Original Article

Introduction

Psoriasis is a common inflammatory skin disease involving altered keratinocyte differentiation, for example (Gandarillas et al., 1999; Hobbs et al., 2004), accumulation of dendritic cells (DCs) (Lowes et al., 2005), as well as T-cell activation (Boyman et al., 2004; Kohlmann et al., 2004). Genetically, psoriasis exhibits a polygenic as well as heterogeneous inheritance pattern (reviewed in Rahman and Elder, 2005). Clinically and histologically, however, the disease is quite uniform, despite case-to-case variability, suggesting that the interplay of susceptibility alleles and exogenous factors terminates in a common set of molecular changes.

Expression profiling has been used in several studies to characterize transcriptional changes in psoriasis (e.g., Oestreicher et al., 2001; Nomura et al., 2003; Zhou et al., 2003; Kulski et al., 2005; Quekenborn-Trinquet et al., 2005). When profiling the transcriptome in whole skin samples, a sizable fraction, if not the majority, of dysregulated transcripts might be expected to constitute "noise" due to inhomogenous sampling. Therefore, microarray studies in psoriasis to date have analyzed lists of transcriptional changes by a candidate approach, assigning relevance to specific transcripts based on their involvement in well-established aspects of the disease, such as keratinocyte differentiation, IFN signaling, or influx of inflammatory cells. By this approach, however, a large portion of transcriptional changes may be dismissed as artefacts, precluding the identification of unrecognized relevant signaling pathways.

The transcription factor peroxisome proliferator-activated receptor (PPAR) contributes to the regulation of adipogenesis, glucose metabolism, myogenesis, and macrophage function, and has previously been shown to be upregulated in psoriasis (Westergaard et al., 2003). This transcription factor is expressed in basal and suprabasal epidermal keratinocytes (Westergaard et al., 2001), is induced by IFN-, stress-activated kinase, or tumor necrosis factor- (TNF) via activator protein 1 (AP1)-mediated transcription, and contributes to the so-called "regenerative skin phenotype" characterizing the late phase of wound healing (Tan et al., 2001; Di-Poi et al., 2003). Furthermore, PPAR regulates keratinocyte differentiation, blocks apoptosis (reviewed in Icre et al., 2006), and induces angiogenesis (Piqueras et al., 2006), adding to an activity profile highly consistent with a role in psoriasis.

To be activated, all PPAR isoforms require binding of fatty acid ligands shuttled by fatty acid-binding proteins. One of these, psoriasis-associated fatty acid-binding protein (=FABP5), specifically interacts with PPAR (Tan et al., 2002), is upregulated in psoriasis (Madsen et al., 1992), and is enriched in the highly proliferative transit amplifying keratinocyte subset (O'Shaughnessy et al., 2000). Moreover, the retinoic acid shuttle cellular retoinic acid-binding protein II is overexpressed in psoriasis (Siegenthaler et al., 1992), possibly activating PPAR, since retinoic acid itself is a high-affinity PPAR ligand (Shaw et al., 2003). Furthermore, all PPAR isoforms require heteromerization with retinoid acid X-receptor for full activity (Kojo et al., 2003; Tan et al., 2005).

The precise effects of PPAR on PPAR/ activity, and vice versa, have not been precisely defined in vivo to date because of superimposed ligand and co-regulator effects; but the outcome of PPAR activation is clearly distinct from that of PPAR/ activation. Specifically, PPAR may act as a direct antagonist to PPAR (Zuo et al., 2006), or be effected in opposite fashion to PPAR by the same ligand (Jarvis et al., 2005). Furthermore, unliganded PPAR itself can act as a dominant negative to the ligand-containing isoform (Lee et al., 2006). In light of these data, it is not surprising that activation of PPAR has an inhibitory effect on psoriasis, whereas this is not the case with PPAR (Kuenzli and Saurat, 2003; Malhotra et al., 2005).

Here, we show that the vast majority of all transcriptional changes in psoriatic skin lesions are highly reproducible across two independent studies. While all expected changes, such as DC- and IFN-related gene upregulation, were observed, a large group of altered genes was involved in fatty acid signaling, with a profile suggestive of PPAR activation. Heparin-binding EGF-like growth factor (HB-EGF), an established inducer of epidermal hyperplasia (Kimura et al., 2005), was identified as a direct target gene of PPAR. Our data suggest that activation of PPAR is a key element in psoriasis and identify deranged intermediary metabolism as a hitherto unrecognized feature of the disease.

Results

Transcriptional dysregulation in psoriasis is highly reproducible

Expression profiling from lesional versus non-lesional psoriatic skin was performed in five patients with active disease, as detailed in Materials and Methods. Several types of analyses were carried out to rule out significant distortions and to confirm that non-lesional and lesional samples are clearly distinguishable in their expression profiles, that is that microarray analysis predicts disease status with certainty (Figure S1). Approximately, 3.6% of the 12,600 transcripts were up- and 2.1% downregulated by twofold or more when applying a threshold of P<0.05 in a Welsh-weighted t-test (see Supplementary Material "Charité all transcripts"). To estimate the extent of artificial alterations due to inhomogenous sampling, we quantified the changes reported in a previous study employing a comparable experimental platform (Affimetrix U95 microarray; Zhou et al., 2003). Those data include 326 transcripts upregulated by twofold or more in lesional skin. Applying the same threshold to the subset of validated translated genes, we were in fact able to confirm 80% of all tested transcripts (Table 1). When lowering the cutoff to 1.5-fold upregulation, almost 90% of upregulated transcripts were observed in both data sets. Only four transcripts, not confirmed in the current data set, were significantly upregulated in the previous data set (3-fold): IL-7 receptor, CLUAL protein, RASGRP2, RGS1. Three of these transcripts (IL-7 receptor, RASGRP2, RGS1) are T-cell associated, and two (RASGRP2 and RGS1) upregulated in CD4⁺ T cells, suggesting that the discrepancy may be in fact be explained by uneven sampling of CD4⁺ T cells known to localize preferentially subepidermally in psoriatic lesions. Likewise, approximately, 70% of transcripts reported to be downregulated by Zhou et al. (2003) were also downregulated in the current data set. Only 10 genes significantly downregulated (2.5-fold) could not be confirmed. Thus, independent expression profiles comprising data from 17 patients exhibit an unexpected level of concordance. Other available microarray data sets (Oestreicher et al., 2001; Itoh et al., 2005; Quekenborn-Trinquet et al., 2005) were not included in the present meta-analysis because of differences in experimental procedure, but collectively confirm the notion that transcriptional changes in psoriatic lesions are highly reproducible despite sampling complex cell-type mixtures. Thus, the majority of these transcripts do not represent random noise, but are probably related to disease pathogenesis, and must consequently be considered in data analysis.

