The urinary bladder and associated tract is lined by the urothelium. Once considered as just an impermeable epithelium, it is becoming evident that the urothelium not only functions as a volume-accommodating urinary barrier but has additional roles, including sensory signaling. Lack of access to normal human urothelium has hampered physiological investigation, and although cell culture systems have been developed, there has been a failure to demonstrate that normal human urothelial (NHU) cells grown in vitro retain the capacity to form a functional differentiated urothelium. The aim of this study was to develop a biomimetic human urothelium from NHU cell cultures. Urothelial cells isolated from normal human urothelium and serially propagated as monolayers in serum-free culture were homogeneous and adopted a proliferative, nondifferentiated phenotype. In the presence of serum and physiological concentrations of calcium, these cells could be reproducibly induced to form stratified urothelia consisting of basal, intermediate, and superficial cells, with differential expression of cytokeratins and superficial tight junctions. Functionally, the neotissues showed characteristics of native urothelium, including high transepithelial electrical resistance of >3,000 Ω·cm2, apical membrane-restricted amiloride-sensitive sodium ion channels, basal expression of Na+-K+-ATPase, and low diffusive permeability to urea, water, and dextran. This model represents major progress in developing a biomimetic human urothelial culture model to explore molecular and functional relationships in normal and dysfunctional bladder physiology.
- cell culture
the urinary bladder and associated urinary tract are lined by the urothelium, an epithelium that is highly specialized to accommodate changes in bladder volume and provide a permeability barrier to urine (8). The urothelium has also been proposed to have a sensory role (9), dysregulation of which may be important in the pathogenesis of dysfunctional bladder syndromes, such as interstitial cystitis (41). Our current understanding of normal human urothelial cell physiology and specifically the relationship between morphological differentiation and functional specialization is hampered by the lack of suitable cell culture models (reviewed in Ref. 23).
The urothelium is a transitional epithelium and displays a regular architecture, increasing in morphological complexity from basal cells, through a variable number of intermediate cells, to the highly differentiated superficial or umbrella cells (20). The superficial cell layer is primarily responsible for providing the permeability barrier (30); the cells are interconnected by tight junctional complexes, which restrict paracellular ion transport and polarize the cell by limiting diffusion of transport proteins between the apical and basolateral membranes (12). In addition, superficial cells show a unique specialization of the apical plasma membrane, with thickened plaques of asymmetric unit membrane (AUM) decorating up to 90% of the luminal surface (16). These plaques are composed of a number of component proteins, the uroplakins (UPs) (7, 47, 48), which can be used as objective markers of terminal urothelial cytodifferentiation in many species, including man (31). The critical role of the AUM plaque in limiting transcellular permeability has been demonstrated in the UPIIIa−/− transgenic mouse, which developed a “leaky” urothelium in association with incomplete plaque formation (17, 18).
Consistent with its barrier properties, the transurothelial electrical resistance (TER) of the urothelium is one of the highest recorded for any tissue (2, 18, 21, 22, 24, 30). Although the urothelium is relatively impermeable, there is ion flux across the epithelium (8, 24). Sodium is the principal transported ion (25), by a mechanism that is modulated by a variety of molecular and physical factors (4, 10, 26, 46). It has been proposed that transurothelial sodium ion flux via mechanosensitive ion channels located in the apical membrane of the superficial cells may have a sensory role in normal micturition (9). Thus it is evident that there is an important relationship between molecular differentiation and function of the urothelium.
We have previously described a cell culture system to propagate normal human urothelial (NHU) cells (37, 38). In culture, NHU cells acquire a proliferative, regenerative phenotype but do not express markers of urothelial differentiation (29). NHU cells can be induced to express uroplakin genes (44, 45) and will form a stratified and partially differentiated urothelium when seeded on a deepithelialized urothelial stroma in organ culture (35). Although these observations suggest that cultured NHU cells retain the potential to undergo cytodifferentiation and are not compromised by propagation in vitro, it has yet to be demonstrated that the cells are capable of forming a functional barrier ex vivo. The purpose of this study was therefore to explore the capacity of in vitro propagated NHU cells to generate a functional barrier urothelium, with the longer term objective of using this model to explore the relationship between cytodifferentiation and (patho)physiology.
MATERIALS AND METHODS
Chemicals and Reagents
Unless specified otherwise, all chemicals were of analytical reagent grade or tissue culture grade, as appropriate, and were obtained from Sigma (Gillingham, UK).
