Abstract
Because some investigational peroxisome proliferator-activated receptors (PPAR) agonists cause tumors in the lower urinary tract of rats, we compared normal human and rat urothelium in terms of PPAR and retinoid X receptor (RXR) expression and proliferation-associated phenotypes. In situ, few human but most rat urothelial cells were Ki67 positive, indicating fundamental differences in cell cycle control. Rat and human urothelia expressed all 3 PPAR and the RXRα and RXRβ isoforms in a predominantly nuclear localization, indicating that they may be biologically active. However, immunolocalization differences were observed between species. First, whereas PPARα and PPARβ/δ were expressed throughout the human bladder or ureteric urothelium, in the rat urothelium PPARα was primarily, and PPARβ/δ exclusively, restricted to superficial cells. Second, RXRβ was restricted to intermediate and superficial layers of the human urothelium but tended to be absent from the rat superficial cells. Third, PPARγ expression was present throughout the urothelia of both species but was most intense in the superficial human urothelium. Species differences were also observed in the expression of PPAR and RXR isoforms between cultured rat and human urothelial cells and in the smooth muscle. Our findings highlight the unique coexpression of multiple PPAR and RXR isoforms by urothelium and suggest that species differences in PPAR function between rat and human urothelia may be explored in an in vitro setting.
Introduction
Peroxisome proliferator-activated receptors (PPAR) are nuclear hormone receptors, which function as transcription factors to regulate a diverse range of functions (Kersten et al., 2000). Activation of PPAR results in heterodimerization with the 9-cis retinoic acid receptor (RXR), which binds to peroxisome proliferator response elements (PPRE) to activate transcription of target genes. Three PPAR isoforms exist, which are products of distinct genes. These isoforms, termed PPARα, PPARβ/δ, and PPARγ, vary in their tissue distribution and transcriptional activities (Dreyer et al., 1992; Kersten et al., 2000).
Selective PPARα ligands (fibrates) are used clinically as antilipidemic agents, and selective PPARγ agonists (thiazolidinediones; TZDs) are used as insulin sensitizers (Staels and Fruchart, 2005; Leiter, 2006). Selective PPARβ/δ agonists are also being explored for treatment of type II diabetes (Chang et al., 2007). Capitalizing on the beneficial effects of selective PPAR agonists in diabetes, dual-acting PPARα/γ agonists (-glitazars) have been developed, which exhibit improved insulin sensitivity and lipid lowering effects over subtype-specific agonists (Lohray et al., 2001; Brand et al., 2003; Larsen et al., 2003; Ye et al., 2003; Saad et al., 2004; Pickavance et al., 2005).
The continued development of TZDs and dual-acting “-glitazars” has been complicated by carcinogenic effects in rodents (El-Hage, 2004; Cohen, 2005). These included hemangiosarcomas, liposarcomas, hepatomas, and transitional cell carcinomas in the urothelium of the urinary bladder/renal pelvis (El-Hage, 2004). PPAR agonists tested to date have all shown to be nongenotoxic (El-Hage, 2004). Thus, the cancers seen in PPAR agonist–treated rodents likely arise through nongenotoxic mechanisms. Accordingly, the hepatocarcinogenesis induced by PPARα agonists in rodents has been shown to be mediated by the murine PPARα (Peters et al., 2005). Furthermore, the hepatocarcinogenic effect in rodents is not of human relevance, due to structural differences between human and mouse PPARα proteins and functional differences between the human and mouse liver (Morimura et al., 2006).
Dual-acting “-glitazars” appear more potent than PPARγ agonists at inducing tumors associated with the lower urinary tract (LUT) in rats. Some studies suggested an indirect effect, in which bladder cancer resulted from regenerative responses to damage induced by precipitated urinary deposits in male rats (Cohen, 2005; Dominick et al., 2006; Tannehill-Gregg et al., 2007). However, others have suggested a direct effect of PPAR agonists on the rat bladder epithelial (urothelial) lining, based on rapid changes in urothelial gene expression and intracellular signaling, as well as early urothelial hypertrophy in treated rats (Egerod et al., 2005; Oleksiewicz et al., 2005).
To discriminate between direct and indirect PPAR agonist effects, knowledge about PPAR isoform expression in the rat urothelium is required. Furthermore, to extrapolate from the rat findings, it may be helpful to understand the equivalence between the rat and human urothelium with respect to the expression of receptors and response to PPAR signaling. As the first step toward this aim, we have compared the expression and localization of the known PPAR and RXR receptors in normal rat and normal human urothelial tissues in situ and in vitro. Furthermore, we have extended the previously described method for culturing normal human urothelial (NHU) cells to rats (Southgate et al., 1994; Southgate et al., 2002) and compared PPAR and RXR receptor localization patterns between short-term cultures of rat and human urothelial cells.