Table 1 - Concordance of transcriptional regulation in lesional psoriatic skin in two independent data sets¹.

Full table

Psoriasis-specific gene regulation

In addition to identifying genes exhibiting consistent changes in independent data sets, we sought to uncover changes specific to psoriasis. To this end, we performed expression profiling of transcriptional changes in lesional versus non-lesional skin from atopic dermatitis and contact allergy, processing all samples identical to the psoriasis samples. We then filtered the entire data set to include genes regulated either exclusively in psoriasis or at least twofold less in the latter diseases. These genes (shown in Table 2) define a disease-specific pattern of gene dysregulation. Unexpectedly, the largest group of genes was related to fatty acid signaling. Thus, highly reproducible and disease-specific expression profiling suggests that aberrant fatty acid signaling is a central feature of psoriatic pathogenesis (see below).

Table 2 - Genes-regulated disease – specifically in psoriasis¹.

Full table

DC-related genes

As expected, numerous type 1 IFN-induced genes were found upregulated in lesional psoriasis (Table 2, second group). While the best characterized source of IFN- is the plasmacytoid DC subset (Gary-Gouy et al., 2002), a myeloid CD11c⁺ DC subset was recently characterized in psoriatic lesional skin (Lowes et al., 2005). Therefore, we screened the present microarray data for transcripts allowing indirect fingerprinting of DC subsets. This analysis, summarized in Table 3, revealed upregulation of transcripts highly indicative of the myeloid CD14 CD1⁺ DC subset as defined in Ahn et al. (2002) and consistent with the DC phenotype reported by Lowes et al. (2005), suggesting that this myeloid DC subset, and not plasmacytoid DCs, may in fact be the primary source of type 1 IFN secretion in psoriatic epidermis. Furthermore, the upregulation of kynuroninase and indoleamine dioxygenase points to a potential mechanism for the relative scarcity of CD4⁺ T cells in psoriatic epidermis, as those genes are involved in DC-mediated inhibition of CD4⁺ T cells (Marteau et al., 2005). Finally, the strongly upregulated chemokine CCL20 has been identified as critical chemoattractant for CD11c+ DCs into the skin (Le Borgne et al., 2006). Thus, the upregulated transcripts listed in Table 3, which are largely confirmed in independent studies, as a group defines a molecular profile of the expanded mature myeloid DC subset in psoriasis.

Table 3 - Upregulation of dendritic cell-related transcripts in psoriatic lesional skin.

Full table

Dysregulation of genes related to fatty acid signaling in psoriasis

Transcripts congruently regulated in two independent data sets were grouped into functional categories. Although this procedure is somewhat arbitrary, it clearly showed that a large group of dysregulated genes were related to fatty acid signaling (Figure 1). Several of these were related to adipogenesis, fatty acid shuttling, the peroxisomal oxidative response, and sebocyte regulation, and are detailed in Table 4. All these processes are regulated by PPAR, which was itself one of the most strongly upregulated transcripts (10.1-fold; P=0.017), confirming previous reports (Westergaard et al., 2003). Additional transcripts dysregulated in both data sets were associated with other processes regulated by PPAR, that is myogenesis and glucose turnover (listed in the Supplementary Material "gene categories"). Taken together, a large fraction of dysregulated transcripts in lesional psoriasis strongly suggests activation of PPAR, which, in addition, is also a key regulator of keratinocyte differentiation and apoptosis (Tan et al., 2003, 2004; Schmuth et al., 2004).

Figure 1.

Functional distribution of transcripts with altered expression in lesional psoriatic skin in two independent data sets. (Shown are current data and Zhou et al., 2003; total number of genes classified: n=175). Genes with known roles in multiple categories were simultaneously assigned to all relevant categories. For an annotated list of all changes, see Supplementary Material "Charité-gene categories". "Miscellaneous" contains the following categories: IL-1 related; stem cell differentiation associated; glucose metabolism; neutrophil related.

Full figure and legend (66K)

Table 4 - Genes dysregulated in psoriatic lesions related to fatty acid signaling¹.

Full table

Expression of two PPAR isoforms in vitro

Previously, a mixed nuclear and cytosolic expression of PPAR has been shown in psoriatic epidermis on the protein level (Westergaard et al., 2003). To study whether overexpression of PPAR is maintained in vitro, and to study the subcellular distribution of PPAR in vitro, we cultivated primary keratinocytes and prepared nuclear and cytosolic extracts. As shown in Figure 2a, PPAR was expressed at widely varying levels in keratinocytes, which expanded from lesional or non-lesional psoriatic skin as well as from control skin. Thus, the established overexpression of PPAR in lesional psoriasis does not represent a cell-autonomous phenotype of keratinocytes from psoriatic individuals. Interestingly, two isoforms were detected using an N-terminally directed antibody, migrating approximately 2–3 kDa apart at the expected size of 55 kDa. The shorter of these isoforms was clearly predominant in the nucleus. HaCaT keratinocytes only exhibited the smaller isoform both in cytosol and the nucleus (blot shown on right). The specificity of both bands was confirmed using peptide competition (Figure 2b). Next, we activated PPAR by addition of the synthetic ligand L-165041 in primary keratinocytes from three donors. As shown in Figure 2c, while PPAR was again found to range widely, addition of synthetic ligand did not result in consistent translocation to the nucleus. Whether the two isoforms represent differentially phosphorylated isoforms, analogous to PPAR/, is currently under investigation. Taken together, these data show that, as yet uncharacterized, PPAR isoforms are enriched in different subcellular compartments and that the subcellular localization of PPAR is independent of its activation status.

