Expression and Localization of Peroxisome Proliferator-Activated Receptors and Nuclear Factor κB in Normal and Lesional Psoriatic Skin

ORIGINAL ARTICLE

Expression and Localization of Peroxisome Proliferator-Activated Receptors and Nuclear Factor jB in Normal and Lesional Psoriatic Skin Majken Westergaard, Jeanette Henningsen, Claus Johansen,w So¢e Rasmussen, Morten Lyhne Svendsen,w U¡e Birk Jensen,z Henrik Daa Schrder,y Bart Staels, Lars Iversen,w Lars Bolund,z Knud Kragballe,w and Karsten Kristiansen Department of Biochemistry and Molecular Biology and yDepartment of Pathology, University of Southern Denmark, Odense, Denmark; wDepartment of Dermatology, Marselisborg Hospital, and zDepartment of Human Genetics, University of Aarhus, Denmark; INSERM U545, Institut Pasteur de Lille and FaculteŁ de Pharmacie, UniversiteŁ Lille, France

Abnormal epidermal proliferation and di¡erentiation characterize the in£ammatory skin disease psoriasis. Here we demonstrate that expression of PPARd mRNA and protein is markedly upregulated in psoriatic lesions and that lipoxygenase products accumulating in psoriatic lesions are potent activators of PPARd. The expression levels of NF-jB p50 and p65 were not signi¢cantly altered in lesional compared with nonlesional psoriatic skin. In the basal layer of normal epidermis both p50 and p65 were sequestered in the cytoplasm, whereas p50, but not p65, localized to nuclei in the suprabasal layers, and this distribution was maintained in lesional psoriatic skin. In normal human keratinocytes PPAR agonists neither impaired IL-1b-induced translocation of p65 nor IL-1b-induced NF-jB DNA binding. We

show that PPARd physically interacts with the N-terminal Rel homology domain of p65. Irrespective of the presence of agonists none of the PPAR subtypes decreased p65-mediated transactivation in keratinocytes. In contrast p65, but not p50, was a potent repressor of PPAR-mediated transactivation. The p65-dependent repression of PPARd- but not PPARa- or PPARcmediated transactivation was partially relieved by forced expression of the coactivators p300 or CBP. We suggest that de¢cient NF-jB activation in chronic psoriatic plaques permitting unabated PPARd-mediated transactivation contributes to the pathologic phenotype of psoriasis. Key words: IjBa/immunohistochemistry/ keratinocytes/NF-jB p50/NF-jB p65. J Invest Dermatol 121:1104 ^1117, 2003

P

of psoriasis due to their pro-di¡erentiating and antiproliferative e¡ects on keratinocytes and their ability to modulate T cell in¢ltrates (Berth-Jones and Hutchinson, 1992; Itin et al, 1994). Several lines of evidence have indicated that members of the peroxisome proliferator-activated receptor (PPAR) and liver X receptor subfamilies of the nuclear hormone receptors are of importance for epidermal di¡erentiation and barrier formation (Hanley et al, 1997; 1998; 2000; Komuves et al, 1998; Rivier et al, 1998; 2000), and it has been suggested that PPARa plays an important role in lipid metabolism in reconstructed epidermis (Rivier et al, 2000). Interestingly, systemic administration of the PPARg-selective ligand troglitazone was reported to improve psoriasis and reverse the abnormal phenotype of transplanted psoriatic skin (Ellis et al, 2000). Three PPAR subtypes, a, d (b), and g have been identi¢ed, each characterized by distinct tissue distribution and functions (Willson et al, 2000). PPARs regulate gene expression by binding with their heterodimeric partner, retinoid X receptors, to speci¢c peroxisome proliferator response elements (PPRE) in responsive genes. Fatty acids and their derivatives, including hydroxylated eicosanoids and prostaglandins, as well as several synthetic ligands have been shown to activate the di¡erent PPAR subtypes with variable e⁄cacy and potency (Forman et al, 1995; Kliewer et al, 1995). All three PPAR subtypes are expressed in mouse epidermis during fetal development, but shortly after birth expression of PPARs become undetectable in the interfollicular epidermis

soriasis is an in£ammatory skin disease that a¡ects approximately 2% of the population in Europe and Northern America. Major characteristics of psoriasis are hyperproliferation and accumulation of numerous in£ammatory mediators in the involved areas (Hammarstrom et al, 1975; Kragballe and Fallon, 1986; Stern, 1997). Cytokines and in£ammatory mediators contribute to the establishment and maintenance of the pathologic phenotype of psoriasis lesions by promoting lymphocyte in¢ltration and activation, and by directly modulating keratinocyte proliferation and di¡erentiation (Uyemura et al, 1993; Terui et al, 2000). Classical nuclear hormone receptors such as the retinoid and vitamin D receptors are critically involved in the modulation of epidermal di¡erentiation and function (Pillai and Bikle, 1991; Li et al, 2000a). Accordingly, naturally occurring retinoids and vitamin D or synthetic analogs have been used widely in the treatment Manuscript received December 31, 2002; revised May 13, 2003; accepted for publication May 22, 2003 Reprint requests to: Karsten Kristiansen, Department of Biochemistry and Molecular Biology, University of Southern Denmark, Campusvej 55, DK-5230 Odense M, Denmark. Email: [email protected] Abbreviations: CBP, CREB-binding protein; EMSA, electrophoretic mobility shift assay; ERK, extracellular signal-regulated kinase; HODE, hydroxyoctadecadienoic acid; iNOS, inducible nitric oxide synthetase; NF-kB, nuclear factor kB; PPAR, peroxisome proliferator-activated receptor; PPRE, peroxisome proliferator-activated receptor response element.

0022-202X/03/$15.00 . Copyright r 2003 by The Society for Investigative Dermatology, Inc. 1104

VOL. 121, NO. 5 NOVEMBER 2003

(Michalik et al, 2001). Interestingly, expression of PPARa and PPARd is induced in response to skin injury, where PPARa was shown to be important during the initial in£ammatory phase and PPARd was shown to play a pivotal role for cell migration and survival during re-epithelialization (Michalik et al, 2001; Tan et al, 2001; Di-Po|« et al, 2002). In human skin all PPAR subtypes are expressed also in the adult interfollicular epidermis with PPARd being by far the most abundantly expressed PPAR subtype (Rivier et al, 1998; Matsuura et al, 1999; Westergaard et al, 2001). Furthermore, treatment of normal human keratinocytes in vitro with a PPARd-selective ligand induces expression of keratinocyte di¡erentiation marker genes suggesting that PPARd plays a role in di¡erentiation of human keratinocytes (Westergaard et al, 2001). The ubiquitously expressed transcription factor nuclear factor kB (NF-kB) is a key player in the regulation of immune and in£ammatory responses (Barnes and Karin, 1997; Baeuerle, 1998), and numerous ¢ndings have indicated that NF-kB also plays a pivotal role in epidermal biology (Kaufman and Fuchs, 2000; Fuchs and Raghavan, 2002). Direct or indirect disruption/modulation of components involved in NF-kB activation (Seitz et al, 1998; 2000a; 2000b; van Hogerlinden et al, 1999; Hu et al, 1999; Li et al, 1999; Takeda et al, 1999) was reported to exert a profound in£uence on epidermal proliferation, di¡erentiation, and/or homeostasis, and in the absence of proper activation of NF-kB, terminal di¡erentiation was perturbed. Interestingly, mice with an epidermis-speci¢c deletion of IKK2 (IKKb) were born with a histologically normal skin, but epidermal hyperplasia and in£ammation with impaired di¡erentiation developed within the ¢rst week after birth (Pasparakis et al, 2002). The epidermal hyperplasia was shown not to result from cell-autonomous hyperproliferation of keratinocytes, but appeared to be dependent on a reaction driven by the innate immune system (Pasparakis et al, 2002). In addition, the recent and surprising ¢nding that IKK1 (IKKa) regulates epidermal di¡erentiation in a manner independently of NF-kB via production of a secreted di¡erentiation-inducing factor (Hu et al, 2001) indicates that the way that components of the NF-kB systems coordinate cell cycle withdrawal, di¡erentiation, and protection from apoptosis is far more complex than previously anticipated. Because recent evidence points to a role of PPARd in cellular proliferation (He et al, 1999; Jehl et al, 2000; Hansen et al, 2001) and an earlier report (Rivier et al, 1998) suggested that expression of PPARd was increased in lesional psoriatic skin, we decided to investigate whether expression of the di¡erent PPAR subtypes was aberrantly regulated in hyperproliferative lesional psoriatic skin. In addition, as substantial evidence suggests that at least PPARa and PPARg, apart from playing important roles in lipid homeostasis, are involved in the modulation of in£ammatory responses via negative cross-talk with transcription factors such as NF-kB and activating protein 1 (Devchand et al, 1996; Chinetti et al, 1998; Jiang et al, 1998; Ricote et al, 1998; Staels et al, 1998; Delerive et al, 1999; Su et al, 1999; Straus et al, 2000), we investigated the interplay between PPAR and NF-kB in keratinocytes. MATERIALS AND METHODS Biopsies Keratome biopsies from lesional and nonlesional psoriatic skin in patients with psoriasis vulgaris were obtained as described previously (Fogh et al, 1989). In addition, excision biopsy specimens from lesional psoriatic skin before and after treatment at the Dead Sea were obtained. Biopsies were taken in local anesthesia 1% lidocain cum epinephrine. Protocols were approved by the local ethical committee before biopsies were taken and written consent was provided by patients before experiments were initiated. Cell culture and interleukin 1b (IL-1b) stimulation Keratinocytes from adult humans (controls) were isolated from human skin obtained after plastic surgery. First passage keratinocytes were grown in Keratinocyte-SFM (Gibco BRL/Life Technologies, Gaithersburg, MD)

