such as cigarette smoke and oxidising pollutants, or from activated infl ammatory cells in the lung [1,3]. Increased ROS levels infl uence the infl ammatory response by activating redox-sensitive transcription factors, which induce pro-infl ammatory gene expres-sion [4]. ROS have been shown to activate transcrip-tion factors such as nuclear factor (NF)- κ B and activation protein 1 (AP-1), which are key regulators of several cellular processes, including chromatin remodelling and the expression of different pro-infl ammatory genes [5,6].

As a transcriptional regulator, AP-1 is involved in the control of cellular processes comprising cell

Piclamilast inhibits the pro-apoptotic and anti-proliferative responses of A549 cells exposed to H 2 O 2 via mechanisms involving AP-1 activation

MANUEL MATA 1,2,3 , FEDERICO PALLARDO 2 , ESTEBAN JES Ú S MORCILLO 2,3 & JULIO CORTIJO 1,2,3

1 Research Foundation of the University General Hospital of Valencia, Spain, 2 University of Valencia, Spain, and 3 Centro de Investigaci ó n Biom é dica en Red (CIBER) de Enfermedades Respiratorias, Valencia, Spain

(Received date: 24 December 2011 ; Accepted date: 20 February 2012 )

ABSTRACT Aims. Reactive oxygen species (ROS) are involved in the pathogenesis of many infl ammatory diseases such as chronic obstructive pulmonary disease (COPD). They can alter the expression of genes involved in cellular damage by activating transcription factors, including the NF- κ B and the activator protein 1 (AP-1). Phosphodiesterase type 4 (PDE4) inhibitors have anti-infl ammatory and antioxidant effects, as described in in vivo and in vitro COPD models. This study analysed the effects of piclamilast, a selective PDE4 inhibitor, on modulating the global gene expression profi le in A549 cells exposed to H 2 O 2 . Main methods . Changes in gene expression were analysed using high-density Affymetrix microarrays and validated by RT-PCR. Cell proliferation was studied using BrdU incorporation. Apoptosis was assessed by fl ow cytometry using annexin V-fl uorescein isothiocyanate. C-Jun phosphorylation and AP-1 activation were determined by ELISA and luciferase assay, respectively. Key fi ndings . Our results indicate that H 2 O 2 modifi ed the expression of several genes related to apoptosis, cell cycle control and cell signalling, including IL8 , FAS , HIG2 , CXCL2 , CDKN25 and JUNB . Piclamilast pre-treatment sig-nifi cantly inhibited the changes in 23 genes via mechanisms involving AP-1 activation and c-Jun phosphorylation at Ser63. Functional experiments confi rmed our results, suggesting new targets related to the antioxidant properties of PDE4 inhibi-tors. Signifi cance . This is the fi rst study to demonstrate antioxidant effects of a selective PDE4 inhibitor at the global gene expression level, and the results support the importance of AP-1 as a key regulator of the expression of genes involved in the infl ammatory response of epithelial cells to oxidative damage.

Keywords: oxidative stress , apoptosis , cAMP , microarrays , AP-1

Correspondence: Manuel Mata, Research Foundation of the University General Hospital of Valencia, Avda Tres Cruces s/n, 46014. Valencia, Spain. Tel: +34961972146. Fax: +34961972145. E-mail:mata_manroi@gva.es

Free Radical Research, May 2012; 46(5): 690–699

ISSN 1071-5762 print/ISSN 1029-2470 online © 2012 Informa UK, Ltd. DOI: 10.3109/10715762.2012.669040

ORIGINAL ARTICLE

Introduction

Chronic obstructive pulmonary disease (COPD) is a chronic respiratory disease associated with an abnor-mal infl ammatory response of the lungs to noxious particles, mainly cigarette smoke. This leads to chronic bronchitis-bronchiolitis (small airway disease) and emphysema, resulting in partially irreversible airfl ow limitation [1]. COPD is the fi fth leading cause of death worldwide, accounting for more than 2.5 million deaths annually (WHO World Health Report 2002).

Reactive oxygen species (ROS) are involved in the pathogenesis of COPD [2]. ROS can come from exter-nal sources, for example, inhaled airborne oxidants

Free

Rad

ic R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

yu M

edic

al C

ente

r on

12/

07/1

4Fo

r pe

rson

al u

se o

nly.

cAMP modulates gene expression changes induced by ROS 691

proliferation, survival and death [7]. AP-1 comprises multiple homodimeric and heterodimeric protein com-plexes that include members of the Jun and Fos families, with c-Jun being the major component [8]. The activity of c-Jun is controlled by the phosphorylation state of specifi c amino acids: phosphorylation at Ser63 or Ser73 activates c-Jun, and dephosphorylation at Thr239 inhib-its c-Jun [7]. Phosphorylation at Ser63 and Ser73 is mediated by ERK1/2 and JNK kinases [9,7] and results in transcriptionally active c-Jun [10]. Both ERK1/2 and JNK kinases are activated by ROS [11].

Phosphodiesterase type 4 (PDE4) inhibitors increase the cAMP level by inhibiting the enzyme responsible for its hydrolysis. The anti-infl ammatory effects of PDE4 inhibitors have been well documented in animal and cellular models of COPD [12]. PDE4 inhibitors markedly reduce ROS generation in neutrophils, air-way epithelial cells such as A549 type II cells and human pulmonary smooth muscle cells [12]. They can alter the effects of ROS on the upregulation of pro-infl ammatory molecules at the transcriptional level. For example, the PDE4 inhibitor rofl umilast has been reported to inhibit tumour necrosis factor- α and inter-leukin (IL)-1 β induction via mechanisms involving the inhibition of NF- κ B and the inhibition of p38 MAP kinase and JNK phosphorylation [13].

Alveolar type II cells play important roles in the pathogenesis of COPD. They secrete surfactant, which maintains alveolar and airway stability, and pro-infl ammatory molecules including Granzyme A and B. They are also involved in alveolar epithelial barrier repair [14]. Studies have described the tran-scriptional modifi cations occurring in airway epithe-lial cells in response to oxidative stress, particularly those in alveolar type II cells. In particular, the induc-tion of anti-proliferative and pro-apoptotic genes after oxidant insult has been demonstrated.