Materials and Methods
Antibodies
Primary antibodies used were selected as cross-reacting with human and rat tissues from the suppliers’ datasheets. The only exceptions were the antibodies against human PPARα and β/δ, which were raised against highly conserved peptides and in preliminary studies were found to show appropriate immunolabeling patterns on rat and human tissues. Controls were included in all immunochemistry experiments and included use of no primary and irrelevant primary antibody controls.
PPARα rabbit affinity–purified antipeptide antibody from Affinity Bioreagents (Golden, CO; catalogue number PA1-822A) was used at 1 μg.ml−1 for immunohistochemistry and 5 μg.ml−1 for immunofluorescence.
PPARβ/δ rabbit affinity purified antipeptide antibody from Affinity Bioreagents (catalogue number PA1-823A) was used at 1 μg.ml−1 for immunohistochemistry and 5 μg.ml−1 for immunofluorescence.
PPARγ mouse monoclonal antibody clone E8 from Santa Cruz Biotechnology (supplied by Autogen Bioclear, Colne, UK; catalogue number SC-7273) was used at 100 ng.ml−1 for immunohistochemistry. PPARγ rabbit monoclonal clone 81B8 from Cell Signaling Technology (supplied by New England Biolabs UK, Hitchin, UK; catalogue number 2443) was used at 500 ng.ml−1 for immunofluorescence.
RXRα rabbit affinity–purified antipeptide antibody code D-20 from Santa Cruz Biotechnology (catalogue number SC-553) was used at 100 ng.ml−1 for immunohistochemistry and 5 μg.ml−1 for immunofluorescence.
RXRβ rabbit immunoglobulin from Upstate Biotechnology (Chandlers Ford, UK; catalogue number 06-527) was used at 625 ng.ml−1 for immunohistochemistry. RXRβ affinity-purified antipeptide antibody code C-20 from Santa Cruz Biotechnology (catalogue number SC-831) was used at 4 μg.ml−1 for immunofluorescence.
Cytokeratin 20 (CK20) mouse monoclonal clone Ks20.8 from Novacastra (Newcastle Upon Tyne, UK; catalogue number NCL-CK20) was used for immunohistochemistry at 5μg.ml−1.
Cytokeratin 7 (CK7) mouse monoclonal clone LP1K, a gift from Cancer Research UK (London), was used at 1:2000 for immunohistochemistry.
UPIIIa mouse monoclonal clone AU1 from Progen Biotechnik (Heidelberg, Germany; catalogue number 651108) was used at 1:100 for immunohistochemistry.
Ki67 mouse monoclonal clone MM1 from Novocastra (catalogue number NCL-L-Ki67) was used at 500 ng.ml−1.
Tissues
Normal Human Urothelium
The collection of surgical specimens was approved by the relevant Local Research Ethics Committees and had full, informed patient consent. Surgical specimens of normal urothelium were obtained from patients with no histological evidence of urothelial dysplasia or malignancy (Table 1). Tissues were collected in transport medium, consisting of Hank’s balanced salt solution (HBSS) containing 10 mM HEPES pH 7.6 and 20 KIU aprotinin (Trasylol, Bayer plc, Newbury, UK), as described previously (Southgate et al., 1994; Southgate et al., 2002). Tissues were either fixed in 10% formalin for 24 hours and transferred to 70% (v/v) ethanol before processing into paraffin wax for immunohistology or used to establish finite cell lines of normal human urothelial (NHU) cells.
NHU Cell Culture
NHU cell lines of finite lifespan were established from resection specimens of ureteric urothelium and maintained in complete keratinocyte-serum free medium (KSFMc) consisting of KSFM supplemented with bovine pituitary extract, epidermal growth factor at the manufacturer’s recommended concentrations (Invitrogen, Paisley, UK), and cholera toxin (30 ng.ml−1, Sigma Aldrich, Poole, UK). The preparation, maintenance, and characterization of NHU cell cultures has been previously detailed, including comparison of cultures derived from ureteric and bladder sources (Southgate et al., 1994; Southgate et al., 2002). NHU cell lines were used between passages 2 and 5.
