Long-Acting β2-Agonists Increase Fluticasone Propionate-Induced Mitogen-Activated Protein Kinase Phosphatase 1 (MKP-1) in Airway Smooth Muscle Cells

Mitogen-activated protein kinase phosphatase 1 (MKP-1) represses MAPK-driven signalling and plays an important anti-inflammatory role in asthma and airway remodelling. Although MKP-1 is corticosteroid-responsive and increased by cAMP-mediated signalling, the upregulation of this critical anti-inflammatory protein by long-acting β2-agonists and clinically-used corticosteroids has been incompletely examined to date. To address this, we investigated MKP-1 gene expression and protein upregulation induced by two long-acting β2-agonists (salmeterol and formoterol), alone or in combination with the corticosteroid fluticasone propionate (abbreviated as fluticasone) in primary human airway smooth muscle (ASM) cells in vitro. β2-agonists increased MKP-1 protein in a rapid but transient manner, while fluticasone induced sustained upregulation. Together, long-acting β2-agonists increased fluticasone-induced MKP-1 and modulated ASM synthetic function (measured by interleukin 6 (IL-6) and interleukin 8 (IL-8) secretion). As IL-6 expression (like MKP-1) is cAMP/adenylate cyclase-mediated, the long-acting β2-agonist formoterol increased IL-6 mRNA expression and secretion. Nevertheless, when added in combination with fluticasone, β2-agonists significantly repressed IL-6 secretion induced by tumour necrosis factor α (TNFα). Conversely, as IL-8 is not cAMP-responsive, β2-agonists significantly inhibited TNFα-induced IL-8 in combination with fluticasone, where fluticasone alone was without repressive effect. In summary, long-acting β2-agonists increase fluticasone-induced MKP-1 in ASM cells and repress synthetic function of this immunomodulatory airway cell type.


Introduction
Asthma is a chronic airway disease characterized by excessive inflammation. Corticosteroids are widely-used to repress inflammation in asthma and control of asthma symptoms can be improved by combination therapy with long-acting b 2 -agonists [1,2,3]. There are a number of mechanisms by which this enhanced anti-inflammatory effect might be achieved at the cellular level: these include control of histone acetylation, increased function of glucocorticoid receptor (GR) or the b 2adrenergic receptor, and upregulation of anti-inflammatory proteins (reviewed in [4]). Mitogen-activated protein kinase phosphatase 1 (MKP-1) is one such anti-inflammatory protein whose upregulation is increasingly recognized as an important molecular mechanism responsible for the improved anti-inflammatory action of corticosteroids and long-acting b 2 -agonists in combination.
MKP-1 is an immediate-early gene [5] that is corticosteroidresponsive [6] and contains a cAMP-responsive element (CRE) in its 59-promoter [7,8]. We and others have demonstrated that the upregulation of MKP-1 by corticosteroids is responsible, in part, for their anti-inflammatory actions in clinically-relevant airway cell types such as airway smooth muscle (ASM) and bronchial epithelial cells [9,10,11,12,13,14]. Moreover, MKP-1 mRNA expression and protein upregulation can be enhanced by longacting b 2 -agonists [12,15]. The increased expression of MKP-1 may explain the beneficial effects of b 2 -agonists/corticosteroid combination therapies in the repression of inflammatory gene expression in asthma [12,16]. However the upregulation of this critical anti-inflammatory protein by long-acting b 2 -agonists and clinically-used corticosteroids has been incompletely examined to date.
To address this gap in knowledge, the aim of this study is to examine MKP-1 mRNA expression and protein upregulation induced by two long-acting b 2 -agonists (salmeterol and formoterol), alone or in combination with the corticosteroid fluticasone propionate (abbreviated as fluticasone). We will assess the antiinflammatory function of MKP-1 induced by these combination therapies on the synthetic function of primary human ASM cell in vitro by measuring repression of interleukin 6 (IL-6) and interleukin 8 (IL-8); two cytokines known to be upregulated in a MAPK-mediated manner [10,17,18].

Cell Culture
Human bronchi were obtained from patients undergoing surgical resection for carcinoma or lung transplant donors in accordance with procedures approved by the Sydney South West Area Health Service and the Human Research Ethics Committee of the University of Sydney. ASM cells were dissected, purified and cultured as previously described by Johnson et al. [19]. A minimum of three different ASM primary cell cultures established from individual patients were used for each experiment.
Unless otherwise specified, all chemicals used in this study were purchased from Sigma-Aldrich (St. Louis, MO).