The present study analysed changes in the mRNA expression profi le of alveolar type II A549 cells in response to subtoxic levels of intracellular stress pro-duced by exposure to H 2 O 2 , and investigated the effect of piclamilast, a selective PDE4 inhibitor, on these changes. We studied the effect of piclamilast on apop-tosis and proliferation, and the role of c-Jun phospho-rylation and AP-1 activation in these processes.

Materials and methods

Cells and culture

Human A549 type II lung epithelial cells were obtained from the American Tissue Type Collection and cultured as described [15]. Before treatment, the cells were washed with 1 � phosphate-buffered saline (PBS).

A549 cells grown at confl uence in 6-well culture plates were pre-treated with or without piclamilast 1 μ M for 30 minutes and exposed to H 2 O 2 200 μ M

for 4 or 24 hours. Culture medium was removed, and the cells were processed for expression, proliferation, apoptosis or c-jun activation as described later.

This concentration of H 2 O 2 was chosen as subtoxic concentration based on previous studies. Toxicity was assessed by the MTT assay (data not shown).

Total RNA extraction and real-time RT-PCR

Total RNA was extracted from cultured cells using an RNeasy Mini kit (Qiagen, Hilden, Germany) accord-ing to the manufacturer ’ s instructions. The concen-tration and integrity of the extracted RNA were assessed using a 2100 Bioanalyser (Agilent, Santa Clara, CA, USA). Preparations with an RNA integrity number (RIN) � 9.5 were discarded.

Real-time RT-PCR was used to analyse gene expres-sion, as described previously [15]. cDNA was gener-ated with TaqMan reverse transcription reagents (P/N N808 - 0234; Applied Biosystems, Foster City, CA, USA). Assays-on-Demand (Applied Biosystems) were used to analyse selected genes (Table I).

Affymetrix gene expression analysis

Cells were cultured in medium containing 200 μ M H 2 O 2 without and with 1 μ M piclamilast for the indi-cated times. The cells were harvested by trypsinisation, and total RNA was extracted, as described above. The experimental procedures were as described in the GeneChip Expression Analysis Technical Manual (Affyme-trix, Santa Clara, CA, USA). The microarray chips were washed and labelled in an FS400 Hybridization Wash Station (Affymetrix, Santa Clara, CA, USA) and a GeneChip Hybridisation Oven 645 (Affymetrix, Santa Clara, CA, USA), respectively. The microarrays were scanned in a GeneChip Scanner 3000 (Affyme-trix, Santa Clara, CA, USA). Images (DAT fi les) were acquired using GeneChip Operating Software (GCOS), and digitalised images (CEL fi les) were generated.

DNA Chip Analysis Software (Cheng Li laboratory, Department of Biostatistics, Harvard University, Bos-ton, MA, USA) was used to analyse the data. The CEL fi les were normalised by the invariant sets method [16], and model-based expression values were obtained using the perfect match/mismatch difference model. Images were inspected for imperfections, and the qual-ity of the data was verifi ed with the outlier detection algorithm described by Li and Wong [16].

Analysis of variance (ANOVA) was used to test for signifi cant differences between control cells and cells exposed to H 2 O 2 in the presence or absence of piclamilast. The false discovery rate tool included in dCHIP was used to detect false positives.

Signifi cantly increased probe sets compared with baseline were identifi ed using the following crite-ria: � 1.8-fold change; difference of � 100 between baseline and experimental means; and detection call

Free

Rad

ic R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

yu M

edic

al C

ente

r on

12/

07/1

4Fo

r pe

rson

al u

se o

nly.

692 M. Mata et al.

Table I. Piclamilast prevented H 2 O 2 -induced changes in gene expression levels in A549 cells. A549 cells were cultured and exposed to 200 μ M H 2 O 2 in the absence or presence of 1 μ M piclamilast, for 4 and 24 hours. The following experimental groups were included: control cells, cells stimulated with 200 μ M H 2 O 2 for 4 and 24 hours, and cells pre-treated with 1 μ M piclamilast and then stimulated with 200 μ M H 2 O 2 (PI) for 4 and 24 hours. Total RNA was extracted, and the expression profi les were analysed using an Affymetrix human U133A expression microarray chip. Genes whose expression was altered after H 2 O 2 exposure and signifi cantly modulated by piclamilast were selected. Fold change (FC) changes in H 2 O 2 columns were calculated using control cells as baseline, while FC changes in PI (4 and 24 hours) columns were calculated using H 2 O 2 (4 and 24 hours respectively) as baseline.

Probe set Gene titleGene

symbolGO biological

process

H2O2 4 hours

FC

H2O2 24 hours

FC

PI 4 hours

FC

PI 24 hours

FC

202540_s_at 3-hydroxy-3-methylglutaryl-Coenzyme A reductase

HMGCR Apoptosis � 1.92 NS 2.59 NS

200800_s_at heat shock 70kDa protein 1A /// heat shock 70kDa protein 1B

HSPA1A /// HSPA1B

Apoptosis � 4.25 � 1.86 2.67 NS

209230_s_at nuclear protein 1 NUPR1 Apoptosis 3 NS � 5.5 NS204962_s_at centromere protein A CENPA Apoptosis � 3.24 NS 3.29 NS218145_at tribbles homolog 3 (Drosophila) TRIB3 Apoptosis 2.18 NS � 2.73 NS204803_s_at Ras-related associated with diabetes RRAD Cell

communication2.48 NS � 2.09 NS

204695_at cell division cycle 25A CDC25A Cell cycle NS � 1.81 2.31 NS209714_s_at cyclin-dependent kinase inhibitor 3 CDKN3 Cell cycle NS 1.97 NS � 2.2204240_s_at structural maintenance of chromosomes 2 SMC2 Cell cycle NS � 1.95 NS 2.38202147_s_at interferon-related developmental regulator 1 IFRD1 Cell cycle 1.84 NS � 2.03 NS217988_at cyclin B1 interacting protein 1 CCNB1IP1 Cell cycle 1.9 NS � 1.93 NS204203_at CCAAT/enhancer binding protein (C/EBP),

