The differential impact of PDE4 subtypes in human lung fibroblasts on cytokine-induced proliferation and myofibroblast conversion

The Differential Impact of

PDE4 Subtypes in Human Lung Fibroblasts on Cytokine-Induced Proliferation and Myofibroblast Conversion

JENS SELlGE, ARMIN HATZELMANN,

^AND

TORSTEN DUNKERN*

Department

of

In-Vitro Biology, Nycomed GmbH, Konstanz, Germany

Lung fibroblast proliferation and differentiation into myofibroblasts are pathological key events during development of lung fibrosis. Cyclic nucleotide signaling is described as a negative modulator of these cellular processes, and cyclic nucleotide degrading type 4

phosphodiesterases (PDE4) are important regulators of these pathways. In this study, we elucidated expression and the role of individual subtypes of PDE4 in primary normal human lung fibroblast (NHLF) in controlling cytokines-induced proliferation and conversion to myofibroblasts by short-interfering RNAs (siRNAs) induced knockdown. We verified the expression of PDE4A, B, and D, while PDE4C was only minor or even not expressed in NHLF. An efficient liposome-mediated transfection method for mRNA silencing and a knockdown of the expressed PDE4 subtypes was achieved in these cells. This knockdown was further validated by PDE4 protein expression analysis and PDE4 activity measurements. Functionally, the knockdown of PDE4A and PDE4B inhibited proliferation induced by the cytokine combination of bFGF and IL-I

13,

whereas knockdown of PDE4D was ineffective. In contrast, TGF-fl induced differentiation into myofibroblasts was affected by knockdown of PDE4B and PDE4D, but not by PDE4A knockdown. In summary, our data allow to assign different PDE4 subtypes to distinct functions of human lung fibroblasts and highlight the predominant role of PDE4B in controlling pathophysiological processes of human lung fibroblasts. This provides a scientific rationale for focused therapeutic targeting of PDE4B to

treat respiratory diseases with fibrotic lesions in the lung.

Fibrotic alterations in the lung are a pathological feature of pulmonary diseases like asthma, idiopathic pulmonary fibrosis (IPF) or chronic obstructive pulmonary disease (COPD).

Although, in contrast to IPF, fibrosis may not be the dominant aspect in asthma and COPD, it is assumed that progression of such structural changes have a strong impact on long-term outcome of these diseases (Racke et aI., 2008). In asthma and COPD the activation of fibroblasts leads to an excessive deposition of extracellular matrix proteins that modulate the development of subepithelial fibrosis that affects the large and small airways of the lung (Du et aI., 1999; Fahy et aI., 2000). In IPF proliferating lung fibroblasts participate in the formation of

"fibroblastic foci." This initial inflammatory driven remodeling process is thought to lead to the characteristic thickening of alveolar septae and the collapse of normal lung architecture (Chambers, 2008).

During initiation and progression of fibrosis, lung fibroblast activities are regulated by cytokines released from for example epthelial and inflammatory cells (Meneghin and Hogaboam, 2007).

Thus, an intial phase of migration activity is followed by intensive proliferation and finally completed by differentiation into myofibroblasts (De Boer et aI., 1998; Bartram and Speer, 2004). Myofibroblasts are characterized by the expression of alpha-smooth muscle actin (a-SMA) and a high level of collagen and cytokine production. Therefore, the TGF-fl-induced myofibroblast conversion is expected to have a key role in perpetuating inflammation, deposition of connective tissue, thus affecting lung functions (Leask and Abraham, 2004).

Four different genes encode the type 4 phosphodiesterases (PDE4), which are the major cAMP degrading enzymes in lung fibroblasts. Multiple transcriptional start sites and alternative mRNA splicing induce different PDE4 isoforms (Lugnier, 2006), which can be grouped in long, short, and the super short forms (Houslay et aI., 2005).

Numerous in vitro and in vivo studies using selective inhibitors for PDE4 demonstrated their therapeutical potential as anti-inflammatory and immunomodulatory drugs

(Bundschuh et aI., 200 I; Hatzelmann and Schudt, 200 I; Cortijo et aI., 2009). However, the use of PDE4 inhibitors in patients is limited because of side effects such as diarrhea, headache, nausea, and emesis (Spina, 2008). The exact mechanisms how PDE4 specific inhibitors cause these side effects are unclear, but it is assumed that they are associated with the non-selective inhibition of all four PDE4 subtypes (PDE4A, B, C, and D). This hypothesis is corroborated by cell-based investigations using small-interfering RNA (siRNA)-mediated knockdown of specific PDE4 subtypes and by knockout of distinct PDE4 subtype genes in mice. Dissection of the therapeutic beneficial functions of PDE4 SUbtypes from side effects might provide the scientific rational for developing next generation PDE4 inhibitors with improved therapeutic window and/or higher efficiency.

Many anti-inflammatory effects are uniquely linked to the function of PDE4B. It was for example shown in murine macrophages, that PDE4B but not PDE4A or PDE4D regulates LPS-induced TNF-a release Uin et aI., 2005). Furthermore,

*Correspondence to: Dr.Torsten

Byk-Gulden Str. 2, 78467 Konstanz. Germany.

E-mail: torsten.dunkern@nycomed.com

1970

DOI: 10.1002/jcp.22529

Konstanzer Online-Publikations-System (KOPS) URN: http://nbn-resolving.de/urn:nbn:de:bsz:352-opus-135908

sUbtype PDE4B was found to be the predominant expressed subtype in human neutrophils and in unstimulated or LPS-stimulated monocytes (Wang et aI., 1999). In human T-Iymphocytes it was shown that subtype PDE4D controls . T-cell functions such as IL-2, IL-5, and IFN-'Y secretion or also

T-cell proliferation (Peter et aI., 2007). Studies with PDE4D- knockout mice demonstrated the essential role of PDE4D in cAMP homeostasis and cholinergic stimulation of the airways resulting in the development of bronchial hyperreactivity.

Furthermore, the phenotype of those knockout mice indicated the central role of PDE4D in balancing relaxant and contracting cues in airway smooth muscles (Hansen et aI., 2000).

PDE4D seems also to be responsible for the known central nervous system-related and cardiac-related side effects of PDE4 selective inhibitors observed in mice (Robichaud et aI., 2002; Lehnart et aI., 2005).

Although there is strong evidence that elevated cAMP levels in human lung fibroblasts inhibit proliferation or differentiation into myofibroblasts, the exact role of the PDE4 subtypes remains unclear. Therefore, we established a PDE4-subtype specific siRNA-mediated knockdown method and investigated the effects of the individual knockdowns on the

pathophysiologically relevant functions such as proliferation known to be induced by cytokines as bFGF and IL-I ~ (Selige et aI., 20 10) and the differentiation into myofibroblasts induced by TGF-~ (Dunkern et aI., 2007).

Overall our data provide evidence that SUbtypes of PDE4 are functionally non-redundant and control distinct functions of lung fibroblasts. We suggest that PDE4A and PDE4B are involved in proliferation of NHLF, while the participation of PDE4D in that case can be excluded. In contrast, the SUbtypes PDE4B and PDE4D are significantly involved in differentiation into myofibroblasts, whereas PDE4A seems to be irrelevant.

Materials and Methods Materials

The selective PDE4 inhibitor piclamilast RP7340 I (benzamide, 3-cyclopentyloxy-4-methoxy-N-(3,S-dichloro-4-pyridyl»

(Raeburn et aI., 1994) and roflumilast N-oxide (Hatzelmann and Schudt, 200 I) were synthesized at the chemical facilities of Nycomed GmbH. Dilutions were prepared from a 10 fLM solutions in 100% DMSO. A final DMSO concentration ofO. I % was used in all incubations and did not interfere with the investigated fibroblast functions. Neither PDE4 inhibitors nor the solvent affected viability of the lung fibroblasts. Human recombinant TGF-~ I and human recombinant Il-I ~ were purchased from R&D Systems GmbH, Wiesbaden, Germany. Human recombinant basic FGF (FGF2) was obtained from Sigma-Aldrich (St. Louis, MO). All other materials are mentioned in the subsequent passages.

