Expression profiling of G-protein-coupled receptors in human urothelium and related cell lines

Authors


Dr Peter Ochodnický, Meibergdreef 15, 1105 AZ Amsterdam, The Netherlands. e-mail: p.ochodnicky@amc.uva.nl

Abstract

What's known on the subject? and What does the study add?

Urothelium emerged as a crucial integrator of sensory inputs and outputs in the bladder wall, and urothelial G-protein-coupled receptors (GPCRs) may represent plausible targets for treatment of various bladder pathologies. Urothelial cell lines provide a useful tool to study urothelial receptor function, but their validity as models for native human urothelium remains unclear. We characterize the mRNA expression of genes coding for GPCRs in human freshly isolated urothelium and compare the expression pattern with those in human urothelial cell lines.

OBJECTIVES

  • • To characterize the mRNA expression pattern of genes coding for G-protein-coupled receptors (GPCRs) in human freshly isolated urothelium.
  • • To compare GPCR expression in human urothelium-derived cell lines to explore the suitability of these cell lines as model systems to study urothelial function.

MATERIALS AND METHODS

  • • Native human urothelium (commercially sourced) and human urothelium-derived non-cancer (UROtsa and TERT-NHUC) and cancer (J82) cell lines were used.
  • • For mRNA expression profiling we used custom-designed real-time polymerase chain reaction array for 40 receptors and several related genes.

RESULTS

  • • Native urothelium expressed a wide variety of GPCRs, including α1A, α1D and all subtypes of α2 and β adrenoceptors. In addition, M2 and M3 cholinergic muscarinic receptors, angiotensin II AT1 receptor, serotonin 5-HT2A receptor and all subtypes of bradykinin, endothelin, cannabinoid, tachykinin and sphingosine-1-phosphate receptors were detected. Nerve growth factor and both its low- and high-affinity receptors were also expressed in urothelium.
  • • In all cell lines expression of most GPCRs was markedly downregulated, with few exceptions.
  • • In UROtsa cells, but much less in other cell lines, the expression of β2 adrenoceptors, M3 muscarinic receptors, B1 and B2 bradykinin receptors, ETB endothelin receptors and several subtypes of sphingosine-1-phosphate receptors was largely retained.

CONCLUSIONS

  • • Human urothelium expresses a wide range of receptors which enables sensing and integration of various extracellular signals.
  • • Human urothelium-derived cell lines, especially UROtsa cells, show comparable mRNA expression to native tissue for several physiologically relevant GPCRs, but lose expression of many other receptors.
  • • The use of cell lines as model systems of human urothelium requires careful validation of suitability for the genes of interest.
Abbreviations
NGF

nerve growth factor

GPCR

G-protein-coupled receptors

S1P

sphingosine-1-phosphate

INTRODUCTION

While classical concepts of the regulation of urinary bladder function have focused on detrusor smooth muscle cells and their innervation, in recent years the urothelium has emerged as a crucial integrator of sensory inputs and outputs in the bladder wall [1,2]. Thus, the urothelium represents much more than a passive barrier between the bladder lumen and the detrusor. Rather, it actively secretes various mediators including acetylcholine [3], ATP [4], nitric oxide [5] and possibly nerve growth factor (NGF) [6], which in turn can act as paracrine modulators of smooth muscle and/or afferent nerve function. Such mediator release from urothelium is under the control of various stimuli originating from urothelium itself or from other cell types. Many of these stimuli, including noradrenaline, acetylcholine or substance P, activate specific G-protein-coupled receptors (GPCRs) on the cell surface to modulate urothelial function [4,5,7]. Therefore, GPCRs expressed on urothelium may represent plausible targets for treatment of various bladder pathologies, including overactive bladder syndrome. Recently, it has been suggested that modulation of urothelial GPCRs could contribute to the beneficial effects of several treatment modalities, including muscarinic receptor antagonists and β3 adrenoceptor agonists [8,9]. However, the specific mechanisms linking urothelial GPCRs to mediator release and the regulation of such receptors under normal and diseased conditions remain largely unknown. A major obstacle in the exploration of these processes is the limited access to freshly isolated human urothelial cells in many laboratories.