PPARs AND NF-kB IN NORMAL AND PSORIATIC SKIN

1105

supplemented with 50 mg per ml bovine pituitary extract, 5 ng per ml human recombinant epidermal growth factor, and 5 mg per ml gentamycin and replated in 75 cm2 culture £asks (Nunc, Roskilde, Denmark). Cells were grown at 5% CO2 and 371C. In experiments including stimulation with IL-1b, pretreatment with Keratinocyte-SFM containing selective PPAR ligands was initiated when the cultures were at approximately 30% con£uency. Control cells not treated with PPAR ligands received similar volumes of vehicle (0.1% dimethylsulfoxide (DMSO)). Cells were incubated for 24 h (for electrophoretic mobility shift assay (EMSA) and immunostaining) or 2 h (for RT-PCR) with the ligands before the addition of IL-1b (10 ng per ml). The cells were harvested 30 min (for EMSA and immunostaining) or 3 h (for RT-PCR) after addition of IL-1b. HaCaT cells were cultured in Dulbecco’s modi¢ed Eagle’s medium (Gibco BRL/Life Technologies) with 10% fetal bovine serum and antibiotics (100 U per mL penicillin, 1 mg per mL streptomycin sulfate) in a humidi¢ed atmosphere of 5% CO2 at 371C. Medium was changed every second day. BRL49653 was kindly provided by Dr J. Fleckner (Novo Nordisk, Bagsv
rd, Denmark) and L165041 was kindly provided by Dr D.E. Moller (Merck Research Laboratories, Rahway, NJ). Wy14643 was purchased from Calbiochem (San Diego, CA). Plasmids pcDNA3.1( þ )hPPARa, pcDNA3.1( þ )hPPARd, and pcDNA3.1( þ )hPPARg1 were constructed as described previously (Westergaard et al, 2001). The pcDNA3.1( þ )hPPARa(DAB) expression plasmid was constructed by ampli¢cation of the hPPARa cDNA from pCR2.1hPPARa by PCR using XbaI/BamHI tagged primers (upstream primer, 50 -AATTGGATCCACCATGGGAGCATTGAACATCGAATG-30; downstream primer, 50 -CTAGTCTAGAAACTCAGTACACGTCCCTG30) and inserted into the XbaI/BamHI sites of pcDNA3.1( þ ). The pcDNA3.1( þ )hPPARd(DAB) expression plasmid was constructed by ampli¢cation of the hPPARd cDNA from pCR2.1hPPARd by PCR using XbaI/BamHI tagged primers (upstream primer, 50 -AATTGGATCCACCATGGGCAGCCTCAACATGGAGTGC-30; downstream primer, 50 -CTAGTCTAGACCGTTAGTACATGTCCTTG-30) and inserted into the XbaI/BamHI sites of pcDNA3.1( þ ). pcDNA3.1( þ )hPPARg(DAB) expression plasmid was constructed by ampli¢cation of the hPPARg cDNA from pCR2.1hPPARg by PCR using Asp718/XbaI tagged primers (upstream primer, 50 -CTAGTCTAGACTGCTAGTACA-AGTCCTTG-30; downstream primer, 50 -CTAGTCTAGACCGTTA-GTACATGTCCTTG30) and inserted into the Asp718/XbaI sites of pcDNA3.1( þ ). The pM1hPPARa(LBD)/Gal4(DBD), pM1-hPPARd(LBD)/Gal4(DBD), and pM1hPPARg(LBD)/Gal4(DBD) constructs were kindly provided by Dr J. Fleckner. The PPREx3 -tk-luc reporter containing three copies of the PPRE from the acyl-CoA oxidase promoter and the UASx4luc reporter were kindly provided by Dr R.M. Evans (Howard Hughes Medical Institute, La Jolla, CA). CMV-b-galactosidase (Clontech, Palo Alto, CA) and the SV40-b-galactosidase (Promega, Madison, WI) vectors were used for normalization. The reporter construct containing three tandem copies of the kB site (p(kB)3 -luc þ , Stratagene, La Jolla, CA) has been described previously (Delerive et al, 1999). p65-CMV2-Flag and p50-CMV2-Flag were kindly provided by Dr A. Baldwin Jr (University of North Carolina at Chapel Hill, NC). pGEX, pGEXp65(1^305), and pGEXp65(354^551) used in pull-down experiments were kindly provided by Dr H. Sakurai (Tanabe Seiyaku, Japan) and have been described elsewhere (Sakurai et al, 1999). Transactivation assays All transfections were carried out in 12-well plates, and the amounts given refer to the amount of plasmid added per well. Transfection studies with full-length hPPAR were performed in HaCaT cells as described elsewhere (Westergaard et al, 2001) using the BES-CaCl2 method. Transfections with hPPAR(DAB) were carried out using 0.75 mg PPREx3 -tk-luc in combination with expression vectors for hPPARa (0.01 mg), hPPARd (0.05 mg), hPPARg1 (0.05 mg), hPPARa(DAB) (0.01 mg), hPPARd(DAB) (0.05 mg), or hPPARg(DAB) (0.05 mg). When using PPAR-Gal4 fusion constructs, 0.75 mg UASx4luc was used in combination with hPPARa(LBD)/Gal4(DBD) (0.125 mg), hPPARd(LBD)/ Gal4(DBD) (0.125 mg), or hPPARg1(LBD)/Gal4(DBD) (0.125 mg). Vehicle (0.1% DMSO) or ligands were added as indicated. 8(S)HETE, LTB4, 12(S)HETE, 12(R)HETE, 15(S)HETE, and 13(S)HODE were obtained from Cayman Chemical (Ann Arbor, MI). Normal human keratinocytes were transiently transfected using E¡ectene (Qiagen, Valencia, CA). Two series of transfection studies with hPPARa, hPPARd, or hPPARg1, and p65 of NF-kB were performed using reporter constructs containing (1) three tandem copies of the kB binding site (p(kB)3 -luc þ ) or (2) three tandem copies of the PPRE (PPREx3 -tk-luc), as stated in the ¢gure legends. Transfection studies using the reporter construct p(kB)3 -luc þ (50 ng) were performed in the presence of 6 ng p65 in combination with 18 and 60 ng hPPARa, hPPARd, or hPPARg1. 110 ng of the SV40-b-

1106

WESTERGAARD ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

galactosidase construct were included for normalization. Transfections with the PPAR-responsive promoter construct were carried out using 160 ng PPREx3 -tk-luc in the presence of 0.625 ng hPPARa, 62.5 ng hPPARd, or 62.5 ng hPPARg1 in combination with 0.1, 0.5, 2.5, and 12.5 ng p65 expression vector. 110 ng of the SV40-b-galactosidase construct were added for normalization. Transfection studies with the PPAR and the p50 construct were performed using the following combinations of PPREx3 tk-luc, hPPAR constructs, and p50 expression vector: hPPARa (0.625 ng), PPREx3 -tk-luc reporter (160 ng), and p50 (0.1, 0.5, and 2.5 ng); hPPARd (20 ng), PPREx3 -tk-luc reporter (130 ng), and p50 (3.2, 16, 80 ng); or hPPARg1 (40 ng), PPREx3 -tk-luc reporter (102.5 ng), and p50 (6.4, 32, 160 ng). 110 ng of the SV40-b-galactosidase construct were added for normalization. In all transfections empty expression vector was added to ensure equal promoter load. Six hours following transfection fresh medium was added and cells were subsequently incubated for about 20 h in Keratinocyte-SFM containing either vehicle (0.1% DMSO), 10 mM Wy14643, 5 nM L165041, or 1 mM BRL49653. In the IL-1b stimulation experiments, normal human keratinocytes were transfected with 75 ng p(kB)3 -luc þ and 225 ng SV40-b-galactosidase construct. Seven hours following transfection the cells were pretreated with vehicle (0.1% DMSO), 5 nM L165041, or 1 mM BRL49653 for 2 h and subsequently incubated for another 3 h with IL-1b (10 ng per mL). In the coactivator titration experiment 293 cells were transiently transfected using the calcium phosphate coprecipitation protocol. 750 ng PPREx3 -tk-luc reporter were used with di¡erent combinations of expression vectors: hPPARa (1.25 ng), hPPARd (62.5 ng), hPPARg1 (62.5 ng), 12.5 ng p65 as indicated, and CPB or p300 (125, 375, 1125 ng). 50 ng of the SV40-bgalactosidase construct were cotransfected for normalization. Medium was changed after 5 h, and the cells were subsequently incubated for approximately 16 h with medium containing Wy14643 (30 mM), L165041 (500 nM), BRL49653 (1 mM), or vehicle (0.1% DMSO). The luciferase and b-galactosidase activities in cell lysates were determined by standard techniques. Luciferase activities were normalized to b-galactosidase activities. All transfections were performed in duplicate or triplicate, measured in duplicate, and repeated two to four times.

Western blotting Whole cell extracts were prepared using the TRIZOL reagent according to the recommendation of the manufacturer (Gibco BRL/Life Technologies). Blotting, enhanced chemiluminescence, and stripping of membranes were performed as described previously (Westergaard et al, 2001). For separation of PPARd, PPARg1, and PPARg2, 40 mg of protein were separated in urea-containing minigels. Detection of p50, p65, and IkBa in whole cell extracts was done by separating 30 mg of protein in standard minigels. Cytoplasmic IkBa was detected by separating 20 mg of protein on 8:16% gradient gels (Novex, Frankfurt, Germany). Primary antibodies used were as follows: mouse anti-PPARg (sc-7273, 1:1000) (this antibody is raised against the human PPARg C-terminus, but recognizes all three PPAR subtypes), rabbit antiIkBa (sc-1643, 1:200), rabbit anti-NF-kB (p65) (sc-109, 1:200), rabbit antiNF-kB (p50) (sc-7178, 1:500), rabbit anti-TATA binding protein (TBP, sc273, 1:2000), and rabbit anti-TFIIB (sc-225, 1:2000), all Santa Cruz Biotechnology, Santa Cruz, CA. Secondary antibodies were horseradish peroxidase conjugated antimouse or antirabbit IgG antibodies (DAKO, Glostrup, Denmark), diluted 1:2000.

RNA puri¢cation TRIZOL reagent was used for the isolation of total RNA as recommended by the manufacturer (Gibco BRL/Life Technologies). The integrity of all RNA samples was con¢rmed by electrophoresis in denaturing formaldehyde-containing gels.

In vitro protein^protein interaction assay (GST pull-down) GST pull-down assays were performed essentially as described previously (Kussmann-Gerber et al, 1999). Brie£y, for expression of the GST fusion proteins Escherichia coli BL21-GROESL harboring the pGEX, pGEXp65(1^305), or pGEXp65(354^551) expression constructs were used. The GST fusion protein bound to glutathione-sepharose 4B beads was incubated with 8 mL of in vitro translated 35S-methionine-labeled protein in the presence of L165041 (1 mM) dissolved in DMSO or DMSO alone in a total volume of 100 mL of incubation bu¡er (20 mM Tris
HCl pH 8.0, 100 mM NaCl, 10 mM ethylenediamine tetraacetic acid, 0.5% Nonidet P- 40, 1 mM Dithioerythritol (DTE), 0.5 mM phenylmethylsulfonyl £uoride, 1% skimmed milk, and protease inhibitors) and rotated for 2 h at 41C. Following centrifugation the beads were washed four times in incubation bu¡er with skimmed milk and once in incubation bu¡er without skimmed milk. Beads were resuspended in 100 mL sodium dodecyl sulfate (SDS) sample bu¡er, boiled for 5 min, and centrifuged. The supernatant was loaded on an SDS-polyacrylamide gel. After drying, the gels were exposed to a Phosphoimager screen and analyzed using the ImageQuant program version 5.