Figure 2.

Expression of PPAR and p65/RelA in primary adult keratinocytes. (a) Western blot of nuclear and cytosolic fractions from keratinocytes expanded from non-lesional, or lesional skin, as well as from healthy control skin, as indicated. Twenty micrograms of nuclear and cytosolic fractions were loaded per lane. PonceauS staining is shown on the bottom as control for even loading. (b) Cytosolic and nuclear fractions of primary keratinocytes from healthy control skin were prepared and processed exactly as in panel (a), except that anti-PPAR was pre-incubated with 10 g/ml of antigenic peptide (+peptide) for 30 minutes before incubation of the blot with antibody. Non-specific bands not competed by peptide are marked by (^*). (c) Western blot of nuclear and cytosolic fractions from keratinocytes from three control individuals stimulated with the PPAR-specific ligand L-165041 (LD) or vehicle (con) for 48 hours. Nuclear extract from HaCaT keratinocytes was included as positive control (rightmost lane). (d) The blot shown in b was re-probed using p65/RelA antibody.

Full figure and legend (120K)

Localization of NF-B is independent of PPAR

Since p65/RelA and PPAR have previously been reported to interact in a ligand-dependent manner (Westergaard et al., 2003), we examined localization of p65 at steady-state conditions, as well as upon activation of PPAR. As shown in Figure 2d, p65/RelA exhibited predominantly cytosolic localization in primary keratinocytes. Activation of PPAR did not induce nuclear accumulation of p65/RelA. We conclude that in adult epidermal keratinocytes NF-B translocation is not regulated by PPAR ligation. Therefore, heterodimerization of p65/RelA and PPAR is unlikely to contribute to PPAR activity in the epidermis.

PPAR is regulated through AP1 activation in keratinocytes

The transcriptional regulation of PPAR is complex; several trans-acting factors have been described to date. In murine keratinocytes, a major transcriptional activation has been shown to initiated by IFN-, stress-activated kinase, and TNF, all of these pathways being integrated by an AP1 site in the PPAR promoter (Tan et al., 2004). To assess the contribution of this cis-acting element, we blocked expression of two AP1 subunits active in adult epidermis, c-jun and junB, by RNA interference in HaCaT keratinocytes. As shown in Figure 3a, knockdown of junB greatly reduced steady-state PPAR expression, whereas c-jun knockdown had only a minor effect. By contrast, knockdown of -catenin, which has been reported to strongly regulate PPAR expression in colon carcinoma cells (He et al., 1999), had a negligible effect on PPAR expression (Figure 3b), both in normal and in calcium-free medium. This was not due to incomplete cell infection, as shown in Figure 3c. Thus, PPAR expression in human keratinocytes is primarily regulated via AP1 activation, specifically by junB, whereas canonical Wnt signaling has no effect.

Figure 3.

Transcriptional regulation of PPAR expression in human keratinocytes. (a) Human HaCaT keratinocytes were transfected in the absence (con) or presence of siRNA specific for junB or c-Jun, as indicated. Forty-eight hours after transfection, RNA was analyzed by RT-PCR (top panel), and protein by Western blot (bottom) for the expression of PPAR. (b) HaCaT cells grown under standard conditions or in calcium-free medium were infected with empty control lentivirus (pLL3.7) or lentivirus harboring a -catenin-specific siRNA sequence, as indicated in the figure. Forty-eight hours after infection, RNA was analyzed by RT-PCR (top) and protein by Western blot (bottom) for the expression of -catenin and PPAR. (c) FACS analysis of the E-green fluorescent protein reporter for the control virus.

Full figure and legend (79K)

PPAR enhances proliferation in keratinocytes

Keratinocyte proliferation is greatly increased in psoriasis. Since PPAR can either block or enhance cellular proliferation, depending on cellular context, we examined the effect of PPAR on proliferation in human keratinocytes. To this end, HaCaT keratinocytes, which are fast cycling under steady-state conditions, were infected with a lentivirus containing an RNAi sequence recognizing PPAR or control lentivirus. As shown in Figure 4a, the infection ratio for both vectors, quantified by FACS analysis of the green fluorescent protein reporter was approximately 94% for both vectors. However, 13 days after infection, the percentage of green fluorescent protein-positive cells had decreased significantly only after infection with PPAR^RNAi, but not after control virus, indicating a proliferative disadvantage in the absence of PPAR. In a complementary approach, adult primary epidermal keratinocytes, which are proliferating much slower than HaCaT cells, were exposed to the PPAR-specific ligands L-165041 or GW-501516. As shown in Figure 4b and c, stimulation with either ligand caused a highly significant increase in cell number in keratinocytes expanded from three independent donors. These data indicate that PPAR augments, rather than inhibits, keratinocytes proliferation, suggesting that this effect may contribute to keratinocyte hyperproliferation in psoriasis.

Figure 4.