Cell culture

Normal human lung fibroblasts (NHlF, Cambrex Bio Science, Wakersville, MD) were cultured in Fibroblast Basal Medium plus FGM SingleQuots^jtj (both Lonza, Wakersville, MD) in tissue culture flasks at 37°C in an atmosphere containing 5% CO2 up to the eight passage. Passaging and medium replacement was performed every 4 days. Determination of cell count and mortality of NHLF was achieved by either trypan blue (0.4%, Sigma-Aldrich) staining and using a hemacytometer, or by usage of the Vi_CeIl™

(Beckman Counter, Krefeld, Germany).

Treatment of human lung fibroblasts with TFG-JlI

NHLF cells (5 x 10⁴) were plated into six-well cell culture plates and incubated for 24 h at

3 r c.

Thereafter, cells were washed with PBS, original medium was replaced with medium containing no fetal calf serum and cells were cultured for further 24 h until treatment with TGF-j3 I.

The used PDE4 inhibitor was applied 30 min before TGF-j3 I treatment. Forty-eight hours after TGF-j3 I treatment cells were lysed (View cell lysis) and stored at -80°C.

RNA interference of PDE4 subtypes with siRNA

Synthetic siRNA oligonucleotides specific for regions of the PDE4 SUbtype A, B, and D mRNAs were designed and synthesized by Invitrogen (Paisley, UK) Stealth RNAi and Dharmacon (Lafayette, CO) On-Target plus Duplex siRNAs. The silencing effects of several siRNA oligonucleotides were screened and tested initially for their ability to silence the expression of different PDE4 subtypes. The most active of these oligonucleotides for PDE4A (On-Target plus DuplexJ-007647 _I 4: target sequence: 5'-3': CAG GAG UCG UUG GAA GUU A), subtype PDE4B (Stealth RNAi:

HSSI07718: 5'-3': UUA GAA GCC AUC UCA CUG ACA GAC C) and SUbtype PDE4D (On-Target plus DuplexJ-004757_17:

target sequence: 5'-3': GAG UUC UUC UUC UUG AUA A) blocked the expression in a range of 70-90% (determined by RTq-PCR, related to the corresponding control). NHLF were transfected with siRNA oligonucleotides by either

Electroporation/Nucleofection (NF) (Amaxa Nucleofector TM,

Koeln, Germany) or Lipofection (Lipofectamine ™ RNAiMAX, Invitrogen).

NF of NHlF was carried out using the nucleofector kit for primary mammalian fibroblasts (Amaxa), according to the instructions of the manufacturer. Different siRNA concentrations were tested (0.03-2 fLM).

To test transfection efficiency, different concentrations of FITC-Iabeled non-targeting siRNA (siGlO Green, Dharmacon) and different NF-programs were tested and the cells were analyzed by fluorescence microscopy. The NF-program U-23 created the best results. Immediately after electroporation, the samples were removed from the cuvette with I ml prewarmed medium and transferred to six-well plates (Corning Costar, VWR International GmbH, Darmstadt, Germany) containing I ml of prewarmed medium. For electroporation controls, only I x siRNA buffer, but no siRNA was added to the cell suspension prior transfection. For untreated controls, cells were resuspended in nucleofection solution, but no electroporation was carried out.

Lipofection of NHlF was carried out using Lipofectamine ™ RNAiMAX according to the instructions of the manufacturer for reverse transfection. Different siRNA concentrations were tested (0.015-150 nM). For lipofection controls, cells were treated with RNAiMAX without siRNA. As siRNA negative control Silencer')}' Select Negative Control #1 siRNA (Ambion/Applied Biosystems, Darmstadt, Germany) was used in a concentration of 5 nM. In order to confirm the achieved mRNA knockdown and siRNA specificity, we performed gene expression assays (RT q-PCR), as described in the parts "RNA samples," "cDNA-synthesis" and

"Real-time-quantitative PCR." Alternative, we used the TaqMan,f/:

Gene Expression Cells-to-CT ™ Kit (Ambion, Austin, TX) to validate the mRNA knockdown in parallel of the functional investigations.

RNA samples

RNA samples from primary human lung fibroblasts were prepared according to the instructions of the manufacturer using the High Pure RNA isolation kit (Roche Applied Science, Mannheim, Germany). RNA levels were quantified with the Nanodrop ND- I 000 spectrophotometer (PEQLAB Biotechnologie GmbH, Erlangen, Germany) and then stored at -80°C.

cDNA synthesis

One microgram RNA was reverse-transcribed using random hexanucleotide primers (Roche Applied Science), dNTPs (PCR 3 Mix, larova, Teltow, Germany) and avian myeloblastosis virus (AMV) reverse transcriptase (Roche Diagnostics GmbH, Penzberg, Germany) at 4r C for I h. cDNAs were diluted in I mM Tris

0.1 mM EDT A (pH S.O) to a final concentration of 2 ng/ml and stored at -20°C until further use.

Real-time quantitative PCR (TaqMan)

All primers and 6-carboxyfluorescein (FAM)- or VIC-labeled probes were obtained from Applied Biosystems. For normalization we used primers and VIC-labeled probes for detection of ISS ribosomal RNA. The following endogenous ISS rRNA control primers and probe were used: sense 5' -CGG CT A CCA CAT CCA AGG AA-3', antisense 5'-GCT GGA ATT ACC GCG GCT-3', probe 5'-VIC-TGC TGG CAC CAG ACTTGC CCT C-TAMRA- 3'.

For analyzing the gene expression of PDE4 we used designed primer and probe sets as follows: PDE4A (NM_006202) sense, 5'- GTG GCT CCG GAT GAG TTC TC-3', antisense, 5'-GGG CTG CTG TGG CTT ACA G-3', probe 5'-FAM-CCG GGA GGA A TT CGT GGT-minor groove binder-3' (MGB); PDE4B (NM_002600) sense, 5'-AGCAGCACAAAGACGCTTTGT -3', antisense, 5'-TCA GTC TCT CCC AGG GAA TCT C-3', probe 5'-FAM- TGA TTG ATC CAG AAA AC-MGB-3'; PDE4D (NM_006203) sense, 5'-GGC AGe; GTC AAA CTG AGA AAT T-3', antisense 5'-TGA CTG CCA CTG TCC TTT TCC-3', probe 5'-FAM-TAG AGG AAG ATG GTG AGT CAG-MGB-3'. The PDE4 SUbtype primer/probe sets have been designed to bind to the 3' end of the open reading frame of each subtype, at positions where individual splice variants of one subtype are identical.

TaqMan PCRs were run on ABI 7900 HT Sequence Detection Systems (Applied Biosystems). Each PCR reaction was performed in a total volume of 25 JLI in 96-well plates, containing 2.5 JLI cDNA (5 ng), 12.5 JLI qPCR Mastermix Plus (Eurogentec, Seraing, Belgium), 1.25 JLI of the primer/probe set, and nuclease-free water.

ISS rRNA primers and probe were used at a concentration of 50 nM each. All PCRs were performed in triplicate for each sample.

Relative expression units (REUs) were calculated from LlCt = CtFAM - CtVIC according to the equation

REU = 2-ACt x 107. The REU unit describes the expression of a target RNA relative to the ISS rRNA internal standard of the respective sample (Herrmann et aI., 2007).

