peroxisome proliferator-activated receptor-γ agonists inhibit the replication of respiratory...

6) 335–346www.elsevier.com/locate/yviro

Virology 350 (200

Peroxisome proliferator-activated receptor-γ agonists inhibit the replicationof respiratory syncytial virus (RSV) in human lung epithelial cells

Ralf Arnold ⁎, Wolfgang König

Institute of Medical Microbiology, Otto-von-Guericke-University, Leipzigerstr. 44, 39120 Magdeburg, Germany

Received 14 December 2005; returned to author for revision 13 February 2006; accepted 9 March 2006Available online 17 April 2006

Abstract

We have previously shown that peroxisome proliferator-activated receptor-γ (PPARγ) agonists inhibited the inflammatory response of RSV-infected human lung epithelial cells. In this study, we supply evidence that specific PPARγ agonists (15d-PGJ2, ciglitazone, troglitazone, Fmoc-Leu) efficiently blocked the RSV-induced cytotoxicity and development of syncytia in tissue culture (A549, HEp-2). All PPARγ agonists understudy markedly inhibited the cell surface expression of the viral G and F protein on RSV-infected A549 cells. This was paralleled by a reducedcellular amount of N protein-encoding mRNA determined by real-time RT-PCR. Concomitantly, a reduced release of infectious progeny virus intothe cell supernatants of human lung epithelial cells (A549, normal human bronchial epithelial cells (NHBE)) was observed. Similar results wereobtained regardless whether PPARγ agonists were added prior to RSV infection or thereafter, suggesting that the agonists inhibited viral geneexpression and not the primary adhesion or fusion process.© 2006 Elsevier Inc. All rights reserved.

Keywords: RSV; PPARγ; Ciglitazone; Troglitazone; 15d-PGJ2; PPARα; Bezafibrate; NHBE; A549; HEp-2

Introduction

Respiratory syncytial virus (RSV) is the leading cause ofserious respiratory tract infections in infants and young childrenthroughout the world (Hall andMcCarthy, 2000). It accounts forapproximately 125,000 pediatric hospitalizations and up to 450deaths of children each year in the United States (Shay et al.,2001). It has been recognized as a major cause of morbidity andmortality in immunocompromised patients, the elderly, andchildren with underlying lung–heart diseases (McDonald et al.,1982; Falsey et al., 1995; Glezen et al., 2000). Despite nearlyfive decades of intensive RSV research, there exist neither aneffective active vaccination nor a promising anti-inflammatorytherapy (Kimpen, 2001). The prophylactic administration of ahumanized monoclonal anti-RSV antibody is highly effectivebut costly and currently only recommended for high-riskchildren up to 2 years of age, i.e., premature infants and childrenwith bronchopulmonary dysplasia or congenital heart disease(IMPACT-RSV Study Group, 1998). Moreover, the value of

⁎ Corresponding author. Fax: +49 391 6713384.E-mail address: [email protected] (R. Arnold).

0042-6822/$ - see front matter © 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.virol.2006.03.008

ribavirin, the only licensed anti-RSV preparation, has beenincreasingly questioned in the past (American Academy ofPediatrics Committee on Infectious Diseases, 1996). During thelast several years, a variety of RSVantiviral substances has beendescribed (Torrence, 2000; Brooks et al., 2004). Differentpromising inhibitors of the F protein were generated (Cianci etal., 2004; Douglas et al., 2005), but due to the lack of aproofreading mechanism of the RNA polymerase, the selectionof drug-resistant RSV variants is a major problem (Douglas etal., 2005; Zhao and Sullender, 2005). Due to their property toinhibit the virus–host infection process, these drugs would beprimarily useful as a prophylactic agent. Quite recently, asuccessful inhibition of respiratory syncytial virus replicationobtained by nasally administered siRNAwas reported (Bitko etal., 2005). Considerable therapeutic value was observed whenused as a prophylactic as well as a treatment drug after infection.However, currently neither of these substances is approved forthe treatment of RSV infection.

Therefore, it remains an important task to develop noveltherapeutical targets. A better understanding of the cellularsignaling transduction pathways which are intimately involvedin the RSV replication process is of major importance. It is

mailto:[email protected]

http://dx.doi.org/10.1016/j.virol.2006.03.008

336 R. Arnold, W. König / Virology 350 (2006) 335–346

known that the infection of the human airway epithelial cells,the target for RSV infection, results in a prominent activation ofthe transcription factor NF-κB leading to the expression of awide spectrum of immune response genes (Tian et al., 2002).Quite recently, Reimers et al. reported that the viral M2-1protein mediates the activation of NF-κB by direct protein–protein interaction (Reimers et al., 2005). However, theactivation of NF-κB is dependent on RSV replication and notvice versa (Fiedler et al., 1996). In contrast, when NF-κB isactivated either by TNF-α or by IL-1β prior to RSV infection,this transcription factor is capable of signaling an innateantiviral response that is independent of IFN and the JAK/STATpathway (Bose et al., 2003). In addition, the blocking of STAT-1α or Map kinase ERK-1/2 activity attenuates RSV genetranscription and infection (Kong et al., 2003, 2004). Interest-ingly, the nonstructural RSV proteins NS1 and NS2 arediscussed as viral evasion factors. They inhibit the expressionof Stat2 and block thereby the antiviral α/β interferon responseof the target cell (Lo et al., 2005). Nevertheless, despite of thesefirst insights, a detailed knowledge of the activated cellularsignal transduction pathways involved in the RSV replicationprocess in infected human lung epithelial cell is still lacking.

Peroxisome proliferator-activated receptors (PPARs) areligand-activated transcription factors which form a subfamilyof the nuclear receptor gene family consisting of three isotypes:PPARα, PPARβ, and PPARγ (Lee et al., 2003). They regulatethe transcription of distinct genes through heterodimerizationwith the retinoid X receptors (RXR). They were first identifiedas regulators of the lipid and glucose metabolism (Evans et al.,2004), but during the last decade, data accumulated showingthat PPARs may also be involved in the complex regulation ofimmune and inflammatory processes (Daynes and Jones, 2002).The activation of PPARα and PPARγ correlated with theinhibition of inflammatory cell responses in a variety of celltypes and the lung (Clark, 2002; Standiford et al., 2005). Inaddition, it has been shown that some viruses may exploit theseligand-activated transcription factors of the nuclear hormonereceptor family for their own growth advantage. For example,the hepatitis B virus uses the nuclear hormone receptor RXRαand the peroxisome proliferator-activated receptor α (PPARα)for its own replication cycle in liver cells (Tang and McLachan,2001). Also, the HIV controls, at least partly, its replication bythe interaction of the virus encoded Nef protein with the PPARγof the host cell (Otake et al., 2004). Therefore, the objective ofthe present study was to assess whether PPAR agonists mightinfluence the replication of RSV in human lung epithelial cells.We analyzed the following specific PPARγ agonists in ourstudy: the natural ligand 15-deoxy-Δ12,14-prostaglandin J2(15d-PGJ2), the synthetic agonist Fmoc-Leu, as well as therecently FDA approved antidiabetic thiazolidinedione deriva-tives ciglitazone and troglitazone, respectively. Besides thedifferent specific PPARγ agonists, we also included thehypolipidemic drug bezafibrate, as a representative PPARαligand.