Effect of PPAR on keratinocyte proliferation. (a) HaCaT keratinocytes were infected with control virus (pLL3.7) or lentivirus harboring a PPAR-specific siRNA sequence (PPAR^RNAi). The frequency of infected cells was determined by FACS analysis of the E-green fluorescent protein reporter gene. (b) Proliferation of adult primary keratinocytes in vitro grown for 7 days in the presence of 1 M of the PPAR-specific ligand L-165041 (shaded columns) or vehicle (white). Data represent meanSD of experiments performed in duplicate or triplicate with cells from three independent donors. ^*P<0.01. (c) As in panel (b) except that the PPAR-specific ligand GW-501516 was used at 5 M concentration.

Full figure and legend (90K)

Heparin-binding EGF-like growth factor is a target of PPAR

We next sought to identify target genes activated by PPAR activation involved in enhanced proliferation. To this end, primary keratinocytes from four independent donors were grown in the presence of L-165041, or vehicle, for 48 hours. Total gene expression was then analyzed using the PIQUOR SkinPatho array, which contains approximately 1,100 transcripts relevant to epidermal biology. The magnitude of gene induction by PPAR has been found to lie in the range of 1.5–3-fold in previous microarray studies (Tan et al., 2001; Tanaka et al., 2003; Tachibana et al., 2005). Since this moderate induction is close to the background level, we applied a threshold for a minimum fold change of greater than three standard deviations of the mean fold change observed across all genes (illustrated in Figure 5a). Although this conservative approach reduced sensitivity, thus preventing detection of further potential target genes, it largely eliminates identification of false positives. Table 5 summarizes the observed changes. Several of the listed genes pertaining to fatty acid metabolism have previously been identified as targets in other cell types (Tanaka et al., 2003; Tachibana et al., 2005), thus confirming the present data. One gene, HB-EGF was of particular interest because it is also upregulated 3.6-fold in vivo in lesional psoriasis (Table 5) and because it is known to enhance keratinocyte proliferation. We, therefore, confirmed the observed upregulation of HB-EGF by quantitative real-time PCR. As a positive control, we used CIDE-A, an proapoptotic gene previously shown to be indirectly downregulated by PPAR (Tan et al., 2001). Expression of PPAR itself was used as negative control. As shown in Figure 5b, both HB-EGF and CIDE-A were moderately, but statistically significant induced by PPAR activation in four independent patients, whereas expression of PPAR itself did not change, in confirmation of the microarray data. On the protein level, HB-EGF expression was increased on activation of PPAR (Figure 5c, left) and decreased after lentivirus-mediated knockdown of PPAR in primary keratinocytes (Figure 5c, right). Finally, we addressed whether HB-EGF is directly or indirectly induced by PPAR. To this end, primary keratinocytes were incubated in the absence, or presence, of L-165041, with or without cycloheximide. Subsequently, HB-EGF expression was determined by reverse transcription-PCR (RT-PCR). ACADVL, an established PPAR target gene was amplified as positive control. CIDE-A, previously shown to be indirectly induced by PPAR (Tan et al., 2001), was used as a negative control. As shown in Figure 5d, induction of CIDE-A by PPAR activation could be suppressed, whereas this was not the case for ACADVL and HB-EGF. Thus, HB-EGF is a direct target gene of PPAR.

Figure 5.

HB-EGF is a PPAR target gene. (a) Scatter plot of mean fold changes induced by stimulation of primary adult epidermal keratinocytes with the PPAR-specific ligand L-165041. Shaded horizontal lines indicate the threshold applied for significant changes (three SD of the mean change across all genes). Error bars indicate SD for all genes exceeding the threshold. Open symbols denote genes listed in Table 5. (b) Left, quantitative PCR showing gene expression relative to hypoxanthine ribosyltransferase in primary keratinocytes from four donors stimulated with L-165041 for 48 hours. Right, meanSD of the changes for all individuals. ^*P<0.01; ^**P<0.001. (c) Western blot analysis: left, primary keratinocytes were stimulated as in panel (b), whole-cell lysates separated by SDS-PAGE and blots probed with anti-HB-EGF. PonceauS staining is shown as a loading control. Right, primary keratinocytes were infected either with empty control virus (vector) or virus containing a PPAR RNAi-sequence (PPAR^RNAI) for 48 hours before cell harvesting. (d) RT-PCR analysis of primary keratinocytes were stimulated with L-165041, or vehicle (0.5% DMSO), or 5 M cycloheximide (CHX) for 24 hour.

Full figure and legend (119K)

Table 5 - Genes induced by PPAR activation in keratinocytes¹.

Full table

Discussion

Several independent lines of evidence implicate PPAR in psoriasis. PPAR is highly, and specifically, upregulated in psoriasis. The activity profile of this nuclear hormone receptor includes antiapoptotic and proangiogeneic effects, as well as regulation on keratinocyte differentiation. PPAR is activated by TNF, a central mediator of the inflammatory response in psoriasis. Its interaction partners FABP5 and cellular retinoic acid-binding protein II are upregulated in psoriasis (Table 2). To this list, we here add (1) that a large group of dysregulated genes in psoriasis lesions in vivo are related to PPAR function; (2) that PPAR enhances keratinocyte proliferation; (3) that it directly induces HB-EGF. The role of PPAR on proliferation is, in fact, highly controversial and other recent papers report a growth-inhibitory effect (Burdick et al., 2007). However, using primary adult epidermal human keratinocytes, we observe a growth-stimulatory role using two different PPAR ligands. Since the stimulatory time period used was 7 days, the observed increase in cell number is highly unlikely to result exclusively from inhibited apoptosis. Therefore, it appears that PPAR activation in human keratinocytes indeed enhances proliferation. HB-EGF represents a highly relevant target, since it drives keratinocyte proliferation, regulates wound healing, and induces epidermal hyperplasia, thus replicating important elements of the psoriatic phenotype in vivo (Zheng et al., 2003; Kimura et al., 2005; Shirakata et al., 2005). Moreover, we define two regulatory aspects of PPAR activity in keratinocytes: its expression is independent of canonical Wnt signaling and its localization and activity is independent of p65/RelA.

