In contrast, postincubation with the p38MAPK inhibitor 3 or 4 h after TNFα stimulation, which specifically affects the second but not the first p38MAPK activity peak, reduced GM-CSF but not IL-8 and VEGF transcription (Fig. 5, A-C). The data confirm that GM-CSF transcription depends on both TNFα-induced direct p38MAPK activation (first peak) and TNFα-induced, ET-1-dependent p38MAPK reactivation (second peak). Moreover, these data demonstrate that this second peak does not contribute to IL-8 and VEGF transcription and indicate that IL-8 and VEGF transcription is independent from TNFα-induced ET-1 up-regulation. In agreement, the endothelin receptor antagonist bosentan blocks GM-CSF transcription (Knobloch et al., 2009) but not IL-8 and VEGF transcription in response to TNFα (Fig. 5D).
Finally, we investigated whether experimental p38MAPK hyperactivation might restore the resveratrol effects. Anisomycin [(2S,3R,4R)-4-hydroxy-2-(4-methoxybenzyl)pyrrolidin-3-yl acetate] is an inhibitor of protein synthesis but also an p38MAPK and Jnk activator (Hazzalin et al., 1998). At 0.1 μM, at which anisomycin does not affect protein synthesis but activates p38MAPK in HASMCs (Hazzalin et al., 1998; Shan et al., 2006), the drug partially restores p38MAPK activity and IL-8, GM-CSF, and VEGF transcription in TNFα- and resveratrol-treated cells (Fig. 6). In summary, these data demonstrate that resveratrol reduces IL-8, GM-CSF, and VEGF release in part via p38MAPK blockade. They further explain at least partially the cytokine specificity of the COPD influence on resveratrol effects: resveratrol impairs ET-1-dependent p38MAPK reactivation, which is crucial for GM-CSF but not for IL-8 and VEGF, stronger in COPD than in NS and S.
Repeated Resveratrol Addition Reduces Its Effective Concentration.
Resveratrol bioavailability is limited and thus it is disputable whether its anti-inflammatory effects at concentrations of 10−5 M and higher in vitro have any in vivo relevance (Cottart et al., 2010). In therapy of chronic airway diseases drugs are repeatedly administered. Thus, we asked whether repeated addition of resveratrol could reduce its effective concentration in vitro. To this end, HASMCs from COPD were stimulated every 24 h with TNFα in fresh medium with either resveratrol addition before any or only before the first TNFα stimulation. After a total time of 72 h cytokines in the supernatant were quantified, and repeated versus single resveratrol additions were compared for their efficiency in reducing cytokine release. For IL-8 and GM-CSF, but, surprisingly, not for VEGF, repeated resveratrol addition reduces the effective concentration compared with a single addition (Fig. 7). These data increase the in vivo relevance of the study.
We have elucidated here the inflammatory HASMC response comparative between nonsmokers and current smokers with and without COPD. TNFα-exposed HASMCs of smokers with and without COPD show increased IL-8 but decreased GM-CSF and VEGF mRNA levels and releases. Thus, cigarette smoke augments the capability of TNFα to induce IL-8 transcription but counter-regulates the mechanism by which TNFα induces GM-CSF and VEGF transcription. TNFα-induced IL-8, GM-CSF, and VEGF expression in HASMCs all partially depend on MAPKs including p38MAPK (Knobloch et al., 2009; Raidl et al., 2007; this study). TNFα induces repeated p38MAPK activation in HASMCs: an early direct response with a maximum after 15 min and late responses with maxima after 5 and 9 h (Knobloch et al., 2009). These late responses depend on an ET-1 autoregulatory feedback mechanism, which is induced by TNFα and maintains GM-CSF expression (Knobloch et al., 2009; Fig. 8). Consequently, TNFα-induced GM-CSF expression is sensitive to specific blockade of both early and late p38MAPK activation and to the endothelin receptor antagonist bosentan. Late p38MAPK activation is, as GM-CSF expression, by trend decreased in smokers and further in COPD. Thus, the negative impact of smoking/COPD on TNFα-induced GM-CSF release might partially be explained by interference with the ET-1/p38MAPK-dependent feedback mechanism. In contrast, TNFα-induced IL-8 and VEGF expression are insensitive to specific blockade of late p38MAPK activation and to bosentan and thus are independent from the ET-1/p38MAPK-dependent feedback mechanism. This might help to explain that the effects of smoking/COPD on IL-8 and GM-CSF are opposite in TNFα-exposed HASMCs.