In this study, we supply evidence that the synthetic PPARγagonists Fmoc-Leu, ciglitazone, troglitazone, and the naturalPPARγ ligand 15d-PGJ2 profoundly protected epithelial

monolayers (HEp-2, A549) from cytopathic effects up to 48 hpost-infection. After 3 days post-infection, only the syntheticPPARγ agonists were still able to protect the RSV-infected cellmonolayer from detrimental major cytopathic cell damage. TheRSV-infected human lung A549 epithelial cells cultured in thepresence of the PPARγ ligands under study expressed a reducedamount of the cell surface bound viral G and F protein.Concomitantly, a downregulated viral N protein mRNA levelwas observed. This coincided with a markedly reduced releaseof infectious progeny virus from RSV-infected lung epithelialcells (A549, NHBE). In addition, the potential use of thesePPARγ agonists in the light of RSV-infection is discussed.

Results

Expression profile of PPAR isoforms in A549 cells

Since A549 cells are a well-accepted in vitro RSV infectionmodel, we used these cells to analyze which of the PPARisotypes are expressed in these human airway epithelial cells atthe protein level. The cells were permeabilized and stainedintracellularly with isoform-specific polyclonal Abs. Theimmunofluorescence signals were determined by FACS anal-ysis. As shown in Fig. 1, A549 cells expressed high amounts ofall three isoforms of PPAR in a constitutive manner. When theprimary unlabeled Abs were omitted, the cells showed noremaining immunofluorescence signal verifying that the secondlabeled Abs did not stain the cells in an unspecific manner (datanot shown). Additionally, prior to cell staining, an excess ofblocking peptides specific for the used PPARα- and PPARγ-specific Abs was added to the primary Abs. These pretreatedPPAR-specific Abs did not stain A549 cells anymore provingbinding specificity of the used Abs (Figs. 1A and B).

Increased cell viability of RSV-infected epithelial cellmonolayer exposed to PPARγ agonists

To assess whether activation of PPARα and PPARγ mighthave some antiviral activity, we infected and cultured confluentA549 cell monolayer in the presence of specific PPAR agonists.The following PPARγ-specific ligands were used in our study:the naturally occurring compound 15d-PGJ2, a metabolite ofprostaglandin D2, the synthetic antidiabetic thiazolidinedionederivatives ciglitazone and troglitazone, and a N-protectedleucine analog designated as Fmoc-Leu (Rocchi et al., 2001).Two molecules of Fmoc-Leu interact with one PPARγmoleculein a highly specific manner. In addition, the PPARα-specificagonist bezafibrate was included in this study to determinewhether activation of PPARα might also interfere with RSVinfection in airway epithelial cells.

The cells were preincubated with the PPAR agonists for30 min, infected with RSVand then cultured for 72 h still in thepresence of the agonists. Thereafter, the cytopathic effect ofRSV was assessed by light microscopy (Fig. 2). As can be seenfrom Fig. 2b, when compared with the medium control (Fig.2a), the previously confluent cell monolayer was totallydestroyed by RSV infection. The infected monolayer showed

Fig. 1. PPARα, PPARβ, PPARγ expression in human A549 cells. Data are shown as relative fluorescence intensity (on a logarithmic scale, x axis) and number of cells(y axis). The relevant isotype control is shown by the dotted line. The overlaid distribution histograms show a constitutive intracellular expression of all three PPARisoforms (A–C, thick lines). The antibody/protein complex formation was blocked by prior incubation of the PPARα and PPARγ polyclonal Abs with specific PPARαand PPARγ blocking peptides (A and B, thin lines). Representative histograms are shown (n = 2).

337R. Arnold, W. König / Virology 350 (2006) 335–346

an extensive loss of adherent cells. However, RSV-infectedmonolayers preincubated with ciglitazone, troglitazone, andFmoc-Leu, respectively, were totally protected from RSV-induced cell damage (Figs. 2c–e). The natural PPARγ ligand15d-PGJ2 showed a slight increase of adherent A549 cells whencompared to the RSV-infected monolayer cultured in medium

Fig. 2. Reduced cytopathic effects in RSV-infected A549 cells cultured in the presenceciglitazone (20 μM) (c), troglitazone (20 μM) (d), Fmoc-Leu (100 μM) (e), 15d-PGJ2(h) for 30 min. Thereafter, cells were infected with RSV (b–h) and cultured for 72cytopathic changes in A549 cells 72 h post-infection. Images were taken at a magn

alone (Figs. 2f and b). In contrast, the PPARα agonistbezafibrate and the vector control DMSO had no protectiveimpact on the RSV-induced cell damage (Figs. 2g and h).

Since HEp-2 cells are well known for their extensivesyncytia formation following RSV infection, we also analyzedthe cytopathic changes in RSV-infected HEp-2 monolayers

of PPAR agonists. Confluent monolayers were preincubated with medium (a, b),(20 μM) (f), bezafibrate (100 μM) (g), and the vector control DMSO (0.2% (v/v))h still in the presence of the agonists. Photomicrographs are shown illustratingification of ×100.


cultured in the presence of PPAR agonists. Since the formationof syncytia proceeded faster and more efficiently in HEp-2monolayer than in A549 cells, we evaluated the cell monolayersby light microscopy already 48 h post-RSV infection. Similar toA549 cells, pretreatment of the monolayers with ciglitazone,troglitazone, and Fmoc-Leu (Figs. 3c–e) totally abolished theRSV-induced syncytium formation (Fig. 3b). Furthermore,when analyzed 48 h post-infection, the natural PPARγ ligand15d-PGJ2 protected the HEp-2 monolayer from syncytiumformation (Fig. 3f). Again, the PPARα agonist bezafibrate andthe vector control DMSO did not influence the RSV-inducedcell fusion (Figs. 3g and h).

Similar results were obtained in A549 and HEp-2 cellmonolayers, when PPARγ ligands were added post-RSVinfection (data not shown). Taken together, all PPARγ ligandsunder study fully protected the RSV-infected airway epithelialcell monolayer from RSV-induced cell damage at least up to48 h post-infection.

Inhibitory effect of PPARγ agonists on viral protein and mRNAexpression

To analyze whether the PPARγ ligands might have a generalinhibitory effect on the cellular protein synthesis, we deter-mined the Toll-like receptor 2 (TLR2) expression pattern andthe release of dehydrogenases from RSV-infected cells in thepresence or absence of the agonists. As shown in Fig. 4A, theinfection with RSV led to a significant upregulation of TLR2 onA549 cells. However, neither of the PPARγ ligands under studymodulated the RSV-induced TLR2 expression pattern. More-over, the enzyme activity of secreted dehydrogenases deter-mined in the harvested cell supernatants of RSV-infected A549cells was independent of the presence of PPARγ ligands (Fig.4B). In summary, these data suggest that the PPARγ agonistsneither inhibit the cellular protein synthesis nor have any

Fig. 3. Reduced cytopathic effects in RSV-infected HEp-2 cells cultured in the prese(a, b), ciglitazone (20 μM) (c), troglitazone (20 μM) (d), Fmoc-Leu (100 μM) (e), 1(0.2% (v/v)) (h) for 30 min. Thereafter, cells were infected with RSV (b–h) and cultuillustrating cytopathological changes in HEp-2 cells 48 h post-infection. Images we

general inhibitory effect on cell viability and/or cell metabolismwhich might account for the reduced viral cytopathic effect.