A discussion of all disease-specific changes in psoriasis (Table 3) is beyond the scope of the current report. However, the genes involved in fatty acid signaling as a group (Table 3, top section) constitute a disease-specific determinant of PPAR activity in psoriasis. Thus, FABP5 and cellular retoinic acid-binding protein II regulate PPAR activity by fatty acid ligands. Moreover, FABP5 enhances keratinocyte motility in wounds (Kusakari et al., 2006). 12R-Lipoxygenase catalyzes the formation of 12R-HETE, which specifically accumulates in psoriasis and has been shown to activate PPAR (Westergaard et al., 2003). Lipocalin 2 is junB induced, modulates adipocyte differentiation, and wound healing (Florin et al., 2006). Dio2 regulates brown fatty tissue homeostasis (Christoffolete et al., 2004). RARRES1, downregulated in psoriasis (Table 3), blocks keratinocyte proliferation (DiSepio et al., 1998). Carnitine acyltransferase regulates fatty acid -oxidation. PPP1R3C (PTG) regulates glycogen metabolism (Printen et al., 1997). The serine/threonine kinase PRK1, found to be downregulated, inhibits phosphorylation of PKB/AKT by lipids (Wick et al., 2000; Lim et al., 2005), which is activated by PPAR. Taken together, these genes underscore the complex regulation of adipogenesis, lipid, and glycogen metabolism exerted by PPAR, and reinforce the concept of psoriasis as a dysregulated wound response program.

The unexpectedly large group of dysregulated genes involved in fatty acid metabolism positions intermediary metabolism as a central aspect of the disease, connected to, but distinct from its inflammatory nature. Thus, increased body mass index is an independent risk factor for psoriasis (Naldi et al., 2005) and may have an effect on disease severity (Marino et al., 2004), as well as TNF-related inflammatory activity (Hamminga et al., 2006). Since the prevalence of metabolic syndrome is increased in psoriasis (Sommer et al., 2006), a combination of insulin resistance, obesity, or chronic inflammation may trigger the expression of PPAR, which in turn contributes to a non-terminated regenerative skin phenotype. This disease mechanism would be expected to be aggravated by acute inflammation, or stress via the induction of PPAR by TNF and stress-activated kinase (Tan et al., 2001). Clinically, these factors are precisely the most potent triggers of disease flares. Thus, activation of PPAR in psoriasis offers a pathogenetic concept unifying various clinical aspects as well as transcriptional changes in psoriasis. Therapeutically, since caloric restriction can decrease the expression level of PPAR isoforms (Masternak et al., 2005), trials to this end appear warranted to explore the efficacy of caloric restriction as a treatment option in appropriately selected psoriasis patients. Conversely, currently on-going trials employing PPAR agonists for the treatment of metabolic syndrome and obesity should be carefully scrutinized for increased incidence of psoriasis.

Mechanistically, PPAR may act as an antagonist to other nuclear hormone receptor heterodimers. Thus, the RXR/RAR heterodimer activates, rather than inhibits apoptosis (Chiba et al., 1997; Monczak et al., 1997). Another competitive antagonism may be exerted toward PPAR, since activation of this isoform directly inhibits STAT3 (Wang et al., 2004) which, when overexpressed, causes a psoriasis-like phenotype in vivo (Sano et al., 2005). Furthermore, the antiapoptotic activity of PPAR may be most pronounced in the context of type 1 IFN activity, since PPAR is expressed in DCs (Jakobsen et al., 2006). In fact, we have observed that PPAR can be induced by IFN- (N.a.Y and J.F., paper in preparation), adding yet further support for a role of PPAR in psoriasis.

The upregulation and activity of PPAR in lesional psoriasis suggests how the psoriasis-like phenotype in K14-vascular endothelial growth factor transgenic mice relates to the human disease, since vascular endothelial growth factor is a target gene of PPAR (Xia et al., 2003; Sano et al., 2005). It is also intriguing that, in contrast to human skin, PPAR is not expressed in adult interfollicular mouse epidermis, perhaps explaining why psoriasis form dermatitis does not occur spontaneously in mice. Finally, it should be pointed out that the PPAR gene is located 4.2 Mb telomeric of the putative PSORS1 locus at HLA-C on chromosome 6p21.2. Since earlier genome-wide scans were performed using microsatellites located actually closer to the PPAR locus than to HLA-C and yielding a positive signal (Nair et al., 1997; Trembath et al., 1997; Enlund et al., 1999), PPAR formally constitutes a candidate susceptibility gene for the disease.

Materials and Methods

Patients

All work with patient materials presented in the current report was carried out in accordance with the Declaration of Helsinki Principles. All patients undergoing biopsy gave prior written consent. Psoriasis and atopic dermatitis patients sampled for expression profiling were undergoing in-patient treatment for active disease at the time of biopsy. Contact allergy was elicited by standard patch testing for 72 hours on the back.

Expression profiling

Full-thickness lesional and non-lesional 5-mm punch biopsies were snap-frozen in liquid nitrogen. RNA was purified using the RNeasy kit (Qiagen, Hilden, Germany). Reverse transcription, second-strand synthesis, cRNA labeling, and hybridization to the Affimetrix U95A array was performed according to Mishra et al. (2002). Microarray processing is detailed in the supplement (legend to Figure S1).