Western Blotting
MKP-1 was quantified by Western blotting using a rabbit polyclonal antibody against MKP-1 (C-19: Santa Cruz Biotechnology, Santa Cruz, CA), compared to a-tubulin used as the loading control (mouse monoclonal IgG 1 , DM1A: Santa Cruz Biotechnology, Santa Cruz, CA). Primary antibodies were detected with goat anti-mouse or anti-rabbit HRP-conjugated

Real-time RT-PCR
Total RNA was extracted using the RNeasy Mini Kit (Qiagen Australia, Doncaster, VIC, Australia) and reverse transcription performed by using the RevertAid First strand cDNA. MKP-1, IL-6 and IL-8 mRNA levels were measured using real-time RT-PCR on an ABI Prism 7500 (Applied Biosystems, Foster City, CA) with the MKP-1 (DUSP1: Hs00610256_g1), IL-6 (Hs00174131_m1) or IL-8 (Hs00174103_m1) TaqManH Gene Expression Assays and the eukaryotic 18S rRNA endogenous control probe (Applied Biosystems) subjected to the following cycle parameters: 50uC for 2 min, 1 cycle; 95uC for 10 min, 1 cycle; 95uC for 15 s, 60uC for 1 min, 40 cycles and mRNA expression (fold increase) quantified by delta delta Ct calculations.

ELISAs
Cell supernatants were collected and stored at 220uC for later analysis by ELISA. IL-6 and IL-8 ELISAs were performed according to the manufacturer's instructions (BD Biosciences Pharmingen, San Diego, CA).

MKP-1 siRNA Knock-down
Protein levels of MKP-1 were reduced by RNA interference and the effect on cytokine secretion examined. ASM cells were transiently transfected using nucleofection with 1 mg MKP-1specific SMART pool siRNA, consisting of a pool of four individual siRNA from Dharmacon (Dharmacon, Lafayette, CO) or a scrambled siRNA control, using methods established in our previous publication [10]. Briefly, ASM cells were transfected with the Nucleofector (Amaxa, Koln, Germany), using the basic kit for primary smooth muscle cells with the manufacturer's optimized protocol of P-024. ASM cells were plated for 16 h after transfection, before being growth-arrested for a further 24 h. Cells were pretreated for 1 h with formoterol+fluticasone (along with relevant controls) before stimulation for 24 h with TNFa (10 ng/ml). Supernatants were then removed for IL-6 and IL-8 protein secretion by ELISA and lysates utilized for MKP-1 Western blotting (compared to a-tubulin as a loading control).

Statistical Analysis
Statistical analysis was performed using either Student's unpaired t tests, repeated measured ANOVA or two-way ANOVA followed by Bonferroni's multiple comparison tests. P values ,0.05 were sufficient to reject the null hypothesis for all analyses.

Long-acting b 2 -agonists Increase Fluticasone-induced MKP-1 Protein Upregulation.
Although MKP-1 is known to be corticosteroid-responsive [10,20] and increased by cAMP-mediated signalling [20] in ASM cells, the upregulation of this critical anti-inflammatory protein by long-acting b 2 -agonists and clinically-used corticosteroids has been incompletely examined to date. To address this, we examine here for the first time, the effect of treating ASM cells with fluticasone, alone or in combination with the long-acting b 2 -agonistssalmeterol (100 nM: Figure 1) or formoterol (10 nM: Figure 2). These concentrations of b 2 -agonists were chosen for the current study as they had been used previously in in vitro investigations of MKP-1 gene expression by Kaur et al. [12].
We first examined the temporal kinetics of MKP-1 protein upregulation in ASM cells treated for up to 24 h with salmeterol or fluticasone, alone or in combination (compared to vehicle-treated controls). As shown in Figure 1A, salmeterol alone induced rapid, but transient upregulation of MKP-1 with a peak at 1 h. In contrast, fluticasone alone induced a sustained increase in MKP-1 protein upregulation; in accordance with our earlier report [10]. This was confirmed by densitometry (as shown in Figure 1B We then compared the effect of formoterol on fluticasoneinduced MKP-1 ( Figure 2). In corroboration of our earlier report [20], formoterol alone induces a transient increase in MKP-1 upregulation with a peak at 1-2 h (Figure 2A). Importantly, Western blotting ( Figure 2A) revealed that when formoterol and fluticasone are added in combination, the peak of MKP-1 upregulation that appears at 1-2 h after the long-acting b 2 -agonist and enhances the sustained pattern of MKP-1 upregulation observed with fluticasone treatment alone. This is supported by densitometric analysis where formoterol significantly increased fluticasone-induced MKP-1 at 1 h (11.362.1-fold) and 2 h (10.161.4-fold), compared to fluticasone alone at the same time points (4.460.9-and 3.960.7-fold, respectively) (P,0.05).  Effect of Long-acting b 2 -agonists on IL-6 and IL-8 mRNA Expression and Secretion As the final aim of the current study was to demonstrate the anti-inflammatory function of MKP-1 on ASM synthetic function by measuring repression of IL-6 and IL-8 secretion induced by TNFa (see next section), it was first important to examine the effects of b 2 -agonists alone on secretion of these cytokines by ASM cells. Interestingly, IL-6 (like MKP-1) is cAMP-responsive. In this way, we show that salmeterol and formoterol significantly enhance IL-6 mRNA expression at 1 h ( Figure 4A) and that formoterol increases IL-6 secretion from ASM cells at 24 h ( Figure 4C) (P,0.05). These protein secretion results are in confirmation of our earlier study [21] and in accord with the relative ability of these b 2 -agonists to elevate cAMP in ASM cells [21]. The involvement of cAMP/adenylate cyclase-mediated pathway in IL-6 mRNA expression is further substantiated by treatment of ASM cells with the cell-permeable cAMP analogue, dibutyryl cAMP, and the adenylate cyclase activator, forskolin. As shown in Figure 4E, dibutyryl cAMP and forskolin increased IL-6 mRNA expression by 2.460.4-fold and 18.861.7-fold, respectively, compared to vehicle control (P,0.05). IL-8, in contrast, is not regulated by cAMP and thus long-acting b 2 -agonists do not increase IL-8 mRNA expression ( Figure 4B), nor protein secretion ( Figure 4D) in ASM cells. Moreover, dibutyryl cAMP and forskolin are without effect on IL-8 mRNA expression ( Figure 4F).