The viral attachment (G) protein and fusion (F) protein,glycoproteins incorporated in the viral envelope, are responsiblefor the initial attachment and fusion of the viral envelope withthe target cell membrane. Moreover, the cytopathic effect intissue culture is mediated by the epithelial cell surfaceexpression of the viral F protein. We, therefore, analyzedwhether the observed protective effect of PPARγ ligands onRSV-infected human lung epithelial cells might be due to areduced expression of these viral envelope proteins. The cellswere pretreated with PPARγ agonists, infected with RSV, andthen cultured for 36 h. As shown in Fig. 5, the expression ofboth viral G and F protein was markedly reduced when cellswere cultured in the presence of ciglitazone, 15d-PGJ2, andFmoc-Leu, respectively. Similar to ciglitazone, troglitazone(20 μM) markedly reduced the expression of both proteins (datanot shown). Again, the PPARα agonist bezafibrate (100 μM)did not modulate the expression level of viral proteins on RSV-infected cells (data not shown). To exclude that the PPARγagonists primarily interfere with the viral adhesion and fusionprocess or may simply kill the virus, we analyzed the F proteinexpression on RSV-infected A549 cells that were eitherpreincubated with the agonists or exposed to the agonistssubsequent to the RSV infection. Our data show that the cellsurface F protein expression was even downregulated byPPARγ agonists which were added to the infected cells 8 h post-infection (Fig. 6).

Next, we asked whether the diminished viral proteinexpression might be associated with a reduced viral mRNAlevel. For that purpose, we analyzed the cellular amount of Nprotein-specific mRNA by real-time RT-PCR. Our data showthat A549 cells expressed a high amount of N protein-encodingmRNA 36 h post-infection (Fig. 7). However, when cells werecultured in the presence of PPARγ agonists, the cellular amount

nce of PPAR agonists. Confluent monolayers were preincubated with medium5d-PGJ2 (20 μM) (f), bezafibrate (100 μM) (g), and the vector control DMSOred for further 48 h in the presence of the agonists. Photomicrographs are shownre taken at a magnification of ×100.

Fig. 4. PPARγ agonists do not affect cell protein synthesis and viability. (A) The RSV-induced expression of Toll-like receptor 2 (TLR2) is not downregulated byPPARγ agonists. Cells were pretreated either with ciglitazone (20 μM), troglitazone (20 μM), 15d-PGJ2 (20 μM), Fmoc-Leu (100 μM), bezafibrate (100 μM), thevehicle DMSO (0.2% (v/v)), or medium, respectively for 30 min, then infected with RSV, washed, and incubated with freshly supplied agonists for 48 h. Results aremeans ± SEM (n = 3). (B) PPARγ agonists did not reduce cell viability. Cells were incubated with the abovementioned agonists for 48 h. Cell viability/proliferationafter treatment with PPAR agonists was assessed by WST-1 assay. Data are shown as means ± SD of a representative experiment (n = 3) performed in quadruplicate.


of N protein-encoding mRNAwas significantly downregulated.All PPARγ agonists under study reduced dose dependently thecellular N protein mRNA level (Figs. 7A and B). Also, areduced G protein mRNA level was demonstrated by RT-PCR(data not shown).

Fig. 5. Reduced expression of RSV G and F protein on human A549 cells cultured wi(left column) and F protein (right column) expression. Cells were pretreated with the afor another 36 h. Expression of viral G and F protein on RSV-infected A549 cells cu15d-PGJ2 (20 μM) (c, d), and Fmoc-Leu (100 μM) (e, f) (thick line), respectively. Balines. The results are representative of multiple independent experiments (n = 3).

PPARγ agonists diminish the replication of RSV in A549 andNHBE cells

To investigate whether the PPARγ agonist-mediated inhibi-tion of viral protein and gene expression results in a diminished

th PPARγ ligands. Flow cytometric analysis of cell surface bound viral G proteingonists for 30 min, RSV-infected, and cultured still in the presence of the agonistsltured in medium alone (thin line) or pretreated with ciglitazone (20 μM) (a, b),ckground fluorescence signals of stained noninfected cells are shown by dotted

Fig. 6. F protein expression is diminished by PPARγ agonists added pre- andpost-RSV infection. Flow cytometric analysis of cell surface bound viral Fprotein expression 48 h post-infection. A549 cells were incubated with thePPARγ agonists troglitazone (50 μM), 15d-PGJ2 (20 μM), Fmoc-Leu (200 μM),and the vehicle control DMSO (0.2% (v/v)), respectively. The agonists wereeither added 30 min prior to RSV infection, directly thereafter (0 min post-infection), or 4 h and 8 h post-infection. A representative experiment is shown(n = 2).


production of infectious progeny virus, we determined theamount of infectious particles in the cell supernatants harvested48 h post-infection. The cells were pretreated with different

Fig. 7. PPARγ agonists downregulated the cellular amount of N protein-encoding mRNA in RSV-infected A549 cells. (A) Cells were pretreated for 30min with 15d-PGJ2 (10–20 μM), ciglitazone (20–40 μM), troglitazone (20–40 μM), or with (B) Fmoc-Leu (200–400 μM) and the vehicle control DMSO(0.1, 0.2, 0.3% (v/v)), respectively, for 30 min. Pretreated cells were infectedwith RSV and cultured in the presence of the agonists for another 36 h. Thecellular amount of N protein mRNAwas quantitatively determined by real-timeRT-PCR. Results are means ± SD of samples analyzed in triplicate; in case errorbars are not visible, they are smaller than symbols shown; The results arerepresentatives of multiple independent experiments (n = 3).

concentrations of the PPARγ agonists as well as the PPARαagonist bezafibrate. Our data show that all used PPARγ agonistssignificantly reduced the amount of infectious RSV particles inthe cell supernatants in a dose-dependent manner (Figs. 8A andB). Especially ciglitazone and Fmoc-Leu showed the mostprominent inhibitory effect on viral replication, i.e., the virustiter decreased by nearly 1000-fold. Furthermore, cellsimmediately exposed to PPARγ agonists post-RSV infectionshowed a comparable diminished release of infectious progenyvirus (data not shown). In contrast, neither the PPARα agonistbezafibrate nor the vector control DMSO (0.05–0.3% (v/v)) didaffect the viral replication rate (Fig. 8B).

The A549 line originated from type II alveolar adenocarci-noma cells. Although this cell line is a well-accepted in vitromodel for analyzing RSV infection in human lung epithelialcells, we next verified whether the PPARγ agonists ciglitazone,Fmoc-Leu, and 15d-PGJ2 might also interfere with RSVreplication in primary human bronchial epithelial cells. Similarto a more clinical setting, we first infected NHBE cells andcultured them thereafter for 48 h in the presence of the PPARγagonists. Our data show that ciglitazone, troglitazone, and theendogenous ligand 15d-PGJ2 inhibited the replication of RSV in

Fig. 8. PPARγ agonists diminished the release of infectious progeny virus fromA549 cells. (A) Cells were pretreated with ciglitazone (5–50 μM), troglitazone(5–50 μM), 15d-PGJ2 (5–20 μM) for 30 min. (B) Cells were pretreated withFmoc-Leu (50–300 μM), bezafibrate (50–300 μM), and the vehicle DMSO(0.05, 0.1, 0.2, 0.3% (v/v)) for 30 min. Thereafter, cells were infected with RSV(MOI = 3) and cultured for 48 h. Supernatants were harvested and immediatelyanalyzed by plaque assay on HEp-2 cells. Results are means ± SEM (n = 8); incase error bars are not visible, they are smaller than symbols shown; *P < 0.01,significant vs. nonpretreated RSV-infected cells.