The collection of surgical specimens was approved by the relevant Local Research Ethics Committees and had full patient consent. Urothelial tissue samples were obtained at surgery from the upper and lower urinary tract of adult and pediatric patients with no history of urothelial dysplasia or neoplasia. Tissue samples were transported at room temperature from surgery in Hanks' balanced salt solution (HBSS; GIBCO, Paisley, UK) containing 10 mM HEPES, pH 7.6 (GIBCO) and 20 kallikrein-inhibiting units (KIU)/ml of aprotinin (Trasylol; Bayer Pharmaceuticals, Newbury, UK). To document tissue integrity on arrival at the laboratory and for comparison with subsequent cultured cells and tissues, representative portions of each sample were processed into paraffin wax for histology and immunohistochemistry. The remaining sample was cut into ∼1-cm2 pieces, placed into Ca2+- and Mg2+-free HBSS, containing 10 mM HEPES, pH 7.6, 20 KIU/ml Trasylol, and 0.1% (wt/vol) EDTA and incubated at 4°C overnight to release urothelial cell sheets. The isolated urothelium was used to establish finite NHU cell lines as previously described (37, 38).
NHU cell cultures were established and maintained in keratinocyte serum-free medium (KSFM), containing recombinant epidermal growth factor and bovine pituitary extract at the manufacturer's recommended concentrations (Invitrogen, Paisley, UK) and supplemented with 30 ng/ml cholera toxin to improve cell plating and attachment (19). KSFM fully supplemented with the aforementioned factors will be referred to as KSFM complete (KSFMc). The cultures were maintained at 37°C in a humidified atmosphere of 5% CO2 in air. The medium was replaced after 24 h and subsequently on alternate days for all experiments. Primary urothelial cells were propagated in Primaria tissue culture flasks (Becton Dickinson, Cowley, UK) and subcultured using a method described in detail elsewhere (37, 38).
The studies reported here are based on NHU cell lines established from 22 independent donors (1 bladder, 16 ureter, 5 renal pelvis; 15 men; mean age 48.2 ± 21.2 yr).
To develop a biomimetic urothelium (Cross WR and Southgate J, 2004; Biomimetic Urothelium; Patent Application W02004/011630), urothelial cell cultures from NHU cell lines at passages 1- 3 were split into two groups: one set was maintained in KSFMc and the other was switched to KSFMc supplemented with 5% (vol/vol) fetal bovine serum (FBS; Harlan Sera-Lab, Loughborough, UK). At confluence, the cells were harvested from Primaria flasks and seeded onto 1-cm2 permeable Snapwell membranes (Costar, High Wycombe, UK) at a density of 1 × 106 cells/cm2. After 24 h, the exogenous calcium concentration of the medium in one-half of both sets of cultures was increased from 0.09 mM (in KSFMc) to 2 mM, by use of a 1 mM CaCl2 stock solution (37); the addition of 5% FBS to KSFMc was found to raise the Ca2+ concentration to 0.2 mM, and this was taken into account when KSFMc containing both 5% FBS and 2.0 mM CaCl2 was prepared.
NHU cell cultures were maintained in one of the following culture media: 1) serum-free KSFMc (0.09 mM Ca2+); 2) serum-free KSFMc (2.0 mM Ca2+); 3) KSFMc supplemented with 5% FBS (0.2 mM Ca2+); and 4) KSFMc supplemented with 5% FBS (2.0 mM Ca2+). Electrophysiological, permeability, and histological studies were performed on the cultures 7 days after they were seeded onto Snapwell membranes. Assessment of the mechanisms of transcellular ion transport was limited to cultures with a high transepithelial electrical resistance (TER; >1,000 Ω·cm2).
NHU cell lines were routinely monitored for contamination by Mycoplasma spp by scrutinizing for extranuclear fluorescence after staining cultures with the DNA-intercalating fluorochrome bisbenzimide (Hoechst 33258; Calbiochem, Nottingham, UK; see section below on immunofluorescence).
Functional Properties of NHU Cell Cultures
Electrophysiological properties of NHU cell cultures.