The Effect of Long-acting b 2 -agonists and Fluticasone, Alone or in Combination, on Repression of TNFa-induced IL-6 and IL-8 Secretion
Finally, the effect of MKP-1 upregulation induced by b 2agonists, alone or in combination with the corticosteroid fluticasone, was demonstrated by pretreating ASM cells for 1 h before stimulation with TNFa. As shown in Figure 5A, and explained by the cAMP-responsiveness of IL-6 (described earlier), formoterol significantly enhanced TNFa-induced IL-6 secretion. Despite this, when b 2 -agonists were added in combination with fluticasone they still significantly repressed TNFa-induced IL-6 (P,0.05). IL-8, on the other hand, is not cAMP-responsive; thus, b 2 -agonists did not significantly increase TNFa-induced IL-8 ( Figure 5B). Interestingly, fluticasone alone did not repress TNFainduced IL-8, while salmeterol, or formoterol, in combination with fluticasone significantly inhibited IL-6 secretion ( Figure 5B). The combination of the long-acting b 2 -agonists with fluticasone, has a statistically greater effect than fluticasone alone on TNFa-induced IL-8 secretion (P,0.05). Taken together, our results indicate that long-acting b 2 -agonists increase fluticasone-induced MKP-1 in an additive manner and in this way MKP-1 acts to repress cytokine secretion from ASM cells.
In preliminary studies, we then knocked-down MKP-1 with siRNA in an attempt to reverse the repression of cytokine secretion by formoterol and fluticasone, in combination. Although MKP-1 siRNA successfully reduced protein levels of MKP-1 in ASM cells ( Figure 6A) we were unable to reverse the inhibition of TNFainduced IL-6 ( Figure 6B) or IL-8 ( Figure 6C) secretion by formoterol and fluticasone. Investigations are now ongoing to examine the hypothesis that in the absence of MKP-1, the p38 MAPK-dependent RNA destabilizing factor -tristetraprolin (TTP) -may serve in a compensatory manner to repress inflammation. However further studies will be required to fully understand how these critical anti-inflammatory molecules are intrinsically intertwined and act together as a fail-safe mechanism to regulate inflammation in health and disease.