By using plasmid DNA with PDE4 subtype specific sequence inserts (PDE4 subtypes A4, A I 0, B I, B2, C I, and D4 and D5) the specificity of the used PDE4 SUbtype specific primer/probe sets was verified and identical amplification efficacy was shown.

Measurements of phosphodiesterase isoenzyme activities and preparation of cellular extracts

Cells (1-3 x 10⁶) were washed twice in phosphate buffered saline (PBS) and resuspended in I ml homogenization buffer (137 mM NaCl, 2.7 mM KCI, S.I mM Na2HP04, 1.5 mM KH 2P04, 10 mM HEPES, I mM EGTA, I mM MgCI2, I mM 13-mercaptoethanol, 5 mM pepstatin A, 10 mM leupeptin, 50 mM phenyl methyl-sulfonyl fluoride, 10 mM soybean trypsin inhibitor, 2 mM benzamidine, pH S.2). Thereafter, cells were disrupted by sonification and PDE activity was assessed as described by Thompson et al. (1979) with some modifications (Bauer and Schwabe, 19S0). The assay mixture (final volume 200 JLI) contained: Tris-HCI 30 mM; pH 7.4, MgC!2 5 mM, 0.5 JLM of cyclic AMP as substrate including eH]cAMP and 0.1 mM EGT A. Reactions were performed for 30 min at 37°C in 96-well plates and terminated by adding 50 JLI 0.2 M HCI per well.

Assays were left on ice for 10 min and then 25 JLg 5' -nucleotidase (Crotalus atrox) was added. Following incubation for 10 min at 37°C assay mixtures were loaded onto QAE-Sephadex A25 columns and subsequently eluted with 30 mM ammonium formiate (pH 6.0). Thereafter, radioactivity in the eluate was measured.

Results were corrected for blank values (measured in the presence of denatured protein) that were below 2% of total radioactivity.

Cyclic AMP degradation did not exceed 25% of the amount of substrate added. The final DMSO concentration was 0.3% (v/v) in all assays. Piclamilast (RP73040 I) was used to determine the total

activity of PDE4 in NHLF after knockdown of PDE4 subtypes A, B, and D. The difference of PDE4 activities was calculated in the presence and absence of I JLM piclamilast. At this concentration PDE4 is completely blocked without interfering with activities from other PDE families.

Cell lysis

For protein expression analysis, cells were lysed using the following buffer: 50 mM Hepes, pH 7.5, 150 mM NaC!, 1.5 mM MgCI20 10%

glycine, 1% Triton X-IOO, 0.5% Desoxycholate, 0.2% SDS, I mM ethylene glycol tetraacetic acid (EGTA), 100 mM NaF,

10 mM Na4P207. Just before usage, I mM Na-Vanadat, I mM phenylmethylsulfonylfluorid (PMSF) and 50 U/ml Benzonase (Merck KGaA, Darmstadt, Germany) was added to the buffer.

Phosphatase Inhibitor Cocktail tablet (Complete Mini from Roche Molecular Biochemicals) in 10 ml buffer was added to maximize protein protection.

Lysis procedure. Cells were washed twice in PBS, trypsinated and centrifuged at 500g for 10 min. Afterwards, the cell pellet was resuspended in 150 JLllysis buffer and stored at -SO°C for further use.

Protein concentration was determined by the BCA Protein Assay (Pierce Biotechnology, Rockford, IL). Photometric measurements were performed with the microplate reader Sunrise from Tecan Group Ltd (Mannedorf, SWitzerland).

Western Blot Analysis

Proteins (6 JLg for detection of a-SMA and ca. 60 JLg for PDE4 subtypes) of the samples were separated on a 10% SDS polyacrylamide gel. Thereafter, proteins were blotted onto a nitrocellulose transfer membrane (Protran nitrocellulose transfer membrane; Schleicher & Schuell Bioscience GmbH, Dassel, Germany) for 2-3 h. Membranes were blocked for 2 h in 5% (w/v) milk powder in PBS containing 0.1 % Tween-20 (PBT), and then incubated for 4 h with the primary antibody in a dilution of I: I ,000.

The rabbit monoclonal antibodies against PDE4A and B were generously provided by Prof. Marco Conti (Stanford University School of Medicine, Stanford, USA).

For detection of a-smooth muscle actin, we used anti-o.SMA from Sigma-Aldrich in a dilution of I :2,000. After three continuing washing steps in PBT, samples were incubated for 2 h with a horseradish peroxidase-coupled secondary antibody (I: I 0,000 dilution, Jackson ImmunoResearch, Inc., West Grove, PAl. After final washing with PBT (three times for 10 min each) blots were developed by using a chemiluminescence detection system (Lumi Light Plus, Roche Applied Science) and a luminescent image analyzer. To control homogenous protein loading, blots were additionally hybridized with a primary anti-GAPDH antibody (I :2,000 dilution, Calbiochem'J.!', Darmstadt, Germany).

By using the AIDA image analyzer software the densitometric analyses of the blots were performed. Each signal was normalized to the overall signal of the corresponding blot. Results are expressed as arbitrary units related to the GAPDH signal (%).

eH)Thymidine incorporation

NHLF cells (3 x 104) were plated into 24-wellcell culture plates and incubated for 24 h at 37°C. Thereafter, cells were starved for 24 h in serum-free Fibroblast Basal Medium. Following

preincubation with piclamilast (I JLM), and IL-I 13 (10-100 pg/ml) for 30 min, fibroblasts were stimulated with 10 ng/ml bFGF over 24h.

For DNA-syntheSiS measurement (methyl-3H)thymidine (I JLCi/well) was added to the cells for the last 7 h of the incubation period. Culture supernatants were removed, adherent cells were washed twice in PBS and exposed to ice-cold 10% (w/v) trichloroacetic acid (TCA) for 30 min. Thereafter TCA was removed and the fixed cells dissolved in 0.2 N NaOH. DNA- incorporated radioactivity was counted and expressed as cpm per well (LS6500, Beckman Counter). Data are presented as fold of non-stimulated cells (index of bFGF or bFGF and IL-I 13 stimulated

cells to starved cells). Control cells (not transfected cells) represent I 00%.

Data analysis and software

Data are presented as means ± SD. For statistical analysis, one way ANOVA (GraphPad Prism 5.0; GraphPad Prism Software,lnc., San Diego, CAl was used. Outliers were identified by Grubb's test.

Densitometric analysis of Western blots were performed with the Advanced Image Data Analyzer (AIDA 3.22; Raytest). The primer and probe sets for quantitative PCR analysis were designed by using Primer Express 2.0 (Applied Biosystems).

Results

The aim of the present study was to investigate the role of individual PDE4 SUbtypes in distinct cellular functions in primary human lung fibroblasts. For this purpose, first the gene expression profile of PDE4 subtypes was elucidated by RTq-PCR.

All primer/probe sets for each of the four PDE4 subtypes that we used have been designed by our group (Peter et aI., 2007) and confirmed to result in similar amplifications rates in standardized RT q-PCR using plasmid DNAs with PDE4 SUbtype specific sequence inserts as templates. Thereby we ensured that using these primers will allow a quantitative comparison of the mRNA expression levels of different PDE4 subtypes. Under unstimulated conditions PDE4A was revealed as the major expressed SUbtype followed by PDE4B in NHLF. Whereas mRNA transcripts of the subtype PDE4D were still in a detectable range, the expression of PDE4 subtype C was very low and close to the limit of detection (Fig. I). Therefore, PDE4C was not further considered in the subsequent siRNA experiments.