The electrophysiological properties of urothelial cultures were measured using a World Precision Instruments DC1100 electronic volt-ohmmeter (EVOM) and vertical modified Ussing chambers specifically designed to accept Snapwell membranes and glass Ag-AgCl electrodes. All experiments were performed in modified Krebs solution (in mM: 118 NaCl, 25 NaHCO3, 4.74 KCl, 1.19 MgSO4, 1.17 CaCl2 and 1 glucose) kept at a constant 37°C and equilibrated with 95% O2-5% CO2.
Measurement of transepithelial potential difference and short circuit current.
Urothelial cells cultured on Snapwell membranes were placed in vertical Ussing chambers with 5 ml modified Krebs solution in both apical and basal hemichambers. The spontaneous potential difference (V) and short-circuit current (I) across the urothelial cell layers (1-cm2 culture membrane) were measured via electrodes connected to the EVOM and recorded on a computer, which was interfaced via an analog-to-digital converter. The TER was calculated from V and I using Ohm's law (I = V/R). The measured TER was corrected by subtracting the mean resistance of three blank Snapwell filters.
Assessment of transurothelial sodium ion transport.
Transcellular sodium ion transport was investigated in urothelial cell cultures following determination of TER. Five microliters of the sodium ion channel inhibitor amiloride was added to the apical hemichamber (final concentration 5 μM-25 nM), and the transurothelial potential difference and short-circuit current were recorded until both parameters had stabilized. As the control, amiloride was added to the basal hemichamber.
The presence and membrane location of the ion pump Na+ -K+-ATPase were delineated using ouabain. As above, the TER of the urothelial culture was determined initially, and both the transurothelial potential difference and short-circuit current were recorded throughout the experiment.
Determination of urea and water permeability of human urothelial cell cultures.
Diffusive urea and water permeability coefficients were determined by measuring radioisotopic fluxes. After the TER of the urothelial culture had been assessed, 25 μl of [3H]water (200 μCi/ml; Sigma) and 25 μl of [14C]urea (200 μCi/ml; Amersham, Little Chalfont, UK) were added to the apical hemichamber. During the next 60 min, duplicate 100-μl aliquots were taken from both the apical and basal hemichambers at 15-min intervals and placed into 5-ml scintillation vials (PerkinElmer, Beaconsfield, UK) containing 4 ml Ultima Gold XR scintillation fluid (PerkinElmer). After sampling, the aliquoted volume was replaced with fresh Krebs solution and the TER was checked to confirm that the urothelial culture had not been physically disturbed. The number of counts of the individual isotopes within the samples was determined using a Packard Tri-carb 2700TR liquid scintillation counter.
The measured diffusive urea and water permeabilities (PD) were calculated using the flux equation: PD = Φ/(A)(ΔC), where Φ is the flux of the isotope tracer across the membrane, calculated from the net increase of the tracer in the basal hemichamber, A is the area of the membrane, and ΔC is the concentration gradient of the isotope across the membrane and is calculated from the mean concentration of the isotope in each chamber for the sampling period (30). In all flux measurements, corrections were made for sample dilution. To determine the permeability of the urothelial cultures, it was necessary to exclude the resistance to flow of water and urea exerted by the unstirred layers and the Snapwell membrane. The diffusive permeability of the urothelial cultures [PD(urothelium)] was calculated by measuring the mean permeability of three blank Snapwell membranes [PD(Snapwell)] using the following formula (30): 1/PD(urothelium) = 1/PD(measured) − 1/PD(Snapwell).
Measurement of dextran permeability of human urothelial cell cultures.
Permeability assays were performed using dextran (molecular weights 4,400 and 9,500) conjugated to fluorescein isothiocyanate (FITC). At the start of the experiments, the medium in the apical compartment of the Snapwell membrane was replaced with 500 μl of the appropriate growth medium containing one of the tracers at 1 mg/ml. The basal compartment was replaced with 1000 μl of tracer-free growth medium. The urothelial cells were returned to the incubator for 3 h, and then duplicate 350-μl samples were taken from the basal compartment and the amount of FITC-dextran was determined using a MFX microtiter plate fluorometer (Dynex, Worthing, UK). The amount of diffused dextran was calculated from a titration curve of known concentration (3.1–200 mg/ml).
Characterization of Native and Cultured Urothelial Cell Phenotypes
Paraffin wax and cryostat sections.