Normal Rat Urothelium
Normal rats were anesthetized with medical grade CO2 and euthanized by cervical dislocation, in accordance with UK Home Office regulations. Thereafter, urinary bladders were rapidly excised and collected either into 10% (v/v) formalin for 24 hours and transferred to 70% (v/v) ethanol before processing into paraffin wax or into ice-cold transport medium for cell culture. The rodent tissues used in this study are detailed in Table 1.
Rat Urothelial Cell Culture
Under aseptic conditions, rat bladders from male 9-month Wistar rats were carefully dissected into smaller pieces and placed in stripping medium consisting of HBSS (without Ca2+ or Mg2+) with 10 mM HEPES pH 7.6, 20 KIU Trasylol, and 0.1% (w/v) EDTA. After 4 hours rotation at 37°C, the urothelium was carefully stripped as sheets from the underlying stroma, collected by centrifugation at 400 g for 5 minutes, resuspended in HBSS containing 250 mg.ml−1 collagenase type IV (from Clostridium histolyticum; Sigma Aldrich) and incubated at 37°C for 30 minutes. Urothelial sheets were collected and disaggregated with gentle pipetting, and following centrifugation, the urothelial cells were resuspended in KSFMc and seeded at 0.25 × 106 cells per 25 cm2 Primaria® tissue culture flasks (BD Biosciences, Oxford, UK). Growth medium was changed every 3 days, and cultures were passaged at 80% to 90% confluence, as described for NHU cells (Southgate et al., 1994; Southgate et al., 2002). As described previously for NHU cell cultures, the separation of the urothelium from the basement membrane as an intact cell sheet limits the potential for contamination of the primary culture by stromal-derived cells, and the use of a serum-free medium developed for keratinocyte cell culture further promotes epithelial but not stromal cell growth (Southgate et al., 1994; Southgate et al., 2002). Thus, the cultures are of urothelial derivation (Nicholls et al., in preparation). Each primary culture of normal rat urothelial (NRU) cells was established from 6 pooled bladders, which were seeded initially into 2 × 25 cm2 flasks and used between passages 1 and 3.
Immunofluorescence Microscopy
Cultured human or rat urothelial cells were grown to 70% to 80% confluence on 12-well glass slides, fixed in a 1:1 mixture of methanol and acetone, air-dried, and incubated overnight at 4°C with titrated primary antibodies or no antibody controls. After extensive washing, slides were incubated in Alexa 488-conjugated goat anti-mouse IgG (5 μg.ml−1; Invitrogen) or goat anti-rabbit IgG (4 μg.ml−1; Invitrogen) for 30 minutes at ambient temperature, before washing in PBS containing 0.25% Tween 20. 0.1 μg.ml−1 Hoechst 33258 (Sigma Aldrich) was added to the last wash to visualize nuclei. Slides were examined under epifluorescence illumination on an Olympus BX60 microscope.
Immunohistochemistry
Sections (5 μm) of paraffin wax–embedded tissue were dewaxed in xylene and rehydrated through ethanol. Endogenous peroxidase activity was blocked by incubation with 3% (v/v) hydrogen peroxide for 10 minutes. Antigen retrieval was performed by digestion of sections for 1 minute in 0.1% (w/v) trypsin in 0.1% (w/v) CaCl2, pH 7.6, followed by boiling for 10 minutes in 10 mM citric acid buffer, pH 6.0 in a microwave oven. Trypsinization was not required for anti-PPARα, RXRα, RXRβ, UPIIIa, and Ki67 antibodies. Endogenous avidin-binding sites were blocked using an avidin/biotin blocking kit (Vector laboratories, Peterborough, UK) according to the manufacturer’s protocol. A 5-minute blocking step was included to prevent nonspecific binding of secondary antibodies (using either 10% rabbit or goat serum in 10 mM Tris-buffered saline, pH 7.6). Slides were incubated with primary antibody overnight at 4°C, washed 3 times in TBS, and incubated with appropriate secondary antibody for 30 minutes at room temperature (biotinylated rabbit anti-mouse at 1:400 and goat anti-rabbit at 1:800, Dako Cytomation Ltd, Ely, UK).
To improve the sensitivity of detection of PPAR and RXR antigens, tyramide-based catalyzed signal amplification was used (Dako Cytomation Ltd; Stahlschmidt et al., 2005). All other primary antibodies were visualized by avidin-biotin-peroxidase detection using a StreptABComplex kit (Dako Cytomation Ltd) according to the manufacturer’s instructions, with 3,3’-diaminobenzidine as chromogen (Sigma Aldrich). Slides were counterstained with Mayer’s hematoxylin, dehydrated through ethanol into xylene, and mounted in DPX (Fisher, Loughborough, UK).