gammaCEBPG Immune response 1.99 NS � 1.95 NS

209774_x_at chemokine (C-X-C motif) ligand 2 CXCL2 Immune response 2.1 NS � 2.09 NS201163_s_at insulin-like growth factor binding protein 7 IGFBP7 Immune response 2.01 2.66 � 2.16 NS202859_x_at interleukin 8 IL8 Immune response 7.46 1.86 � 3.17 � 2.88218976_at DnaJ (Hsp40) homolog, subfamily C,

member 12DNAJC12 Metabolism NS 2.1 � 1.83 NS

209398_at histone 1, H2ac HIST1H1C Metabolism NS 1.81 � 2.76 NS201473_at jun B proto-oncogene JUNB Regulation of

transcription2.18 NS � 2.43 NS

215719_x_at Fas (TNF receptor superfamily, member 6) FAS Response to stress 4.17 2.08 � 2.08 NS218507_at hypoxia-inducible protein 2 HIG2 Response to stress 2.62 1.98 NS NS204178_s_at RNA binding motif protein 14 RBM14 Response to stress � 2 NS 1.87 NS212190_at serpin peptidase inhibitor, clade member 1 SERPINE2 Response to stress NS 2.18 NS � 2.42

FC, fold change; PI, piclamilast pre-treated; NS, no signifi cative.

of ‘ Present ’ in the experimental group. Signifi cantly decreased probe sets met the following criteria: � 1.8-fold change; difference of � 100 between baseline and experimental means; and detection call of ‘ Present ’ in the baseline group. Only those genes that passed the ANOVA fi lter were considered.

Non-agglomerative two-dimensional hierarchical clustering was used to analyse the data expression profi les obtained after applying fi lters, as described previously [17]. The Euclidean distance was used to generate clusters, and probe sets were grouped accord-ing to similar expression values. The two-dimensional non-supervised principal components analysis (PCA) algorithm was used to classify the data [16].

Finally, a gene expression browser tool (NetAffx Analysis Centre, Affymetrix, Santa Clara, CA, USA) was used to categorise genes according to their functionality.

Transcription factor binding sites prediction

Putative transcription factor binding sites were pre-dicted using PROMO software [18,19]. Reference

sequences for selected genes were imported from GenBank (National Center for Biotechnology, Bethesda, MD, USA). The 500 base pairs upstream from the transcription origin (5 ¢® 3¢ ) were selected and imported into the PROMO database. Only fac-tors common to all sequences were considered. The Universal Protein Resource (UniProt) database (http://www.uniprot.org) was interrogated regarding proteins obtained with this approach.

Cell proliferation assay

Human A549 cell proliferation was measured colori-metrically based on BrdU incorporation during DNA synthesis, using a BrdU cell proliferation enzyme-linked immunosorbent assay kit (Roche, Mannheim, Germany) according to the manufac-turer ’ s protocol.

Apoptosis measurement

Apoptosis was detected in human A549 cells as previ-ously described [20]. Briefl y, cells were cultured in

Free

Rad

ic R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

yu M

edic

al C

ente

r on

12/

07/1

4Fo

r pe

rson

al u

se o

nly.

cAMP modulates gene expression changes induced by ROS 693

12-well plates and exposed to 200 μ M H 2 O 2 in the presence or absence of 1 μ M for 24 hours. The cells were trypsinised, stained with annexin V-fl uorescein isothiocyanate and propidium iodide (PI) following the manufacturer ’ s protocol (Annexin-V-Fluos; Roche Applied Science, Barcelona, Spain), and analysed by fl ow cytometry (CyAn ADP; Dako A/S, Glostrup, Denmark).

c-Jun phosphorylation

To investigate c-Jun phosphorylation, A549 cells were growth in T25 fl asks and exposed to 200 μ M H 2 O 2 in the presence or absence of 1 μ M piclamilast for 1 hours. The culture medium was removed, and the cells were rinsed once with ice-cold PBS. Then, 0.5 ml of cell lysis buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM β -glycerophosphate, 1 mM Na 3 VO 4 and 1 μ g/ml leu-peptin) was added, and the cells were scraped and sonicated. The lysates were clarifi ed by centrifugation at 14,000 � g for 10 min at 4 ° C, and the supernatants were stored at � 80 ° C until analysed. The lysate pro-tein concentration was measured using BCA (Thermo Scientifi c, Rockford, IL, USA), and phospho-c-Jun in the lysate was assessed using a PathScan ® Phospho-c-Jun (Ser63) Sandwich ELISA Kit II (Cell Signaling Technology, Boston, MA, USA), following the manufacturer ’ s instructions.

Transient transfection and luciferase assay

The activity of AP-1 was measured using a luciferase reporter assay kit (BD Bioscience Clontech, Palo Alto, CA, USA). To analyse the expression of the AP-1-Luc reporter plasmid, cells were plated in 24-well plates (5 � 10 4 cells/well) and transfected the following day with 1 μ g of plasmid, using TransFast transfection reagent (Promega, Madison, WI, USA), according to the manufacturer ’ s protocol. After a 2-hour incubation with the plasmid – lipid suspension, the medium was changed, and the cells were grown for an additional 12 hours. The cells were incubated with 200 μ M H 2 O 2 for 4 hours, lysed and subjected to the luciferase assay, following the manufacturer ’ s instructions.

Statistical analysis

Affymetrix data analysis was performed as described above. Data from the other experiments are expressed as means � SEM of n independent experiments. Sta-tistical analysis consisted of one-way ANOVA fol-lowed by the Bonferroni correction for multiple comparisons. Signifi cance was accepted for p � 0.05.