Fig. 9. PPARγ agonists reduced the release of progeny virus and the cellular amount of RSV N protein mRNA in NHBE cells. The cells were infected with RSV(MOI = 3) for 2 h, washed, and cultured for 48 h in the presence of the PPAR ligands ciglitazone (20 μM), troglitazone (20 μM), 15d-PGJ2 (10 μM), Fmoc-Leu(100 μM), bezafibrate (100 μM), and the vehicle control DMSO (0.2% (v/v)), respectively. (A) Supernatants were harvested and immediately analyzed by plaque assayon HEp-2 cells. Results are means ± SEM (n = 4); *P < 0.01, significant vs. nonpretreated RSV-infected cells. (B) Cells were harvested, and the cellular amount ofmRNA encoding for viral N protein was quantitatively determined by real-time RT-PCR. Results are means from two independent experiments.


NHBE cells in a comparable manner. The addition of Fmoc-Leureduced the release of progeny virus most efficiently by nearly1000-fold (Fig. 9A). Moreover, similar to A549 cells, thereduced production of infectious virus particles was paralleledby a downregulated N protein mRNA level, suggesting thatPPARγ agonists might also interfere with RSV gene expressionin NHBE cells (Fig. 9B). Similar to A549 cells, the PPARαagonist bezafibrate and DMSO had no effect on the release ofinfectious progeny virus and the cellular mRNA level encodingfor RSV N protein (Figs. 9A and B).

Discussion

In the present study, we supply evidence that the four PPARγagonists ciglitazone, troglitazone, 15d-PGJ2, and Fmoc-Leu,respectively, suppress the replication of RSV in an in vitro RSVinfection model. Human lung epithelial cells of the cell lineA549 as well as primary bronchial epithelial cells showed asignificantly reduced release of infectious progeny virus. Incontrast, activation of PPARα by means of the specific agonistbezafibrate had no impact on RSV replication.

We determined a constitutive expression of PPARα, β, and γin human lung A549 cells by FACS analysis. With regard to theexpression of PPARγ, our data are in accordance with datapreviously published (Michael et al., 1997; Wang et al., 2001;Pawliczak et al., 2002). However, Wang and Pawliczak arguedagainst a prominent expression of PPARα in human NIH-A549,A549, and NHBE cells. In contrast, rabbit corneal epitheliumexpresses PPARα and β but not γ mRNA (Bonazzi et al.,2000). Whether our contrasting PPARα expression pattern isdue to the different antibodies used, the different sensitivities ofthe detection systems, i.e., FACS analysis vs. immunoblottechnique, or the analyzed cell type or cell charge remains to bedetermined.

To analyze the inhibitory effect of PPARγ agonists on RSVproduction in more detail, we determined the viral protein andmRNA expression pattern. The reduced release of infectiousprogeny virus was paralleled by a downregulation of the viral Fand G protein expressed on the cell surface of the infected A549epithelial cells. These data suggest that the observed reducedrelease of progeny virus might be a direct consequence of an

impaired viral protein synthesis and is obviously not due to analtered budding process from the infected cell. It is known thatthe F protein by interacting with the small GTPase RhoA issolely responsible for the cytopathic effect primarily observedin cell culture (Pastey et al., 1999). Therefore, the fact that allfour PPARγ agonists significantly reduced the expression of theviral F protein is fully sufficient to account for the increasedviability and reduced cytopathic effect observed in A549, HEp-2, and NHBE cell monolayers. When assayed 48 h post-infection the natural ligand 15d-PGJ2 inhibited the release ofinfectious progeny virus and fully protected RSV-infected HEp-2 cell monolayer from cell damage. However, when RSV-infected A549 cell monolayers were assayed 72 h post-infection, the natural ligand 15d-PGJ2 protected the monolayersto a lesser degree than the synthetic PPARγ agonistsciglitazone, troglitazone, and Fmoc-Leu, respectively, whichcannot be explained at the present moment. The reason for thatis not known. However, little is known about the biologicalactivity and turnover of 15d-PGJ2 in vivo as well as in vitro lungepithelial cell cultures up to an incubation time of 72 h. It isassumed that 15d-PGJ2 secreted into the microenvironment ofthe inflamed tissue is effectively metabolized by macrophagesand other resident cells. In line with this assumption is theobservation made by Bell-Parikh et al. (2003) that theconcentrations of 15d-PGJ2, measured in the joint fluids ofpatients suffering from rheumatoid arthritis, are too low toactivate PPARγ. We, therefore, hypothesize that the loss ofprotection observed in RSV-infected lung epithelial cellmonolayers cultured up to 72 h post-infection might be due toa more effective metabolism/inactivation of 15d-PGJ2 com-pared to the other PPARγ agonists under study. Detailed timekinetic studies have to address this point in future.

We observed a significant inhibition of progeny virus releasewhen PPARγ agonists were used either as a prophylactic drug inA549 cells or as therapeutic drug in A549 and NHBE cells, i.e.,pre- and post-infection. By analyzing the viral protein expres-sion, we determined that the PPARγ agonists still supply anantiviral effectwhen added 8 h post-infection. Thus, the analyzedPPARγ agonists did not primarily interfere with the viraladhesion and fusion process. Our data concerning the reduced NproteinmRNA expression level suggest that the lifecycle of RSV


might be, at least partly, inhibited at the transcriptional level. Thedetailed molecular mechanisms leading to the reduced viral geneexpression pattern are still not known. However, a generalinhibitory effect on the cellular protein and/or mRNA synthesisrate can be ruled out since the PPARγ agonists did not interfereeither with cell viability or the RSV-induced upregulation ofTLR2.

Recently, the direct activation of PPARγ in A549 cells eitherby 15d-PGJ2 (1–5 μM) or ciglitazone (5–25 μM) wasdemonstrated by Wang et al. (2001). Therefore, we did notreproduce these data. However, in agreement with otherpublished results analyzing the anti-inflammatory and antiviralcapacities of thiazolidinediones and 15d-PGJ2, we also usedwith respect to their PPARγ-binding constants relatively highconcentrations of the agonists to influence RSV production(Hayes et al., 2002; Kwak et al., 2002). One may argue thatnatural and synthetic PPARγ agonists also exert part of theiranti-inflammatory activity by interfering with a variety of othersignal transduction pathways in a PPARγ-independent manner.Especially, the modulation of the Janus/STAT and MAP kinasesignaling pathways has been reported (Takeda et al., 2001; Chenet al., 2003). In this regard, it was recently published thatactivation of ERK1 and 2 is a prerequisite for a successfulreplication of RSV (Kong et al., 2004). Therefore, PPARγagonist-mediated modulation of MAP kinase activity mightcontribute to the observed antiviral effect. In addition, whetherthe RSV-induced inhibition of the antiviral cellular α/β-interferon response, mediated by the NS1 and NS2 proteins,might be counter-regulated by the PPARγ agonists should alsobe addressed in future experiments (Bossert et al., 2003; Spannet al., 2004; Lo et al., 2005).