Results
Urothelial Morphology
Histological integrity and differentiated phenotype of human and rat urothelia was confirmed by cytokeratin and uroplakin immunohistochemistry, as illustrated in Figure 1. CK7 was expressed by all layers of both human and rat urothelia. UPIIIa was present along the superficial luminal edge of all human and rat urothelial samples tested and further extended into the intermediate cell layers in the rat urothelium only. The observation that uroplakin expression is localized at the apical edge of the superficial cell in the human urothelium, but is less restricted in the rodent urothelia, is in agreement with previous reports (Mo et al., 2005). The intermediate filament protein, CK20, was expressed by all human and rat tissue samples, with expression limited to the superficial cells.
To determine the proliferative status of the urothelia, the expression of Ki67, a nuclear proliferation marker present during active cell cycle (G1, S, G2, and M phase) and absent in resting (G0) cells was assessed. In the human urothelium, very few Ki67-positive cells were observed in either the ureter (n = 7) or bladder (n = 4) and, where present, labeling was restricted to a few, predominantly basal cells (Figure 1). By contrast, the rat urothelium exhibited strong nuclear Ki67 labeling of all basal and intermediate cells, whereas superficial cells were negative. The same Ki67 expression pattern was seen consistently in both Sprague Dawley and Wistar rat strains and in both young (5–8 weeks; n = 4) and mature (9 months; n = 2) rats (see Table 1).
Expression and Distribution of PPAR and RXR in Rat and Human Urothelial Tissues
The expression and localization patterns are summarized in Table 2 and illustrated on tissue sections by immunohistochemistry for urothelium (Figure 2) and smooth muscle (Figure 3) and in urothelial cell cultures by immunofluorescence (Figure 4). No differences were noted in PPAR or RXR immunolocalization on the human urothelium from the bladder (n = 4) or ureter (n = 7).
PPARα
The human urothelium displayed prominent nuclear PPARα labeling throughout, with weaker diffuse labeling of the cytoplasm. The rat urothelium exhibited nuclear PPARα labeling, which was most intense in the superficial cells and accompanied by weak, diffuse cytoplasmic labeling throughout the urothelium. PPARα was also detected in the human bladder and ureteric smooth muscle, where there was intense nuclear and minimal cytoplasmic labeling. Diffuse cytoplasmic labeling was also present in the rat bladder detrusor smooth muscle. Cultured urothelial cells from human and rat origins exhibited weak nuclear and cytoplasmic PPARα immunoreactivity.
PPARβ/δ
In the human urothelium, PPARβ/δ expression was restricted primarily to the nuclei of superficial and intermediate cells, with less intense labeling of the basal cells. In the rat urothelium, PPARβ/δ was restricted almost exclusively to nuclei of superficial cells, with very little if any expression evident in the other urothelial layers. In the human bladder and ureteric smooth muscle, PPARβ/δ immunolabeling was weak, diffuse, and cytoplasmic, with no nuclear component. PPARβ/δ immunoreactivity was absent from the rat detrusor smooth muscle. In cultured cells from both human and rat urothelia, PPARβ/δ immunoreactivity was predominantly nuclear. Within the nuclei, the labeling was intense, punctate, and excluded from nucleolar regions.
PPARγ
Intense nuclear PPARγ immunoreactivity was present in all layers of human and rat urothelia, with no cytoplasmic component. There was a tendency for labeling to be most intense in the superficial cells of the human urothelium. Nuclear PPARγ was also present in the human bladder and ureteric smooth muscle but was completely absent from the rat detrusor smooth muscle. In the human urothelial cell culture, localization of PPARγ varied according to the state of confluency, being primarily nuclear in confluent culture but diffusely cytoplasmic and less nuclear in subconfluent cultures; this was consistent with previous observations (Varley, Stahlschmidt, Lee, et al., 2004). In confluent NHU and in NRU cell cultures, PPARγ showed an intense nuclear localization, which was punctate and excluded from nucleolar regions.
RXRα
Intense nuclear RXRα immunoreactivity was observed in all urothelial cells within human and rat urothelial tissues. Although RXRα expression was negative in smooth muscle from the human ureter, it was nuclear in a majority of cells in the human bladder detrusor smooth muscle. In the rat detrusor smooth muscle, nuclear RXRα expression was detected in a proportion of smooth muscle cells. By immunofluorescence on urothelial cell cultures, RXRα was intense and nuclear in NHU cells but showed weaker nuclear labeling of NRU cells; in both species, there was also diffuse cytoplasmic immunoreactivity.