Multiplex RT-PCR Multiplex RT-PCR was carried out as described previously (Westergaard et al, 2001). All primer sets used have been described previously (Westergaard et al, 2001) with the exception of primer sets matching the inducible nitric oxide synthetase (iNOS) and IL-6 genes: iNOS upstream primer, 50 -CAGCGCTACAACATCCTGGAGG30; iNOS downstream primer, 50 -GAGGGACCAGCCAA-ATCCAGTC-30 (iNOS, 30 cycles, 241 bp); IL- 6 upstream primer, 50 -CCTTCGGTCCAGTTGCCTTCTC-30; IL- 6 downstream primer, 50 -TTTTCTGCCAGTGCCTCTTTGC-30 (IL- 6, 25 cycles, 241 bp). All reactions contained the TBP primer set as an internal standard together with one additional primer set. Preparation of cytoplasmic and nuclear extracts and EMSA Preparation of cytoplasmic and nuclear extracts from normal human keratinocytes, treated or not with IL-1b and PPAR agonists as indicated, and EMSA were performed as described previously (Johansen et al, 2000). The NF-kB consensus oligonucleotide (50 -AGTTGAGGGGACTTTCCCAGGC-30) and the TFIID consensus oligonucleotide (50 GCAGAGCATATAAGGTGAGGTAGGA-30) were labeled as described by the manufacturer (Promega, Madison, WI). Labeled oligonucleotides were puri¢ed on a Nick Spin column (G-50, Pharmacia, Uppsala, Sweden). Production of anti-PPARa and anti-PPARd antibodies For production of the PPARa polyclonal antibody, 500 mg puri¢ed Histagged PPARa was mixed with 250 mL incomplete Freund’s adjuvant and injected subcutaneously into a rabbit. Injections were repeated six times with intervals of 14 d. Anti-PPARa antibodies were a⁄nity puri¢ed from the antiserum. In short, immune serum was absorbed against GST immobilized on glutathione-sepharose beads, and anti-PPARa antibodies were subsequently a⁄nity-puri¢ed by binding to a GST-PPARa-A/B domain fusion protein immobilized on CNBr-activated sepharose. Bound antibodies were eluted with 0.1 mM glycine
HCl (pH 2.8) and immediately neutralized in phosphate-bu¡ered saline (pH 7.6). The speci¢city of the a⁄nity-puri¢ed anti-PPARa antibodies was assessed by western blotting. Production and puri¢cation of polyclonal PPARd antibody was performed as described elsewhere (Westergaard et al, 2001).

Immunostaining Immunostaining of NF-kB (p65 or p50) in normal human keratinocytes was performed as described previously (Westergaard et al, 2001). The antibodies used were rabbit anti-NF-kB (p65) antibody (sc109, 1:100), rabbit anti-NF-kB (p50) (sc-7178, 1:500), and Alexa 488 conjugated goat antirabbit IgG (1:400) (Molecular Probes, Eugene, OR). Nuclei were visualized by staining with Hoechst 33258 (Molecular Probes). Immunohistochemistry Immunohistochemistry was performed as described previously (Westergaard et al, 2001). Detection of PPARa, PPARd, and PPARg was performed as described by Westergaard et al (2001) using a⁄nity-puri¢ed PPARa- and PPARd-speci¢c antibodies and a commercially available PPARg antibody (sc- 6285), respectively. To detect NF-kB (p65 or p50), sections were microwaved for 15 min in TEG bu¡er. Blocking was performed overnight and sections were incubated with antibodies at 41C overnight. The antibodies used were rabbit anti-NF-kB (p65) antibody (sc-109, 1:500) and rabbit anti-NF-kB (p50) (sc-7178, 1:500).

Quantitation and statistical analysis TBP was used as the internal standard for multiplex RT-PCR and expression from individual genes was normalized to TBP. Quantitation was performed by phosphorimaging using the ImageQuant program version 5 (Molecular Dynamics). For western blotting TFIIB was used as the internal standard and the level of individual proteins was normalized to TFIIB. Quantitation was performed by laser scanning densitometry using a Ultroscan XL 2222 and the GelScan XL software package (Pharmacia LKB). Results are expressed as mean7SD. For statistical analysis a one-way ANOVA was performed followed by a Student t test. A probability of po0.05 was considered as statistically signi¢cant.

RESULTS Expression of PPAR in nonlesional and lesional psoriatic skin We have shown that PPARd may play a role in control of proliferation of basal cells and suggested that it serves as a molecular switch involved in the initiation of keratinocyte di¡erentiation (Westergaard et al, 2001). In addition, it was reported that expression of PPARd mRNA tended to be

VOL. 121, NO. 5 NOVEMBER 2003

upregulated in psoriatic skin (Rivier et al, 1998). Therefore, we compared expression of the PPARs in nonlesional and lesional skin in psoriasis patients. The skin samples were obtained either as keratome biopsies of lesional and nonlesional skin in patients with untreated psoriasis vulgaris, or as excision biopsies of lesional psoriatic skin in patients before and after treatment at the Dead Sea. Total RNA was isolated and expression of PPAR mRNA was determined by RT-PCR (Fig 1). These analyses revealed a 3^ 5-fold upregulation of PPARd mRNA expression in lesional psoriatic skin compared to nonlesional psoriatic skin (3.7fold70.67, po0.01) (Fig 1a). Similarly, robust upregulation of PPARd mRNA expression was observed in RNA isolated from biopsies of untreated lesional skin compared to post-treatment biopsies (3.2-fold70.54, po0.01) (Fig 1b). No consistent changes in the levels of PPARa mRNA were observed. The level of PPARg mRNA tended to be slightly elevated in lesional psoriatic skin. The levels of PPARg mRNA in all samples were close to the detection limit, however. In untreated lesional psoriatic skin the level of PPARg mRNA was higher in excision biopsies than in keratome biopsies. This di¡erence may be due to contaminating dermal tissue in the excision biopsies. This may also explain the very low level of aP2 mRNA expression observed in excision biopsies. The expression of PPAR proteins was examined in lesional psoriatic skin by western blotting. As shown in Fig 1(c), 1(d), paralleling the results obtained by RT-PCR (Fig 1a, b), a 4^5fold upregulation of the PPARd protein was observed in lesional psoriatic skin compared to nonlesional psoriatic skin (in the matched pairs of lesional and nonlesional skin 4.4 -fold70.87, po0.01; in the biopsies from pre- and post-treatment patients 4.1fold70.73, po0.01). In four of seven biopsy sets tested, we also observed slightly elevated levels of the PPARg protein. The level of PPARa was below the detection limit using this particular antibody. Immunohistochemical analyses of PPAR in nonlesional and lesional psoriatic skin The distribution of PPARa, PPARd, and PPARg in normal skin and nonlesional and lesional

Figure 1. Expression of PPAR subtypes in lesional and nonlesional psoriatic skin. Total RNA and protein were prepared from (a), (c) biopsies of lesional and nonlesional psoriatic skin from patients with psoriasis vulgaris or (b), (d) biopsies of psoriasis patients before and after treatment at the Dead Sea. (a), (b) Multiplex RT-PCR analysis of the expression of PPARa, PPARd, PPARg1, PPARg2, and aP2. TBP served as an internal control. (c), (d) Expression of PPARd and PPARg was analyzed by western blotting. Forty micrograms of protein were loaded in each lane. Protein extracts from 293 cells overexpressing PPARg2 or PPARd were used as positive controls. The antibody used for detection (sc-7273) is directed against the C-terminal region of human PPARg, but recognizes also PPARd (and PPARa). Expression of endogenous PPARd is clearly detectable in 293 cells. Equal loading/transfer was con¢rmed by amido-black staining of membranes and by immuno-detection of TFIIB.

PPARs AND NF-kB IN NORMAL AND PSORIATIC SKIN

1107

psoriatic skin was determined by immunohistochemical analyses (Fig 2). Most available PPAR antibodies exhibit subtype crossreactivity. Consequently, we tested a battery of di¡erent antiPPAR antibodies against 293 cells overexpressing each of the human PPAR subtypes. Using this test the previously described anti-PPARd antibody (Westergaard et al, 2001) was shown to speci¢cally recognize PPARd. In addition, we found that a polyclonal rabbit anti-PPARa antibody generated by immunization with puri¢ed his-tagged full-length PPARa speci¢cally recognized PPARa in immunohistochemistry. Finally, the commercially available anti-PPARg antibody sc- 6285 (Santa Cruz Biotechnology) was found to be PPARg speci¢c in immunohistochemistry (results not shown). In normal skin, PPARa and PPARg exhibited a clear cytoplasmic localization in the basal layer. In stratum spinosum few cells exhibited nuclear localization of PPARa and PPARg, whereas in stratum granulosum PPARa and PPARg predominantly localized to nuclei. PPARd was strongly expressed in basal and suprabasal layers. Interestingly, in the basal layer, PPARd intracellular localization varied in that PPARd exhibited a cytoplasmic localization in a subset of cells and was present in the nuclei of the remaining cells. In suprabasal cells, PPARd localized predominantly to the nuclei both in stratum spinosum and in stratum granulosum. The intracellular distribution of the PPAR in nonlesional skin from psoriasis patients resembled normal skin. In the thickened epidermis of lesional psoriatic skin, the cytoplasmic localization of PPARa and PPARg extended through most of the suprabasal layers with only the outermost layer of suprabasal cells exhibiting a nuclear localization. In contrast, PPARd localized to nuclei throughout the entire suprabasal part of lesional psoriatic skin, but maintained the mixed cytoplasmic and nuclear localization in the basal layer. Activation of PPAR subtypes by eicosanoids accumulating in lesional psoriatic skin Psoriatic skin lesions are characterized by the accumulation of numerous lipid chemoattractants (in£ammatory mediators) comprising di¡erent classes of oxygenated fatty acid derivates, in particular lipoxygenase

1108

WESTERGAARD ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure 2. Immunohistochemical detection of PPARs in normal, lesional, and nonlesional psoriatic skin. Punch biopsies from lesional and nonlesional psoriatic skin were analyzed for expression and localization of PPARa, PPARd, and PPARg using subtype-speci¢c anti-PPAR antibodies as described in Materials and Methods. All sections were counterstained with hematoxylin.