Urothelium-derived cell lines could represent a valuable alternative, as they can be easily maintained and propagated, genetically manipulated, and readily used as experimental models. However, the validity of urothelial cell lines as models for native human urothelial cells remains unclear. To this end, we have characterized the expression pattern of various GPCRs, including adrenergic, cholinergic muscarinic, bradykinin, endothelin-1, cannabinoid, tachykinin, angiotensin II and sphingosine-1-phosphate (S1P), several serotonin receptors and some other relevant urothelial regulators, including sphingosine kinases, NGF and its receptors in freshly isolated human urothelial tissue. The primary aim of the current study was to compare the expression patterns in native urothelial tissue with those in three human urothelium-derived cell lines (immortalized but non-cancer UROtsa and TERT-NHUC cells and human urothelial cancer J82 cells).

MATERIALS AND METHODS

FRESHLY ISOLATED TISSUE AND CELL LINES

Normal urothelial human tissue was obtained post mortem (up to 48 h) from bladders of patients (four males, 16–66 years) without known urological disease. Tissue procurement was performed by Asterand plc (Royston, Herts, UK) and the UK Human Tissue Bank (De Monfort University, Leicester, UK) following approval by the applicable ethical committees. Urothelium was cleaned and microdissected from the dome region of the bladders using fine sprung scissors under a dissection microscope and stored in RNAlater (Sigma-Aldrich, St Louis, MO, USA) to prevent RNA degradation.

UROtsa cells are a cell line derived from normal human urothelium by immortalization using simian virus 40 large T antigen gene construct [10]. They were a kind gift from Dr Rene C. Krieg and Dr Corinne Henkel (University of Aachen, Germany). The cells were cultured in RPMI 1640 medium (Invitrogen, Carlsbad, CA, USA) supplemented with 5% fetal bovine serum and penicillin/streptomycin (100 U/mL; 100 µg/mL) until 80%–90% confluence and starved for 16 h before proceeding with RNA isolation.

TERT-NHUC cells are also derived from normal human urothelial cells but had been immortalized by retrovirus-mediated transduction of human telomerase reverse transcriptase [11]. They were a kind gift from Professor Margaret Knowles (University of Leeds, UK). TERT-NHUC cells were cultured in keratinocyte serum-free medium (Invitrogen) supplemented with bovine pituitary extract, epidermal growth factor and penicillin/streptomycin. The cells were starved for 16 h by excluding growth factors from the medium before RNA isolation.

J82 cells are a bladder cancer cell line derived from transitional cell carcinoma and were obtained from American Type Culture Collection (LGC Standards, Wesel, Germany). They were cultured in Dulbecco's modified Eagle's medium supplemented with penicillin/streptomycin, 10% fetal bovine serum and 2 mm l-glutamine until confluent.

RNA ISOLATION AND RT-PCR ARRAY

Total RNA was isolated from the cells and the tissue following lysis in guanidine-thiocyanate-containing buffer using RNeasy (Qiagen, Hilden, Germany) RNA purification technology based on the specific RNA interactions with silica-based membrane, followed by on-column DNAse treatment. Subsequently, the quality of the RNA was verified using Experion RNA analyser based on automated capillary electrophoresis (Biorad, Hercules, CA, USA) and RNA concentration and purity were quantified spectrophotometrically (Nanodrop, Wilmington, DE, USA). All samples provided intact total mRNA as determined by an RNA quality indicator (1–10) greater than 9. RNA was additionally pretreated with DNase I (Invitrogen) to prevent genomic DNA contamination. Total RNA (500 ng) was then reverse transcribed to cDNA using iScript cDNA Synthesis Kit (Biorad) with RNAse inhibitor, based on the mix of oligo(dT) and random hexamer primers. Diluted cDNA corresponding to 20 ng RNA input was used per individual real-time PCR performed in a custom-designed real-time PCR array (SABiosciences, Frederick, MD, USA). The genes analysed with specific predesigned primers (PCR product lengths and reference position of the PCR products in the transcripts are listed in Table 1) included GPCRs, some other relevant urothelial regulators, urothelial markers, four reference genes (ROCK2, RPLP0, HPRT, GAPDH) and negative controls to control for the quantity and specificity of the reactions. Each of the predesigned primer sets met the criterion of 90% RT-PCR efficiency as tested by the manufacturer and calculated as described before [12]. Four samples per experimental group were assayed in 96-well plates in a real-time PCR array using an iCycler IQ (Biorad), each array representing an individual patient sample or an independent cell culture experiment. The reaction conditions were as follows: initial activation 10 min at 95 °C; 40 cycles of denaturation for 15 s at 95 °C; annealing/extension for 1 min at 60 °C. Melt curve analysis confirmed single PCR product formation for each individual primer set.