Treatment with bFGF (10 ng/ml) and/or IL-I ~ (10 pg/ml), which has been shown to induce proliferation in NHLF (Selige et aI., 20 I 0), did not alter the PDE4 subtype mRNA expression pattern. In contrast, treatment with IL-I ~ at concentrations 250 pg/ml (±bFGF) increased the gene expression of PDE4B

350

,...,

::> 300

w ~ _c 250

.9

(J) (J)

200

Q) '-

150

a.

X Q)

<.(

100

z 0:: _E 50

0

~~ ~~ ^~v ~

fV«) fV«}

q,Q q,Q q, q,

Fig. I. PDE4 subtype gene expression of unstimulated NHLF. Data are expressed as relative mRNA expression units (REU, mean ± SD;

n=9)

and D but did not affect PDE4A and C mRNA (Fig. 2A-D). Since PDE4B and PDE4D genes contain a cAMP-responsive element in their promotor region (Vicini and Conti, 1997; D'Sa et aI., 2002), we assume that their induction is due to the described ability of IL-I ~ to induce cAMP synthesis via the COX-2/PGE₂ pathway. We proved this hypothesis by analyzing the additional effect of a PDE4 inhibitor treatment under these experimental conditions. As expected the PDE4 inhibitor piclamilast seemed to potentiate the IL-I ~ induced increase in PDE4B and D mRNA, which was statistical significant (P

<

0.00 I) and more pronounced for PDE4D. Treatment with TGF-<x (0.5 ng/ml), which has been shown to induce differentiation into

myofibroblasts (Dunkern et aI., 2007), had no influence on the PDE4 subtype mRNA expression pattern (data not shown).

To elucidate the question whether defined SUbtypes of PDE4 correlate with determined functions in lung fibroblast, we tested different siRNA approaches to selectively knock down the different PDE4 SUbtypes. For transfection, we compared the Amaxa nucleofection (NF, electroporation) technique (Gresch et aI., 2004) with a liposome-mediated transfection (Feigner et aI., 1987). The liposome-mediated transfection (LF) turned out to be superior versus the nucleofection technique.

Thus, NF substantially induced mortality in NHFL up to approximately 30% and also induced the expression of our target genes PDE4A (P ~ 0.0 I) and PDE4B (P ~ 0.00 I) (Fig. 3A,B). This was not observed by LF. In addition, initial siRNA transfection experiments revealed that for obtaining maximum knockdown micromolar siRNA concentrations had to be applied by using NF, whereas by using LF already nanomolar concentrations were sufficient (data not shown).

Out of a panel of about 25 commercial available siRNA's, we selected those with best efficacy and specificity. By that we identified a set of PDE4 subtype specific siRNAs, which by using LF induced an efficient knockdown of 60-90% in NHLF (Fig. 4).

We tested different siRNA concentrations and showed that a significant mRNA knockdown was already achieved at a siRNA concentration of 0.15 nM. Maximal knockdown was obtained for all PDE4 subtypes at a minimized siRNA concentration of

1.5 nM (Fig. 4, left side). The chosen siRNAs showed high SUbtype specificity up to a concentration of 150 nM, as each one down-regulated only the corresponding target and not the mRNA expression of the other two PDE4 SUbtypes. Further, we tested whether the knockdowns were stable over a period of96 h (Fig. 4, right side). Forthese experiments we chose for all knockdowns of the subtypes of PDE4 the same siRNA concentration of 5 nM.

mRNA knockdowns became significant 24 h after LF and were stable over time for all PDE4 SUbtypes. All tested siRNAs were highly specific. However, we observed that the down regulation of subtype PDE4B induced PDE4D gene expression, which became significant after 72 and 96 h (P ~ 0.0 I and P ~ 0.05).

Validation of siRNA-induced PDE4 subtype knockdown in NHLF by protein expression and PDE4 activity analysis

In order to further validate the siRNA induced knockdown besides mRNA expression analysis we investigated the protein expression of the PDE4 subtypes by Western blot analysis (Fig. 5).

We used a pan-PDE4A specific antibody, which detected a long form of PDE4A (106-109 kDa). The signal intensity of the band was already reduced at 24 h after siRNA transfection already (Fig. SA).

The pan-PDE4B specific antibody detected two different isoforms of PDE4B. The strongest signal related to a long form of PDE4B (~I 07 kDa) whereas we detected a short form of PDE4B (around 65-67 kDa). Densitometric quantification

A

bFGF PDE4i

15

PDE4A

no IL-1P +

IL-ll3 [10pg/ml]

+ +

+

C

₁

PDE4C

- C .Q

g ~ 10 c: I!!

o a.

o x

0"

-0< ₅ -z .£la:

S

IL-1P [50pg/ml]

+ +

+

o n ,-., _{no IL-1P} ,..-, IL-1P

--

IL-1P

..

[10 pg/ml) [50 pg/ml]

bFGF + + + - ⁺ +

PDE4i - ⁺ - ⁺

B ^PDE4B

15 _.2 c ,g ~ 10

§ _{o x}

l!

'0"

32<

.2~ S

no IL-1P

bFGF + + + + +

PDE4i + +

0 ^PDE4D

15 _ 0 c

e·~ 10

8

_a

.,,< ~!1 -z 5 ,2",

§.

IL-1Jl [50 pg/mO

bFGF + + + + +

PDE4i + +

Fig. 2. Gene expression ofPDE4A(A), PDE4B (B), PDE4C (C), and PDE4D (D) after stimulation with bFGF (I Ong/ml) ± IL-I J3 in NHLF. Data are expressed as fold of control from relative mRNA expression units (mean ± SD; n = 3), compared to untreated control: *p < 0.05, **p< 0.0 I,

***P < 0.00 I.

demonstrated a time-dependent down regulation of PDE4B after siRNA transfection (Fig. 5B).

Although we tested different antibodies for subtype PDE4D and optimized the Western blot procotols, we could not detect PDE4D protein expression in NHLF. Since also mRNA expression of PDE4D was much lower than for PDE4A and B, we assume very low protein expression of 4D isoforms, which is below the limit of detection by Western blot.

Furthermore, we investigated whether the siRNA transfection also reduced PDE4 activity besides mRNA and protein expression (Fig. 6A). A time-dependent decrease of

A

CJcontrol

EE]LF

~ ^_NF

~ 300 c - .Q

f---ti--1 o VI

... VI

g

~ ²⁰⁰

u ~

0« ~ ~ 100

S

0

~'I'

~Q ~<;,

~Q ~Q

~Q

PDE4 activity after PDE4A knockdown was observed, which became significant compared to vehicle control after 96 h (reduction by 24.6 ± 3.0%, P":; 0.05). Also after knockdown of PDE4B, PDE4 activity decreased in a time-dependent manner.

This decrease became significant at 72 h (reduction by 51.3 ± 9.1 %, P":; 0.0 I) and 96 h (reduction by 53.5 ± 3.8%,

P":;O.OOI).

We observed a tendency of reduction in total PDE4 activity after the knockdown of the mRNA of SUbtype PDE4D (reduction by 11.6 ± 2.7% at 24 hand 21.8 ± 4.8% at 96 h).

However, this reduction did not become significant, which is in

B

:,!!

0

.:

~ ~

"- 0

E a;

(J

e~ ~ ^",e

.s

^~'I'

w

^:i!:-v

..