products. Thus, LTB4, the 5-, 8-, 9-, 11-, 12-, 15-hydroxyeicosatetraenoic acid (HETE), and 9- and 13 hydroxyoctadecadienoic acid (HODE) have been reported to accumulate in psoriatic skin with 12-HETE, 9-HODE, and 13 HODE being the most abundant (Camp et al, 1983; Fogh et al, 1989). It has previously been shown that several of these lipoxygenase products are able to enhance PPAR-mediated transactivation (Yu et al, 1995; Devchand et al, 1996; Forman et al, 1997; Kliewer et al, 1997; Krey et al, 1997), and recently it was shown that PPAR-mediated keratinocyte di¡erentiation was impaired by lipoxygenase inhibitors (Thuillier et al, 2002). Consequently, we examined the ability of the abundant lipoxygenase products to activate the individual human PPAR subtypes. As ligand selectivity and e⁄cacy may exhibit cell-type speci¢city, ligand-dependent transactivation was tested in the spontaneously transformed human keratinocyte HaCaT cell line, previously shown to be an appropriate substitute for normal human keratinocytes in transfection analyses (Westergaard et al, 2001). Similar to normal human keratinocytes, HaCaT cells have a high basal level of activated extracellular signal-regulated kinase 1 (ERK1) and ERK2 resulting in a blunted ligand-dependent activation of PPARg (Westergaard et al, 2001). In agreement with previous observations, ligands only enhanced transactivation by full-length hPPARa or hPPARa(DAB) 2^3 -fold (Westergaard et al, 2001), whereas ligand-dependent activation of the hPPARa(LBD)/Gal4(DBD) fusion was much more pronounced (Fig 3). All the tested HETEs modestly

enhanced transactivation by full-length hPPARa and hPPARa(DAB). As reported earlier (Yu et al, 1995; Forman et al, 1997; Kliewer et al, 1997; Krey et al, 1997), 8(S)-HETE was found to be a potent PPARa agonist. In addition, 12(R)-HETE and 15(S)-HETE markedly enhanced hPPARa(LBD)/Gal4(DBD)mediated transactivation. Interestingly, 13(S)-HODE potently enhanced transactivation by full-length PPARa without having any e¡ect on transactivation mediated by hPPARa(DAB) and hPPARa(LBD)/Gal4(DBD). Activation of the ERK-mitogenactivated protein (ERK-MAP) kinases a¡ects PPAR-mediated transactivation in part by phosphorylation of target residues in the N-terminal A/B domains (Hu et al, 1996; Shalev et al, 1996; Adams et al, 1997; Camp and Tafuri, 1997; Camp et al, 1999; JugeAubry et al, 1999; Barger et al, 2000). Phosphorylation of PPARa has been reported to enhance (Shalev et al, 1996; Juge-Aubry et al, 1999) or repress (Barger et al, 2000) PPARa-mediated transactivation. Similarly, 13(S)-HODE has been reported to activate the ERK-MAP kinases (Hsi et al, 2001) or to repress activation of the ERK2-MAP kinase (Mani et al, 1998). Hence, it is possible that the 13(S)-HODE-induced transactivation by fulllength hPPARa is related to an e¡ect of 13(S)-HODE on ERKMAP kinase activation rather than 13(S)-HODE being a bona ¢de hPPARa ligand, or alternatively that a combination of the two properties dictates the e¡ect of 13(S)-HODE on PPARamediated transactivation. PPARd was e¡ectively activated by 8(S)-, 12(S)-, 12(R)-, and 15(S)-HETE. In addition, 13(S)-HODE was a weak activator,

VOL. 121, NO. 5 NOVEMBER 2003

PPARs AND NF-kB IN NORMAL AND PSORIATIC SKIN

1109

Figure 3. Di¡erential activation of PPARa, PPARd, and PPARc by eicosanoids present in lesional psoriatic skin. The subtype selectivity of the selected eicosanoids was determined by transient transfections in HaCaT cells using either vectors encoding full-length PPAR, PPAR with deleted AB domains, or Gal4 -PPAR(LBD) fusions with the appropriate luciferase reporter constructs. For analysis using full-length or PPAR with deleted AB domains the following combinations (given per well) were used: hPPARa (0.01 mg), hPPARd (0.05 mg), hPPARg1 (0.05 mg), hPPARa(DAB) (0.01 mg), hPPARd(DAB) (0.05 mg), or hPPARg(DAB) (0.05 mg) with PPREx3 -tk-luc reporter (0.75 mg) and a CMV-b-galactosidase construct (0.1 mg). For analysis using Gal4 fusions the following combinations (given per well) were used: hPPARa(LBD)/Gal4(DBD) (0.125 mg), hPPARd(LBD)/Gal4(DBD) (0.125 mg), or hPPARg(LBD)/Gal4(DBD) (0.125 mg) with UASx4 -tk-luc (0.75 mg) and a CMV-b-galactosidase construct (0.1 mg). In all transfections, empty expression vector was added to ensure equal promoter load. Cells were incubated with medium containing 5 mM of 8(S)-HETE, 12(S)-HETE, 12(R)-HETE, 15(S)HETE, 13(S)-HODE, LTB4, or 100 mM Wy14643, 1 mM BRL49653, 10 nM L165041, or vehicle (0.1% DMSO) for approximately 16 h prior to harvest. Reporter activity was normalized to b-galactosidase values. All transfections were performed independently a minimum of four times in duplicate. Results are presented as mean7SD. An asterisk indicates signi¢cant deviation from the vehicle-treated controls (po0.05).

whereas LTB4 did not enhance PPARd-mediated transactivation. Finally, 8(S)-, 12(S)-, 12(R)-, and 15(S)-HETE, and LTB4 modestly induced PPARg1-mediated transactivation. Similar results were obtained using PPARg2 (results not shown). As noted above, the high levels of activated ERK-MAP kinases blunted ligand-induced activation of the full-length PPARg, and deletion of the AB domain enhanced ligand-induced transactivation. Ligand-dependent PPAR-mediated transactivation depends on interaction with cofactors (Xu et al, 1999). We examined the expression of a panel of known PPAR coactivators, p300, CREB-binding protein (CBP), and steroid receptor coactivator 1, and corepressors, silencing mediator for retinoid and thyroid hormone receptors and nuclear receptor corepressor, to determine whether their expression was altered in lesional psoriatic skin compared to nonlesional psoriatic skin. No signi¢cant changes in the expression of mRNA encoding these coactivators and corepressors were found (results not shown). Recently, we showed that addition of the PPARd-selective agonist L165041 to normal human keratinocytes induced expression of the endogenous CD36/fatty acid transporter (FAT) gene (Westergaard et al, 2001). The ability of PPARd-selective ligands to induce CD36 expression in normal keratinocytes combined with the increased levels of PPARd and PPARd ligands in lesional psoriatic skin suggested that CD36 expression might be elevated in lesional psoriatic skin. In agreement with this notion, multiplex RT-PCR revealed a pronounced upregulation of CD36 mRNA expression in the biopsies of lesional skin compared with matched biopsies of nonlesional skin (8.1-fold73.4, po0.004) (Fig 4a). In the biopsies collected

before and after treatment at the Dead Sea a less pronounced but still signi¢cant upregulation was also observed (2.4 fold70.84, po0.002) (Fig 4b). For comparison, expression of a known marker for psoriatic lesional skin, E-FABP, was also analyzed. Re£ecting the expression pattern of CD36, we found a greater upregulation of E-FABP mRNA expression in the biopsies of lesional skin compared with the matched biopsies of nonlesional skin (3.6 -fold70.86, po0.004) (Fig 4a) than in the biopsies collected before and after treatment at the Dead Sea (2.2-fold70.63, po0.002) (Fig 4b). The upregulation of CD36 expression in lesional psoriatic skin is compatible with an enhanced PPARd-dependent induction of CD36 expression. As the HETEs and HODEs that accumulate in psoriatic lesional skin are also potent activators of the other PPAR subtypes (Fig 3), however, contribution from PPARa and PPARg cannot be excluded. It is also evident that numerous other factors may play important roles in the upregulation of CD36 expression in psoriatic lesional skin. Analysis of NF-jB expression and intracellular localization NF-kB has been implicated as a central player in the control of epidermal proliferation and di¡erentiation, and based on the phenotypes of transgenic knockout mice it was suggested that NF-kB is involved either in the initiation of the di¡erentiation process or in the maintenance of cell cycle withdrawal and the di¡erentiated state (Klement et al, 1996; Hu et al, 1999; Li et al, 1999; Makris et al, 2000; Schmidt-Supprian et al, 2000). In addition, IKKa appears to be critically involved in epidermal di¡erentiation in a NF-kB-independent manner (Hu et al, 2001).

1110

WESTERGAARD ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure 4. Elevated expression of CD36 mRNA in lesional psoriatic skin. Total RNA was prepared from (a) biopsies of lesional and nonlesional psoriatic skin from patients with psoriasis vulgaris or (b) biopsies from psoriasis patients before and after treatment at the Dead Sea. The expression of CD36 (FAT) was determined by multiplex RT-PCR. The expression of epidermal fatty acid binding protein (E-FABP) served as a positive control for lesional psoriatic skin (Madsen et al, 1991), and expression of TBP served as an internal control.

No comprehensive analysis of NF-kB expression and intracellular localization in normal, nonlesional, and lesional psoriatic skin has been presented, and given the reported e¡ects of PPARs and PPAR ligands on NF-kB-dependent processes, we analyzed whether NF-kB expression and intracellular localization were altered in lesional psoriatic skin compared with nonlesional and normal skin. NF-kB in skin consists predominantly of p50/p50 homodimers or p50/p65 heterodimers, with p50/p50 homodimers prevailing (Budunova et al, 1999; results not shown). Western blotting of whole cell extracts revealed no signi¢cant overall alterations in the expression of the p65 and p50 subunits of NF-kB and IkBa in lesional psoriatic skin compared with nonlesional skin. By densitometric scanning the ratio between expression in lesional and nonlesional skin was determined to be 0.9670.10 for p50, 1.1070.39 for p65, and 1.0370.23 for IkBa (Fig 5). Immunohistochemical analyses revealed that normal interfollicular human skin displays a characteristic intracellular distribution of p50 and p65 (Fig 6). In the basal layer there is a cytoplasmic accumulation of p50 with few cells exhibiting nuclear localization of p50. In the suprabasal layers most cells contain nuclear p50. In contrast, p65 exhibits exclusively a cytoplasmic localization in both the basal layer and the suprabasal layers. In total, specimens from six di¡erent individuals with no history of psoriasis were analyzed, and all displayed this characteristic distribution. Next specimens of skin from ¢ve patients with chronic psoriasis were analyzed. In these ¢ve specimens we found virtually the same intracellular distribution of p50 and p65 in nonlesional skin. In lesional skin strong nuclear staining of p50 was observed in the outermost suprabasal layer whereas fewer cells in the lower parts of the suprabasal layer exhibited a nuclear localization of p50. A priori we would have expected a nuclear localization of p65 in lesional skin due to in£ammation and accumulation of proin£ammatory cytokines known to induce nuclear translocation of NF-kB. In none of the ¢ve specimens did we observe nuclear accumulation of p65 in lesional skin, however. Thus, our results clearly demonstrate that persistent nuclear localization of p65 is not a hallmark of psoriasis. This conclusion was corroborated by gel shift analyses showing that nuclear extracts from normal, nonlesional, and lesional psoriatic skin only revealed DNA binding of the p50/p50 homodimer (Johansen et al, manuscript in preparation). Physical and functional interaction between PPAR and NF-jB Substantial evidence has accumulated that PPARa and PPARg or their ligands directly or indirectly are involved in the modulation of NF-kB activity (Poynter and Daynes, 1998; Delerive et al, 1999; 2000; Su et al, 1999; Chung et al, 2000; Li et al, 2000b). PPARd expression is upregulated in psoriatic lesions, and even though PPARd predominantly localized to nuclei in the