Cell culture

Expansion of adult epidermal keratinocytes on feeder layer was performed modified according to Rheinwald and Green (1975). The detailed protocol is specified in the Supplementary Material. Briefly, full-thickness punch biopsies were incubated with dispase solution overnight at 4°C. Epidermal sheets were carefully removed and incubated in trypsin solution for 15 minutes at 37°C. Trypsin was inactivated by addition of 5 ml keratinocyte medium (DMEM low glucose; HAM's F12 adenine PenStrep plus HICE cocktail (hydrocortisone, insulin, cholera toxin, and EGF)). Cells were re-suspended and seeded on plates with prepared feeder layer (consisting of mitomycin C-treated adult fibroblasts). Medium was first replaced after 2 days. At confluence (7–9 days), cells were split 1:2 or 1:3, depending on confluence, and re-seeded on fresh feeder layer. After a second passage 7 days later (splitting ratio 1:2), cells were subjected to experiments. HaCaT keratinocytes were cultured in RPMI 1640-containing GLUTAMAX™-I (substituted on a molar equivalent basis for L-glutamine), 100 U/ml Pen/Strep (Invitrogen-Gibco, Karlsruhe, Germany), 10% fetal calf serum, and 0.4 mM calcium.

Synthetic ligand stimulation

Cells were seeded 5 10⁵ cells/58 cm² dish and incubated overnight at 37°C in a humidified atmosphere of 5% CO₂. The selective PPAR agonists L-165041 and GW-501516 (Calbiochem, Darmstadt, Germany) were diluted and aliquoted in DMSO under N₂ atmosphere. Cells were stimulated with 1 M (L-165041) or 5 M (GW-501516) for the time span specified in Results. Medium with fresh ligand was added every 24 hours. Control cells received identical volume of vehicle (0.05% v/v final DMSO).

Western blotting

Nuclear and cytoplasmic protein fractions were prepared using NE-PER reagents (Pierce, Rockford, IL, no. 78833). For HB-EGF blots, whole cell extracts were prepared with buffer containing CHAPS exactly as described in Klose (1999). Protein extracts were aliquoted, shock-frozen in liquid N₂ and stored at -80°C. Protein concentrations were determined using BCA Protein Assay kit (Pierce, no. 23225). Twenty micrograms of protein extract was loaded per lane, separated on 7.5% acrylamide-SDS gels, and blotted onto cellulose-nitrate membranes (Whatman Schleicher & Schell, Dassel, Germany, no. 10 402 594). Blots were blocked for 30 minutes with 4% non-fat milk in Tris-buffered saline with 0.05% Tween 20, and subsequently incubated with primary and secondary antibodies in 4% non-fat milk/Tris-buffered saline with 0.05% Tween 20 as follows. PPAR: rabbit anti-human PPAR polyclonal purified IgG (Cayman Chemical Europe, Tallinn, Estonia, no. 101720; antigenic peptide corresponds to aa 39-54 (SSSYTDLSRSSSPPSL)), diluted at 1:300 overnight at 4°C; p65/RelA: (Santa Cruz Biotechnology, Santa Cruz Biotech, CA, no. sc-109G) 1:1,000, overnight at 4°C. HB-EGF: (R&D Systems, Minneapolis, MN, AF-259-NA) 1:500, overnight at 4°C. Secondary antibodies were anti-rabbit-HRP (Amersham Biosciences, Buckinghamshire, England, no. NA934) at 1:2,000, anti-goat-HRP (Santa Cruz Biotechnology, no. SC-2922) 1:6,000. Incubation with appropriate secondary antibody was carried for 1 hour at room temperature. Blots were developed with ECL Plus Western Blotting Detection system (Amersham Biosciences, no. RPN2 132).

RT-PCR

RNA extraction was performed using the NucleoSpin II extraction kit (Machery-Nagel, Düren, Germany, no. 740 955). All reagents used for cDNA synthesis (SuperScript II Reverse Transcriptase, dNTP, RnaseOut, Oligo(dT)_12–18, dithiothreitol) were purchased from Invitrogen-Gibco (Karlsruhe, Germany). GoTaq DNA Polymerase (Promega, Mannheim, Germany, no. M3001) was used at 1 U/25 l for amplification. The annealing temperature for all PCR reactions was 55°C except for PPAR (54°C). Primer sets, number of cycles, and amplicon lengths are detailed in the Supplementary Methods. For quantitative real-time PCR, cDNA synthesis was performed using Moloney Murine Leukemia Virus Reverse Transcriptase (MMLV-RT). Amplification was performed in a TaqMan SDS7700, using Assay-on-Demand kits (ABI) for hypoxanthine ribosyltransferase (Hs99999909_m1); CIDE-A (Hs00154455_m1), HB-EGF (Hs00181813_m1), and PPAR (Hs00602622_m1) according to the manufacturer's instructions. Reactions were performed in triplicate and expression levels normalized to hypoxanthine ribosyltransferase.

PPAR target gene identification

Primary keratinocyte culture, stimulation with L-165041, and RNA preparation was performed as detailed above. Samples were labeled either with Cy3 or Cy5, hybridized to Piquor SkinPatho microarrays (Memorec, Cologne, Germany), and scanned by the Charité in-house microarray service facility. Dye swab experiments were performed to rule out uneven green/red intensities. Data with at least twice background intensity and %CV of less than 30% between the four replicas on each array were further analyzed as detailed in Table 5.

Lentiviral transfer vectors

pLL3.7 [14] was a kind gift of Luk van Parijs. Short hairpin RNAi sequence for PPAR: (5'-tGTCGTACGATCCGCATGAAttcaagagaTTCATG-CGGATCGTACGACttttttc) or -catenin: (5'-tGCTGAAACATGCAGTTGTAttcaagaga-TACAACTGCATGTTTCAGCttttttc; target sequence underlined) were synthesized in both strands, annealed, and cloned immediately downstream of the U6 promoter using HpaI/XhoI restriction sites. Parent plasmid served as control. An alternative shRNAi sequence for PPAR (CCACTACGGTGTTCATGC) was synthesized analogously with MluI and ClaI restriction sites and cloned between under the control of H1 promoter of pLVTH (Szulc et al., 2006), which was kindly provided by Didier Trono, as well as pPAX2 and pMD2.G for second generation packaging. Production of VSV-G pseudotyped replication-deficient lentivirus was carried out according to Mitta et al. (2005). Lentivirus infection was performed at a multiplicity of infection (MOI) of three for 12 hours at 37°C.