Results

Effect of piclamilast on transcriptional changes induced by H 2 O 2 , analysed with Affymetrix microarrays

The changes in the gene expression profi le of human epithelial cells after exposure to H 2 O 2 were determined using Affymetrix technology. A549 cells were grown to confl uence in 6-well culture plates and exposed to 200 μ M H 2 O 2 in the presence or absence of 1 μ M piclamilast. At 4 or 24 hours after H 2 O 2 stimulation, the culture medium was removed, and total RNA was extracted. Human U133A arrays were processed as described in the Methods, and CEL fi les were generated using GCOS software. The digitalised images were nor-malised, and after ANOVA and fi ltering, a total of 171 genes passed the fi ltering criteria described in the Meth-ods (data not shown). The false discovery rate was � 0.9%, indicating a low number of false positives.

The expression levels of these genes were also anal-ysed in cultures pre-treated with 1 μ M piclamilast and then exposed to the same H 2 O 2 concentration. The expression levels of 22 genes were signifi cantly altered by pre-incubation with piclamilast (Table II). A func-tional analysis indicated that these genes are involved in apoptosis, the cell cycle, oxidative stress and infl am-mation. DNA Chip analyser software was used to clus-ter the expression values of selected genes and samples (Figure 1, panel A). The results concurred with those obtained by unsupervised PCA analysis (Figure 1, panel B), which graphically represents the differential distribution of samples according to the covariance of the expression values obtained for the fi ltered genes.

To validate the microarray data, real-time RT-PCR was performed using specifi c primers to target the same sequences used for the Affymetrix probes (Table I). We included two more experimental groups: A549 cells exposed to 1 μ M piclamilast for 4 or 24 hours, but not exposed to H 2 O 2 . There were no signifi cant differences in gene expression between either of these two groups and the control group (data not shown). The changes in gene expression revealed by real-time RT-PCR were similar to those identifi ed in the microar-ray data, supporting the statistical analysis used in this study. The only differences observed are due to differ-ences in statistical signifi cance of the two technologies that were used.

Piclamilast prevents apoptosis induction in A549 cells exposed to H 2 O 2

Confl uent A549 cells were exposed to 200 μ M H 2 O 2 for 24 hours in the presence or absence of 1 μ M piclamilast and assayed for apoptosis. Exposure to H 2 O 2 increased the percentage of apoptotic cells by 166.08%, and this increase was almost completely prevented by co-incubation with 1 μ M piclamilast (Figure 2). As expected, treatment of cells with piclamilast alone had no effect.

Free

Rad

ic R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

yu M

edic

al C

ente

r on

12/

07/1

4Fo

r pe

rson

al u

se o

nly.

694 M. Mata et al.

Table II. Real-time RT-PCR validation of the microarray data. The comparative DDCt method was used to compare expression levels of selected genes in cells at 4 or 24 hours after exposure to 200 μ M H 2 O 2 and in unexposed cells. The expression levels of these genes were also analysed at 4 and 24 hours in cultures pre-treated with 1 μ M piclamilast (PI) compared to H 2 O 2 stimulated cells at 4 and 24 hours respectively.

Gene symbol Assay IDH 2 O 2 4

hours FCH 2 O 2 24 hours FC

PI 4 hours FC

PI 24 hours FC

HMGCR Hs00168352_m1 � 2.01 NS 2.39 NSHSPA1A Hs00359163_s1 � 7.21 � 2.13 3.15 NSNUPR1 Hs00202610_m1 2.89 NS � 3.05 NSCENPA Hs00156455_m1 � 6.25 NS � 6.50 NSTRIB3 Hs01082394_m1 3.21 NS � 3.11 NSRRAD Hs00188163_m1 3.46 NS � 3.27 NSCDC25A Hs00947994_m1 NS � 3.25 NS 1.99CDKN3 Hs00193192_m1 NS 2.13 NS � 2.01SMC2 Hs00197593_m1 NS � 2.98 NS 3.15IFRD1 Hs00912593_m1 1.89 NS � 2.40 NSCCNB1IP1 Hs01565445_m1 3.45 NS � 3.25 NSCEBPG Hs00156454_m1 3.51 NS � 2.99 NSCXCL2 Hs00601975_m1 6.23 NS � 7.00 NSIGFBP7 Hs00266026_m1 1.89 3,65 � 2.07 � 1.52IL8 Hs00174103_m1 12.24 2.03 � 8.23 � 1.15DNAJC12 Hs01111685_m1 NS 1.89 � 1.13 � 2.01HIST1H1C Hs00271185_s1 NS 2.25 NS � 2.54JUNB Hs00197370_m1 1.98 NS � 2.00 NSFAS Hs00531110_m1 4.08 6.25 � 3.09 � 4.89HIG2 Hs00203383_m1 3.57 3.25 � 1.57 � 3.59RBM14 Hs00197061_m1 � 3.51 NS � 3.71 NSSERPINE2 Hs00385730_m1 NS 4.32 NS � 3.22

FC, fold change; NS, ;PI, propidium iodide.

Figure 1. Piclamilast prevented H 2 O 2 -induced changes in gene expression levels in A549 cells. A549 cells were cultured and exposed to 200 μ M H 2 O 2 in the absence or presence of 1 μ M piclamilast, for 4 and 24 hours. The following experimental groups were included: control cells (C), cells stimulated with 200 μ M H 2 O 2 for 4 (H4) and 24 (H24) hours, and cells pre-treated with 1 μ M piclamilast and then stimulated with 200 μ M H 2 O 2 for 4 (P4) and 24 (P24) hours. Total RNA was extracted, and the expression profi les were analysed using an Affymetrix human U133A expression microarray chip. Genes whose expression was altered after H 2 O 2 exposure and signifi cantly modulated by piclamilast were selected. Supervised non-agglomerative hierarchical cluster analysis (panel A) and non-supervised Principal Components Analysis (panel B) were conducted using the DNA chip analyser software. All experiments were performed in triplicate. Green and red colours (panel A) made reference to Euclidean distances calculated using the dCHIP analyzer software.

Piclamilast restores the proliferation of epithelial cells

BrdU incorporation in A549 epithelial cells was mea-sured at 24 hours. As expected, cultures exposed to H 2 O 2 had a low growth rate, 50% of that in the unexposed cells. In cells pre-treated with piclamilast, the impairment of cell growth by H 2 O 2 was partially prevented (Figure 3).