^~

~e ... e

~ ~

«x

~o v

Fig. 3. Comparison ofNucleofection (NF) and Lipofection (LF). A: Gene expression ofPDE4A, B, and D is increased 24 h after NF but not after LF in non-transfected NHLF. Dataareexpressed aspercentofnon-treated cells (mean ± SD;n = 6), **p < 0.0 I, ***p< 0.00 I. B: Cell mortalityofNHLF was enumerated directly after NF or in case of LF, 24 h after treatment. Data are expressed as percent of total enumerated cells. "Non-treated"- cells and "LF" + siRNA": n = 8 (mean ± SD), "NF", NF vehicle", and "LF": n = 2.

siRNA concentrations

A

(silencing PDE4A)

300

B

(silencing PDE4B)

c o _ 300

m

C5 200

~E !j 8

<0 Z # 100 Cl:_

E

C

~ ···!···~·· ···~

.... ·· ... ···1

(silencing PDE4D)

300

[nM)

[nM)

c 0 _

300

300

.~ '0200

"'"

~ c

§i-8

:0

_{Z # 100}

Cl:_

E

300

Time response

T ^-+-PDE4A

··.··PDE4B ____ PDE4D

.... [h)

~ T '~

~

-+-PDE4A ... ··PDE48 __PDE4D

[h)

-+-PDE4A

··.··PDE48 -.-PDE4D

[h)

Fig. 4. Gene expression of the different POE4 subtypes A, Band 0 in NHLF 24 h after concentration-dependent (left) and time-dependent (right) siRNA-mediated knockdown of either POE4A (A) POE4B (B) or POE40 (C) by lipofection (RNAiMAX). Data (mean ± SO; n = 4-5) are shown as percent of control.

accordance with the low expression detected by RT q-PCR and Western blot methods.

Figure 6B visualizes the sum of different proportions of PDE4 sUbtype activities 96 h after the siRNA treatments. The sum of all proportions reached 99.9% oftotal PDE4 activity. Although this design approach does not consider the standard deviations within the experiments, the fact that there is no residual PDE4 activity left provides further evidence (besides the RTq-PCR results) that under the investigated conditions the sUbtype PDE4C is not expressed in NHLF to a substantial extent.

The effect of PDE4 subtype knockdowns on cytokine induced proliferation in NHLF

Recently we have characterized the effect of bFG F and I L-I ~ on proliferation of NHLF (Selige et aI., 20 I 0). We demonstrated,

that the bFGF-induced proliferation is potentiated by co-treatment with I L-I ~ at concentrations of ::::: 10 pg/ml.

Higher concentrations (~50 pg/ml) of IL-I ~ induced COX-2 expression and subsequent PGE₂ synthesis, which triggered an increase in cAMP. Under this condition of high cAMP, proliferation of NHLF was inhibited and pan-inhibitors of PDE4 potentiated the antiproliferative effect of IL-I ~.

To determine individual PDE4 subtypes, that mediate the antiproliferative effect, we transfected NHLF with sUbtype- specific siRNAs. After LF, we cultured NHLF for 24 h. After additional 24 h of starvation, proliferation was stiml!lated for 24 h by bFGF (10 ng/ml) ± IL-I ~ (10 pg/ml) treatment.

Thereafter, proliferation was measured by means of eH)thymidine incorporation.

The effect of the individual siRNA tranfections on cytokine- induced proliferation was related to non-transfected cells. As

A

₁₀₀

POE4 A long fonn -.!:

E:J PDE4A long form (106-109 kDa) o Q) C 75

~ e·~

---

GAPOH (36 kDa) o 0. VI

u <! ~ 50

-~o.

I 24 48 72 96 PDE4AsiRNA

transfection

II 24 48 72 96 I correspondent

controls

[h]

B

< ~ _ l--. :..-.. _ POE4 B long from

l...J I - : '--' - '...., """"' (-107 kDa) POE4 B short form

_. - -

(65-67 kDa) GAPOH (36 kDa)

I 48 72 96 II 48 72 96 I [h]

PDE4B siRNA correspondent transfection ~ontrols

Ow x

~C<D

o 0. 25

100

control 24h 48h 72h 96h

control

c::J PDE48 long form ImIlI PDE48 short form

Fig. 5. PDE4A (A) and PDE4B (B) protein expression in NHLF and densitometric quantification 24-96 h after siRNA-mediated knockdown (lipofection with RNAiMAX). Quantitative data of one representative blot are shown and expressed as percent of the correspondent control. Two PDE4 long forms of PDE4A and PDE4B and a short form of PDE4B were identified and down regulated after siRNA-transfection.

shown in Figure 7, treatment with the lipofection agent (indicated as sLvehicle) itself marginally inhibited proliferation by maximum 20%.

Tranfection with siRNA targeting PDE4A resulted in a significant reduction of proliferation induced by bFGF or in combination with IL-I Jj (Fig. 7 A.B;

P:::;

0.00 I).

LF with PDE4B targeting siRNA inhibits proliferation induced by bFGF apparently, which was further pronounced and statistically significant by co-treatment with IL-I Jj

(Fig. 7B; P :::; 0.00 I). T ransfections with siRNA targeting PDE4D showed no effects on cytokine stimulated proliferation of NHLF. We assume that the procedure with this siRNA is equivalent to experiments with siRNA negative controls.

However, in addition we tested a scrambeld siRNA, which is described not to target any eukaryotic genes as a negative control for our investigations.

A

150

Influence of PDE4 subtype knockdowns on TGF-Q induced lung fibroblast to myofibroblast conversion Lung fibroblast to myofibroblast differentiation is blocked by treatment with PDE4 specific inhibitors (Dunkern et aI., 2007;

Sabatini et aI., 20 I 0). To ascertain which individual PDE4 subtype is involved in this effect, we transfected NHLF with PDE4 subtype-specific siRNAs, starved them for 24 hand stimulated with TGF-Jj (0.5 ng/ml) for a period of 48 h. Because myofibroblasts have been described to express the marker a- SMA whereas lung fibroblasts do not, we measured the extent of myofibroblast differentiation by Western blot analyses for a- SMA and quantified the blots by densitometry (Fig. 8A.B). To equalize different exposure times of the Western blots, each sample was calculated as the percentage of total signal intensity on the blot. Furthermore, we normalized each a-SMA signal to

B

PDE4 activity _{- -conlrol}

~ E o

~ 100 '0

:c

~

'0

?fl. 50

J

--vehicle

•· ... ··PDE4A -.- PDE4B -+-PDE4D

Fig. 6. Total PDE4 activity is time-dependently down regulated after siRNA-liposome (RNAiMAX)-mediated knockdown in NHLF. A: Under unstimulated normal cell culture conditions, PDE4B represents the most pronounced proportion of PDE4 enzyme activity and reaches a maximum 72 haftersiRNAtransfection. Data are expressed aspercentofvehiclecontrol (mean ± SD; n = 5). B: Pie chartofPDE4 activityafter96 h in NHLF (as seen in A). The sum of PDE4 subtype specific activities is close to 100%.

A

g

^c: ₀

e

_{C 0} ^{13 :i: Q)} _Q)

^< ""

^w ₀ ^{co UJ} 0 ^{<:t 0 UJ}

""

0 ^controls :2

^-=

^{OJ OJ ~:c ~OJ "'0 E Q) <}

""

^{UJ 0 III}

...

^{UJ 0 C}

""

^{UJ 0 < z 0::} 'w

>1 ^~

> ,

_~I ^0..

1 ~I

U U >1 0..

1 0..