RXRβ
In the human urothelium, RXRβ was nuclear and present predominantly in the superficial and intermediate cells, with weaker or absent nuclear labeling of the basal cells. Intense nuclear RXRβ was present in rat urothelial cells from basal and intermediate cell layers but absent from many superficial cells. Nuclear RXRβ immunoreactivity was evident in human ureteric and rat bladder smooth muscle but was only expressed by some cells in the human bladder smooth muscle. RXRβ was intensely nuclear and punctate in NRU cell cultures but was less intensely nuclear in NHU cells, where there was also a cytoplasmic component.
Discussion
The transitional epithelium that lines much of the LUT is a mitotically quiescent tissue with a constitutively low rate of turnover, yet it maintains a high regenerative potential in response to a range of insults, such as damage, injury, and infection (Hicks, 1975). Ki67 is a proliferation marker that is synthesized early in G1 phase and is absent from cells that have withdrawn from the cell cycle. Accordingly, in both rat and human urothelia, the terminally differentiated superficial cells that coexpressed UPIIIa and CK20 were Ki67 negative. However, whereas very few human urothelial cells were Ki67 positive, most basal and intermediate rat urothelial cells were Ki67 positive, irrespective of age or strain (Wistar or Sprague Dawley; data not shown). Others have reported lower levels of Ki67 labeling of the rat bladder urothelium (Nguyen, 2007). However, our findings are supported by flow cytometric cell cycle analysis of freshly dissociated rat urothelial cells, which revealed an unusual combination of very low S-phase percentages (<1%) and disproportionately high G2/M percentages (9.12%; Kaneko et al., 1984; Oleksiewicz et al., 2005). By contrast, freshly dissociated human urothelial cells exhibited a high G0/G1 population that was released rapidly into cell cycle after seeding in culture (Varley et al., 2005). We believe that it is highly unlikely that the Ki67-positive cells in the rat urothelium are actively proliferating, as rat urothelial cells have a very low S-phase percentage and very low BrdU labeling indices (Oleksiewicz et al., 2005; Dominick et al., 2006). Rather, we suggest that the Ki67-positive cells represent a pool of proliferation-capable cells, providing the well-known rapid regenerative potential of normal urothelium. The difference in Ki67 labeling suggests a fundamental difference in how urothelial regeneration is regulated in the rat and human. It is tempting to speculate that this difference may contribute to the susceptibility of the rat to urolithiasis-mediated bladder cancer (Cohen, 2005).
Chronic activation of PPARγ and PPARα has been implicated in transitional cell carcinoma development in rats, as evidenced by rapid changes in urothelial gene expression and intracellular signaling, as well as early urothelial hypertrophy in treated rats (Egerod et al., 2005; Oleksiewicz et al., 2005). The major prerequisite for this mechanism is urothelial coexpression of PPARγ and PPARα, a nontrivial assumption, as the PPARγ and PPARα isoforms generally exhibit nonoverlapping expression patterns (Chang et al., 2007). In situ hybridization studies have demonstrated transcripts for all 3 PPAR genes in human and rabbit urothelia (Guan et al., 1997), and PPARγ transcripts are also expressed in the mouse urothelium, as well as in the presumptive urothelium of the urogenital sinus (Jain et al., 1998). However, transcript expression may not necessarily relate to expression of functional receptors. In this study, we have shown for the first time that all 3 PPAR isotypes and the RXRα and RXRβ heterodimerization partners are expressed at the protein level in urothelial cells. Thus, in the rat LUT, there appears to be a correlation between PPAR isoform expression patterns and susceptibility to the carcinogenic effect of some PPAR agonists (El-Hage, 2004; Egerod et al., 2005; Oleksiewicz et al., 2005). However, it remains to be determined whether there is a causal effect of PPARs in bladder cancer development in rats.