Predicting putative transcription factor binding sites

The microarray analysis revealed 22 genes whose expression was altered signifi cantly (17 upregulated and 5 downregulated) as an effect of piclamilast compared with the expression levels in H 2 O 2 -stimulated cells. To examine whether common

Free

Rad

ic R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

yu M

edic

al C

ente

r on

12/

07/1

4Fo

r pe

rson

al u

se o

nly.

cAMP modulates gene expression changes induced by ROS 695

Figure 2. H 2 O 2 induced apoptosis in A549 cells. A549 cells were cultured and exposed to 200 μ M H 2 O 2 for 24 hours in the presence or absence of 1 μ M piclamilast. Apoptosis was measured by fl ow cytometry using annexin V-fl uorescein isothiocyanate and PI. Representative scatter-plots are shown in panel A. H 2 O 2 induces an increase of FITC fl uorescence (left), which is reverted in piclamilast pre-treated cultures (right). Fluorescence mean intensities of three independent experiments are represented in panel B. Results are expressed as % of positive cells vs. untreated for each experimental group. * p � 0.05 compared with control. # p � 0.05 compared with H 2 O 2 -exposed cells.

Figure 3. H 2 O 2 inhibited cell proliferation in A549 cells. A549 cells were cultured and exposed to 200 μ M H 2 O 2 for 8, 12 and 24 hours in the presence or absence of 1 μ M piclamilast. Cell proliferation was assayed by measuring BrdU incorporation during DNA synthesis. Data represent the percentage of BrdU incorporation compared with that in control cultures at 24 hours. Experiments were performed in triplicate, and the data represent the results of three independent experiments. * p � 0.05 compared with control. # p � 0.05 compared with H 2 O 2 -exposed cells.

transcription regulatory elements were involved, the 500-bp regions upstream from the transcription ori-gins of the upregulated genes were characterised. The sequences were submitted to the PROMO database, and the maximum matrix dissimilarity rate was fi xed at 30. Considering only transcription factors common to all of the submitted sequences, we found 12 puta-tive binding sites for transcription factors (Table III), including the transcription factor AP-1.

Piclamilast prevents AP1 activation in A549 cells exposed to H 2 O 2

To examine the role of AP1 in regulating the tran-scription of selected genes, A549 cells were cultured in 24-well plates and transfected with AP-1-Luc plas-mid, a luciferase reporter plasmid containing the AP-1 sequence. The cells were exposed to 200 μ M H 2 O 2 for 4 hours in the presence or absence of 1 μ M piclamilast, and the luciferase activity was measured.

The exposure to H 2 O 2 strongly induced luciferase activity (2.43-fold change compared with mock-treated cells; Figure 4, panel A). This increase was signifi cantly inhibited by 1 μ M piclamilast, reducing the luciferase activity to a level near that of the control cultures (1.25-fold change compared with mock-treated cells). We did not see luciferase activity in cul-tures pre-treated with 1 μ M piclamilast (0.98-fold change compared with mock-treated cultures).

Piclamilast prevents c-Jun phosphorylation in A549 cells exposed to H 2 O 2

c-Jun is the major component of AP-1, and its activ-ity is regulated by the phosphorylation of specifi c amino acid residues. Next, we analysed whether c-Jun activation was involved in the regulatory effect of piclamilast. A549 cells were cultured in T25 culture fl asks and exposed to 200 μ M H 2 O 2 in the presence or absence of 1 μ M piclamilast for 30 minutes. The cells were lysed, and c-Jun phosphorylation was anal-ysed by ELISA. c-Jun was markedly phosphorylated at Ser63 (2.37-fold change compared with unstimu-lated cultures), and this was inhibited by piclamilast pre-treatment (1.51-fold compared with unstimu-lated cultures) (Figure 4, panel B). No effect of piclamilast alone was observed. The results support a role of c-Jun in the observed activation of AP-1.

Discussion

This article presents an Affymetrix-based gene expres-sion profi le analysis of human type II A549 epithelial

Free

Rad

ic R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

yu M

edic

al C

ente

r on

12/

07/1

4Fo

r pe

rson

al u

se o

nly.

696 M. Mata et al.

Table III. Putative transcription factor binding sites in genes upregulated by oxidative stress. The 500-bp regions upstream from the transcription origins of the genes whose expression levels were upregulated by 200 μ M H 2 O 2 stimulation in A549 cells and modulated by 1 μ M piclamilast were analysed using PROMO. We selected only those sequences that were present in all of the genes analysed. The maximum acceptable matrix dissimilarity rate was 30. The Multi-genome Analysis of Positions and Patterns of Elements of Regulation (MAPPER) bioinformatics tool was used to calculate consensus sequences using the Hidden Markov Model.

Factor name Factor symbol HMM consensus

General transcription factor II-I TFII-I * - � agaggtagg � - * Signal Transducer and Activator of Transcription 4 STAT4 * - � ttcccggaa � - * Protein C-ets-1 c-Ets-1 * - � acttcctg � - * Transcriptional repressor protein Ying and Yang 1 YY1 * - � ccatttttagac � - * Retinoid X receptor alpha RXR-alpha * - � gatg.cctt.ctgaccc.ctt � - * CCAAT/enhancer-binding protein beta C/EBPbeta * - � TtgCa.aat.a � - * Glucocorticoid receptor beta GR-beta * - � tgtcct � - * Cellular tumour antigen p53 p53 * - � ggacatgcccgggcatgtcc � - * X-box-binding protein 1 XBP1 * - � atcacgcg � - * Transcription factor AP-1 AP1/c-Jun * - � TgActca.aggt � - * Glucocorticoid receptor alpha GR-alpha * - � acatgctg � - * activating enhancer-binding protein 2 alpha AP-2alphaA * - � gcccccaggcggggc � - *

Figure 4. Piclamilast inhibited the oxidative-dependent activation of AP-1 and c-Jun in A549 cells. A549 cells were cultured and transfected with AP-1-Luc plasmid before stimulation with 200 μ M H 2 O 2 in the presence or absence of 1 μ M piclamilast. After 4 hours, the cells were lysed, and the luciferase activity was measured (panel A). For c-Jun activation, the cells were cultured and exposed to 200 μ M H 2 O 2 in the presence or absence of 1 μ M piclamilast for 30 minutes. c-Jun phosphorylated at Ser63 was determined by ELISA. Experiments were performed in triplicate, and the data represent the results of three independent experiments. * p � 0.05 compared with control. # p � 0.05 compared with H 2 O 2 -exposed cells.

cells exposed to subtoxic concentrations of H 2 O 2 in the presence or absence of piclamilast, a selective PDE4 inhibitor.