Figure 5. Analysis of NF-jB and IjBa expression in lesional and nonlesional psoriatic skin. Whole cell protein extracts were prepared from lesional and nonlesional psoriatic skin from patients with psoriasis vulgaris, and analyzed for expression of the NF-kB subunits p65 and p50 and IkBa by Western blotting. Thirty micrograms of total protein were applied per lane and equal loading/transfer was con¢rmed by amido-black staining of membranes and by immuno-detection of TFIIB.

suprabasal layer of lesional psoriatic skin, clear cytoplasmic staining was also observed (Fig 2). This fact combined with the ¢ndings that several of the hydroxylated fatty acids that accumulate in psoriatic lesions are potent PPARd agonists prompted us to investigate whether PPARd like PPARa and PPARg interacts with p65 and whether the accumulation of PPARd ligands might interfere with cytokine-induced nuclear accumulation of NF-kB. To address these questions we ¢rst employed GST pull-down experiments to determine if PPARd also interacts with the p65 subunit of NF-kB. Figure 7 shows that full-length PPARd interacted strongly and speci¢cally with the N-terminal part of p65 containing the Rel homology region mediating DNA binding, dimerization, and IkBa interaction. Furthermore, deletion of the AB domain of PPARd did not impair interaction with p65. Interaction was ligand independent (results not shown). Thus it appears that all PPAR subtypes are able to interact directly with p65. Next we investigated whether PPAR agonists a¡ected IL-1binduced translocation and DNA binding of NF-kB human keratinocytes (Fig 8). In growing nontreated cells p50 (Fig 8a) and p65 (Fig 8b) exhibited a cytoplasmic localization similar to that observed in proliferating basal cells. IL-1b treatment induced nuclear translocation of both p50 and p65. Preincubation with PPARa-, PPARd-, or PPARg-selective agonists for 24 h or 1 h (results not shown) did not signi¢cantly a¡ect the IL-1binduced nuclear translocation of p50 and p65. Accordingly, IL-1b-induced degradation of IkBa was not a¡ected by

VOL. 121, NO. 5 NOVEMBER 2003

PPARs AND NF-kB IN NORMAL AND PSORIATIC SKIN

1111

Figure 6. Localization of NF-jB p50 and p65 in normal, lesional, and nonlesional psoriatic skin. Immunohistochemical analysis of NF-kB (p50 and p65) expression and localization in punch biopsies from normal, lesional, and nonlesional psoriatic skin. In total specimens from six control patients with no history of psoriasis and ¢ve patients with chronic psoriasis were analyzed. A uniform pattern of p50 and p65 distribution was found in all control specimens with cytoplasmic sequestering of p50 and p65 in the basal layer, whereas nuclear localization of p50 was observed in the suprabasal layers. Unexpectedly, no nuclear accumulation of p65 was observed in the epidermis of patients with chronic psoriasis. The paraf¢n-embedded sections were counterstained with hematoxylin.

Figure 7. In vitro interaction between hPPARd and the p65 subunit of NF-jB. Full-length and DAB hPPARd were synthesized in the presence of [35S]methionine by in vitro translation and analyzed with either GST-p65 (residues 1^305) or GST-p65 (residues 354^551) bound to glutathione-sepharose beads as baits in pull-down assays (see Materials and Methods for details).

administration of PPAR agonists (Fig 9a). It has been reported that PPAR agonists may interfere with NF-kB activation without interfering with nuclear translocation by impairing NF-kB DNA binding (Delerive et al, 2000). In keratinocytes this appears not to be the case. Treatment with IL-1b induced a rapid and transient increase in NF-kB DNA-binding activity, and preincubation for 24 h or 1 h (results not shown) with PPAR agonists did not reduce IL-1b-induced NF-kB binding ^ if anything, it appeared that preincubation with the PPARa agonist enhanced NF-kB DNA binding. Binding to an oligonucleotide containing a consensus TATA box was una¡ected by IL-1b and PPAR ligands (Fig 9b). Physical interaction and reciprocal transcriptional regulation have been described between NF-kB and PPARa and PPARg in macrophages, COS-1, and HeLa cells (Delerive et al, 1999; Chung et al, 2000; Li et al, 2000b). Consequently, we analyzed whether reciprocal functional interaction between PPARd and p65 or p50 could be observed in primary human keratinocytes. In the ¢rst set of experiments a reporter construct containing three copies of the kB site was transfected with di¡erent combinations of vectors expressing p65 or PPAR. Figure 10 shows that p65 strongly activated the reporter. Either PPAR subtype alone had no e¡ect, and, interestingly, even a 30-fold molar excess of the PPAR-expressing vectors did not reduce the

p65-dependent transactivation whether or not a PPAR-selective agonist was present. On the contrary, coexpression of PPARs tended to slightly increase p65-mediated transactivation in a concentration-dependent manner. Essentially the same results were obtained using a reporter construct containing part of the proximal IL- 6 promoter with coexpression of p65 and PPARd (results not shown). Forced expression of p50 did not result in activation of the NF-kB-responsive reporter construct, neither in the presence nor in the absence of coexpressed PPARs (results not shown). Contrasting the e¡ect of PPARs on p65-driven transactivation, PPAR-mediated ligand-dependent and ligandindependent activation of a reporter driven by three copies of the PPRE from the acyl-CoA oxidase gene were strongly and concentration-dependently repressed by expression of p65 (Fig 11a). Noteworthy, p50 coexpression did not repress PPARmediated transactivation (Fig 11b). The p65-dependent repression of PPAR-mediated transactivation takes place in the absence of known kB sites in the reporter construct. The e¡ect of p65 is not general as expression of b-galactosidase from the normalizing vector was una¡ected, and furthermore p65 did not exert the same repressive action on SREBP-1a-mediated transactivation (results not shown). As coexpression of PPARs and/or administration of PPAR agonists did not repress p65-mediated transactivation, PPAR

1112

WESTERGAARD ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure 8. The e¡ect of PPAR agonists on IL-1b-induced nuclear translocation of NF-jB. Immuno£uorescence analysis of the intracellular localization of (a) the p50 and (b) the p65 subunit of NF-kB in normal human keratinocytes preincubated for 24 h with Wy14643 (10 mM), L165041 (5 nM), BRL49653 (1 mM), or vehicle (0.1% DMSO) prior to IL-1b treatment for 30 min as described in Materials and Methods. The p50 and p65 subunits of NF-kB were visualized by indirect immuno£uorescence microscopy.

agonists should not decrease IL-1b-induced expression of an NFkB-responsive reporter gene in keratinocytes. To test this, cells were transfected with the NF-kB-responsive reporter, left for 7 h, and then pretreated or not for 2 h with BRL49653 or L165041 before IL-1b was added. The cells were harvested after a further 3 h of incubation. Figure 12(a) shows that treatment with IL-1byielded a robust induction of reporter gene expression. This induction was not signi¢cantly repressed by pretreatment with L165041 or BRL49653. As the e¡ect of PPAR ligands on NFkB-responsive genes may be chromatin and/or promoter context dependent (Delerive et al, 2002), we analyzed the e¡ect of PPAR agonists on IL-1b-induced expression of two endogenous NFkB-responsive genes, the iNOS gene and the IL- 6 gene. Figure 12(b) shows that expression of both genes is strongly induced by IL-1b, and preincubation with the PPARd- or the PPARgselective agonist did not impair induction. In conclusion, PPAR expression or administration of PPAR agonists did not impair the ability of the proin£ammatory cytokine IL-1b to induce p65-dependent transactivation in keratinocytes. In contrast, p65 is a potent inhibitor of ligand-independent as well as liganddependent transactivation mediated by all three PPAR subtypes. p65-dependent repression of PPAR-mediated transactivation is not unique to keratinocytes but is also observed in other cell types examined, including COS-1 cells and 293 cells (Delerive et al, 1999; S. Rasmussen, J. Henningsen, M.Westergaard, K. Kristiansen, unpublished results). To examine if sequestering of coactivators contributes to the p65-dependent repression of PPAR-mediated transactivation, we investigated the e¡ect of increasing expression of CBP and p300 on p65-dependent repression of PPAR-mediated transactivation. Cotransfection with an equimolar amount of the CBP expression vector slightly increased PPARa-, PPARd-, and PPARg-mediated transactivation, whereas greater amounts of the cotransfected CBP expression vector reduced transactivation mediated by all PPAR subtypes. Cotransfection with a vector expressing p300 repressed PPARa-

and PPARg-mediated transactivation even at the lowest amount of cotransfected p300 expression vector. In contrast, coexpression of p300 enhanced PPARd-mediated transactivation. Interestingly, forced expression of either CBP or p300 in a concentrationdependent manner partially rescued PPAR-mediated transactivation in the presence of coexpressed p65, with rescue of PPARd- and PPARa-mediated transactivation being the most prominent (Fig 13). The interpretation of coactivator titration experiments is di⁄cult as the endogenous levels of the ectopically expressed coactivators as well as the levels of numerous other cofactors that interact with the PPARs and p65 dictate the response. Our results indicate, however, that competition for coactivators contributes to, but is not the sole phenomenon responsible for, p65-dependent repression of PPAR-mediated transactivation. DISCUSSION

It is well established that classical nuclear receptors play important roles in the epidermis, and during the last couple of years members of the PPAR family have been added to the growing list of nuclear receptors involved in keratinocyte di¡erentiation and function (Hanley et al, 1997; 1998; 1999; Rivier et al, 1998; Komuves et al, 2000a; 2000b). Recently, PPARa and PPARd were shown to play pivotal roles in controlling in£ammation, migration, survival, and di¡erentiation of keratinocytes during wound healing in mice (Michalik et al, 2001; Tan et al, 2001). Interestingly, all PPAR subtypes are expressed during fetal development of the skin in mice, but expression in the interfollicular epidermis ceases shortly after birth (Michalik et al, 2001). Expression of PPARa and PPARd in the adult epidermis is induced in response to in£ammation, however (Michalik et al, 2001). In contrast, all PPAR subtypes are expressed in interfollicular human epidermis, with PPARd being by far the predominant subtype (Rivier et al, 1998; Westergaard et al, 2001; this paper).