Table 1.  List of the studied transcripts
UniGeneGenBankSymbolDescriptionPCR product lengthPCR product reference position in the transcript
Hs.709175NM_033303ADRA1AAlpha-1A adrenergic receptor1901012
Hs.368632NM_000679ADRA1BAlpha-1B adrenergic receptor122369
Hs.557NM_000678ADRA1DAlpha-1D adrenergic receptor145388
Hs.249159NM_000681ADRA2AAlpha-2A adrenergic receptor1911150
Hs.247686NM_000682ADRA2BAlpha-2B adrenergic receptor1493080
Hs.123022NM_000683ADRA2CAlpha-2C adrenergic receptor191386
Hs.99913NM_000684ADRB1Beta-1 adrenergic receptor1201609
Hs.591251NM_000024ADRB2Beta-2 adrenergic receptor1571433
Hs.2549NM_000025ADRB3Beta-3 adrenergic receptor171412
Hs.632119NM_000738CHRM1Muscarinic acetylcholine receptor M11842584
Hs.535891NM_000739CHRM2Muscarinic acetylcholine receptor M2109184
Hs.7138NM_000740CHRM3Muscarinic acetylcholine receptor M3108808
Hs.248100NM_000741CHRM4Muscarinic acetylcholine receptor M41231323
Hs.584747NM_012125CHRM5Muscarinic acetylcholine receptor M51911080
Hs.477887NM_031850AGTR1Type-1 angiotensin II receptor1681887
Hs.405348NM_000686AGTR2Type-2 angiotensin II receptor107328
Hs.525572NM_000710BDKRB1B1 bradykinin receptor135349
Hs.654542NM_000623BDKRB2B2 bradykinin receptor184296
Hs.183713NM_001957EDNRAEndothelin-1 receptor A1211690
Hs.82002NM_000115EDNRBEndothelin-1 receptor B931290
Hs.75110NM_016083CNR1Cannabinoid receptor 1128690
Hs.73037NM_001841CNR2Cannabinoid receptor 2185151
Hs.633301NM_001058TACR1Substance P receptor134649
Hs.88372NM_001057TACR2Substance K receptor1891744
Hs.942NM_001059TACR3Neuromedin K receptor1891394
Hs.247940NM_000524HTR1A5-hydroxytryptamine receptor 1A791248
Hs.654586NM_000621HTR25-hydroxytryptamine receptor 2A1873144
Hs.154210NM_001400S1PR1Sphingosine-1-phosphate receptor 11662082
Hs.655405NM_004230S1PR2Sphingosine-1-phosphate receptor 2189387
Hs.585118NM_005226S1PR3Sphingosine-1-phosphate receptor 3127549
Hs.662006NM_003775S1PR4Sphingosine-1-phosphate receptor 4180448
Hs.501561NM_030760S1PR5Sphingosine-1-phosphate receptor 51721648
Hs.68061NM_021972SPHK1Sphingosine kinase 1146857
Hs.528006NM_020126SPHK2Sphingosine kinase 2951379
Hs.2561NM_002506NGFNerve growth factor (beta polypeptide) (NGF)100890
Hs.415768NM_002507NGFRNerve growth factor receptor p75NTR106718
Hs.406293NM_002529NTRK1Neurotrophic tyrosine kinase. receptor 1, trkA1642297
Hs.710330NM_006760UPK2Uroplakin II93338
Hs.411501NM_005556KRT7Keratin 7150638

DATA ANALYSIS

Relative gene expression was analysed for each well using the ΔΔCt method, calculating the expression of each individual gene relative to the reference gene expression (2–ΔCt) and a fold upregulation or downregulation (2–ΔΔCt) of the specific gene normalized to the average expression observed in native tissue. These analyses were based on ROCK2 as the reference gene, since it displayed similar Ct values across all cell sources and the lowest variability among all experimental groups (data not shown) compared with other tested reference genes. Based on the assay design and variability, all transcripts showing a Ct value higher than 35 were considered to be unexpressed (below the detection limit). The significance of the changes given in Table 2 was based on unpaired Student t test comparisons of 2–ΔCt values.