The effects of H 2 O 2 on the viability of A549 cells were previously studied in the laboratory with a 3-(4,5-dimethylthiazol-2-yl)-2,5 diphenyltetrazolium bromide (MTT) test (data not shown). The following doses of H 2 O 2 were tested: 100, 200, 250, 500, 750 and 1000 μ M. The highest dose that did not induce

a signifi cantly release of MTT was 200 μ M. This is the reason why we selected this dose of H 2 O 2 . In this study, we were interested in analysing both anti-infl ammatory and anti-apoptotic effects of piclami-last. For this reason, two different time points have been selected for microarray analysis: 4 (for infl am-matory alterations) and 24 (for apoptosis alterations) hours after H 2 O 2 addition. Using the selected dose of H 2 O 2 , the increase in the expression of IL8 (selected

Free

Rad

ic R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

yu M

edic

al C

ente

r on

12/

07/1

4Fo

r pe

rson

al u

se o

nly.

cAMP modulates gene expression changes induced by ROS 697

as marker of infl ammation) was maximal at 4 hours after stimulation of cells, while the annexin V staining (selected as marker of apoptosis) was maximal at 24 hours after stimulation of cells (data not shown).

Once the dose and time points were fi xed, microar-ray experiments were performed. H 2 O 2 exposure affected the expression of several genes related to apoptosis, proliferation and oxidative stress responses. These changes were minimised by piclamilast and were correlated with cellular measurements of apop-tosis and proliferation based on cytometry and BrdU incorporation, respectively. Our results support the involvement of the transcription factor AP-1 and the phosphorylation of its major component, the protein c-Jun, at Ser63.

Oxidative stress is involved in the pathogenesis of various airway diseases, including asthma, interstitial fi brosis and COPD. There is evidence implicating the antioxidant role of PDE4 inhibitors in cells in COPD and fi brosis.

Oxidative stress inhibits the proliferation of epithe-lial cells, and at least three of the genes identifi ed here (Table II) are important cell cycle regulators. Cyclin B1 interacting protein 1 (CCNB1IP1) is a member of the E3 ubiquitin ligase family and its expression reduces cyclin B1 levels inhibiting, the G2/M transi-tion stimulated by the CDK/cyclin B complex [21]. CCNB1IP1 expression was higher in H 2 O 2 -exposed cells compared with control levels, and this effect was abolished by piclamilast. Cell division cycle 25 A (CDC25A) is an essential regulator of the cell cycle via the dephosphorylation and activation of the CDK2/cyclin E and A complex, which regulates the G1/S and G2/M transitions [22]. Oxidative stress induced by H 2 O 2 decreased CDC25A expression in our cultures, and this was reversed by piclamilast. Similar behaviour has been observed for the cyclin-dependent kinase inhibitor CDKN3, which is involved in the dephosphorylation and inactivation of CDK2 kinase [23]. These changes are consistent with the results of the BrdU incorporation assays, which indi-cate a signifi cant decrease in the proliferation of cells exposed to H 2 O 2 and reversal of this effect by 1 μ M piclamilast.

Many researchers have reported the pro-apoptotic effects of oxidative stress in epithelial cells [2] and the relevance of oxidative stress in the development of infl ammatory respiratory diseases such as COPD [24]. One of the best-characterised apoptotic systems is the FAS / FAS ligand receptor-mediated death-signalling pathway. Stimulation of the FAS receptor results in its trimerisation and subsequent activation of caspase 3, which culminates in effective cell death [25]. Our data confi rmed the upregulation of FAS as a result of H 2 O 2 exposure; FAS upregulation was inhibited by piclamilast treatment and correlated with the results obtained with Annexin V (Figure 2). The anti-apoptotic effect of rolipram, a selective PDE4

inhibitor, has been described in a model of apoptosis induction by gal-8 in Jurkat cells [26], but the present study is the fi rst to report an anti-apoptotic effect of a PDE4 inhibitor in epithelial cells exposed to ROS.

Two other important apoptosis-related genes are centromere protein A, which is involved in chromo-some stabilisation, and heat shock 70-kDa protein 1, a direct inhibitor of caspase 3-induced apoptosis [27]. Their expression was decreased after a 4-hour expo-sure to H 2 O 2 and increased in cells pre-treated with piclamilast. TRIB3 is a potent negative regulator of the pro-survival kinase Akt and is involved in the sup-pression of cell survival by mixed linage kinase 3 [28]. TRIB3 expression was strongly increased after expo-sure to H 2 O 2 for 4 hours, and its expression level fell below baseline in cells pre-treated with piclamilast.