The fact that PPARγ agonists are able to modulate theactivity of other transcription factors is well demonstrated(Straus et al., 2000; Përez-Sala et al., 2003). Recently, weobserved that all PPARγ agonists under study profoundlyreduced the binding activity of NF-κB (p65/p50) in RSV-infected cells (Arnold and König, in press). Quite recently,Bailey and Ghosh (2005) published data that PPARγ agonists,which induce sumoylation of PPARγ, prevent the degradationof the NF-κB-bound coreceptor complex thereby excluding itsexchange by a coactivator complex. This leads to a transrepres-sion of the NF-κB-dependent promoters. However, no data areavailable which support the notion that a reduced NF-κBactivity interferes with the replication of RSV (Fiedler et al.,1996). Nevertheless, the reduced NF-κB binding activitycorrelated with a diminished release of proinflammatorycytokines (IL-1α, IL-6, TNF-α) and chemokines (CXCL8 andCCL5) from RSV-infected lung epithelial cells (Arnold andKönig, in press). Furthermore, a reduced cell surface expressionof ICAM-1 and MHC-I was observed (our own unpublisheddata). Currently, there exist no satisfactory anti-inflammatorytreatment for RSV infection. Corticosteroids have been shownto have little to no effect in the overall outcome of RSVinfection (Bonville et al., 2001). In contrast, several studies haveclearly demonstrated that PPARγ agonists have additional anti-inflammatory effects compared to corticosteroids. In this regard,Patel et al. (2003) reported that PPARγ agonists, but not

dexamethasone, inhibited the release of G-CSF from humanairway smooth muscle cells. Furthermore, the accumulation ofhuge amounts of PMN in the RSV-infected lung is a hallmark ofsevere lower respiratory tract disease (Everard et al., 1994).Quite recently, it was published that the neutrophil infiltration ofthe lung was significantly reduced by the application of PPARγagonists during bleomycin-induced mouse lung injury (Genov-ese et al., 2005). Moreover, it has been shown in a murine LPS-induced airway inflammation model that the thiazolidinedionecompound rosiglitazone, in contrast to dexamethasone, reducedthe neutrophil number in the lung tissue when administeredafter the LPS insult (Birrell et al., 2004). A direct anti-inflammatory effect of PPARγ agonists on neutrophil andeosinophil cell function was also reported (Woerly et al., 2003;Imamoto et al., 2004).

In summary, proinflammatory immune effector mechanismsas well as direct cytopathic effects contribute to the overalloutcome of RSV-induced lower respiratory tract disease. Ourdata presented herein, together with our previously publishedanti-inflammatory capacity of PPARγ agonists on the nativeimmune response of the RSV-infected lung epithelial cell,suggest that PPARγ agonists might have a beneficial effect inthe course of an acquired RSV infection.

Materials and methods

Compounds

The PPAR agonists ciglitazone, troglitazone, bezafibrate,Fmoc-Leu (Merck Biosciences, Schwalbach, Germany), and15-deoxy-Δ12,14-prostaglandin J2 (15d-PGJ2) (Biomol, Ham-burg, Germany) were stored in DMSO (Sigma, Deisenhofen,Germany) at −80 °C. At the day of experiment, compoundswere freshly diluted in basal medium (DMEM or BEGM) andadded to the cells with a final DMSO concentration of 0.05–0.3% (v/v).

Cell culture

Human A549 pulmonary type II epithelial cells (passages 4–20) and HEp-2 cells (American Type Culture Collection(ATCC), Rockville, MD, USA) were cultured in Dulbecco'smodified eagle medium (DMEM) (4500 mg/l D-glucose)supplemented with 2 mM glutamine, 5% (v/v) inactivatedfetal calf serum (FCS), streptomycin (100 μg/ml), and penicillin(100 IU/ml). Normal human bronchial epithelial cells (NHBE)(Cambrex/Clonetics, Verviers, Belgium) were cultured incomplete BEGM medium (5 μg/ml insulin, 0.5 μg/mlhydrocortisone, 10 μg/ml transferrin, 6.5 ng/ml triiodothyro-nine, 0.5 μg/ml epinephrine, 0.5 ng/ml human epidermal growthfactor, 0.1 ng/ml retinoid acid, 50 μg/ml gentamicin, and 52 μg/ml bovine pituitary extract). NHBE cells were cultured andexpanded according to the instructions of the manufacturer. Tosupport cell attachment and growth of NHBE cells, tissueculture flasks and plastic plates (Greiner, Frickenhausen,Germany) were precoated with fibronectin (10 μg/ml)(Sigma) for 30 min at 37 °C. Cells of passages 3–7 were


seeded on 24-well plates and cultured until confluency incomplete medium. All cell cultures and prepared virus stockswere free of mycoplasmic contamination routinely verified by acommercially available mycoplasma detection kit (RocheDiagnostics, Mannheim, Germany).

Virus growth and preparation

The Long strain of RSV (ATCC) was propagated and titratedin HEp-2 cells. The Long strain was used, as this strain iscapable of causing severe disease in human infants. For viruspropagation, confluent monolayers were infected with RSV(MOI = 0.1) for 3 h in DMEM without FCS. The monolayerswere washed, overlayed with DMEM (0.5% FCS), andincubated at 37 °C in 5% CO2 atmosphere until cytopathiceffect reached ∼80%. Thereafter, the supernatants wereharvested, and cellular debris was removed by centrifugation(5000 × g, 10 min). RSV was concentrated by polyethyleneglycol precipitation (10%) and purified by means of discontin-uous sucrose gradient centrifugation (Ueba, 1978). To stabilizethe purified virus particles, they were resolved in 20% sucrose/NT-buffer (150 mMNaCl, 50 mM Tris–HCl, pH 7.5) and storedat −80 °C. No contaminating cytokines, including IL-1α, TNF-α, IL-6, IL-8, RANTES, GM-CSF, and interferonα/β, weredetected in these sucrose-purified viral preparations. Also, virusstocks and media were free of lipopolysaccharide routinelyassayed by the Limulus hemocyanin agglutination assay(Sigma).

RSV titration

For virus titration prepared RSV stock solutions andharvested A549 cell supernatants were serially tenfold dilutedonto confluent HEp-2 monolayers cultured in 96-well flat-bottomed plates. The virus titer was quantified as previouslydescribed (Arnold et al., 2004). Briefly, infected cells werecultured for 48 h under methyl cellulose. Then monolayers werefixed with paraformaldehyde (2%) and stained with mouse anti-P protein mAb (clone: 3C4, kindly supplied by Dr. H. Werchau,Department of Medical Virology, Ruhr-University, Bochum,Germany). Binding of the primary Ab was detected by a secondstaining with HRP-conjugated rabbit anti-mouse Ab (Dako,Hamburg, Germany). Thereafter, RSV plaques were visualizedby colorimetric staining, counted microscopically andexpressed as log10 plaque forming unit (pfu)/ml. The stocktiter of the used virus pool was 108 pfu/ml.

Cell experiments

Epithelial cells (A549 and NHBE cells) were seeded in 24-well (1 × 105 cells/well) culture plates and cultured overnight inDMEM and BEGM, respectively. If not stated otherwise,confluent monolayers were washed with basal medium andpretreated with PPAR agonists for 30 min prior to RSVadsorption. The cells were treated with varying doses ofciglitazone (5–50 μM), troglitazone (5–50 μM), 15d-PGJ2 (5–20 μM), Fmoc-Leu (50–300 μM), bezafibrate (50–300 μM),

and solvent (DMSO; 0.05–0.3% (v/v)), respectively, in avolume of 1 ml. Thereafter, the medium was reduced to avolume of 200 μl, and cells were infected with RSV for 2 h stillin the presence of the used PPAR agonists. The virus stock wasdiluted in culture medium to a defined multiplicity of infection(MOI) of 3. An equivalent amount of 20% sucrose/NT-bufferadded to noninfected cells served as mock control. Then, cellswere washed and incubated with fresh medium (DMEM (2%FCS) or supplemented BEGM in case of NHBE cells) in thepresence of freshly supplied agonists for another 36 h–72 h. Inaddition, RSV-infected cells were incubated with PPAR agonistspost-RSV infection. Thereafter, cells were detached andharvested as previously described (Arnold and König, 2005).As was determined by a modified 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, designated asWST-1 assay (Roche), the added agonists induced no significantcell death (>3%) at the concentrations used (data not shown).