Table 2.  Comparison of relative mRNA expression of G-protein-coupled receptors and urothelial markers in human urothelial cell lines and native urothelial tissue
DescriptionNative tissueUROtsaTERT-NHUCJ82
Average Ct values ±semRelative expression 2^(−ΔCt) ±semExpression of corresponding receptor type (%)Relative expression 2^(−ΔCt) ±semFold change vs native tissue (P < 0.05)Relative expression 2^(−ΔCt) ±sem Fold change vs native tissue (P < 0.05) Relative expression 2^(−ΔCt) ±sem Fold change vs native tissue (P < 0.05)
  1. Data are means ±sem of four experiments. ND, not detected/below detection limit.

α adrenergic receptors          
 Alpha-1A adrenergic receptor33.3 ± 0.70.003 ± 0.0011%0.002 ± 0.001+1.49ND <−2.14 ND <−6.99
 Alpha-1B adrenergic receptor>35NDNDNDNDND ND ND ND
 Alpha-1D adrenergic receptor29.8 ± 0.50.035 ± 0.00918%0.003 ± 0.001 −11.11 0.004 ± 0.001 −9.35 ND <−38.46
 Alpha-2A adrenergic receptor28.2 ± 0.70.103 ± 0.05453%ND <−66.67 ND <−71.43 0.003 ± 0.001 −35.71
 Alpha-2B adrenergic receptor30.1 ± 0.60.029 ± 0.00915%ND <−18.18 ND <−20.41 ND <−32.26
 Alpha-2C adrenergic receptor30.3 ± 1.10.025 ± 0.01813%ND <−15.38 0.004 ± 0.002 −7.04 0.008 ± 0.001 −3.21
β adrenergic receptors          
 Beta-1 adrenergic receptor30.3 ± 1.20.025 ± 0.0159%0.028 ± 0.003+1.140.004 ± 0.001 −6.45 ND <−22.73
 Beta-2 adrenergic receptor27.0 ± 0.40.235 ± 0.09488%0.602 ± 0.122 +2.56 0.145 ± 0.044 −1.62 0.070 ± 0.012 −3.34
 Beta-3 adrenergic receptor32.1 ± 1.00.007 ± 0.0053%0.002 ± 0.001−4.37ND −5.13 ND <−7.94
Cholinergic muscarinic receptors          
 Muscarinic acetylcholine receptor M1>35NDNDNDNDND ND ND ND
 Muscarinic acetylcholine receptor M227.3 ± 0.50.193 ± 0.04788%0.002 ± 0.001 −111.11 ND −125.00 ND <−200.00
 Muscarinic acetylcholine receptor M330.4 ± 1.10.023 ± 0.01110%0.038 ± 0.009+1.650.002 ± 0.001 −12.50 ND <−25.64
 Muscarinic acetylcholine receptor M434.1 ± 0.50.002 ± 0.0011%0.005 ± 0.001 +2.84 ND <−1.23 0.004 ± 0.001 +2.21
 Muscarinic acetylcholine receptor M534.1 ± 0.70.002 ± 0.0011%ND<−1.07ND <−1.29 ND <−1.51
Angiotensin II receptors          
 Type-1 angiotensin II receptor27.6 ± 0.50.158 ± 0.044100%0.002 ± 0.001 −83.33 ND −111.11 ND −200.00
 Type-2 angiotensin II receptor>35NDNDNDNDND ND ND ND
Bradykinin receptors          
 B1 bradykinin receptor28.7 ± 0.60.076 ± 0.02124%0.119 ± 0.021+1.5610.014 ± 0.003 +1.56 0.011 ± 0.001 −7.25
 B2 bradykinin receptor27.0 ± 0.50.238 ± 0.04776%0.098 ± 0.025 −2.42 0.032 ± 0.005 −7.46 0.034 ± 0.006 −6.94
Endothelin receptors          
 Endothelin-1 receptor A25.5 ± 0.20.685 ± 0.52231%0.141 ± 0.038−4.851.304 ± 0.326 +1.90 0.010 ± 0.001 −71.43