We have previously demonstrated expression of PPARγ and its heterodimerization partner RXRα in NHU cells (Stahlschmidt et al., 2005; Varley, Stahlschmidt, Smith, et al., 2004) and have reported that PPARγ signaling initiates differentiation of NHU cells (Varley, Stahlschmidt, Lee, et al., 2004; Varley, Stahlschmidt, Smith, et al., 2004; Varley et al., 2006). In the present study, we confirmed expression of PPARγ and RXRα by human urothelium in situ and in vitro. Biological activity of the PPARγ and RXRα transcription factors in the human urothelium in situ was indirectly supported by the exclusively nuclear localization patterns. Furthermore, the finding that the human as well as rat urothelium expresses all 3 PPAR isoforms and both the RXR α and β (but not RXRγ; unpublished data) isoforms supports the hypothesis that PPAR signaling is a key, phylogenetically conserved constituent of urothelial biology. While PPARα and RXRα expression was seen in all urothelial layers in the rat as well as human, expression of PPARγ, PPARβ/δ, and RXRβ appeared to show some correlation with differentiation stage (Table 2), as would be predicted from previous studies in NHU cell cultures (Stahlschmidt et al., 2005; Varley, Stahlschmidt, Lee, et al., 2004; Varley, Stahlschmidt, Smith, et al., 2004; Varley et al., 2006). Intriguingly, rat and human urothelia exhibited some differences in the distribution of PPAR and RXR isoforms (Table 2), the relevance of which is as yet unknown, but which may indicate species-specific differences in urothelial responses to PPAR signaling. The expression and localization of receptors was generally equivalent between urothelia in situ and in vitro, although some differences were noted in the distribution between nuclear and cytoplasmic compartments. This is likely to reflect modulating influences of the different environments; for example, the availability of ligand or the influence of other signaling pathways. For example, we have shown previously that autocrine activation of the epidermal growth factor receptor in subconfluent NHU cell cultures results in phosphorylation of PPARγ and sequestration in the cytoplasmic compartment (Varley, Stahlschmidt, Lee, et al., 2004).
Although RXRs are important regulators of PPAR function, prior to this study, very little was known about the expression of RXRs in the lower urinary tract. We have shown that a major point of difference between human and rat urothelia in situ was in the pattern of expression of RXRα and RXRβ, which were also differentially expressed in vitro. Agonist-bound PPARs heterodimerize with RXRs to bind specific PPREs, activating transcription of target genes (Desvergne and Wahli, 1999; Berger and Moller, 2002). PPARs can form heterodimers with all RXRs, and specific combinations can influence the recognition of target gene promoters (Juge-Aubry et al., 1997; Feige et al., 2006). In our opinion, the observed colocalization between PPARs and RXRs in urothelial cells of humans as well as rats supports a biological function for PPAR signaling in urothelial biology. The significance of RXR isoform expression to differential species responses is at present unknown, but our study raises the possibility that it could be addressed in the in vitro setting.
Finally, it should be mentioned that we observed PPAR and RXR expression in smooth muscle cells of the lower urinary tract, the implication of which is as yet unknown. Two particular points of interest were the differential expression of RXRα and RXRβ by human bladder and ureteric smooth muscle, respectively, and the observation that in both rats and humans, PPARs were more highly coexpressed in urothelium than in smooth muscle, again supporting a unique role for PPAR signaling in urothelial biology (Table 2).
In summary, the present study has described the expression of PPAR and RXR receptors in human and rat urothelium and detrusor smooth muscle. This study has confirmed that the urothelium is a potential target tissue for PPAR signaling and has indicated a number of significant differences in expression and distribution of PPARs and RXRs between species. These differences may underlie a differential response to PPAR agonists via the assembly and activity of specific PPAR/RXR heterodimers. The differences in expression of the proliferation marker, Ki67, between the species further suggests that rat and human urothelia may respond differentially following an insult/infection. Although PPARγ signaling is implicated in proliferation and differentiation (Varley, Stahlschmidt, Lee, et al., 2004; Varley, Stahlschmidt, Smith, et al., 2004; Varley et al., 2005, 2006), the role of PPARα and PPARβ/δ signaling in the urothelium has not been investigated. In other epithelial tissues, PPARα has been shown to affect hepatocellular proliferation (Shah et al., 2007) and to inhibit vascular smooth muscle cell proliferation (Zahradka et al., 2003, 2006), whereas activation of PPARβ/δ can induce terminal differentiation with concomitant inhibition of cell proliferation in keratinocytes (Kim et al., 2005) and Paneth cells via hedgehog signaling (Varnat et al., 2006). As our study shows that human and rat urothelial cell cultures retain the in situ expression patterns of PPAR and RXR isotypes, an in vitro experimental approach may clarify the role of PPARα and PPARδ signaling in the urothelium and provide a route to bridging the cross-species barrier.
Footnotes
Acknowledgments
We thank our Urology colleagues for their assistance with the collection of urological specimens. JS holds a research chair funded by York Against Cancer. This study was funded by Novo Nordisk.