Quantification of cell proliferation and cell viability

Confluent A549 cell monolayers (5 × 104 cells/well) placedin 96-well flat-bottomed plates were incubated with the agonistsfor 48 h. For control, cells were incubated with the vehicleDMSO alone. Cell viability and cell proliferation wasdetermined by analyzing the cleavage of the tetrazolium saltWST-1 (4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazo-lio]-1,3-benzene disulfonate) to formazan by mitochondrialdehydrogenases (Roche). A reduction in the number of viablecells results in a decrease in the overall activity of mitochondrialdehydrogenases in the cell supernatant. This reduced enzymeactivity leads to a reduced amount of formazan dye formed,which directly correlates to the number of metabolically activecells in the culture. Briefly, 10 μl of a ready to use solution wasadded to 100 μl (1:10 final dilution) for the last 1 h ofincubation. The plate was directly read at 450 nm against areference wavelength of 620 nm on a SpectraFluorPlus Reader(Tecan, Crailsheim, Germany).

FACS staining

The constitutive intracellular expression of PPARs innoninfected A549 cells was analyzed by flow cytometry. Byusing the Fix and Perm kit for intracellular staining (BDBiosciences (BD), Heidelberg, Germany), the cells were fixedand permeabilized according to the manufacturer's instructions.Thereafter, the cells were washed two times in permeabilizationwashing buffer and analyzed with the primary unlabeledpolyclonal rabbit anti-PPARα-Ab, rabbit anti-PPARγ-Ab (Cay-man Chemicals, Ann Arbor, MI, USA), rabbit anti-PPARβ Ab(Santa Cruz Biotechnology, Santa Cruz, CA, USA), andunspecific rabbit IgG (Sigma), respectively, for 1 h at 4 °C.Then, primary bound Abs were detected by secondary stainingwith PE-labeled AffiniPure F(ab′)2 fragment of goat anti-rabbitIgG (H + L) (Dianova, Hamburg, Germany) for 1 h at 4 °C. Todetermine binding specificity of the used primary anti-PPARαand anti-PPARγ Abs, we blocked antibody/protein complexformation by using PPARα- and PPARγ-specific blocking


peptides (Cayman Chemicals). As recommended by themanufacturer, prior to cell staining, the blocking peptides(0.5 mg/ml) were mixed with Abs in a 1:1 (v/v) ratio andincubated for 60 min. To determine RSV protein synthesis inRSV-infected A549 cells, the cells were analyzed for cell surfaceexpression of viral G and F protein (Arnold and König, 2005).Binding of monoclonal mouse anti-G protein and anti-F proteinAb (Biotrend, Köln, Germany) was visualized by a secondstaining with Cy3-labeled AffiniPure goat anti-mouse IgG(H + L) (Dianova). The expression of TLR2 was determinedwith a PE-labeled monoclonal mouse anti-TLR2 Ab(eBioscience, San Diego, USA). The corresponding unlabeledand labeled mouse IgG subtypes served as isotype controls(BD). After cell washing, the cell bound fluorescence signal wasdetermined by FACSCalibur (BD), and data were analyzed bymeans of CellQuest software (BD). The mean fluorescenceintensity (MFI) of 10,000 cells was determined and corrected bysubtraction of background fluorescence of the isotype control.

RNA extraction and real-time RT-PCR

The total cellular RNA from noninfected and RSV-infectedepithelial cells (5 × 105) was extracted using the QiAmp 96 viralRNA Biorobot kit on the Biorobot 3000 System from Qiagen(Düsseldorf, Germany). Reverse transcription (RT)-PCR wasperformed with M-MLV RT-buffer components (Invitrogen,Karlsruhe, Germany). Prepared total RNA (2 μg) was added to50 μl PCR reaction mix consisting of 50 mM Tris–HCl buffer(pH 8.3), 3 mM MgCl2, 75 mM KCl, 10 mM DTT, 500 μMdNTPs, 2 ng/μl oligo d(T)(12–18), 1 U/μl RNAase inhibitor(Applied Biosystems, Foster City, CA, USA), and 4 U/μl M-MLV. The synthesis of cDNA was performed at 37 °C for60 min. The cDNA was stored at −20 °C.

The amount of RSV N protein mRNA was quantified byTaqman real-time RT-PCR as previously reported (Dewhurst-Maridor et al., 2004). Briefly, using the Gene Amp 5700Sequence Detector (Applied Biosystems) the RNA (100 ng) wasreverse transcribed to 2.5 μl cDNA and amplified in a volume of25 μl by means of TaqMan Universal PCRMaster Mix (AppliedBiosystems) containing specific primers for RSV N protein anda fluorescently labeled probe synthesized by Metabion (Pla-negg-Martinsried, Germany): forward, 5′-CTCAATTTCCT-CACTTCTCCAGTGT-3′; reverse, 5′-CTTGATTCCTC-GGTGTACCTCTGT-3′; probe, 5′-6-carboxy-fluorescein(FAM)-TCCCATTATGCCTAGGCCAGCAGCA-6-carboxy-tetramethyl-rhodamine-(TAMRA)-3′. The cDNAwas denaturedat 95 °C for 10 min and amplified by 40 cycles (95 °C, 15 s and60 °C, 1 min). Concentrations of primers and probes wereoptimized to obtain a reaction yield at the lowest threshold cycleand a high cycle to cycle increase of the resulting fluorescentsignal (ΔRn). The primers were used at a concentration of300 nM together with the probe diluted to 200 nM. The primer/probe set for the housekeeping gene human GAPDH wassupplied as a predeveloped kit by Qiagen. The reactioncomponents were mixed and the amplification profile settledaccording to the instructions of the manufacturer. Negativecontrols were carried out with water instead of cDNA. The

cDNA prepared from RSV-infected cells and diluted up to 105

served as a positive control validating that the efficiencies of theRSV N protein and GAPDH PCR were approximately equal.Expression of GAPDH gene was not significantly altered duringthe time of incubation with RSV, drugs and vehicle. Therefore,the relative mRNA expression of each gene was normalized tothe level of GAPDH in the same RNA preparation, i.e., thecomparative CT method was used to analyze the relativequantities of RSV N protein mRNA in cell samples. The relativeRNA amount was calculated by using the following equation:2−ΔΔCT were ΔΔCT = ΔCT,q − CT,cb. CT is defined as the CT

value for RSV N protein minus the CT value for GAPDH for agiven sample, q is the unknown sample, and cb is the calibrator(noninfected sample).

Statistics

If not stated otherwise, data were expressed as themean ± SEM. For statistical significance, analysis of data wasperformed using Student's t test (two-sided). A value ofP < 0.05 was considered significant.

Acknowledgment

The authors wish to thank Ms. Bettina Polte for her excellenttechnical assistance.

References

American Academy of Pediatrics Committee on Infectious Diseases, 1996.Reassessment of the indications for ribavirin therapy in respiratory syncytialvirus infections. Pediatrics 97, 137–140.