 Endothelin-1 receptor B24.3 ± 0.41.540 ± 0.47169%ND <−1000 ND <−1000 0.424 ± 0.032 −3.64
Cannabinoid receptors          
 Cannabinoid receptor 130.1 ± 0.50.029 ± 0.00793%0.002 ± 0.001 −17.54 ND <−20.41 0.023 ± 0.002 −1.25
 Cannabinoid receptor 233.9 ± 0.50.002 ± 0.0017%0.002 ± 0.001+1.01ND <−1.01 ND <−3.29
Tachykinin receptors          
 Substance P receptor27.8 ± 0.50.134 ± 0.04131%0.006 ± 0.002 −23.81 ND −100.00 ND −100.00
 Substance K receptor29.8 ± 0.80.035 ± 0.0188%0.003 ± 0.002 −10.42 0.010 ± 0.002 −3.46 0.012 ± 0.001 −3.08
 Neuromedin K receptor26.9 ± 0.50.259 ± 0.19861%ND <−166.67 ND −166.67 ND <−333.33
Serotonin receptors          
 5-hydroxytryptamine receptor 1A>35ND0NDNDND ND ND ND
 5-hydroxytryptamine receptor 2A32.7 ± 0.50.005 ± 0.003100%ND−3.05ND <−3.44 ND <−5.32
Sphingosine-1-phosphate receptors          
 Sphingosine-1-phosphate receptor 125.2 ± 0.20.837 ± 0.15068%ND −500.00 0.151 ± 0.027 −5.56 0.080 ± 0.007 −10.53
 Sphingosine-1-phosphate receptor 227.7 ± 0.50.147 ± 0.03112%0.034 ± 0.010 −4.31 0.125 ± 0.050 −1.18 0.070 ± 0.019 −2.09
 Sphingosine-1-phosphate receptor 327.5 ± 0.80.170 ± 0.09814%0.029 ± 0.004 −5.78 0.029 ± 0.016 −5.81 0.196 ± 0.003 +1.15
 Sphingosine-1-phosphate receptor 431.5 ± 0.80.011 ± 0.0051%0.003 ± 0.002−3.220.002 ± 0.001 −6.02 ND <−9.17
 Sphingosine-1-phosphate receptor 528.8 ± 0.80.067 ± 0.0415%0.675 ± 0.065 +10.04 0.495 ± 0.033 +7.37 0.331 ± 0.050 +4.93
Other          
 Sphingosine kinase 129.7 ± 0.80.038 ± 0.0120.744 ± 0.131 +19.80 17.63 ± 5.19 +469.57 0.603 ± 0.113 +16.06
 Sphingosine kinase 227.9 ± 0.60.127 ± 0.0340.264 ± 0.048 +2.09 0.364 ± 0.127 +2.88 0.833 ± 0.049 +6.58
 Nerve growth factor (NGF)34.2 ± 0.60.002 ± 0.0010.004 ± 0.011+2.22ND ND 0.012 ± 0.002 +12.46
 Nerve growth factor receptor p75NTR29.6 ± 0.50.040 ± 0.0090.281 ± 0.020 +6.98 0.006 ± 0.002 −7.09 0.007 ± 0.004 −6.21
 Neurotrophic tyrosine kinase receptor trkA34.0 ± 1.00.002 ± 0.0010.002 ± 0.001+1.03ND <−1.44 ND <−1.47
Urothelial markers          
 Uroplakin II25.1 ± 0.70.917 ± 0.6430.008 ± 0.003 −111.11 0.010 ± 0.005 −100.00 0.002 ± 0.001 −500.00
 Keratin 722.8 ± 0.64.385 ± 1.3096.169 ± 0.228+1.4145.46 ± 9.55 +10.37 50.68 ± 6.96 +11.56

RESULTS

EXPRESSION OF UROTHELIAL MARKERS

Confirming the urothelial origin of the human native tissue samples cytokeratin 7, normally specifically expressed in all urothelial cell subtypes (i.e. umbrella, intermediate and basal cells), was the highest expressed transcript among all genes tested (Table 2). The urothelial origin of the cells was also confirmed by the high abundance of uroplakin II mRNA, a specific marker for umbrella cells.