1

~I

^{'w .2 "u; rn 'w}

_'"

^{til Q)}

'iii 'iii 'iii ^0.+ c:

bFGF (10 ng/mij

-

^{+ + + + + + + +}

bFGF [10 ng/ni] + + + + + IL-1 J3 (10 pg/ml(

-

+ + + + + + + I

Fig. 7. Knockdown of PDE4 subtypes inhibits cytokine-induced proliferation in NHLF. A: ('H)thymidine incorporation as measurement of proliferation was induced by bFGF ( I 0 ng/ml) and inhibited after siRNA-liposome (RNAiMAX)-mediated knockdown of PDE4A. Proliferation data areexpressedaspercentofcontrol (mean ± SD;n

=

5-8). B: ('H)thymidine incorporation was induced in NHLF(n

=

12)bybFGF(1 0 nglml)plusIL- 113 (10 pg/ml) and inhibited after siRNA-liposome (RNAiMAX)-mediated knockdown of PDE4A (n = I I) and PDE4B (n = 8). The liposome (RNAiMAX) itselfhasasignificantinhibitoryeffecton proliferation (n = 9). The knockdown ofPDE4Areached the maximal antiproliferative effect ofthe panPDE4 inhibitor (n

=

3) piclamilast (111M) in this setup. The knockdown of PDE4D (n

=

12), as well as the transfection with a scrambled siRNA as neg. control (n = 4) have no effect on proliferation compared to the vehicle control. Data are expressed as percent offold of control (mean ± SD), *p < 0.05, ***p < 0.00 I.

A B

.. __ . __ '. ____ R_. '_ .. ___ ... __ .. ___ ... " .. ___ . ___ ^~;~_~_ .. _!

i

^{81011 I}

i

^GAPDH

I

i

---.---j

, aSMA i

i

^{81012 i}

!

^GAPDH

I

- - - - --_..:..:...:...-'-'''---j

aSMA

i

I - - - -- - _ GAPDHi

i ·-·--·--·---·-··-·-·---·---·-·-·~·----·-·--i

i ... . .

aSMA

I

! 81014 . :

i _ ... -.._ . . . . GAPDHI

i !

! ---·---··-.. ·--·---·-·-···--·--·---····--1

i

!

! . . . - - --,... aSMA

I

i

81015 ;

! - - - -

^GAPDHi

I control <m~ UI I

~ ~ ~ .~

I

%i %i ~

i

+ + + + +

i

⁸¹⁰¹³

I

_,

TGF

,

-~1

control

<

^{co 0 (ii}

"" "" ...

^~

UJ W w

'E

0 0 0 til

0.. 0.. 0.. 13

'iii 'iii 'iii '0.

+ + + +

TGF-131 +

Fig. 8. Knockdown of PDE4 subtypes inhibits TGF-I3-induced myofibroblast differentiation of NHLF. A: .. SMA protein expression after stimulation with TGF-131 (0.5 ng/ml) in normal (control) and RNAiMAX-transfected NHLF to knockdown PDE4A, B, and D was detected.

Additionally, GAPDH protein expression as endogenous control was measured. B: Densitometric quantification of the Western blots shows that the knockdown of PDE4B and D as well as the PDE inhibitor piclamilast (111M) inhibit .. SMA expression as measurement of myofibroblast differentiation. The knockdown ofPDE4Adoes not reduce .. SMA protein expression. Data are expressed as percent ofTGF-I3-control (mean ± SD, n = 5), **p < 0.0 I, ***p < 0.00 I.

the corresponding endogenous protein loading control GAPDH. T ransfection reagent and scrambled siRNA controls have been shown before to be without effect on at-SMA protein expression (data not shown). Forty eight hours after stimulating NHLF with TGF-[3 the silencing of sUbtype PDE4A, LF caused no effect on TGF-[3 induced at-SMA expression, which is in contrast to the efficacy seen with this siRNA in the proliferation experiments. However, silencing of sUbtype PDE4B and PDE4D provoked a significant decrease in at-SMA expression (Fig. 8B; sLPDE4B: -64% P:s 0.0 I; si_PDE4D: -40%, P:s 0.05).

As expected, treatment with the panPDE4 specific inhibitor piclamilast reduced the TGF-[3 induced at-SMA expression to a similar extent (-47%, P:s 0.05).

Discussion

The present study identifies the role of different PDE4 sUbtypes in human lung fibroblasts in specific functions as proliferation and myofibroblast conversion. These functions are of pathophysiological relevance in progression of fibrosis in lung diseases such as COPD, IPF, or asthma. By increasing cyclic AMP, PDE4 inhibitors act anti-inflammatory but are also capable of preventing and reducing fibrotic and vascular remodeling in the lung, as clinical trials or investigations in vitro and in animals attested (Rabe et aI., 2005; Shepherd, 2006;

Cortijo et aI., 2009). In order to mimic the pathophysiological conditions relevant to humans, we examined primary human lung fibroblasts and stimulated them with cytokines, which have been described to be increased in patients with asthma, COPD, or IPF.

Several investigations by our group already proved the presence of PDE4 in lung fibroblasts of various species by gene and protein expression analysis as well as isoenzyme activity determination (Selige et aI., 20 I 0). However, the PDE4 sUbtype profile was not investigated quantitatively before. Subtype PDE4A is most dominant in NHLF, followed by PDE4B and PDE4D. In contrast the RTq-PCR signal for PDE4C was at the limit of detection and together with the PDE activity measurements, PDE4C was considered to be absent in these cells under the given experimental conditions, as also described by Spina (2008).

In a previous publication, our group examined the complex interplay of the cytokines IL-I[3 and bFGF on proliferation of human lung fibroblasts. IL-I[3 plays a central role due to its concentration-dependent ability to either potentiate bFGF-induced proliferation or to induce cAMP via the COX-2/

PGE2 pathway, which than exerts an inhibitory effect on proliferation (Selige et aI., 20 I 0).

Since one of the objectives of this study was to examine the effects of PDE4 subtype-specific knockdowns on the IL-I[3/

bFGF-induced proliferation, we first examined whether treatment with these cytokines changes the PDE4 subtype expression pattern in comparison to the already described pattern observed under non-stimulated conditions. Under conditions where NHLF start to proliferate (bFGF and co-treatment with IL-I[3 at 10 pg/ml) the PDE4 subtype expression did not change. However, co-treatment using IL-I[3 at a concentration of SO pg/ml, at which cAMP also will increases (Selige et aI., 20 I 0), resulted in increased PDE4B and PDE4D mRNA expression. This can be explained by the fact that both genes (PDE4B and PDE4D) have a cAMP-regulated intronic promotor (Vicini and Conti, 1997; D'Sa et aI., 2002). Under the same conditions no changes in PDE4A and PDE4C gene expression were observed. TGF-[3 at a concentration of 0.5 ng/ml, 48 h after treatment did not influence the mRNA expression in NHLF (data not shown). This result stands in line with a previous work of our group demonstrating that treatments with TGF-[3 does not alter the PDE4 activity of primary human lung fibroblasts (Dunkern et aI., 2007).

Since no PDE4 subtype-specific inhibitors exist, the mRNA- knockdown approach represents an appropriate method to investigate PDE4 subtype specific functions in vitro. An electroporation method turned out to be inferior versus a liposome-mediated tranfection since the former one induced mortality to a,high degree and also induced the expression of all PDE4 genes.

The lipofection of NHLF with individual siRNAs versus PDE4A, PDE4B, and PDE4D showed a very stable knockdown of mRNA over time, tested in a range from 24 to 96 h. A decrease in PDE4A and B protein was already observed 24 h after transfection and was maximum between 72 and 96 h, indicating that the total turnover of these proteins in NHFL is completed within this time period. This has not been investigated before. However, PDE4D protein was not detectable by Western blot analysis even using different antibodies and protocols or targeting different epitopes of PDE4D. This might be explained by a low expression level of PDE4D in these cells as also suggested by the results of the gene expression analysis. Other groups demonstrated the expression of PDE4D on protein level in lung fibroblast cell lines, but not in primary cells (Martin-Chouly et aI., 2004).