Arnold, R., König, W., 2005. Respiratory syncytial virus infection of humanlung endothelial cells enhances selectively ICAM-1 expression. J. Immunol.174, 7359–7367.

Arnold, R., König, W., in press. Peroxisome–proliferator-activated receptor-γagonists inhibit the release of proinflammatory cytokines from RSV-infectedepithelial cells. Virology.

Arnold, R., König, B., Werchau, H., König, W., 2004. RSV deficient in solubleG protein induced an increased proinflammatory response in human lungepithelial cells. Virology 330, 384–397.

Bailey, S.T., Ghosh, S., 2005. ‘PPAR’ting ways with inflammation. Nat.Immunol. 6, 966–967.

Bell-Parikh, L.C., Ide, T., Lawson, J.A., McNamara, P., Reilly, M., FitzGerald,G.A., 2003. Biosynthesis of 15-deoxy-Δ12,14-PGJ2 and the ligation ofPPARγ. J. Clin. Invest. 112, 945–955.

Birrell, M.A., Patel, H.J., McCluskie, K., Wong, S., Leonard, T., Yacoub, M.H.,Belvisi, M.G., 2004. PPAR-γ agonists as therapy for diseases involvingairway neutrophilia. Eur. Respir. J. 24, 18–23.

Bitko, V., Musiyenko, A., Shulyayeva, O., Barik, S., 2005. Inhibition ofrespiratory viruses by nasally administered siRNA. Nat. Med. 11, 50–55.

Bonazzi, A., Mastyugin, V., Mieyal, P.A., Dunn, M.W., Laniado-Schwartzman,M., 2000. Regulation of cyclooxygenase-2 by hypoxia and peroxisomeproliferators in the corneal epithelium. J. Biol. Chem. 275, 2837–2844.

Bonville, C.A., Mehta, P.A., Krilov, L.R., Rosenberg, H.F., Domachowske, J.B.,2001. Epithelial cells infected with respiratory syncytial virus are resistant tothe anti-inflammatory effects of hydrocortisone. Cell. Immunol. 213,134–140.

Bose, S., Kar, N., Maitra, R., DiDonato, J.A., Banerjee, A.K., 2003. Temporalactivation of NF-κB regulates an interferon-independent innate antiviralresponse against cytoplasmic RNA viruses. Proc. Natl. Acad. Sci. U.S.A.100, 10890–10895.


Bossert, B., Marozin, S., Conzelmann, K.K., 2003. Nonstructural proteins NS1and NS2 of bovine respiratory syncytial virus block activation of interferonregulatory factor 3. J. Virol. 77, 8661–8668.

Brooks, M.J., Sadadeusz, J.S., Tannock, G.A., 2004. Antiviral chemotherapeuticagents against respiratory viruses: where are we now and what's in thepipeline? Curr. Opin. Pulm. Med. 10, 197–203.

Chen, C.W., Chang, Y.H., Tsi, C.J., Lin, W.W., 2003. Inhibition of IFN-γ-mediated inducible nitric oxide synthase induction by the peroxisomeproliferator-activated receptor γ agonist, 15-Deoxy-Δ12,14-prostaglandin J2,involves inhibition of the upstream Janus kinase/STAT1 signaling pathway.J. Immunol. 171, 979–988.

Cianci, C., Yu, K.L., Combrink, K., Sin, N., Pearce, B., Wang, A., Civiello, R., Voss,S., Luo, G., Kadow, K., Genovesi, E.V., Venables, B., Gulgeze, H., Trehan, A.,James, J., Lamb, L., Medina, I., Roach, J., Yang, Z., Zadjura, L., Colonno, R.,Clark, J., Meanwell, N., Krystal, M., 2004. Orally active fusion inhibitor ofrespiratory syncytial virus. Antimicrob. Agents Chemother. 48, 413–422.

Clark, R.B., 2002. The role of PPARs in inflammation and immunity.J. Leukocyte Biol. 71, 388–400.

Daynes, R.A., Jones, D.J., 2002. Emerging roles of PPARs in inflammation andimmunity. Nat. Rev. 2, 748–759.

Dewhurst-Maridor, G., Simonet, V., Bornand, J.E., Nicod, L.P., Pache, J.C.,2004. Development of a quantitative TaqMan RT-PCR for respiratorysyncytial virus. J. Virol. Methods 120, 41–49.

Douglas, J.L., Panis, M.L., Ho, E., Lin, K.Y., Krawczyk, S.H., Grant, D.M., Cai,R., Swaminathan, S., Chen, X., Cihlar, T., 2005. Small molecules VP-14637and JNJ-2408068 inhibit respiratory syncytial virus fusion by similarmechanisms. Antimicrob. Agents Chemother. 49, 2460–2466.

Evans, R.M., Barish, G.D., Wang, Y.X., 2004. PPARs and the complex journeyto obesity. Nat. Med. 10, 1–7.

Everard, M.L., Swarbrick, A., Wrightham, M., McIntyre, J., Duncley, C., James,P.D., Sewell, H.F., Milner, A.D., 1994. Analysis of cells obtained bybronchial lavage of infants with respiratory syncytial virus infection. Arch.Dis. Child. 71, 428–432.

Falsey, A.R., Cunningham, C.K., Barker, W.H., Kouides, R.W., Yuen, J.B.,Menegus, M., Weiner, L.B., Bonville, C.A., Betts, R.F., 1995. Respiratorysyncytial virus and influenza virus A infections in the hospitalised elderly.J. Infect. Dis. 172, 389–394.

Fiedler, M.A., Wernke-Dollries, K., Stark, J.M., 1996. Inhibition of viralreplication reverses respiratory syncytial virus-induced NF-κB activationand interleukin-8 gene expression in A549 cells. J. Virol. 70, 9079–9082.

Genovese, T., Cuzzocrea, S., Di Paola, R., Mazzon, E., Mastruzzo, C., Catalano,P., Sortino, M., Crimi, N., Caputi, A.P., Thiemermann, C., Vancheri, C.,2005. Effect of rosiglitazone and 15-deoxy-Δ12,14-prostaglandin J2 onbleomycin-induced lung injury. Eur. Respir. J. 25, 225–234.

Glezen, W.P., Greenberg, S.B., Atmar, R.L., Piedra, P.A., Couch, R.B., 2000.Impact of respiratory virus infections on persons with chronic underlyingconditions. J. Am. Med. Assoc. 283, 499–505.

Hall, C.B., McCarthy, C.A., 2000. Respiratory syncytial virus. In: Mandell,G.L., Benett, J.E., Dolin, R. (Eds.), Principles and Practice of InfectiousDiseases. Churchill Livingstone, Philadelphia, pp. 1782–1801.

Hayes, M.M., Lane, B.R., King, S.R., Markovitz, D.M., Coffey, M.J., 2002.Peroxisome proliferator-activated receptor γ agonists inhibit HIV-1replication in macrophages by transcriptional and post-transcriptionaleffects. J. Biol. Chem. 277, 16913–16919.

Imamoto, E., Yoshida, N., Uchiyama, K., Kuroda, M., Kokura, S., Ichikawa, H.,Naito, Y., Tanigawa, T., Yoshikawa, T., 2004. Inhibitory effects ofpioglitazone on expression of adhesion molecules on neutrophils andendothelial cells. BioFactors 20, 37–47.

IMPACT-RSV Study Group, 1998. Palivizumab, a humanized respiratorysyncytial virus monoclonal antibody, reduces hospitalization from respira-tory syncytial virus infection in high-risk infants. Pediatrics 102, 531–537.