Conflict of Interest

The authors state no conflict of interest.

References

1. Ahn JH, Lee Y, Jeon C, Lee SJ, Lee BH, Choi KD et al. (2002) Identification of the genes differentially expressed in human dendritic cell subsets by cDNA subtraction and microarray analysis. Blood 100:1742–1754 | PubMed | ISI | ChemPort |
2. Baudry A, Yang ZZ, Hemmings BA (2006) PKBalpha is required for adipose differentiation of mouse embryonic fibroblasts. J Cell Sci 119:889–897 | Article | PubMed | ISI | ChemPort |
3. Bowcock AM, Shannon W, Du F, Duncan J, Cao K, Aftergut K et al. (2001) Insights into psoriasis and other inflammatory diseases from large-scale gene expression studies. Hum Mol Genet 10:1793–1805 | Article | PubMed | ISI | ChemPort |
4. Boyman O, Hefti HP, Conrad C, Nickoloff BJ, Suter M, Nestle FO (2004) Spontaneous development of psoriasis in a new animal model shows an essential role for resident T cells and tumor necrosis factor-alpha. J Exp Med 199:731–736 | Article | PubMed | ISI | ChemPort |
5. Burdick AD, Bility MT, Girroir EE, Billin AN, Willson TM, Gonzalez FJ et al. (2007) Ligand activation of peroxisome proliferator-activated receptor-beta/delta(PPARbeta/delta) inhibits cell growth of human N/TERT-1 keratinocytes. Cell Signal.
6. Butz CE, McClelland GB, Brooks GA (2004) MCT1 confirmed in rat striated muscle mitochondria. J Appl Physiol 97:1059–1066 | Article | PubMed | ISI | ChemPort |
7. Chen H, Weng YC, Schatteman GC, Sanders L, Christy RJ, Christy BA (1999) Expression of the dominant-negative regulator Id4 is induced during adipocyte differentiation. Biochem Biophys Res Commun 256:614–619 | Article | PubMed | ISI | ChemPort |
8. Chiba H, Clifford J, Metzger D, Chambon P (1997) Distinct retinoid X receptor-retinoic acid receptor heterodimers are differentially involved in the control of expression of retinoid target genes in F9 embryonal carcinoma cells. Mol Cell Biol 17:3013–3020 | PubMed | ISI | ChemPort |
9. Christoffolete MA, Linardi CC, de Jesus L, Ebina KN, Carvalho SD, Ribeiro MO et al. (2004) Mice with targeted disruption of the Dio2 gene have cold-induced overexpression of the uncoupling protein 1 gene but fail to increase brown adipose tissue lipogenesis and adaptive thermogenesis. Diabetes 53:577–584 | Article | PubMed | ISI | ChemPort |
10. Di-Poi N, Michalik L, Tan NS, Desvergne B, Wahli W (2003) The anti-apoptotic role of PPARbeta contributes to efficient skin wound healing. J Steroid Biochem Mol Biol 85:257–265 | Article | PubMed | ISI | ChemPort |
11. DiSepio D, Ghosn C, Eckert RL, Deucher A, Robinson N, Duvic M et al. (1998) Identification and characterization of a retinoid-induced class II tumor suppressor/growth regulatory gene. Proc Natl Acad Sci USA 95:14811–14815 | Article | PubMed | ChemPort |
12. Enlund F, Samuelsson L, Enerback C, Inerot A, Wahlstrom J, Yhr M et al. (1999) Psoriasis susceptibility locus in chromosome region 3q21 identified in patients from southwest Sweden. Eur J Hum Genet 7:783–790 | Article | PubMed | ISI | ChemPort |
13. Florin L, Knebel J, Zigrino P, Vonderstrass B, Mauch C, Schorpp-Kistner M et al. (2006) Delayed wound healing and epidermal hyperproliferation in mice lacking JunB in the skin. J Invest Dermatol 126:902–911 | Article | PubMed | ISI | ChemPort |
14. Gandarillas A, Goldsmith LA, Gschmeissner S, Leigh IM, Watt FM (1999) Evidence that apoptosis and terminal differentiation of epidermal keratinocytes are distinct processes. Exp Dermatol 8:71–79 | Article | PubMed | ISI | ChemPort |
15. Gary-Gouy H, Lebon P, Dalloul AH (2002) Type I interferon production by plasmacytoid dendritic cells and monocytes is triggered by viruses, but the level of production is controlled by distinct cytokines. J Interferon Cytokine Res 22:653–659 | Article | PubMed | ISI | ChemPort |
16. Ge L, Gordon JS, Hsuan C, Stenn K, Prouty SM (2003) Identification of the delta-6 desaturase of human sebaceous glands: expression and enzyme activity. J Invest Dermatol 120:707–714 | Article | PubMed | ISI | ChemPort |
17. Gill RK, Saksena S, Alrefai WA, Sarwar Z, Goldstein JL, Carroll RE et al. (2005) Expression and membrane localization of MCT isoforms along the length of the human intestine. Am J Physiol 289:C846–C852 | Article | ISI | ChemPort |
18. Hamminga EA, van der Lely AJ, Neumann HA, Thio HB (2006) Chronic inflammation in psoriasis and obesity: implications for therapy. Med Hypotheses 67:768–773 | Article | PubMed | ISI | ChemPort |
19. He TC, Chan TA, Vogelstein B, Kinzler KW (1999) PPAR is an APC-regulated target of nonsteroidal anti-inflammatory drugs. Cell 99:335–345 | Article | PubMed | ISI | ChemPort |
20. Hellemans K, Michalik L, Dittie A, Knorr A, Rombouts K, De Jong J et al. (2003) Peroxisome proliferator-activated receptor-beta signaling contributes to enhanced proliferation of hepatic stellate cells. Gastroenterology 124:184–201 | Article | PubMed | ISI | ChemPort |