While all cell lines, including J82 cancer cells, retained cytokeratin 7 expression, uroplakin II was markedly downregulated (Table 2), suggesting some degree of de-differentiation and loss of the umbrella cell phenotype.

GPCR EXPRESSION PATTERN IN NATIVE UROTHELIAL CELLS

The expression levels of all investigated transcripts are given in Table 2. As expected, native urothelium expressed a wide variety of GPCRs. α2A was the most abundant subtype of α adrenoceptors, with lower expression of α1D, α2B and α2C. β2 was by far the most prevalent β adrenoceptor subtype, whereas β1 and β3 represented only minor fractions of β adrenoceptor mRNA. Among muscarinic receptors, only highly prevalent M2 and, to a lesser extent, M3 subtypes were reliably detected. Furthermore, native urothelium expressed angiotensin II AT1 receptor, serotonin 5-HT2A receptor and all subtypes of bradykinin, endothelin, cannabinoid, tachykinin and S1P receptors.

Finally, native urothelium also expressed sphingosine kinases. Minimal expression of NGF and its high-affinity receptor trkA were detected, while low-affinity neurotrophin receptor p75NTR was rather abundant, indicating that urothelium could produce or respond to neurotrophins.

GPCR EXPRESSION PATTERN IN HUMAN UROTHELIAL CELL LINES

As shown in Table 2 and Fig. 1, all three cell lines displayed marked downregulation of the majority of the GCPR transcripts with only a few genes remaining unchanged or elevated compared with native urothelium. Virtually no α adrenoceptor subtypes were detected in the cell lines. While the β adrenoceptor expression pattern was largely maintained in UROtsa cells, the other cell lines TERT-NHUC and J82 only expressed the β2 receptor subtype out of all adrenoceptors. UROtsa cells retained the expression of M3 muscarinic receptors but expression of all other muscarinic subtypes (including M2, the most abundant in native urothelium) was markedly downregulated, whereas TERT-NHUC and J82 cells expressed all muscarinic receptor subtypes minimally or not at all. No or minimal expression of angiotensin II, cannabinoid, serotonin or tachykinin receptors was observed in all studied cell lines, with the exception of a prominent cannabinoid CB1 receptor expression in J82 cells. Interestingly, expression of bradykinin receptors and endothelin-1 ETA receptor was maintained in all cell lines, although often at a lower level than in native urothelium, while ETB receptor was only expressed by J82 cells. All cell lines were characterized by a marked change in S1P receptor expression pattern, including a notable upregulation of S1P5 (the only upregulated GPCR in all cell lines) along with S1P1, S1P2 and S1P3 downregulation. Also, both sphingosine kinase isoforms were heavily upregulated. The cell lines differed in their expression of NGF and its receptors: while UROtsa cell line retained a minimal NGF and trkA expression with an upregulation of p75NTR, in TERT-NHUC all components were downregulated or missing, whereas J82 seemed to overexpress NGF.

Figure 1.

Scatter plot showing the relation between relative mRNA expression of the individual transcripts (as log10 of 2−ΔCt) in native tissue and in (A) the UROtsa cell line, (B) the TERT-NHUC cell line and (C) the J82 cell line. Genes downregulated or upregulated more than twofold (grey diagonals) are shown in green and red, respectively, whereas those without any change are given as black dots.

DISCUSSION

Our study shows that native human urothelium expresses a wide variety of GPCRs at the mRNA level, consistent with the idea of GPCRs being prominent regulators of urothelial function. Thus, human urothelium expressed all β adrenoceptor subtypes. While this is in line with mRNA and immunohistochemistry data of other investigators [13,14], the specific quantitative role of each subtype (especially β3) remains to be established. Importantly, stimulation of urothelial β adrenoceptors was shown to lead to formation of nitric oxide in experimental animals [15] and thus represents a potential mechanism contributing to the beneficial effects of β adrenoceptor agonists in the treatment of bladder overactivity [9].

Although mRNA for all muscarinic receptor subtypes has been reported in human urothelium by other authors [16,17], only M2 and M3 were reliably detected in the present study. The prominence of these two subtypes is in agreement with the muscarinic expression pattern observed when the whole bladder tissue is studied. Enhancement of urothelial ATP release upon stimulation of muscarinic receptors is well established [4,17], but the relative role of each receptor subtype remains to be determined.