It should be mentioned that several commercially available siRNA's, which were proposed to be subtype-specific, have been tested in these investigations and turned out to be either inefficient in respect to knockdown or unspecific (data not shown). Since the identified siRNAs were shown to be sUbtype- specific, it can be assumed that each siRNA addressing a distinct subtype of PDE4 can also be used as high homology negative control for the other subtypes, as also described by Peter et al.

(2007).

Noteworthy, the siRNA-induced knockdown of PDE4B conversely and significantly increased the mRNA expression of PDE4D. There was also a tendency of increasing mRNA levels of PDE4B and D by knockdown of PDE4A. This might be considered as a counter-regulation of NHLF as a consequence of increased cAMP levels induced by the knockdown (due to reduced cAMP hydrolytic activity). In the present study cAMP levels after siRNA transfection were not measured but the pattern of PDE4 SUbtype upregulation observed by specific PDE4 SUbtype knockdowns is similar to those after the treatment of NHLF with IL-I[3 at different concentrations, which also demonstrated to increase cAMP (Selige et aI., 20 I 0).

The fact that silencing PDE4D did not increase PDE4B (and also not PDE4A) mRNA expression can be explained by the overall low expression of PDE4D. Thereby its knockdown might increase cAMP only marginally and thus does not reach the cAMP concentration threshold required to induce PDE4B expression.

We confirmed the siRNA-mediated knockdown by overall PDE4 activity measurements. As expected, subtype specific siRNA transfection reduced activity in parallel to the reduction of PDE4 protein. Knockdown of PDE4B caused the most prominent reduction in overall PDE4 enzyme activity under the investigated conditions, followed by PDE4A and PDE4D. This finding does not directly reflect the mRNA expression level of the SUbtypes under untreated conditions. The apparent discrepancy might be explained by a compensatory induction of other PDE4 SUbtypes contributing to overall PDE4 activity and/

or by further regulation of the other PDE4 SUbtypes on the level of translation or by activity-controlling post-translational modifications as, for example phosphorylation (Conti et aI., 2003 or Houslay and Adams, 2003).

The sum of reduction in PDE4 activity induced by the particular knockdown with PDE4A, B, and D specific siRNA's resulted in nearly 100% of total PDE4 activity. Besides the already described gene expression data this supports our supposition that PDE4C is not expressed under our culture conditions in NHLF.

The mRNA knockdown of the sUbtype PDE4A inhibited bFGF ± IL-I ~-induced proliferation of NHLF in a significant manner. Both cytokines have shown to be increased under pathophysiological fibrotic conditions (Zhang et al.. 1993;

Kranenburg et al.. 2005) and might therefore contribute to the excessive proliferation of fibroblasts in these conditions. The anti proliferative effect of PDE4A knockdown reached ~ I 00%

of the effect by the panPDE4 inhibitor piclamilast. administered under the same stimulatory and experimental conditions.

The effect of mRNA silencing of PDE4B on proliferation was dependent on the cytokine treatment used to initiate proliferation. Under condition of bFGF-induced proliferation.

PDE4B knockdown inhibited proliferation slightly without reaching statistical significance. The enhanced efficacy of a PDE4B knockdown on bFGF plus IL-I ~ induced proliferation might be explained by either inhibiting the synergistic proliferation induced by both cytokines and/or by potentiating the antiproliferative action of IL-I ~-induced cAMP release in these cells as demonstrated by pan-PDE4 inhibitors (Sabatini et al.. 20 10). In the latter case we have to assume that the PDE4B knockdown will increase IL-I ~-induced cAMP above a certain threshold level leading to antiproliferative effect. An indication for that increase is our observation that under conditions of PDE4B knockdown the cAMP-regulated PDE4D mRNA expression gets induced. Concerning proliferation. this up-regulation does not seem to compensate functionally the PDE4B depletion. Further. we demonstrate that the knockdown of PDE4D (as well as control transfection with a scrambled siRNA) did not inhibit proliferation and therefore we would exclude a role of this enzyme in controlling proliferation.

This latter finding might be of relevance for the future design of therapeutic PDE4 inhibitors. since inhibition of PDE4D is considered to provoke at least same adverse events in patients (Giembycz. 200 I). Therefore. it is possible to separate a beneficial therapeutic mechanism. namely inhibition of proliferation. from side effects by inhibiting distinct PDE4 sUbtypes.

Besides proliferation. the differentiation into myofibroblast is another important pathophysiological feature in lung fibrosis.

Several studies have demonstrated that PDE4 inhibitors reduce

TGF-~-induced conversion of fibroblasts to myofibroblasts (Dunkern et al.. 2007). Our results now demonstrate that knockdown of the PDE4 subtype Band D. but not PDE4A inhibit TGF-~-induced expression of a-SMA and thus differentiation into myofibroblast. Inhibition by PDE4B knockdown was most effective and comparable to the treatment with the pan-PDE4 inhibitor piclamilast.

In summary. our study assigned PDE4 sUbtypes to pathophysiologically relevant lung fibroblast functions.

Thereby. it supports the concept that by defined pharmacological targeting of PDE4 sUbtypes a therapeutic interference with pathophysiologic functions is achievable. The overall data indicate a pre-dominant role'of PDE4B in controlling both. proliferation and the differentiation into myofibroblast. Although PDE4D has as well an influence on differentiation. it is supposed to be majorly responsible for the narrow therapeutic window of PDE4 inhibitors. Therefore we hypothesize that an ideal therapeutic PDE4 inhibitor designed to target fibrotic conditions in the lung should be selective for PDE4B inhibition.

Acknowledgments

We thank Prof. Dr. Marco Conti for providing the PDE4 antibodies. scientific advices and helpful discussions. We acknowledge the excellent scientific support of Dr. Hermann Tenor and helpful advices of Prof. Dr. Albrecht Wendel. This work was part of the PhD thesis of Dr. Jens Selige and was

supported in parts by the International Research Training Group 1331 (IRTG 1331) funded by the Deutsche Forschungsgemeinschaft (DFG).

Literature Cited

Bartram U, Speer CPo 2004. The role of transforming growth factor beta in lung development and disease. Chest 125:754-765.

Bauer AC, Schwabe U. 1980. An improved assay of cyclic 3' ,5' -nucleotide phosphodiesterases with QAE-Sephadex columns. Naunyn Schmiedebergs Arch PharmacoI311:193-198.

Bundschuh DS, Eltze M, BarsigJ. Wollin L. Hatzelmann A, Beume R. 200 I. In vivo efficacy in airway disease models of roflumilast, a novel orally active PDE4 inhibitor. J Pharmacal Exp Ther 297:280-290.

Chambers RC. 2008. Abnormal wound healing responses in pulmonary fibrosis: focus on coagulation signaling. Eur Respir Rev 17:130-137.

Conti M, Richter W, Mehats C, Livera G, Park JY, Jin C. 2003. Cyclic AMP·specific PDE4 phosphodiesterases as critical components of cyclic AMP signaling. J Bioi Chern 278:5493- 5496.

Cortijo J,lranzo A, Milara X, Mata M, Cerda-Nicolas M, Ruiz-Sauri At Tenor H, Hatzelmann A.

Marcillo EJ. 2009. Roflumilasl. a phosphodiesterase 4 inhibitor, alleviates bleomycin- induced lung injury. Br J Pharmacol 156:534-544.

de Boer WI, van Schadewijk A. Sont JK. Sharma HS, Stolk J. Hiemstra ps. Van Krieken JH.

1998. Transforming growth factor beta I and recruitment of macrophages and mast cells in airways in chronic obstructive pulmonary disease. Am J Respir erit Care Med 158: 195 1-

1957.

D'Sa C, Tolbert LM, Conti M. Duman RS. 2002. Regulation of cAMP.specific phosphodiesterases type 4B and 40 (PDE4) splice variants by cAMP signaling in primary cortical neurons. J Neurochem 81 :745-757.