21. Hemmrich K, von Heimburg D, Cierpka K, Haydarlioglu S, Pallua N (2005) Optimization of the differentiation of human preadipocytes in vitro. Differentiation Research Biol Diversity 73:28–35 | ChemPort |
22. Hobbs RM, Silva-Vargas V, Groves R, Watt FM (2004) Expression of activated MEK1 in differentiating epidermal cells is sufficient to generate hyperproliferative and inflammatory skin lesions. J Invest Dermatol 123:503–515 | Article | PubMed | ISI | ChemPort |
23. Hoepfner D, Schildknegt D, Braakman I, Philippsen P, Tabak HF (2005) Contribution of the endoplasmic reticulum to peroxisome formation. Cell 122:85–95 | Article | PubMed | ISI | ChemPort |
24. Icre G, Wahli W, Michalik L (2006) Functions of the peroxisome proliferator-activated receptor (PPAR) alpha and beta in skin homeostasis, epithelial repair, and morphogenesis. J Invest Dermatol 126:Suppl:30–35 | Article |
25. Itoh K, Kawasaki S, Kawamoto S, Seishima M, Chiba H, Michibata H et al. (2005) Identification of differentially expressed genes in psoriasis using expression profiling approaches. Exp Dermatol 14:667–674 | Article | PubMed | ISI | ChemPort |
26. Jakobsen MA, Petersen RK, Kristiansen K, Lange M, Lillevang ST (2006) Peroxisome proliferator-activated receptor alpha, delta, gamma1 and gamma2 expressions are present in human monocyte-derived dendritic cells and modulate dendritic cell maturation by addition of subtype-specific ligands. Scand J Immunol 63:330–337 | Article | PubMed | ISI | ChemPort |
27. Jarvis MC, Gray TJ, Palmer CN (2005) Both PPARgamma and PPAR influence sulindac sulfide-mediated p21WAF1/CIP1 upregulation in a human prostate epithelial cell line. Oncogene 24:8211–8215 | Article | PubMed | ISI | ChemPort |
28. Kedishvili NY, Popov KM, Jaskiewicz JA, Harris RA (1994) Coordinated expression of valine catabolic enzymes during adipogenesis: analysis of activity, mRNA, protein levels, and metabolic consequences. Arch Biochem Biophys 315:317–322 | Article | PubMed | ISI | ChemPort |
29. Kimura R, Iwamoto R, Mekada E (2005) Soluble form of heparin-binding EGF-like growth factor contributes to retinoic acid-induced epidermal hyperplasia. Cell Struct Funct 30:35–42 | Article | PubMed | ISI | ChemPort |
30. Klose J (1999) Fractionated extraction of total tissue proteins from mouse and human for 2-D electrophoresis. Methods Mol Biol 112:67–85 | PubMed | ChemPort |
31. Kohlmann WM, Urban W, Sterry W, Foerster J (2004) Correlation of psoriasis activity with abundance of CD25+CD8+ T cells: conditions for cloning T cells from psoriatic plaques. Exp Dermatol 13:607–612 | Article | PubMed | ISI |
32. Kojo H, Fukagawa M, Tajima K, Suzuki A, Fujimura T, Aramori I et al. (2003) Evaluation of human peroxisome proliferator-activated receptor (PPAR) subtype selectivity of a variety of anti-inflammatory drugs based on a novel assay for PPAR delta(beta). J Pharmacol Sci 93:347–355 | Article | PubMed | ISI | ChemPort |
33. Kuenzli S, Saurat JH (2003) Effect of topical PPARbeta/delta and PPAR agonists on plaque psoriasis. A pilot study. Dermatology 206:252–256 | Article | PubMed | ISI | ChemPort |
34. Kulski JK, Kenworthy W, Bellgard M, Taplin R, Okamoto K, Oka A et al. (2005) Gene expression profiling of Japanese psoriatic skin reveals an increased activity in molecular stress and immune response signals. J Mol Med 83:964–975 | Article | PubMed | ISI | ChemPort |
35. Kusakari Y, Ogawa E, Owada Y, Kitanaka N, Watanabe H, Kimura M et al. (2006) Decreased keratinocyte motility in skin wound on mice lacking the epidermal fatty acid binding protein gene. Mol Cell Biochem 284:183–188 | Article | PubMed | ISI | ChemPort |
36. Le Borgne M, Etchart N Goubier A, Lira SA, Sirard JC, van Rooijen N et al. (2006) Dendritic cells rapidly recruited into epithelial tissues via CCR6/CCL20 are responsible for CD8+ T cell crosspriming in vivo. Immunity 24:191–201 | Article | PubMed | ISI | ChemPort |
37. Lechpammer S, Epperly MW, Zhou S, Nie S, Glowacki J, Greenberger JS (2005) Adipocyte differentiation in Sod2(-/-) and Sod2(+/+) murine bone marrow stromal cells is associated with low antioxidant pools. Exp Hematol 33:1201–1208 | Article | PubMed | ISI | ChemPort |
38. Lee CH, Kang K, Mehl IR, Nofsinger R, Alaynick WA, Chong LW et al. (2006) Peroxisome proliferator-activated receptor delta promotes very low-density lipoprotein-derived fatty acid catabolism in the macrophage. Proc Natl Acad Sci USA 103:2434–2439 | Article | PubMed | ChemPort |
39. Leonard AE, Bobik EG, Dorado J, Kroeger PE, Chuang LT, Thurmond JM et al. (2000) Cloning of a human cDNA encoding a novel enzyme involved in the elongation of long-chain polyunsaturated fatty acids. Biochem J 350 Part 3:765–770 | Article | PubMed | ISI |