Samples of the native tissue were fixed in 10% (vol/vol) formalin in PBS, dehydrated through graded alcohols, and embedded in paraffin wax. Five-micrometer sections were cut and stained with hematoxylin and eosin. Additionally, 5-mm3 samples of tissue were embedded in Cryo-M-Bed compound (Bright, Burgdorf, Germany) before being frozen on card-ice, quenched in liquid nitrogen, and stored at −80°C. Five-micrometer cryosections were cut and collected onto 12-well Multitest slides (Hendley, Loughton, UK).
Immunofluorescent labeling of cryosections and cultured urothelial cells.
Indirect immunofluorescence was performed as previously described (37). Cryostat sections and urothelial cell cultures grown on Snapwell membranes were labeled with polyclonal and monoclonal antibodies to cytokeratins (CKs), tight junction components, and a membrane transport-associated protein (Table 1). The Snapwell cultures were washed in PBS, fixed in a 1:1 mixture of methanol:acetone for 2 min, then air-dried. Cryosections were used unfixed. Urothelial cell cultures and tissue sections were incubated with appropriately diluted primary antibody for 1 h at room temperature. Excess unbound antibody was removed by two washes in PBS, followed by fixation in a 1:1 mixture of methanol:acetone for 2 min. After air-drying, the pretitrated fluorescence-conjugated secondary antibody (Table 1) was applied for 30 min. Slides were washed twice with 0.25% (wt/vol) Tween 20 (polyoxyethylene sorbiton monolaurate) in PBS, incubated for 5 min in a fluorescent DNA intercalating dye (0.1 μg/ml Hoechst 33258 or 2.5 μg/ml propidium iodide in PBS) to counterstain nuclei and rinsed in distilled water, before air-drying. Sections were mounted in glycerol containing 0.1% (wt/vol) p-phenylenediamine to prevent photobleaching. Tissue sections were viewed through an Olympus BX60 microscope equipped with wide-aperture oil-immersion objectives, epifluorescent illumination, dual and specific FITC, and Texas red filters (Olympus Southall, UK). Labeled urothelial cell cultures were also analyzed using a Nikon Bio-Rad laser confocal microscope equipped with an argon laser. Omission of primary antibodies from the labeling protocol served as negative controls.
Cell cultures were lysed directly into reducing SDS electrophoresis sample buffer, resolved on 8–16% SDS polyacrylamide gradient gels, and electrotransferred onto nitrocellulose membranes. Membranes were incubated with titrated primary monoclonal antibodies against CK13 or CK14 (Table 1) for 16 h at 4°C. Bound antibody was detected with goat anti-mouse IgG Alexa 680 (Molecular Probes, Paisley, UK) and visualized on a LI-COR Odyssey infrared scanner (LI-COR Biosciences UK, Cambridge, UK). To check loading, blots were stripped and reprobed with anti-β-actin monoclonal antibody (Sigma) followed by secondary anti-mouse IgG Alexa 680 (Molecular Probes) and detected as above.
Transmission electron microscopy.
Samples of freshly isolated tissue and urothelial cell cultures propagated on Snapwell filters were fixed in 0.1 M phosphate buffer (pH 7.2) containing 4% (wt/vol) paraformaldehyde and 2.5% (wt/vol) glutaraldehyde for 1–2 h at room temperature. The samples were washed in phosphate buffer and postfixed for 1 h at 20°C with 1% (wt/vol) osmium tetroxide in 0.1 M phosphate buffer. Specimens were dehydrated through graded ethanols, cleared in propylene oxide, and impregnated with increasing ratios of Araldite-Polarbed resin/propylene oxide with final embedding in 100% Araldite-Polarbed resin. Seventy-nanometer sections were prepared on gold mesh grids, stained with 2% uranyl acetate, followed by 0.25% lead citrate in 0.4% NaO. Specimens were viewed at 80 kV in a Jeol 1200EX electron microscope (Joel, Garden City, UK).
Scanning electron microscopy.
Urothelial cell cultures on Snapwell filters were fixed as for transmission electron microscopy, dehydrated through graded ethanols, and critical point dried. Coated specimens were examined in a Hitachi S-2400 scanning electron microscope at various magnifications, with an accelerating voltage of 8 kV. Micrographs were taken using a Pentax A3 Date S camera on Kodak T-max film.
Means and standard deviations or standard errors were used as descriptive statistics. For determination of statistical significance, Instat 3 software (Graphpad.com) was used for analysis of variance using the Kruskal-Wallis test (nonparametric ANOVA). A P value <0.05 was regarded as statistically significant.