Discussion
This study demonstrates that long-acting b 2 -agonists increase fluticasone-induced MKP-1 mRNA expression and protein upregulation in ASM cells and together these asthma drugs have significant effects on cytokine secretion from ASM, an important immunomodulatory cell type in asthma.
Long-acting b 2 -agonists mediate their actions via interaction with the b 2 -adrenergic receptor, activation of adenylate cyclase and production of cAMP. Gene transcription is then induced in genes that contain cAMP-responsive elements (CRE) in their 59promoter. MKP-1 has a CRE-containing promoter [7,8] and we recently reported that the long-acting b 2 -agonist formoterol induced MKP-1 mRNA and protein expression in a cAMPmediated manner via the b 2 -adrenergic receptor-protein kinase A pathway [15]. MKP-1 gene expression and protein upregulation was rapid and peaked by 1 h [15]. We confirm these studies here and extend them to show that another commonly-used long-acting b 2 -agonist, salmeterol, also rapidly upregulates MKP-1 in a similar manner. Taken together with the demonstration that MKP-1 can be upregulated by cAMP analogues [20] these results present a unifying explanation that agonists that act to increase cAMP will increase MKP-1. Indeed, this has proven to be the case for sphingosine 1-phosphate (S1P), a bioactive sphingolipid known to be elevated in asthma [22]. S1P acts via G s -coupled receptors to increase cAMP [22] and we recently demonstrated that S1P increased MKP-1 in a cAMP-dependent manner to serve an important role a negative feedback controller and restrain the extent and duration of pro-inflammatory cellular signalling [20].
Importantly, when long-acting b 2 -agonists are added in combination with the corticosteroid fluticasone they significantly and rapidly upregulated MKP-1 protein and mRNA expression. The temporal pattern of MKP-1 protein upregulation reflects the additive contribution of each asthma drug. That is, b 2 -agonists rapidly, but transiently, upregulates MKP-1 at 1-2 h, while corticosteroids induced a sustained increase in MKP-1 protein.
Thus, when added together they upregulate MKP-1 early (by 1 h) but in a sustained manner. This additive upregulation was supported by mRNA levels. These results suggest a molecular basis for the added anti-inflammatory effects of both asthma drugs in combination.
To support the above assertion that increased expression of MKP-1 may explain the beneficial effects of b 2 -agonists/corticosteroid combination therapies in the repression of inflammatory gene expression in asthma [16], we performed in vitro studies measuring secretion of IL-6 and IL-8 from ASM cells after stimulation with TNFa. It is important to note that IL-6 is a cAMP-responsive gene. The 59-promoter for IL-6 contains a CRE [21] and can be induced by cAMP analogues, as well as shortacting and long-acting b 2 -agonists [21,23]. Herein we confirmed the cAMP-responsiveness of IL-6 by demonstrating significant IL-6 mRNA expression induced by the cell-permeable analogue, dibutyryl cAMP, as well as the adenylate cyclase activator, forskolin. In contrast, cAMP is only a weak activator of IL-8 in ASM cells [24] and, as highlighted by Holden et al. in lung epithelial cells [25], the promoter for IL-8 does not contain CRE consensus sequences [26] and that transcriptional regulation may involve cross-talk between transcription factors. Our current study demonstrates the lack of effect of cAMP-elevating agents on IL-8 expression in ASM cells.
Thus, due to the cAMP-dependence of IL-6 expression, the long-acting b 2 -agonist formoterol significantly enhanced TNFainduced IL-6 secretion. This corroborated our earlier report [21] and underscored the importance of the study by Holden et al. [25] where the authors stated that enhancement of IL-6 (and IL-8) release by b 2 -agonists may contribute to the deleterious effects of b 2 -agonist used monotherapy. Notably, when added in combination with fluticasone, b 2 -agonists significantly repressed IL-6 secretion, but not to a greater extent than fluticasone alone. These results are in agreement with Holden et al. where increased inflammatory cytokine released was reversed by corticosteroid action [25].
The repressive effect on TNFa-induced IL-8 were more clearcut, as long-acting b 2 -agonists added on their own did not increase IL-8 expression. Importantly, we observed that fluticasone alone was without repressive effect on TNFa-induced IL-8 secretion, but that b 2 -agonists added in combination significantly repressed TNFa-induced IL-8. This enhanced repressive effect of b 2agonists and corticosteroids correlates with the increased MKP-1 expression by both asthma drugs in combination. Why the combined effects of LABA/fluticasone on MKP-1 result in an additive suppressive effect on IL-8, but not on IL-6, cannot be completely delineated at this stage, but may reflect the relative MAPK-dependent contribution to IL-6 and IL-8 secretion. However, our in vitro results support those recently published in vivo in mild asthmatics [27], where formoterol in combination with corticosteroids improved clinical symptoms (FEV 1 ) and reduced IL-8 in a manner linked to MKP-1 upregulation. Interestingly, it has been previously shown by Knobloch et al. [28] that the corticosteroid dexamethasone suppresses IL-8 in ASM cells of non-smokers and smokers, but not in those of COPD; whether MKP-1 is involved warrants further investigation.
Our preliminary studies have raised the hypothesis that in the absence of MKP-1 (knocked-down by siRNA), the p38 MAPKdependent TTP may be upregulated and serve as a compensatory anti-inflammatory factor. TTP is an immediate-early gene [29] that is rapidly and robustly expressed via p38 MAPK signalling [30,31,32,33]. Moreover, TTP's function as an mRNA destabilising factor is dictated by phosphorylation on two key serine residues (S52 and S178). Further studies investigating the regulation of the RNA destabilising factor TTP by MKP-1, and visa versa, in ASM cells are warranted.
In summary, long-acting b 2 -agonists increase fluticasoneinduced MKP-1 in ASM cells and together these asthma drugs have significant repressive effects on cytokine secretion from ASM cells. This study further substantiates the anti-inflammatory role played by MKP-1 as an important molecular mechanism responsible for the improved anti-inflammatory action of corticosteroids and long-acting b 2 -agonists in combination in asthma.