Aquaporin expression contributes to human transurothelial permeability in vitro and is modulated by NaCl

Abstract

It is generally considered that the bladder is impervious and stores urine in unmodified form on account of the barrier imposed by the highly-specialised uro-epithelial lining. However, recent evidence, including demonstration of aquaporin (AQP) expression by human urothelium, suggests that urothelium may be able to modify urine content. Here we have we applied functional assays to an in vitro-differentiated normal human urothelial cell culture system and examined both whether AQP expression was responsive to changes in osmolality, and the effects of blocking AQP channels on water and urea transport. AQP3 expression was up-regulated by increased osmolality, but only in response to NaCl. A small but similar effect was seen with AQP9, but not AQP4 or AQP7. Differentiated urothelium revealed significant barrier function (mean TER 3862 Ω.cm(2)), with mean diffusive water and urea permeability coefficients of 6.33×10(-5) and 2.45×10(-5) cm/s, respectively. AQP blockade with mercuric chloride resulted in decreased water and urea flux. The diffusive permeability of urothelial cell sheets remained constant following conditioning in hyperosmotic NaCl, but there was a significant increase in water and urea flux across an osmotic gradient. Taken collectively with evidence emerging from studies in other species, our results support an active role for human urothelium in sensing and responding to hypertonic salt concentrations through alterations in AQP protein expression, with AQP channels providing a mechanism for modifying urine composition. These observations challenge the traditional concept of an impermeable bladder epithelium and suggest that the urothelium may play a modulatory role in water and salt homeostasis.

Figure 1. Immunolocalisation and expression of AQP3 in response to altered medium osmolality.

a) Immunofluorescence labelling of non-differentiated (proliferative) and differentiated NHU cell cultures grown on glass slides showing increased labelling intensity and membrane localisation of AQP3 at high concentrations of NaCl, but not urea, following 72 hours of exposure. Scale bar: 10 µM. b) Immunohistochemistry of differentiated NHU cell constructs grown on permeable Snapwell membranes and exposed for 72 hours to indicated osmolalities. Note no effect of urea, but major increase in AQP3 expression in all layers in response to medium made hyperosmotic (500 mosm/kg) by the addition of NaCl, irrespective of exposure via apical or basal aspect. Scale bar: 50 µM.

Figure 2. Analysis of AQP3 expression by immunoblotting.

a) AQP3 expression by non-differentiated (left panel) and differentiated (right panel) NHU cell cultures following osmotic stress, compared to β-actin (loading control). Immunoblotting analysis revealed an up-regulation of AQP3 protein expression by NHU cell cultures treated for 72 hours with NaCl-supplemented hyper-osmotic medium. In hypo-osmotic culture, expression was reduced by 30% and 38%, respectively, relative to control. Urea had no effect on AQP3 protein expression. b) Exposure of non-differentiated NHU cell cultures to a range of osmotic solutes for 48 hours. NaCl induced a 6.4-fold increase in AQP3 protein expression, but other solutes had minimal effect and hypo-osmotic culture medium reduced expression. c) 48 hour responses of non-differentiated NHU cells to a range of osmotic salt concentrations. d) Time-dependency of non-differentiated NHU cell response to hyperosmotic salt. Each panel shows a representative immunoblot for AQP3 compared to β-actin as loading control. Following normalisation for loading, the fold change in AQP3 expression was calculated relative to 295 mosm/kg control in panels a, b and c, or no treatment in panel d. Compiled results from n = 4 experiments performed using three (a and b) or two (c and d) independent NHU cell lines are presented as bar charts and show the mean fold change ± SD. Significant changes relative to control are marked as * P<0.05; ** P<0.01.

Figure 3. Immunolocalisation and expression of AQPs 4, 7 and 9 in response to altered medium osmolality.

AQP expression and localisation was determined by immunofluorescence microscopy in proliferating NHU cultures (a) and differentiated urothelial constructs (b). AQP4 and 7 protein expression was not affected by culture osmolality. This finding was independent of the phenotype of the cells. AQP9 was expressed in the cytoplasm of differentiated cells grown under physiological conditions, in proliferative and differentiated cells exposed to hyperosmotic conditions, but was undetectable in differentiated cultures subjected to hypoosmotic culture. Scale bar: 50 µM.

Figure 4. Water and urea permeability of differentiated urothelial tissue constructs.

Diffusive permeability coefficients across differentiated urothelial cell cultures are shown for (a) [³H]-water and (b) [¹⁴C]-urea. Results were compiled from a total of 10 urothelial constructs from five independent cell lines. Data shown as mean ± SD.

Figure 5. Effect of adaptation to hyperosmotic NaCl conditions on transurothelial permeability to water and urea.

Differentiated NHU constructs were maintained in normal medium (295 mosm/kg) or preconditioned for 48 hours in medium adjusted to 500 mosm/kg with NaCl. Immediately prior to permeability studies, the medium was replaced with normal medium. Urothelial constructs were investigated for permeability to water (a) and urea (b) in an osmotic equilibrium (to investigate diffusive permeability: Pd) or in the presence of a 100 mosm/kg transepithelial osmotic gradient (to determine permeability flux: Pf). Pd and Pf were determined by measuring radio-isotopic fluxes from the apical compartment across the tissue to the basal compartment. Each bar shows the mean ± SD for n = 4. Adaptation to hyperosmotic conditions resulted in a significant increase in water (P<0.001) and urea (P<0.001) flux across the urothelium in the presence of an osmotic gradient, but did not affect the diffusive permeability of either (P>0.05).

Figure 6. The effect of mercuric chloride on the permeability of differentiated urothelial constructs.

Water and urea permeability coefficients of differentiated urothelial constructs are shown 15 minutes after application of HgCl₂. Results are compiled from 8 independent urothelial constructs. Exposure to HgCl₂ significantly reduced transurothelial permeability to water and urea (* P<0.05; ** P<0.01). The water permeability coefficient P_D diminished by 2.9 fold and the urea permeability coefficient P_D diminished by 3.6 fold following exposure to 300 µM HgCl₂.