Phenotypic Properties of Cultured Human Urothelial Cells
Urothelial cell cultures derived from the renal pelvis, ureter and bladder of different donors displayed similar growth characteristics, morphological, and immunocytochemical characteristics irrespective of the age, sex, and clinical status of the individual.
Primary and passaged urothelial cell cultures established and propagated in KSFMc formed monolayers when grown on Primaria tissue culture plastic. When seeded onto Snapwell membranes, the urothelial cells maintained in KSFMc continued to grow as monolayers, but after 24–48 h, small islands of cellular stratification also developed (Fig. 1B). Urothelial cell cultures transferred onto Snapwell membranes and switched to medium supplemented with calcium and/or FBS exhibited cellular stratification of between 3 and 7 cell layers (Fig. 1, C, E, and G).
At sites of urothelial stratification, prominent intercellular tight junctions were visible between the superficial cells, irrespective of whether the cells had been propagated in KSFMc or KSFMc supplemented with calcium and/or FBS (Fig. 1, D, F, and H). Reflective of the native urothelium, immunocytochemical studies demonstrated that the tight junctions consisted of the cortical protein zonal occludens-1 and the integral proteins occludin and claudin 4 (Fig. 2). Claudin 1 was expressed at the cell border at sites of urothelial stratification (Fig. 2). In addition, the transport protein Na+-K+-ATPase was located inferior to the tight junctions and restricted to the basolateral membrane of the superficial cells (data not shown).
Ultrastructurally, the apical membrane of cultured urothelial cells was flat with numerous small microvilli, irrespective of the medium in which the cells had been propagated. There were no obvious concave thickened regions, characteristic of the AUM plaques of superficial urothelium in situ (Fig. 1A, C, E, and G).
The CK expression profile was used to determine the stage of maturation of the urothelial cultures (36). As in native urothelium, NHU cells grown on Snapwell membranes were positive for CK8, CK13, CK17, CK18, and CK19, irrespective of the culture medium in which they were propagated (Table 2). None of the cultured urothelial cells expressed CK20. However, the composition of the culture medium affected the proportion of cells expressing CK13 and CK14 (Fig. 3, A and B), with an increase in the number of CK13-positive and a decrease in the number of CK14-positive cells in FBS-supplemented medium. Western blot analysis demonstrated a 17-fold increase in CK13 expression and a 1.6-fold decrease in CK14 expression in urothelial cells propagated in medium supplemented with FBS and calcium, relative to control cultures maintained in KSFMc (Fig. 3B).
Functional Properties of Cultured Human Urothelial Cells
The TER of the urothelial cell cultures was significantly affected by the medium in which the NHU cells had been propagated. Cultures established and maintained in KSFMc exhibited a mean TER of 18.5 ± 2.4 Ω · cm2 (Table 3). When cultures were switched to 2 mM calcium KSFMc, the mean TER increased, but not significantly, to 49.4 ± 8.9 Ω·cm2. Urothelial cells passaged into KSFMc supplemented with FBS and either maintained in this medium or transferred to KSFMc supplemented with FBS and calcium, exhibited a significantly increased TER of 2,509.9 ± 172.2 Ω · cm2 (P < 0.001) and 3,023.4 ± 564.4 Ω·cm2 (P < 0.001), respectively, relative to the KSFMc control; there was no statistical difference between these two TER values.
Transurothelial sodium ion transport.
Transcellular transport of sodium ions was demonstrated in NHU cell cultures propagated in KSFMc supplemented with FBS and adjusted to 2 mM calcium. When added to the basal aspect of the culture, amiloride had no effect on the transepithelial potential difference or short-circuit current (Fig. 4). By contrast, apically applied amiloride decreased both measured electrophysiological parameters (Figs. 4 and see 6A), suggesting that the urothelial cells transferred sodium ions transcellularly via an apical membrane-restricted amiloride-sensitive ion channel. The change in the short-circuit current induced by amiloride was dose dependent, and the inhibition constant for amiloride was 340 nM (Fig. 5).
The role of Na+-K+-ATPase in sodium ion transport across in vitro propagated urothelial cultures was investigated using ouabain. The addition of ouabain to the apical aspect of the cultures had minimal effect on the transepithelial potential difference or the short-circuit current (Fig. 6C). However, addition of ouabain to the basal side of the cultures markedly reduced both of the electrophysiological parameters (Fig. 6D), suggesting that sodium ions were transported across the basolateral membrane via an active transport mechanism involving Na+-K+-ATPase.