Cigarette smoking and COPD pathogenesis also augment IL-8 release from AMs (Culpitt et al., 2003b; Koch et al., 2004b). In this context it has been proposed that oxidative and nitrosative stress induced by cigarette smoke impairs the activity of histone deacetylases (HDACs), thereby amplifying inflammatory responses to NF-κB that controls IL-8 expression (Barnes, 2004; Barnes and Adcock, 2009). This could also explain increased IL-8 transcription and release in/from TNFα-exposed HASMCs of smokers/COPD. VEGF expression in TNFα-exposed HASMCs seems to be independent of HDAC (Raidl et al., 2007). Although this could explain why cigarette smoking does not enhance VEGF expression it remains unclear why VEGF expression is reduced in smokers/COPD.
IL-8 and GM-CSF significantly contribute to the establishment of chronic airway inflammation in COPD (Barnes, 2004), predicting IL-8 and GM-CSF reduction to be of therapeutic benefit. VEGF maintains lung homeostasis through its function as a survival factor and by promoting tissue repair (Kanazawa, 2007; Tuder and Yun, 2008). VEGF receptor blockade causes emphysema in animal models, and VEGF levels in sputum and plasma are decreased in COPD with emphysema but increased in COPD without emphysema (Kasahara et al., 2000; Marwick et al., 2006; Kanazawa, 2007; Tuder and Yun, 2008). Thus, VEGF might offer protection from the development of emphysema in COPD. Although further experimental support is required, therapeutic reduction of VEGF levels in COPD might rather be counterproductive.
Current treatment of lung inflammation in stable COPD is insufficient, and the clinical benefit of corticosteroids seems to be controversial (Adcock and Barnes, 2008; Barnes and Adcock, 2009). Corticosteroids have formerly been shown to reduce IL-8 and GM-CSF from TNFα-exposed HASMCs (Saunders et al., 1997; John et al., 1998; Pang and Knox, 2000). However, these studies did not define the characteristics of HASMC donors and did not distinguish between nonsmokers, smokers, and subjects with and without respiratory symptoms. We have shown here that dexamethasone can efficiently reduce IL-8 and GM-CSF release from TNFα-stimulated HASMCs of nonsmokers. However, this effect is already restricted in smokers without COPD and further limited in smokers with COPD. This demonstrates that the release of inflammatory mediators from HASMCs is refractory to corticosteroids in COPD. Corticosteroids are likewise inefficient in reducing IL-8 and GM-CSF release from AM of COPD subjects (Culpitt et al., 2003b; Armstrong et al., 2009). Mechanistically, corticosteroid insensitivity in COPD has been linked to the reduction of HDAC expression and activity caused by cigarette smoke-induced oxidative/nitrosative stress (Ito et al., 2006; Adcock and Barnes, 2008; Barnes and Adcock, 2009). HDAC is required for NF-κB blockade by ligand-bound glucocorticoid receptors. Contrary to IL-8 and GM-CSF, dexamethasone was approximately equally efficient in the reduction of VEGF release from TNFα-stimulated HASMCs in the three cohorts. Summarized, these data demonstrate that the limited utility of corticosteroids in COPD is also attributed to their limited anti-inflammatory effects on HASMCs.
The antiaging drug resveratrol is discussed as an alternative to corticosteroids in COPD therapy (Rahman, 2008). Resveratrol efficiently reduces IL-8 and GM-CSF release from AM and HASMCs of COPD after stimulation with TNFα, IL-1β, and/or cigarette-smoke-saturated medium (Culpitt et al., 2003a; this study), indicating that its anti-inflammatory properties are superior to those of corticosteroids in COPD. In contrast to dexamethasone, resveratrol has low efficiency for VEGF in COPD, suggesting increased protection against the development of emphysema, an indication of accelerated lung aging. Furthermore, resveratrol might counteract accelerated aging in COPD by activating sirtuins, which are key regulators of aging (Barnes, 2008b; Ito and Barnes, 2009). Thus, there is now evidence that the therapeutic utility of resveratrol in COPD might be superior to corticosteroids because of superior anti-inflammatory and antiaging properties.