The Jun gene family is a key component of AP1, a redox-sensitive transcription factor involved in the regulation of several infl ammatory genes [29]. Recently, Chaum et al. [30] reported that H 2 O 2 increased the expression of JUNB in a dose-dependent manner in human epithelial cells. This fi nding is consistent with our data, which show sig-nifi cant upregulation of two important infl ammatory genes, CXCL2 and IL8 , whose expression is con-trolled by AP1 [31,32]. Moreover, PDE4 inhibition reduced the expression of JUNB , CXCL2 and IL8 to near-baseline levels, indicating the importance of AP-1 activation in the infl ammatory response of the epithelium to oxidative injury. The involvement of AP-1 in cellular responses to oxidative stress is fur-ther supported by the identifi cation of 12 putative binding sites for transcription factors, including AP-1, in the 5¢ untranslated regions of genes upreg-ulated by H 2 O 2 treatment in the present study. The strong induction of luciferase activity by H 2 O 2 in A549 cells transfected with an AP-1-Luc reporter plasmid, which is consistent with results of a previ-ous study [33], provides additional evidence indicat-ing a critical role for AP-1 in the oxidative stress response. In addition, the phosphorylation of c-Jun at Ser63, refl ecting c-Jun activation, was observed in H 2 O 2 -exposed cells in the present study; as c-Jun is the major component of AP-1, its activation would presumably result in AP-1 activation [7]. Piclamilast signifi cantly reduced H 2 O 2 -induced luciferase activ-ity and c-Jun phosphorylation, suggesting that the observed effects of this PDE4 inhibitor are mediated via the activation of AP-1. The role of AP-1 activa-tion in the mechanism of PDE4 inhibition has been reported in different in vitro models, including T cells stimulated with IL-4 [34] and A549 cells exposed to oxidants [35]. However, these studies analysed the expression of only IL-5 and IL-8, respectively. Here, we identifi ed 22 genes whose expression levels were altered by piclamilast and that contained putative AP-1 binding sites in their 5 ¢ untranslated regions.

Free

Rad

ic R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

yu M

edic

al C

ente

r on

12/

07/1

4Fo

r pe

rson

al u

se o

nly.

698 M. Mata et al.

In summary, PDE4 inhibitors have demonstrated effi cacy in the treatment of infl ammatory diseases such as COPD, by reducing ROS generation in different cell types, including neutrophils, smooth muscle cells and airway epithelial cells [12]. Our data demonstrate that piclamilast, a PDE4 inhibitor, reduced the expression of genes involved in apoptosis and cell cycle regulation in epithelial cells exposed to subtoxic concentrations of H 2 O 2 . These effects were correlated with functional changes observed by the fl ow cytometric analysis of apoptosis and the determination of BrdU incorpora-tion. Furthermore, the effects of piclamilast are prob-ably mediated by the activation of the transcription factor AP-1. We identifi ed several novel genes that con-tain putative AP-1 binding sites and are likely involved in the antioxidant effect of PDE4 inhibitors in airway epithelial cells. These fi ndings contribute to an under-standing of the molecular mechanisms of the anti-infl ammatory effect of PDE4 inhibitors.

Acknowledgements

This work was supported by grants SAF2005 - 00669/SAF2008 - 03113 (JC), PI10/02294 (MM) and CIBERES (CB06/06/0027) from the Ministry of Sci-ence and Innovation and Health Institute ‘ Carlos III ’ of Spain and by research grants from the Regional Government (GV2007/287 and AP073/10, from Generalitat Valenciana) SAF2009-08913 (EJM) and Prometeo/2008/045 (EJM).

Declaration of interest

The authors report no confl icts of interest. The authors alone are responsible for the content and writing of the paper.

References

Demedts IK, Demoor T, Bracke KR, Joos GF, Brusselle GG. [1] Role of apoptosis in the pathogenesis of COPD and pulmo-nary emphysema. Respir Res 2006;7:53 – 62. Chandra J, Samali A, Orrenius S. Triggering and modulation [2] of apoptosis by oxidative stress. Free Radic Biol Med 2000;29:323 – 333. Malia P, Johnson SL. How viral infections causes exacerba-[3] tions of airway diseases. Chest 2006;130:1203 – 1210. Kirkham P, Rahman I. Oxidative stress in asthma and [4] COPD: antioxidants as a therapeutic strategy. Pharmacol Ther 2006;111:476 – 494. Rahman MS, Yang J, Shan LY, Unruh H, Yang X, Halayko [5] AJ, et al. IL-17R activation of human airway smooth muscle cells induces CXCL-8 production via a transcriptional-dependent mechanism. Clin Immunol 2005;115:268 – 276. Biswas SK, Rahman I. Environmental toxicity, redox signal-[6] ing and lung infl ammation: the role of glutathione. Mol Aspects Med 2009;30:60 – 76. Morton S, Davis RJ, McLaren A, Cohen P. A reinvestigation [7] of the multisite phosphorylation of the transcription factor c-Jun. EMBO J 2003;22:3876 – 3886.