Kimpen, J.L.L., 2001. Management of respiratory syncytial virus infection.Curr. Opin. Infect. Dis. 14, 323–328.

Kong, X., Juan, H.S., Kumar, M., Behera, A.K., Mohapatra, A., Hellermann,G.R., Mane, S., Lockey, R.F., Mohapatra, S.S., 2003. Respiratorysyncytial virus infection activates STAT signalling in human epithelialcells. Biochem. Biophys. Res. Commun. 306, 616–622.

Kong, X., Juan, H.S., Behera, A., Peebles, M.E., Wu, J., Lockey, R.F.,

Mohapatra, S.S., 2004. ERK-1/2 activity is required for efficient RSVinfection. FEBS Lett. 559, 33–38.

Kwak, B.R., Myit, S., Mulhaupt, F., Veillard, N., Rufer, N., Roosnek, E., Mach,F., 2002. PPARγ but not PPARα ligands are potent repressors of majorhistocompatibility complex class II induction in atheroma-associated cells.Circ. Res. 90, 356–362.

Lee, C.H., Olson, P., Evans, R.M., 2003. Minireview: Lipid metabolism,metabolic diseases, and peroxisome proliferator-activated receptors. Endo-crinology 144, 2201–2203.

Lo, M.S., Brazas, R.M., Holtzman, M.J., 2005. Respiratory synytial virusnonstructural proteins NS1 and NS2 mediate inhibition of Stat2 expressionand alpha/beta interferon responsiveness. J. Virol. 79, 9315–9319.

McDonald, N.E., Hall, C.B., Suffin, S.C., Alexon, C., Harris, P.J., Manning,J.A., 1982. Respiratory syncytial virus infection in infants with congenitalheart disease. N. Engl. J. Med. 307, 397–400.

Michael, L., Lazar, M., Mendelson, C., 1997. Peroxisome proliferator-activatedreceptor gamma1 expression is induced during cyclic adenosine monopho-sphate-stimulated differentiation of alveolar type II pneumonocytes.Endocrinology 138, 3695–3703.

Otake, K., Omoto, S., Yamamoto, T., Okuyama, H., Okada, H., Okada, N.,Kawai, M., Saksena, N.K., Fujii, Y.R., 2004. HIV-1 Nef protein in thenucleus influences adipogenesis as well as viral transcription through theperoxisome proliferator-activated complexes. AIDS 23, 189–198.

Pastey, M.K., Crowe, J.E., Graham, B.S., 1999. RhoA interacts with the fusionglycoprotein of respiratory syncytial virus and fascilitates virus-inducedsyncytium formation. J. Virol. 73, 7262–7270.

Patel, H.J., Belvisi, M.G., Bishop-Bailey, D., Yacoub, M.H., Mitchell, J.A.,2003. Activation of peroxisome proliferator-activated receptors in humanairway smooth muscle cells has a superior anti-inflammatory profile tocorticosteroids: relevance for chronic obstructive pulmonary diseasetherapy. J. Immunol. 170, 2663–2669.

Pawliczak, R., Han, C., Huang, X.L., Demetris, A.J., Shelhammer, J.H., Wu, T.,2002. 85-kDa cytosolic phospholipase A2 mediates peroxisome proliferator-activated receptor γ activation in human lung epithelial cells. J. Biol. Chem.277, 33153–33163.

Përez-Sala, D., Cernuda-Morollón, E., Caňada, F.J., 2003. Molecular basis forthe direct inhibition of AP-1 DNA binding by 15-Deoxy-Δ12,14-prostaglan-din J2. J. Biol. Chem. 278, 51251–51260.

Reimers, K., Buchholz, K., Werchau, H., 2005. Respiratory syncytial virus M2-1 protein induces the activation of nuclear factor kappa B. Virology 331,260–268.

Rocchi, S., Picard, F., Vamecq, J., Gelman, L., Potier, N., Zeyer, D., Dubuquoy,L., Bac, P., Champy, M.F., Plunket, K.D., Leesnitzer, L.M., Blanchard, S.G.,Desreumaux, P., Moras, D., Renaud, J.P., Auwerx, J., 2001. A uniquePPARγ ligand with potent insulin-sensitizing yet weak adipogenic activity.Mol. Cell 8, 737–747.

Shay, D.K., Holman, R.C., Roosevelt, G.E, Clarke, M.J., Anderson, L.J., 2001.Bronchiolitis-associated mortality and estimates of respiratory syncytialvirus-associated deaths among US children, 1979–1997. J. Infect. Dis. 183,16–22.

Spann, K.M., Tran, K.C., Chi, B., Rabin, R.L., Collins, P.L., 2004. Suppressionof the induction of alpha, beta, and gamma interferons by the NS1 and NS2proteins of human respiratory syncytial virus in human epithelial cells andmacrophages. J. Virol. 78, 4363–4369.

Standiford, T.J., Keshamouni, V.G., Reddy, R.C., 2005. Peroxisome prolif-erator-activated receptor-γ as a regulator of lung inflammation and repair.Proc. Am. Thorac Soc. 2, 226–231.

Straus, D.S., Pascual, G., Li, M., Welch, J.S., Ricote, M., Hsiang, C.H.,Sengchanthalangsy, L.L., Hhosh, G., Glass, C.K., 2000. 15-deoxy-Δ12,14-prostaglandin J2 inhibits multiple steps in the NF-κB signaling pathway.Proc. Natl. Acad. Sci. U.S.A. 97, 4844–4849.

Takeda, K., Ichiki, T., Tokunou, T., Iino, N., Takeshita, A., 2001. 15-Deoxy-Δ12,14-prostaglandin J2 and thiazolinediones activate the MEK/ERKpathway through phosphatidylinositol 3-kinase in vascular smooth musclecells. J. Biol. Chem. 276, 48950–48955.

Tang, H., McLachan, A., 2001. Transcriptional regulation of hepatitis B virus bynuclear hormone receptors is a critical determinant of viral tropism. Proc.Natl. Acad. Sci. U.S.A. 98, 1841–1846.


Tian, B., Zhang, Y., Luxon, B.A., Garofalo, R.P., Casola, A., Sinha, M., Brasier,A.R., 2002. Identification of NF-κB-dependent gene networks in respiratorysyncytial virus-infected cells. J. Virol. 76, 6800–6814.

Torrence, P.F., 2000. RSV infections: developments in the search for new drugs.Drug News Perspect. 13, 226–233.

Ueba, O., 1978. Respiratory syncytial virus: I. Concentration and purification ofthe infectious virus. Acta Med. Okayama 32, 265–272.

Wang, A.C.C., Dai, X., Luu, B., Conrad, D.J., 2001. Peroxisome proliferator-

activated receptor-γ regulates airway epithelial cell activation. Am. J.Respir. Cell. Mol. Biol. 24, 688–693.

Woerly, G., Honda, K., Loyens, M., Papin, J.P., Auwerx, J., Staels, B., Capron,M., Dombrowicz, D., 2003. Peroxisome–proliferator-activated receptorsalpha and gamma down-regulate allergic inflammation and eosinophilactivation. J. Exp. Med. 198, 411–421.

Zhao, X., Sullender, W.M., 2005. In vivo selection of respiratory syncytialviruses resistant to palivizumab. J. Virol. 79, 3962–3968.

peroxisome proliferator-activated receptor-γ agonists inhibit the replication of respiratory...

Documents