VOL. 121, NO. 5 NOVEMBER 2003

Figure 9. PPAR agonists do not prevent IL-1b-induced degradation of IjBa or NF-jB DNA binding in normal human keratinocytes. (a) Cytoplasmic extracts or (b) nuclear protein extracts were isolated from normal human keratinocytes preincubated for 24 h with vehicle (0.1% DMSO), Wy14643 (10 mM), L165041 (5 nM), or BRL49653 (1 mM) prior to stimulation with IL-1b (see Materials and Methods for details). (a) Twenty micrograms of cytoplasmatic protein were separated in SDS-polyacrylamide gels and the level of IkBa was determined by western blotting. Equal loading/transfer was con¢rmed by amido-black staining (not shown) of membranes and by immuno-detection of TFIIB. The n.s. denotes a nonspeci¢c band. (b) EMSA were performed using a double-stranded oligonucleotide harboring a consensus kB element or a consensus TATA box.

In this study we show that expression of PPARd is dramatically upregulated at the mRNA as well as at the protein level in hyperproliferative lesional skin from psoriasis patients. This upregulation is probably related to the accumulation of proin£ammatory cytokines in the psoriatic lesions in analogy with the in£ammation-dependent upregulation of PPARd expression in mouse skin (Tan et al, 2001). A detailed immunochemical analysis of the intracellular distribution of PPAR subtypes revealed no marked disturbances comparing normal, nonlesional, and lesional psoriatic skin. In particular, PPARd exhibited a nuclear localization in the entire suprabasal region of normal and psoriatic skin, whereas two populations of cells were present in the basal layer, one exhibiting nuclear localization of PPARd and one exhibiting a cytoplasmic accumulation of PPARd. Lesional psoriatic skin is characterized by the accumulation of numerous in£ammatory lipoxygenase products. By transient

PPARs AND NF-kB IN NORMAL AND PSORIATIC SKIN

1113

transfections in HaCaT cells, we showed that human PPARd was potently activated by 8(S)-HETE, 12(S)-HETE, 12(R)-HETE, and 15(S)-HETE. Of these, 12(R)-HETE in particular has been shown to accumulate to high levels in lesional psoriatic skin (Woollard, 1986). Upregulation of PPARd expression combined with high abundance of potent agonists would suggest that PPARd-responsive genes might be upregulated in lesional psoriatic skin. We have presented evidence that CD36/FAT is a PPARdresponsive gene in human keratinocytes (Westergaard et al, 2001). In keeping with this notion, we now demonstrate a signi¢cant upregulation of CD36/FAT expression in lesional psoriatic skin. Thus, our results suggest that PPARd-mediated transactivation is enhanced in lesional psoriatic skin. Members of the NF-kB/Rel family of transcription factors are pleiotropic regulators of cellular growth and homeostasis (Ghosh and Karin, 2002). Recently, NF-kB was also implicated as a pivotal regulator of epidermal proliferation and di¡erentiation (for reviews see Kaufman and Fuchs, 2000; Fuchs and Raghavan, 2002), even though the precise function of NF-kB in epidermal homeostasis remains enigmatic. It is evident that forced expression of an arti¢cial transactivating form of p50 or p65 in the basal layer of mouse epidermis leads to cell cycle arrest, hypoplasia, and accelerated di¡erentiation (Seitz et al, 1998; 2000a). Conversely, epidermis-speci¢c abrogation of NF-kB activation by forced expression of a nondegradable form of IkBa has been shown to be associated with hyperplasia and hyperkeratinization (Seitz et al, 1998). It is unclear, however, whether the observed hyperplasia represents a keratinocyte-cell-autonomous phenomenon or is a secondary e¡ect related to altered immune homeostasis. This aspect was underscored by the interesting ¢nding that mice with an epidermis-speci¢c deletion of IKK2 (IKKb) were born with a normal skin but developed a severe in£ammatory dermatitis during the ¢rst week after birth (Pasparakis et al, 2002). Along the same line, lack of IKK1 (IKKa) resulted in epidermal hyperplasia and severely impaired di¡erentiation, which could be rescued by a kinase-inactive IKK1 (IKKa) mutant, and it was shown that expression of IKK1 (IKKa) led to the synthesis of a polypeptide that stimulated keratinocyte di¡erentiation independently of NF-kB activation (Hu et al, 2001). Taken together these results clearly indicate that the role of the NF-kB system in skin homeostasis is far more complex than ¢rst anticipated. Indeed it is possible that NF-kB activation plays a minor role in the normal development of interfollicular skin. Thus, by using a transgenic mouse model harboring an integrated NF-kB reporter, no NFkB-dependent transactivation was observed in the interfollicular epidermis (Schmidt-Ullrich et al, 2001). A lack of NF-kBmediated transactivation would also be consistent with our ¢nding that only p50 exhibits a nuclear localization in the suprabasal layer of normal skin, and normally, p50/p50 homodimers repress rather than stimulate transcription (Ghosh and Karin, 2002). This said, it should be mentioned that convincing evidence has been presented that NF-kB plays an antiapoptotic role in the epidermis (Seitz et al, 2000b; Qin et al, 2001). The same was shown in the case of PPARd during wound healing and in response to stimulation by proin£ammatory cytokines (Tan et al, 2001; Di-Po|« et al, 2002). NF-kB-mediated transactivation has been shown to be repressed by PPARa and PPARg or PPARa or PPARg agonists in several cell types (Jiang et al, 1998; Ricote et al, 1998; Staels et al, 1998; Delerive et al, 1999; Chung et al, 2000; Li et al, 2000b). Interestingly, it was recently demonstrated that PPARa could positively a¡ect NF-kB-dependent induction of IkBa expression, suggesting that the e¡ect of PPAR on NF-kB-mediated transactivation exhibits cell type and/or promoter dependence (Delerive et al, 2002). Our results demonstrate that all PPAR subtypes in the absence or presence of agonists in keratinocytes are unable to repress p65-mediated activation of an NF-kB-responsive reporter. In fact, a slight concentration-dependent activation was observed. Conversely, we observed a very strong downregulation of ligand-independent and ligand-dependent PPAR-mediated

1114

WESTERGAARD ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure 10. p65-mediated transactivation is not suppressed by coexpression of PPAR in the absence or presence of PPAR agonists in normal human keratinocytes. Normal human keratinocytes were transiently transfected with the p(kB)3 -luc þ reporter (50 ng), with di¡erent combinations of expression vectors p65 (6 ng) and hPPARa, hPPARd, or hPPARg1 (18 ng or 60 ng). Medium was changed after 6 h, and the cells were subsequently treated with medium containing Wy14643 (10 mM), L165041 (5 nM), BRL49653 (1 mM), or vehicle (0.1% DMSO) for approximately 20 h prior to harvest. In all transfections empty expression vector was added to ensure equal promoter load and an SV40-b-galactosidase construct was used for normalization. Reporter activity was normalized to b-galactosidase values and fold induction is presented as the mean7SD. Transfections were performed in triplicate, measured in duplicate, and repeated four times. One representative experiment is presented.

Figure 11. The p65, but not the p50, subunit of NF-jB potently suppresses PPAR-mediated transactivation in normal human keratinocytes. Normal human keratinocytes were transiently transfected with PPREx3 -tk-luc (160 ng) with di¡erent combinations of expression vectors. (a) hPPARa (0.625 ng), hPPARd (62.5 ng), hPPARg1 (62.5 ng), and increasing amounts of p65 (0.1, 0.5, 2.5, and 12.5 ng). (b) As (a) but with increasing amounts of p50 using the following combinations: hPPARa (0.625 ng), PPREx3 -tk-luc reporter (160 ng), and p50 (0.1, 0.5, 2.5 ng); hPPARd (20 ng), PPREx3 -tk-luc (130 ng), p50 (3.2, 16, 80 ng), or hPPARg1 (40 ng), PPREx3 -tk-luc reporter (102.5 ng), p50 (6.4, 32, 160 ng). Medium was changed after 6 h, and the cells were subsequently treated with medium containing Wy14643 (10 mM), L165041 (5 nM), BRL49653 (1 mM), or vehicle (0.1% DMSO) for approximately 20 h prior to harvest. In all transfections empty expression vector was added to ensure equal promoter load and an SV40-b-galactosidase construct was used for normalization. Reporter activity was normalized to b-galactosidase values and fold induction is presented as the mean7range. Transfections were performed in duplicate, measured in duplicate, and repeated four times. One representative experiment is presented.

transactivation by coexpression of p65, but not p50. The fact that forced expression of the coactivators CBP or p300 partially rescues the PPAR-mediated transactivation suggests that coactivator sequestering at least in part is responsible for the p65-dependent repression. Generally, considering cross-talk between PPAR and NF-kB most emphasis has been on the e¡ects of PPAR or PPAR ligands on NF-kB. Our ¢ndings suggest that the reverse situation, namely p65-dependent inhibition of PPAR-mediated trans-

activation, should be considered as an important facet in the regulation of PPAR-dependent processes; and p65-dependent downregulation of PPAR-mediated transactivation is not restricted to keratinocytes but is also observed in several other cell types (Delerive et al, 1999; S. Rasmussen, J. Henningsen, M. Westergaard, K. Kristiansen, unpublished results). The ability of p65 to downregulate PPARd-mediated transactivation is also intriguing and paradoxical considering that PPARdvia transcriptional

VOL. 121, NO. 5 NOVEMBER 2003

Figure 12. PPAR agonists do not prevent IL-1b-induced NFjB-mediated gene expression in normal human keratinocytes. (a) Normal human keratinocytes were transiently transfected with the p(kB)3 -luc þ reporter (50 ng). Medium was changed after 6 h. Then cells were preincubated (2 h) with vehicle (0.1% DMSO), 5 nM L165041, or 1 mM BRL49653 and subsequently incubated for 3 h with IL-1b (10 ng per ml). SV40-b-galactosidase was used for normalization. Reporter activity was normalized to b-galactosidase values and fold induction is presented as the mean7SD. Transfections were performed in triplicate, measured in duplicate, and repeated four times. One representative experiment is presented. (b) Total RNA was prepared from normal human keratinocytes pretreated for 0.5 h with vehicle (0.1% DMSO), 5 nM L165041, or 1 mM BRL49653 and subsequently incubated for 3 h with IL-1b (10 ng per ml). Multiplex RT-PCR analysis of IL- 6 and iNOS expression. TBP served as internal control.

induction of ILK and PDK1 was reported to increase Akt1 activity leading to enhanced NF-kB activity in keratinocytes (Di-Po|« et al, 2002). In normal skin the absence of nuclear p65 allows PPAR-dependent transactivation, and the presence of nuclear p50 does not seem to repress PPAR-dependent gene expression, and hence PPARd may exert its pro-di¡erentiating actions. In chronic psoriatic lesions p65 remains in the cytoplasm in spite of the massive accumulation of proin£ammatory cytokines allowing PPARddependent gene activation. Considering the severe e¡ects on epidermal proliferation and di¡erentiation associated with defects in NF-kB activation in keratinocytes, it is tempting to speculate that the absence of NF-kB activation in chronic psoriatic plaques contributes to the hyperproliferative and di¡erentiation-de¢cient phenotype of psoriasis. We thank Drs J. Fleckner and A. Elbrecht for cDNAs encoding human PPARs, Dr R.M. Evans for reporter constructs, DrA. Baldwin Jr for p65 expression construct, Dr H. Sakurai for GST fusion constructs, and Novo Nordisk and Merck Research Laboratories for providing ligands. This work was conducted within the Center for Experimental BioInformatics and supported by the Danish Biotechnology Program, the Karen Elise Jensen Foundation, the Leo Research Foundation, the Danish Natural Science Research Council, the Danish Medical Research Council, the Institute for Experimental Clinical Research, the University of Aarhus, the Aage Bang Foundation, and the Novo Nordisk Foundation.