Alani R, Brown P, Bin é truy B, Dosaka H, Rosenberg RK, [8] Angel P, et al. The transactivating domain of the c-Jun proto-oncoprotein is required for cotransformation of rat embryo cells. Mol Cell Biol 1991;11:6286 – 6295. Davies SP, Reddy H, Caivano M, Cohen P. Specifi city and [9] mechanism of action of some commonly used protein kinase inhibitors. Biochem J 2000;351:95 – 105. Pulverer BJ, Kyriakis JM, Avruch J, Nikolakaki E, Woodgett [10] JR. Phosphorylation of c-jun mediated by MAP kinases. Nature 1991;353:670 – 674. Hsieh HL, Wang HH, Wu WB, Chu PJ, Yang CM. Transform-[11] ing growth factor- β 1 induces matrix metalloproteinase-9 and cell migration in astrocytes: roles of ROS-dependent ERK- and JNK-NF- κ B pathways. J Neuroinfl amm 2010;7:88 – 104. Hatzelmann A, Morcillo EJ, Lungarella G, Adnot S, Sanjar [12] S, Beume R, et al. The preclinical pharmacology of rofl um-ilast—a selective, oral phosphodiesterase 4 inhibitor in development for chronic obstructive pulmonary disease. Pulm Pharmacol Ther 2010;23:235 – 256. Kwak HJ, Song JS, Heo JY, Yang SD, Nam JY, Cheon HG, [13] et al. Rofl umilast inhibits lipopolysaccharide-induced infl am-matory mediators via suppression of nuclear factor- κ B, p38 mitogen-activated protein kinase, and c-Jun NH2-terminal kinase activation. J Pharmacol Exp Ther 2005;315:1188 – 1195. Zhao CZ, Fang XC, Wang D, Tang FD, Wang XD. Involve-[14] ment of type II pneumocytes in the pathogenesis of chronic obstructive pulmonary disease. Respir Med 2010;104:1391 – 1395. Mata M, Sarri á B, Buenestado A, Cortijo J, Cerd á M, Mor-[15] cillo EJ, et al. Phosphodiesterase 4 inhibition decreases MUC5AC expression induced by epidermal growth factor in human airway epithelial cells. Thorax 2005;60:144 – 152. Li C, Wong WH. Model-based analysis of oligonucleotide [16] arrays: model validation, design issues and standard error application. Genome Biol 2001;2:research 0032.1 – 0032.11. Barco A, Patterson S, Alarcon JM, Gromova P, Mata-Roig [17] M, Morozov A, et al. Gene expression profi ling of facilitated L-LTP in VP16-CREB mice reveals that BDNF is critical for the maintenance of LTP and its synaptic capture. Neu-ron 2005;8:123 – 137. Messeguer X, Escudero R, Farr é D, Nu ñ ez O, Mart í nez J, [18] Alb à MM, et al. PROMO: detection of known transcription regulatory elements using species-tailored searches. Bioin-formatics 2002;18:333 – 334. Farr é D, Roset R, Huerta M, Adsuara JE, Rosell ó L, Alb à [19] MM, et al. Identifi cation of patterns in biological sequences at the ALGGEN server: PROMO and MALGEN. Nucleic Acids Res 2003;31:3651 – 3653. Martinez-Losa M, Cortijo J, Piqueras L, Sanz MJ, Morcillo [20] EJ. Taurine chloramine inhibits functional responses of human eosinophils in vitro . Clin Exp Allergy 2009;39:537 – 546. Toby GG, Gherraby W, Coleman TR, Golemis EA. A novel [21] RING fi nger protein, human enhancer of invasion 10, alters mitotic progression through regulation of cyclin B levels. Mol Cell Biol 2003;23:2109 – 2122. Contour-Galcera MO, Sidhu A, Pr é vost G, Bigg D, Ducom-[22] mun B. What’s new on CDC25 phosphatase inhibitors. Pharmacol Ther 2007;115:1 – 12. Yeh CT, Lu SC, Chao CH, Chao ML. Abolishment of the [23] interaction between cyclin-dependent kinase 2 and Cdk-associated protein phosphatase by a truncated KAP mutant. Biochem Biophys Res Commun 2003;305:311 – 314. Demedts IK, Demoor T, Bracke KR, Joos GF, Brusselle GG. [24] Role of apoptosis in the pathogenesis of COPD and pulmo-nary emphysema. Respir Res 2007;30:7 – 53. De Paepe ME, Gundavarapu S, Tantravahi U, Pepperell JR, [25] Haley SA, Luks FI, et al. Fas-ligand-induced apoptosis of

Free

Rad

ic R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

yu M

edic

al C

ente

r on

12/

07/1

4Fo

r pe

rson

al u

se o

nly.

cAMP modulates gene expression changes induced by ROS 699

respiratory epithelial cells causes disruption of postcanal-icular alveolar development. Am J Pathol 2008;173:42 – 56. Norambuena A, Metz C, Vicu ñ a L, Silva A, Pardo E, Oya-[26] nadel C, et al. Galectin-8 induces apoptosis in Jurkat T cells by phosphatidic acid-mediated ERK1/2 activation sup-ported by protein kinase A down-regulation. J Biol Chem 2009;284:12670 – 12679. Marfe G, Pucci B, De Martino L, Fiorito F, Di Stefano C, [27] Indelicato M, et al. Heat-shock pretreatment inhibits sorbitol-induced apoptosis in K562, U937 and HeLa cells. Int J Can-cer 2009;125:2077 – 2085. Humphrey RK, Newcomb CJ, Yu SM, Hao E, Yu D, Kra-[28] jewski S, et al. Mixed lineage kinase-3 stabilises and func-tionally cooperates with TRIBBLES-3 to compromise mitochondrial integrity in cytokine-induced death of pan-creatic beta cells. J Biol Chem 2010;285:22426 – 22436. Buford TW, Cooke MB, Willoughby DS. Resistance exercise [29] induced changes of infl ammatory gene expression within human skeletal muscle. Eur J Appl Physiol 2009;107:463 – 471. Chaum E, Yin J, Yang H, Thomas F, Lang JC. Quantitative [30] AP-1 gene regulation by oxidative stress in the human

retinal pigment epithelium. J Cell Biochem 2009;108:1280 – 1291. Schmeck B, Gross R, N’Guessan PD, Hocke AC, Ham-[31] merschmidt S, Mitchell TJ, et al. Streptococcus pneumoniae -induced caspase 6-dependent apoptosis in lung epithelium. Infect Immun 2006;72:4940 – 4947. Chen LW, Chang WJ, Wang JS, Hsu CM. Interleukin-1 [32] mediates thermal injury-induced lung damage through C-Jun NH2-terminal kinase signaling. Crit Care Med 2007;35:1113 – 1122. Rahman I, Gilmour PS, Jimenez LA, MacNee W. Oxidative [33] stress and TNF- α induce histone acetylation and NF- κ B/AP-1 activation in alveolar epithelial cells: potential mecha-nism in gene transcription in lung infl ammation. Mol Cell Biochem 2002;234 – 235:239 – 248. Markova TP, Niwa A, Yamanishi H, Aragane Y, Higashino [34] H. Antagonism of IL-4 signaling by a phosphodiesterase-4 inhibitor, rolipram, in human T cells. Int Arch Allergy Immunol 2007;143:216 – 224. Brown DM, Hutchison L, Donaldson K, MacKenzie SJ, [35] Dick CA, Stone V, et al. The effect of oxidative stress on macrophages and lung epithelial cells: the role of phosphodi-esterases 1 and 4. Toxicol Lett 2007;168:1 – 6.

This paper was fi rst published online on Early Online on 14 March 2012.

Free

Rad

ic R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y N

yu M

edic

al C

ente

r on

12/

07/1

4Fo

r pe

rson

al u

se o

nly.