Urea and water permeability of human urothelial cell cultures.
The mean diffusive permeability of urea through the urothelial cell cultures propagated in KSFMc and KSFMc supplemented with calcium and/or FBS ranged from 2.7 × 10−5 to 14.5 × 10−5 cm/s (Fig. 7). The diffusive permeability of urea through the urothelial cell cultures propagated in KSFMc supplemented with calcium was significantly less than that recorded for the cultures maintained in KSFMc (P < 0.05) and KSFMc supplemented with FBS (P < 0.01). There was no statistical difference in the urea permeability of cultures propagated in KSFMc supplemented with calcium and those maintained in KSFMc supplemented with calcium and FBS.
The diffusive permeability of water through the urothelial cell cultures propagated in KSFMc supplemented with calcium was significantly less than that for the cultures maintained in KSFMc plus FBS (P < 0.05).
Dextran permeability of human urothelial cell cultures.
The culture medium had a significant effect on the dextran permeability of the urothelial cultures. Relative to the cells propagated in KSFMc, the urothelial cells switched to serum-supplemented medium had a significantly lower permeability to both species of dextran (P < 0.01; Fig. 8). In addition, urothelial cells cultured in calcium-supplemented medium exhibited a lower permeability to 4,400, but not 9,500, molecular weight dextran. There was no difference in permeability between cells propagated in serum-supplemented medium and those cultured in KSFMc supplemented with both calcium and serum.
Since the first serial cultivation of NHU cells in vitro (34), considerable progress has been made to improve propagation techniques, identify gene and antigenic markers of urothelial phenotype, and demonstrate that cultured urothelial cells retain the capacity to differentiate (reviewed in Ref. 38). However, with few exceptions (28, 40), there has been little focus on developing functional urothelial tissue equivalents from propagated cells. The majority of studies have used primary cultures of nonhuman mammalian urothelium to demonstrate aspects of differentiated urothelial tissue function (42), relying on preexisting rather than de novo differentiation.
This study has investigated the functional potential of NHU cells following removal from the disciplined hierarchy of an in situ tissue and propagated as highly proliferative monocultures in vitro. These cells have previously been shown to retain the capacity to express genes and proteins associated with terminal urothelial differentiation (44), but their capacity to form an integrated functional tissue has not been assessed. As it is possible to propagate large numbers of NHU cells from small surgical biopsies (19), the potential to generate a functionally equivalent urothelium from these cells has important implications for tissue engineering and the development of models for the study of the physiological and pharmacological properties of human urothelium.
According to the definition that leaky epithelia typically have a TER <500 Ω·cm2, whereas “tight” epithelia have resistances >500 Ω·cm2 (11), mammalian urothelium is regarded as a tight epithelium with low ionic permeability. Although not documented, we anticipated that the TER of human urothelium would be of a similar magnitude to that of other mammalian species, as human bladder tissue has comparable permeability properties (8). This study has demonstrated for the first time that it is possible to develop a urothelium from propagated NHU cells that exhibits a high TER (>3,000 Ω·cm2), thus establishing that cultured NHU cells retain the capacity to reestablish the barrier properties of urothelium in situ.
Perrone and colleagues (33) also assessed the electrophysiological properties of human urothelial cells in culture, but they achieved a mean TER of only 576 Ω·cm2, an order of magnitude below that recorded in the present study. Whereas NHU cell lines in this study were obtained from patients with no history or evidence of urothelial pathology, Perrone and colleagues generated an immortalized cell line from an individual with interstitial cystitis. Thus the difference in results may be due to the functional capacity of the cells being compromised by SV40T immortalization and/or other differences in culture conditions. Nevertheless, it has been proposed that increased urothelial permeability has a role in the pathophysiology of interstitial cystitis (32). Thus the difference in TER values obtained in this and Perrone's study (33) could be due to an inherent dysfunction in the urothelium in interstitial cystitis and indicates the potential value of in vitro models for studying interstitial cystitis and other such associated conditions.