PPARs AND NF-kB IN NORMAL AND PSORIATIC SKIN

1115

Figure 13. Forced expression of the coactivators CBP or p300 partially relieves p65-mediated repression of PPAR-dependent transactivation. 293 cells were transiently transfected with 750 ng PPREx3 -tkluc with di¡erent combinations of expression vectors (a) hPPARa (1.25 ng), hPPARd (62.5 ng), hPPARg1 (62.5 ng), 12.5 ng p65 as indicated, and increasing amounts of expression vectors encoding (a) CBP or (b) p300 (125, 375, 1125 ng). Medium was changed after 5 h, and the cells were subsequently treated with medium containing Wy14643 (30 mM), L165041 (500 nM), BRL49653 (1 mM), or vehicle (0.1% DMSO) for approximately 20 h prior to harvest. The molar ratio between PPAR-expressing vectors and the p65-expressing vector was 5:1. The amounts of cotransfected vector expressing CBP or p300 represent a 1-fold, 3 -fold, and 9-fold molar excess relative to the amounts of transfected vector expressing the PPAR subtypes. In all transfection experiments empty expression vector was added to ensure equal promoter load and an SV40-b-galactosidase construct was used for normalization. Reporter activity was normalized to b-galactosidase values and fold induction is presented as the mean7range. Transfections were performed in duplicate, measured in duplicate, and repeated two to four times. One representative experiment is presented.

REFERENCES Adams M, Reginato MJ, Shao D, Lazar MA, Chatterjee VK: Transcriptional activation by peroxisome proliferator-activated receptor g is inhibited by phosphorylation at a consensus mitogen-activated protein kinase site. J Biol Chem 272:5128^5132, 1997 Baeuerle PA: Pro-in£ammatory signaling: Last pieces in the NF-kB puzzle? Curr Biol 8:R19^R22, 1998 Barger PM, Brandt JM, Leone TC, Weinheimer CJ, Kelly DP: Deactivation of peroxisome proliferator-activated receptor-a during cardiac hypertrophic growth. J Clin Invest 105:1723^1730, 2000

1116

WESTERGAARD ET AL

Barnes PJ, Karin M: Nuclear factor-kB: A pivotal transcription factor in chronic in£ammatory diseases. N Engl J Med 336:1066^1071, 1997 Berth-Jones J, Hutchinson PE: Vitamin D analogues and psoriasis. Br J Dermatol 127:71^78, 1992 Budunova IV, Perez P, Vaden VR, Spiegelman VS, Slaga TJ, Jorcano JL: Increased expression of p50-NF-kB and constitutive activation of NF-kB transcription factors during mouse skin carcinogenesis. Oncogene 18:7423^7431, 1999 Camp HS, Tafuri SR: Regulation of peroxisome proliferator-activated receptor g activity by mitogen-activated protein kinase. J Biol Chem 272:10811^10816, 1997 Camp HS, Tafuri SR, Le¡ T: c-Jun N-terminal kinase phosphorylates peroxisome proliferator-activated receptor-g1 and negatively regulates its transcriptional activity. Endocrinol 140:392^397, 1999 Camp RD, Mallet AI, Woollard PM, Brain SD, Black AK, Greaves MW: The identi¢cation of hydroxy fatty acids in psoriatic skin. Prostaglandins 26:431^447, 1983 Chinetti G, Griglio S, Antonucci M, et al: Activation of proliferator-activated receptors a and g induces apoptosis of human monocyte-derived macrophages. J Biol Chem 273:25573^25580, 1998 Chung SW, Kang BY, Kim SH, Pak YK, Cho D, Trinchieri G, Kim TS: Oxidized low density lipoprotein inhibits interleukin-12 production in lipopolysaccharide-activated mouse macrophages via direct interactions between peroxisome proliferator-activated receptor-g and nuclear factor-kB. J Biol Chem 275: 32681^32687, 2000 Delerive P, De B, Besnard S, et al: Peroxisome proliferator-activated receptor a negatively regulates the vascular in£ammatory gene response by negative cross-talk with transcription factors NF-kB and AP-1. J Biol Chem 274:32048^32054, 1999 Delerive P, Gervois P, Fruchart JC, Staels B: Induction of IkBa expression as a mechanism contributing to the anti-in£ammatory activities of peroxisome proliferator-activated receptor-a activators. J Biol Chem 275:36703^36707, 2000 Delerive P, De Bosscher K, Vanden Berghe W, Fruchart JC, Haegeman G, Staels B: DNA binding-independent induction of IkBa gene transcription by PPARa. Mol Endocrinol 16:1029^1039, 2002 Devchand PR, Keller H, Peters JM, Vazquez M, Gonzalez FJ, Wahli W: The PPARaleukotriene B4 pathway to in£ammation control. Nature 384:39^43, 1996 Di-Po|« N,Tan MS, Michalik L,Wahli W, Desvergne B: Antiapoptotic role of PPARb in keratinocytes via transcriptional control of the Akt1 signaling pathway. Mol Cell 10:721^733, 2002 Ellis CN, Varani J, Fisher GJ, et al: Troglitazone improves psoriasis and normalizes models of proliferative skin disease: Ligands for peroxisome proliferator-activated receptor-g inhibit keratinocyte proliferation. Arch Dermatol 136:609^616, 2000 Fogh K, Herlin T, Kragballe K: Eicosanoids in acute and chronic psoriatic lesions: Leukotriene B4, but not 12-hydroxy-eicosatetraenoic acid, is present in biologically active amounts in acute guttate lesions. J Invest Dermatol 92:837^841, 1989 Forman BM,Tontonoz P, Chen J, Brun RP, Spiegelman BM, Evans RM: 15-Deoxyd12, 14 -prostaglandin J2 is a ligand for the adipocyte determination factor PPARg. Cell 83:803^812, 1995 Forman BM, Chen J, Evans RM: Hypolipidemic drugs, polyunsaturated fatty-acids, and eicosanoids are ligands for peroxisome proliferator-activated receptor-a and receptor-d. Proc Natl Acad Sci USA 94:4312^4317, 1997 Fuchs E, Raghavan S: Getting under the skin of epidermal morphogenesis. Nat Rev Genet 3:199^209, 2002 Ghosh S, Karin M: Missing pieces in the NF-kB puzzle. Cell 109:S81^S96, 2002 Hammarstrom S, Hamberg M, Samuelsson B, Duell EA, Stawiski M, Voorhees JJ: Increased concentrations of nonesteri¢ed arachidonic acid, 12L-hydroxy5,8,10,14 -eicosatetraenoic acid, prostaglandin E2, and prostaglandin F2a in epidermis of psoriasis. Proc Natl Acad Sci USA 72:5130^5134, 1975 Hanley K, Jiang Y, Crumrine D, et al: Activators of the nuclear hormone receptors PPARa and FXR accelerate the development of the fetal epidermal permeability barrier. J Clin Invest 100:705^712, 1997 Hanley K, Jiang Y, He SS, et al: Keratinocyte di¡erentiation is stimulated by activators of the nuclear hormone receptor PPARa. J Invest Dermatol 110:368^375, 1998 Hanley K, Komuves LG, Bass NM, et al: Fetal epidermal di¡erentiation and barrier development in vivo is accelerated by nuclear hormone receptor activators. J Invest Dermatol 113:788^795, 1999 Hanley K, Komuves LG, Ng DC, et al: Farnesol stimulates di¡erentiation in epidermal keratinocytes via PPARa. J Biol Chem 275:11484^11491, 2000 Hansen JB, Zhang H, Rasmussen TH, Petersen RK, Flindt EN, Kristiansen K: Peroxisome proliferator-activated receptor d (PPARd) -mediated regulation of preadipocyte proliferation and gene expression is dependent on cAMP signaling. J Biol Chem 276:3175^3182, 2001 He TC, Chan TA, Vogelstein B, Kinzler KW: PPARd is an APC-regulated target of nonsteroidal anti-in£ammatory drugs. Cell 99:335^345, 1999 van Hogerlinden M, Rozell BL, Ahrlund-Richter L, Toftgard R: Squamous cell carcinomas and increased apoptosis in skin with inhibited Rel/nuclear factor-kB signaling. Cancer Res 59:3299^3303, 1999 Hsi LC, Wilson L, Nixon J, Eling TE: 15-lipoxygenase-1 metabolites down-regulate peroxisome proliferator-activated receptor g via the MAPK signaling pathway. J Biol Chem 276:34545^34552, 2001 Hu E, Kim JB, Sarraf P, Spiegelman BM: Inhibition of adipogenesis through MAP kinase-mediated phosphorylation of PPARg. Science 274:2100^2103, 1996