For several urothelial GPCRs included in the current study, such as bradykinin receptors [18] and NK1 substance P receptor [7], a role in regulation of urothelial function has been described, whereas for others including cannabinoid [19] and endothelin-1 receptors [20] urothelial expression was reported without exploration of a functional relevance. In addition, we also detected a prominent expression of angiotensin II AT1, serotonin 5-HT2A and tachykinin substance K and neuromedin K receptors that had not been described in urothelium before.

Finally, it has recently been proposed that urothelium could serve as an important source of neurotrophins including NGF [6]. Our finding of low level expression of NGF and its high-affinity receptor trkA along with a stronger expression of low-affinity p75NTR neurotrophin receptor support the existence of the urothelial NGF system but require future functional analysis. Interestingly, urothelial p75NTR could play a role in the development of bladder dysfunction in rat interstitial cystitis [21]. Taken together, these data underscore the prominent role of the urothelium as an integrator of various neuro-humoral signals involved in the control of bladder function.

While it is a limitation of the present study that expression profiling was performed at the mRNA level only, the above examples provide face validation that the current mRNA data may indeed reflect the presence of functional proteins. In addition, minor contamination of isolated urothelium with underlying lamina propria cells cannot be excluded, although high urothelial marker gene expression indicates that urothelium represents the vast majority of the preparation.

Immortalized human urothelial cell lines retained cytokeratin 7 expression but displayed a marked reduction in uroplakin II expression, indicating some degree of de-differentiation and loss of umbrella cell phenotype. Immortalization and/or culturing have been shown to affect the phenotype of urothelial cells [22], contributing to altered expression profiles between the cell lines and the native tissue composed of three urothelial cell subtypes (i.e. umbrella, intermediate and basal cells). All the cell lines displayed marked downregulation of most GPCRs, suggesting changes in the functional phenotypes of these cell lines as well. Such downregulation questions the usefulness of the respective cell lines as models for the affected receptors. On the other hand, expression of several GPCRs was retained in for example UROtsa cells, rendering such cell lines potential cellular models of normal urothelium. Specifically, mRNA expression of β2 adrenoceptors, M3 muscarinic receptors, B1 and B2 bradykinin receptors, ETB endothelin receptors along with several subtypes of S1P receptors was largely maintained in UROtsa cells. While the present study did not attempt confirmation of mRNA data at the protein or functional level, previous data indicate that functionality of several GPCRs in UROtsa could be maintained. For instance, radioligand binding studies confirmed the presence of β2 adrenoceptor protein in UROtsa cells [23] and functional studies on cAMP formation also support functionality of β adrenoceptors in UROtsa cells, possibly including β3 adrenoceptors [14]. Although UROtsa cells retained the expression of both M2 and M3 receptors, the former was considerably downregulated whereas the latter was largely preserved compared with native urothelium. In accordance with these expression data, recently it has been reported that functional muscarinic agonist responses in UROtsa cells involve both subtypes with dominant contribution of M3 receptors [24]. In J82 cancer cells, a previously reported loss of muscarinic receptor expression and signalling is in line with our current mRNA data [25]. Similarly, the maintained expression of ETA and ETB endothelin receptors and of several subtypes of S1P receptors in J82 cells confirms previous functional findings in this cancer cell line [26,27].

In conclusion, urothelium-derived cell lines, especially UROtsa cells, show comparable mRNA expression to native tissue for some receptors (e.g. β2 adrenergic, M3 muscarinic, bradykinin, endothelin and S1P receptors) and may represent valuable model systems for these particular receptors. On the other hand, these cell lines exhibit markedly downregulated expression of many other GPCRs which are of limited value for studies in these cases. Investigators are advised to carefully compare expression of their receptor of interest when planning to use cell lines as models of normal urothelium. mRNA profiling, as presented in the current study, could provide a useful information basis for functional and interventional studies unravelling the importance of urothelial regulation in bladder physiology and pathology.

ACKNOWLEDGEMENTS

This work was supported in part through Coordination Theme 1 (Health) of the European Community's FP7, grant agreement HEALTH-F2-2008-223234.

CONFLICT OF INTEREST

None declared.

Ancillary