Mammalian transurothelial sodium ion transport, which has been demonstrated both in vivo (8) and in vitro (24), may have important physiological roles in sodium homeostasis (26) and bladder sensation (10). The hydrostatic pressure gradient across urothelial tissue influences transurothelial sodium ion flux and extracellular release of ATP (10). Through interaction with purinergic P2X3 ion channels, extracellular ATP acts as a neurotransmitter, modulating the afferent limb of the micturition reflex (5). Thus the urothelium is implicated as a sensor and transmitter of information from the physical environment, but the precise nature of these interactions, their role in modulating cellular processes and relevance to dysfunctional uropathies are largely unknown (3, 41), particularly in humans. This study has shown that NHU cells in vitro transport sodium ions transcellularly via apical membrane-restricted sodium ion channels and the Na+-K+-ATPase ion pump in the basolateral membrane. In addition, the measured IC50 for amiloride inhibition of the short-circuit current was 340 nM, comparable to that previously reported for native rabbit urothelium (10, 27). Thus the cell culture model described in this study is an ideal and practical tool for investigating physiological mechanisms in normal human urothelium.
It was demonstrated that the phenotype of NHU cells in vitro is influenced by the exogenous calcium concentration of the growth medium. However, unlike previous studies that used comparable culture methodologies (19, 37), urothelial cells were shown to undergo stratification in low-calcium conditions, but only when grown on a permeable growth surface. The induction of cellular stratification may have been precipitated by the composition and architecture of the permeable membrane (13), and/or due to membrane facilitation of improved nutrient exchange through the basal cell layer (14).
Analysis of cytokeratin isotype expression suggested that the differentiation status of the urothelial cells was also influenced by the culture conditions. NHU cells in culture express CK8, CK17, CK18, and CK19, all characteristic of native urothelium (19, 37). However, in agreement with previous reports, NHU cells propagated in KSFMc showed downregulation of CK13 in favor of CK14, a CK isotype associated with squamous metaplasia (15), suggesting that in culture, NHU cells switch to a squamous differentiation “program” (45). On transfer into serum-supplemented conditions, the NHU cells readopted a transitional cell phenotype, demonstrating the reversibility of the squamous phenotype of human urothelial cells in vitro.
Although the model described here showed many of the functional and morphological features of normal urothelium, it is still not complete. The study has revealed that NHU cells propagated in vitro can form a partial permeability barrier to water and urea. However, even in the best case, the permeability to water and urea is greater than expected from in vivo measurements (8). In a recently described knockout mouse, deletion of the UPIIIa gene led to incomplete AUM plaque formation and increased water permeability, despite maintenance of the same TER (18). The lack of AUM plaques in this study suggests that despite a transitional cell CK profile, full terminal cytodifferentiation was not attained, thus potentially explaining the measured permeability values. Further modification to the culture system may be required, for example, the introduction of PPARγ agonists (44, 45), to achieve the later stages of urothelial maturation. In static organ culture, a loss of urothelial differentiation with time was suggested to be due to an absence of urine-derived factors or mechanical stimulation (35), and it may be that functional stimulation of cultures will contribute to urothelial maturation.
Terminal junctional complexes positioned between the umbrella cells also contribute to urothelial barrier function, by limiting flux via the paracellular route. There is mounting evidence that the molecular composition of the tight junction defines the barrier properties of different epithelial tissues (43). The constitution of mammalian urothelial tight junctions has begun to be elucidated (1), but that of human urothelium remains to be documented. In the present study, the localization of the tight junction components was similar to that of native tissue, irrespective of the culture medium in which the cells were propagated, suggesting that the documented differences in the permeability properties of the NHU cell cultures was not due to aberrant tight junction formation.
In conclusion, this study has established a methodology to generate a confluent epithelial tissue from in vitro propagated NHU cells that demonstrates many of the functional and phenotypic properties of native urothelium. It clearly demonstrates that NHU cells are not compromised by in vitro propagation, but they retain the capacity to contribute to a functional, ion-transporting epithelium. The full potential of NHU cells to recapitulate the properties of a urinary barrier epithelium will need to be determined by further study, for example, by investigating the expression and function of receptors implicated in sensory mechanisms, such as TRPV1 and TRPM8 (39). However, we feel that the biomimetic human urothelium has a role in dissecting mechanisms involved in normal human urothelial cell physiology and dysfunctional bladder syndromes and has the potential to generate pertinent urothelial facsimiles for bladder tissue engineering (6).
This work was funded by a fellowship awarded to W. R. Cross from the British Urological Foundation and the Ralph Shackman Trust. J. Southgate is sponsored by York Against Cancer.
The authors acknowledge Meg Stark for assistance with electron microscopy.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society