Resveratrol has multiple mechanisms of action that could underlie its anti-inflammatory effects on HASMCs (Pervaiz and Holme, 2009). Cigarette smoke-induced oxidative stress is not only linked to increased airway and systemic inflammation but also to several comorbidities and accelerated aging in COPD (Barnes, 2008b; Ito and Barnes, 2009). Resveratrol scavenges oxygen-derived free radicals, such as superoxide and hydroxyl radical, which boost inflammatory pathways like NF-κB signaling and suppress sirtuins (Barnes, 2008b; Pervaiz and Holme, 2009). As in pilot experiments the addition of cell-permeable superoxide dismutase did not modulate the impact of TNFα on IL-8, GM-CSF, and VEGF in HASMCs of smokers (J. Knobloch and A. Koch, unpublished results), the antioxidant properties of resveratrol might not be a primary explanation for its anti-inflammatory effects on HASMCs. Resveratrol modulates protein kinase pathways in a cell type- and concentration-dependent manner. including phosphatidylinositol 3-kinase/Akt, MAPKs, and NF-κB signaling (Bhat and Pezzuto, 2002; Pervaiz and Holme, 2009). Activation of IL-8, GM-CSF, and VEGF genes depends on p38MAPK and NF-κB, and resveratrol blocks p38MAPK in HASMCs (Barnes and Adcock, 2009; this study). Anisomycin, a well known inducer of p38MAPK and Jnk signaling (Hazzalin et al., 1998), which increases p38MAPK activity in HASMCs, counteracted the reducing effects of resveratrol on IL-8, GM-CSF, and VEGF expression. Thus, p38MAPK blockade by resveratrol might partially explain its effects on IL-8, GM-CSF, and VEGF expression (Fig. 8). However, compared with smokers and nonsmokers, resveratrol efficiency in COPD is increased for GM-CSF but decreased for VEGF and unchanged for IL-8. How can these differences mechanistically be explained? The influence of COPD on resveratrol effects on cytokine release is reflected on the mRNA level for IL-8 and GM-CSF. In response to TNFα both, IL-8 and GM-CSF mRNA expression depend on early p38MAPK activation, but late p38MAPK activation and ET-1 are required only for GM-CSF expression and are irrelevant for IL-8 (Fig. 8). It is noteworthy that resveratrol reduced TNFα-induced early p38MAPK activation without significant differences between the cohorts. In contrast, reduction of both TNFα-induced late p38MAPK activation and ET-1 expression by resveratrol is stronger in COPD compared with nonsmokers and smokers. This explains why resveratrol reduces GM-CSF but not IL-8 more efficiently in COPD than in nonsmokers and smokers without COPD. The impact of COPD on resveratrol effects on VEGF is not reflected on the mRNA level. Thus, this might be caused by posttranslational modifications that need to be explored in further studies.
After a single administration, plasma concentrations of free resveratrol are low in humans, indicating rapid metabolism and limited bioavailability (Cottart et al., 2010). Because the anti-inflammatory effects on cultivated HASMCs require relatively high resveratrol concentrations this might dispute the therapeutic relevance of our data. However, free resveratrol plasma levels might be seriously underestimated because of large amounts potentially contained in the cellular fraction, which has not been assessed in the corresponding studies. Moreover, repeated resveratrol administration increases half-life and plasma concentration in humans, and great effort is being made to increase resveratrol bioavailability, e.g., by structural modification (Cottart et al., 2010). Finally, compared with a single application, repeated application in fresh medium in vitro decreases the effective resveratrol concentration for inhibiting IL-8 and GM-CSF release from TNFα-exposed HASMCs (this study). This underlines the in vivo relevance of our study.
Summarized, this study resolved three important points relevant for COPD: 1) cigarette smoking modulates cytokine release from HASMCs, thereby probably augmenting IL-8-dependent inflammation and decreasing VEGF levels, a survival factor that might offer protection from emphysema; 2) the release of inflammatory mediators (but not of VEGF) from HASMCs is refractory to corticosteroids; and 3) there is evidence for an increased utility of the antiaging drug resveratrol compared with corticosteroids in COPD therapy, because, contrary to corticosteroids, resveratrol is efficient in reducing IL-8 and GM-CSF release and only moderately reduces VEGF release from HASMCs of COPD.
We thank Katja Müller for excellent technical assistance.
The study was financially supported by the Moritz-Stiftung, Cologne, Germany [Grant 36460040].
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
- chronic obstructive pulmonary disease
- alveolar macrophage
- airway smooth muscle cell
- human ASMC
- granulocyte macrophage-colony stimulating factor
- histone deacetylase
- mitogen-activated protein kinase
- nuclear factor κB
- smokers without COPD
- tumor necrosis factor α
- vascular endothelial growth factor
- reverse transcription-polymerase chain reaction
- enzyme-linked immunosorbent assay
- analysis of variance
- Global Initiative for Chronic Obstructive Lung Disease
- glyceraldehyde-3-phosphate dehydrogenase
- forced vital capacity
- forced expiratory volume in 1 s.
- Received February 2, 2010.
- Accepted August 25, 2010.
- Copyright © 2010 by The American Society for Pharmacology and Experimental Therapeutics