Du M. Wang C. Cao D. Weng X. Wu X, Liu Y. 1999. A pathological study on pulmonary fibrosis caused by chronic obstructive pulmonary disease. Zhonghua Jie He He Hu Xi Za Zhi 22:30-32.

Dunkern TR. Feurstein D, Rossi GA, Sabatini F, Hatzelmann A. 2007. Inhibition of TGFbeta induced lung fibroblast to myofibroblast conversion by phosphodiesterase inhibiting drugs and activators of soluble guanylyl cyclase. Eur J Pharmacol 572: 12-22.

Fahy jV, Corry DB, Boushey HA. 2000. Airway inflammation and remodeling in asthma. Curr Opin Pulm Med 6: 15-20.

Feigner PL, Gadek TR. Holm M, Roman R, Chan HW. Wenz M, Northrop JP, Ringold GM. Danielsen M. 1987. lipofection: A highly efficient, lipid-mediated DNA-transfection procedure. Proc Natl Acad Sci USA 84:7413-7417.

Giembycz MA. 200 I. Cilomilast: A second generation phosphodiesterase 4 inhibitor for asthma and chronic obstructive pulmonary disease. Expert Opin Investig Drug 10:1361-

1379.

Gresch 0, Engel FB. Nesic D. Tran TT. England HM, Hickman ES. Korner I, Gan L. Chen S, Castro-Obregon S, Hammermann R, Wolf j, Muller-Hartmann H, Nix M, Siebenkotten G, Kraus G, Lun K. 2004. New non-viral method for gene transfer into primary cells. Methods 33:151-163.

Hansen G, Jin S, Umetsu DT, Conti M. 2000. Absence of muscarinic cholinergic airway responses in mice deficient in the cyclic nucleotide phosphodiesterase PDE4D. Proc Natl Acad Sci USA 97:6751-6756.

Hatzelmann A, Schudt C. 200 I. Anti-inflammatory and immunomodulatory potential of the novel PDE4 inhibitor roflumilast in vitro. j Pharmacol Exp Ther 297:267-279.

Herrmann M. Selige j, Raffael S. Sachs G, Brambilla A. Klein T. 2007. Systematic expression profiling of the gastric H+/K+ ATPase in human tissue. Scand J Gastroenterol 42: 1275- 1288.

Houslay MD, Adams DR. 2003. PDE4 cAMP phosphodiesterases: Modular enzymes that orchestrate signalling cross-talk, desensitization and compartmentalization. Biochem J 370: 1-18.

Houslay MD, Schafer P, Zhang KY. 2005. Keynote review; Phosphodiesterase-4 as a therapeutic target. Drug Discov Today 10:1503-1519.

Jin SL. Lan L, Zoudilova M. Conti M. 2005. Specific role of phosphodiesterase 4B in lipopolysaccharide-induced signaling in mouse macrophages. J Immunol 175: 1523-1531. Kranenburg AR, Willems-Widyastuti A, Mooi WJ, Saxena PR, Sterk PJ, de Boer WI. Sharma HS. 2005. Chronic obstructive pulmonary disease is associated with enhanced bronchial expression ofFGF·I, FGF·2, and FGFR·I. J Pathol 206:28-38.

Leask A, Abraham DJ. 2004. TGF·beta signaling and the fibrotic response. FASEB J 18:816- 827.

Lehnart SE, Wehrens XH. Reiken S. Warrier S, Belevych AE. Harvey RD. Richter W. Jin SL, Conti M, Marks AR. 2005. Phosphodiesterase 40 defiCiency in the ryanodine receptor complex promotes heart failure and arrhythmias. Cell 123:25-35.

Lugnier C. 2006. Cyclic nucleotide phosphodiesterase (PDE) superfamily: A new target for the development of specific therapeutic agents. Pharmacal Ther 109:366-398.

Martin.Chouly CA, Astier A,Jacob C. Pruniaux MP. Bertrand C, Lagente V. 2004. Modulation of matrix metalloproteinase production from human lung fibroblasts by type 4 phosphodiesterase inhibitors. life Sci 75:823-840.

Meneghin A. Hogaboam CM. 2007. Infectious disease, the innate immune response, and fibrosis. J Clin Invest I 17:530-538.

Peter D,jin SL. Conti M. Hatzelmann A, Zitt C. 2007. Differential expression and function of phosphodiesterase 4 (PDE4) subtypes in human primary CD4+ T cells: Predominant role of PDE4D. J Immunol 178:4820-4831.

Rabe KF. Bateman ED. O'Donnell D, Witte S, Bredenbroker D. Bethke TD. 2005.

Roflumilast- An oral anti-inflammatory treatment for chronic obstructive pulmonary disease: A randomised controlled trial. Lancet 366:563-571.

Racke K, Haag S, Bahulayan A. Warnken M. 2008. Pulmonary fibroblasts, an emerging target for anti-obstructive drugs. Naunyn Schmiedebergs Arch Pharmacol 378: 193-20 I.

Raeburn D, Underwood SL, Lewis SA, Woodman VR. Battram CH, Tomkinson A, Sharma S, jordan R, Souness jE, Webber SE. 1994. Anti-inflammatory and bronchodilator properties of RP 7340 I, a novel and selective phosphodiesterase type IV inhibitor. Br J Pharmacol 113:1423-1431.

Robichaud A. Stamatiou PB, Jin SL, Lachance N. MacDonald D, Laliberte F. liu S, Huang Z.

Conti M, Chan cc. 2002. Deletion of phosphodiesterase 4D in mice shortens alpha(2)·

adrenoceptor-mediated anesthesia. a behavioral correlate of emesis. J Clin Invest 110:1045-1052.

Sabatini F, Petecchia l, Boero S. Silvestri M, Klar j, Tenor H, Beume R, Hatzelmann A. Rossi G.

20 I O. A phosphodiesterase 4 inhibitor, roflumilast N-oxide, inhibits human lung fibroblast functions in vitro. Pulm Pharmacol Ther 23:283-291.

Selige J, Tenor H, Hatzelmann A, Dunkern T. 20 I O. Cytokine-dependent balance of mitogenic effects in primary human lung fibroblasts related to cyclic AMP signaling and phosphodiesterase 4 inhibition. J Cell Physiol 223:317-326.

Shepherd Me. 2006. Phosphodiesterase 4 regulation of cyclic AMP in pulmonary remodelling:

Potential roles for isoform selective inhibitors. Pulm Pharmacal Ther 19:24-.

Spina D. 2008. PDE4 inhibitors: Current status. Br J Pharmacol 155:308--315.

Thompson WJ, T erasaki Wl, Epstein PM, Strada SJ. 1979. Assay of cyclic nucleotide phosphodiesterase and resolution of multiple molecular forms of the enzyme. Adv Cyclic Nucleotide Res 10:69-92.

Vicini E, Conti M. 1997. Characterization of an intronic promoter of a cyclic adenosine 3',S'-monophosphate (cAMP)-specific phosphodiesterase gene that confers hormone and cAMP inducibility. Mol Endocrinol II :839-850.

Wang p, Wu P, Ohleth KM, Egan RW, Billah MM. 1999. Phosphodiesterase 4B2 is the predominant phosphodiesterase species and undergoes differential regulation of gene expression in human monocytes and neutrophils. Mol Pharmacol 56:170-174.

Zhang y, Lee TC, Guillemin B, Yu MC, Rom WN. 1993. Enhanced IL-I beta and tumor necrosis factor-alpha release and messenger RNA expression in macrophages from idiopathic pulmonary fibrosis or after asbestos exposure. J Immunol 150:4188-4196.