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Hu Y, Baud V, Delhase M, et al: Abnormal morphogenesis but intact IKK activation in mice lacking the IKKa subunit of IkB kinase. Science 284:316^320, 1999 Hu Y, Baud V, Oga T, Kim KI, Yoshida K, Karin M: IKKa controls formation of the epidermis independently of NF-kB. Nature 410:710^714, 2001 Itin PH, Pittelkow MR, Kumar R: E¡ects of vitamin D metabolites on proliferation and di¡erentiation of cultured human epidermal keratinocytes grown in serum-free or de¢ned culture medium. Endocrinol 135:1793^1798, 1994 Jehl P, Bastie C, Gillot I, Luquet S, Grimaldi PA: Peroxisome-proliferator-activated receptor d mediates the e¡ects of long-chain fatty acids on post-con£uent cell proliferation. Biochem J 350 Part 1:93^98, 2000 Jiang C, Ting AT, Seed B: PPAR-g agonists inhibit production of monocyte in£ammatory cytokines. Nature 391:82^86, 1998 Johansen C, Iversen L, Ryborg A, Kragballe K: 1a,25-dihydroxyvitamin D3 induced di¡erentiation of cultured human keratinocytes is accompanied by a PKCindependent regulation of AP-1 DNA binding activity. J Invest Dermatol 114: 1174^1179, 2000 Juge-Aubry CE, Hammar E, Siegrist-Kaiser C, et al: Regulation of the transcriptional activity of the peroxisome proliferator-activated receptor a by phosphorylation of a ligand-independent trans-activating domain. J Biol Chem 274:10505^10510, 1999 Kaufman CK, Fuchs E: It’s got you covered. NF-kB in the epidermis. J Cell Biol 149:999^1004, 2000 Klement JF, Rice NR, Car BD, et al: IkBa de¢ciency results in a sustained NF-kB response and severe widespread dermatitis in mice. Mol Cell Biol 16:2341^2349, 1996 Kliewer SA, Lenhard JM, Willson TM, Patel I, Morris DC, Lehmann JM: A prostaglandin J2 metabolite binds peroxisome proliferator-activated receptor g and promotes adipocyte di¡erentiation. Cell 83:813^819, 1995 Kliewer SA, Sundseth SS, Jones SA, et al: Fatty-acids and eicosanoids regulate gene expression through direct interactions with peroxisome proliferator-activated receptor-a and receptor-g. Proc Natl Acad Sci USA 94:4318^4323, 1997 Komuves LG, Hanley K, Jiang Y, Elias PM,Williams ML, Feingold KR: Ligands and activators of nuclear hormone receptors regulate epidermal di¡erentiation during fetal rat skin development. J Invest Dermatol 111:429^433, 1998 Komuves LG, Hanley K, Lefebvre AM, et al: Stimulation of PPARa promotes epidermal keratinocyte di¡erentiation in vivo. J Invest Dermatol 115:353^360, 2000a Komuves LG, Hanley K, Man MQ, Elias PM, Williams ML, Feingold KR: Keratinocyte di¡erentiation in hyperproliferative epidermis. Topical application of PPARa activators restores tissue homeostasis. J Invest Dermatol 115:361^367, 2000b Kragballe K, Fallon JD: Increased aggregation and arachidonic acid transformation by psoriatic platelets: Evidence that platelet-derived 12-hydroxy-eicosatetraenoic acid increases keratinocyte DNA synthesis in vitro. Arch Dermatol Res 278:449^453, 1986 Krey G, Braissant O, L’Horset F, Kalkhoven E, Perroud M, Parker MG, Wahli W: Fatty acids, eicosanoids, and hypolipidemic agents identi¢ed as ligands of peroxisome proliferator-activated receptors by coactivator-dependent receptor ligand assay. Mol Endocrinol 11:779^791, 1997 Kussmann-Gerber S, Kratchmarova I, Mandrup S, Kristiansen K: Microa⁄nity columns for analysis of protein^protein interactions. Anal Biochem 271:102^105, 1999 Li M, Indra AK, Warot X, et al: Skin abnormalities generated by temporally controlled RXRa mutations in mouse epidermis. Nature 407:633^636, 2000a Li M, Pascual G, Glass CK: Peroxisome proliferator-activated receptor g-dependent repression of the inducible nitric oxide synthase gene. Mol Cell Biol 20: 4699^4707, 2000b Li Q, Lu Q, Hwang JY, Buscher D, Lee KF, Izpisua-Belmonte JC, Verma IM: IKK1de¢cient mice exhibit abnormal development of skin and skeleton. Genes Dev 13:1322^1328, 1999 Madsen P, Rasmussen HH, Le¡ers H, et al: Molecular cloning, occurrence, and expression of a novel partially secreted protein ‘‘psoriasin’’ that is highly upregulated in psoriatic skin. J Invest Dermatol 97:701^712, 1991 Makris C, Godfrey VL, Krahn-Senftleben G, et al: Female mice heterozygous for IKKg/NEMO de¢ciencies develop a dermatopathy similar to the human Xlinked disorder incontinentia pigmenti. Mol Cell 5:969^979, 2000 Mani I, Iversen L, Ziboh VA: Upregulation of nuclear PKC and MAP-kinase during hyperproliferation of guinea pig epidermis: Modulation by 13 -(S)-hydroxyoctadecadienoic acid (13 -HODE). Cell Signal 10:143^149, 1998 Matsuura H, Adachi H, Smart RC, Xu X, Arata J, Jetten AM: Correlation between expression of peroxisome proliferator-activated receptor b and squamous differentiation in epidermal and tracheobronchial epithelial cells. Mol Cell Endocrinol 147:85^92, 1999 Michalik L, Desvergne B, Tan NS, et al: Impaired skin wound healing in peroxisome proliferator-activated receptor (PPAR) a and PPARb mutant mice. J Cell Biol 154:799^814, 2001 Pasparakis M, Courtois G, Hafner M, et al: TNF-mediated in£ammatory skin disease in mice with epidermis-speci¢c deletion of IKK2. Nature 417:861^866, 2002 Pillai S, Bikle DD: Epidermal vitamin D metabolism, function, and regulation. Adv Lipid Res 24:321^341, 1991 Poynter ME, Daynes RA: Peroxisome proliferator-activated receptor a activation modulates cellular redox status, represses nuclear factor-kB signaling, and

VOL. 121, NO. 5 NOVEMBER 2003

reduces in£ammatory cytokine production in aging. J Biol Chem 273: 32833^32841, 1998 Qin JZ, Bacon P, Chaturvedi V, Nickolo¡ BJ: Role of NF-kB activity in apoptotic response of keratinocytes mediated by interferon-g, tumor necrosis factor-a, and tumor-necrosis-factor-related apoptosis-inducing ligand. J Invest Dermatol 117:898^907, 2001 Ricote M, Li AC, Willson TM, Kelly CJ, Glass CK: The peroxisome proliferatoractivated receptor-g is a negative regulator of macrophage activation. Nature 391:79^82, 1998 Rivier M, Safonova I, Lebrun P, Gri⁄ths CE, Ailhaud G, Michel S: Di¡erential expression of peroxisome proliferator-activated receptor subtypes during the differentiation of human keratinocytes. J Invest Dermatol 111:1116^1121, 1998 Rivier M, Castiel I, Safonova I, Ailhaud G, Michel S: Peroxisome proliferator-activated receptor-a enhances lipid metabolism in a skin equivalent model. J Invest Dermatol 114:681^687, 2000 Sakurai H, Chiba H, Miyoshi H, Sugita T, Toriumi W: IkB kinases phosphorylate NF-kB p65 subunit on serine 536 in the transactivation domain. J Biol Chem 274:30353^30356, 1999 Schmidt-Supprian M, Bloch W, Courtois G, Addicks K, Israel A, Rajewsky K, Pasparakis M: NEMO/IKKg-de¢cient mice model incontinentia pigmenti. Mol Cell 5:981^992, 2000 Schmidt-Ullrich R, Aebischer T, Hulsken J, Birchmeier W, Klemm U, Scheidereit C: Requirement of NF-kB/Rel for the development of hair follicles and other epidermal appendices. Development 128:3843^3853, 2001 Seitz CS, Lin Q, Deng H, Khavari PA: Alterations in NF-kB function in transgenic epithelial tissue demonstrate a growth inhibitory role for NF-kB. Proc Natl Acad Sci USA 95:2307^2312, 1998 Seitz CS, Deng H, Hinata K, Lin Q, Khavari PA: Nuclear factor kB subunits induce epithelial cell growth arrest. Cancer Res 60:4085^4092, 2000a Seitz CS, Freiberg RA, Hinata K, Khavari PA: NF-kB determines localization and features of cell death in epidermis. J Clin Invest 105:253^260, 2000b Shalev A, Siegrist-Kaiser CA, Yen PM, Wahli W, Burger AG, Chin WW, Meier CA: The peroxisome proliferator-activated receptor a is a phosphoprotein: Regulation by insulin. Endocrinol 137:4499^4502, 1996

PPARs AND NF-kB IN NORMAL AND PSORIATIC SKIN

1117

Staels B, Koenig W, Habib A, et al: Activation of human aortic smooth-muscle cells is inhibited by PPARa but not by PPARg activators. Nature 393:790^793, 1998 Stern RS: Psoriasis. Lancet 350:349^353, 1997 Straus DS, Pascual G, Li M, et al: 15-deoxy-d12,14 -prostaglandin J2 inhibits multiple steps in the NF-kB signaling pathway. Proc Natl Acad Sci USA 97:4844^4849, 2000 Su CG, Wen X, Bailey ST, et al: A novel therapy for colitis utilizing PPAR-g ligands to inhibit the epithelial in£ammatory response. J Clin Invest 104:383^389, 1999 Takeda K, Takeuchi O, Tsujimura T, et al: Limb and skin abnormalities in mice lacking IKKa. Science 284:313^316, 1999 Tan NS, Michalik L, Noy N, et al: Critical roles of PPAR b/d in keratinocyte response to in£ammation. Genes Dev 15:3263^3277, 2001 Terui T, Ozawa M, Tagami H: Role of neutrophils in induction of acute in£ammation in T-cell-mediated immune dermatosis, psoriasis: A neutrophil-associated in£ammation-boosting loop. Exp Dermatol 9:1^10, 2000 Thuillier P, Brash AR, Kehrer JP, et al: Inhibition of PPAR-mediated keratinocyte di¡erentiation by lipoxygenase inhibitors. Biochem J 366:901^910, 2002 Uyemura K, Yamamura M, Fivenson DF, Modlin RL, Nickolo¡ BJ: The cytokine network in lesional and lesion-free psoriatic skin is characterized by a T-helper type 1 cell-mediated response. J Invest Dermatol 101:701^705, 1993 Westergaard M, Henningsen J, Svendsen ML, et al: Modulation of keratinocyte gene expression and di¡erentiation by PPAR-selective ligands and tetradecylthioacetic acid. J Invest Dermatol 116:702^712, 2001 Willson TM, Brown PJ, Sternbach DD, Henke BR: The PPARs: From orphan receptors to drug discovery. J Med Chem 43:527^550, 2000 Woollard PM: Stereochemical di¡erence between 12-hydroxy-5,8,10,14 -eicosatetraenoic acid in platelets and psoriatic lesions. Biochem Biophys Res Commun 136: 169^176, 1986 Xu L, Glass CK, Rosenfeld MG: Coactivator and corepressor complexes in nuclear receptor function. Curr Opin Genet Dev 9:140^147, 1999 Yu K, Bayona W, Kallen CB, et al: Di¡erential activation of peroxisome proliferatoractivated receptors by eicosanoids. J Biol Chem 270:23975^23983, 1995