Fig 1.
Formation of well-polarized epithelial cyst by AQP2 expressing MDCK cells.
Single MDCK cells expressing rat AQP2 plated on matrigel form well-polarized cyst. The apical membrane domain is ringed by the tight junction protein (A) Zona Occludens-1 (green) and highlighted by (F) podocalyxin (green). (B) Integrin β1, (C) Na+/K+-ATPase and (D) β-catenin (green) are located on the basolateral membrane. (E, F), AQP2 (red) is distributed diffusely in the cytosol and in the basolateral region under baseline, non-stimulated conditions. DAPI (blue) stains nuclei. Images are single confocal planes taken through the middle of a spherical cyst. Bars = 10 μm.
Fig 2.
Ultrastructure of MDCK-AQP2 cysts.
The ultrastructure of the cyst closely resembles the polarized architecture of renal tubular epithelium in animal kidney. (A) A differential interference contrast image of a whole cyst. The lumen is clearly visible despite some remnants of cells that were shed in the lumen during lumen formation. The basal membrane is in contact with the Matrigel substrate. Scale = 10 μm. (B) In a TEM cross section of a single cell from a MDCK-AQP2 cyst, microvilli can be seen on the apical membrane facing the lumen (L) while the nucleus (N) is oriented toward the basal membrane. Scale = 2 μm. (C) At the apical domain, neighboring cells interact through tight junctions (TJ) and desmosomes (D) and the presence of clathrin-coated pits (arrow) indicate active endocytosis in the apical and subapical region. Scale = 500 nm. (D) Similarly, clathrin coated vesicles are observed at the basal and lateral membranes of the cyst cells (arrows). Scale = 500 nm. In all images (M) indicates mitochondria.
Fig 3.
Regulated trafficking of AQP2 is intact in MDCK cysts.
AQP2 trafficking in MDCK-AQP2 cysts is intact and staining patterns for total AQP2 are comparable to those observed in Brattleboro, and normal rat kidney. (A) MDCK-AQP2 cysts were incubated in serum free medium for 120 minutes. Addition of AVP, FK, or CPT-cAMP to the medium for 40 minutes resulting in apical membrane accumulation of AQP2 (arrows). Bar = 10 μm. (D) Asterisk denotes that significantly (P < = 0.05) more apical, but not basolateral staining of AQP2, relative to intracellular AQP2, was observed following stimulation with AVP, FK or CPT-cAMP. N = 5 cysts (NT, non-treated), N = 12 cysts (AVP/FK/CPT-cAMP stimulated). The data for AVP, FK and cAMP-treated cysts were pooled together because we saw no statistically significant difference in Apical/Internal total AQP2 or Basolateral/Internal total AQP2 between the treatment modalities. (B) In the Brattleboro rat kidney AQP2 was located mainly in the subapical region while apical membrane accumulation of AQP2 was seen after treatment with dDAVP for 3 days. Bar = 10 μm. (C) Similarly, in a tissue slice culture from normal rat kidney, incubation in medium without VP resulted in AQP2 in the cytosol and subapical region. dDAVP treatment for 20 minutes resulted in AQP2 translocation to the apical membrane, with AQP2 still detectable in the cytosol. (E) In transmission electron micrographs, AQP2 in the MDCK-AQP2 cyst is labeled with 15nm gold particles. AQP2 gold particles distributed diffusely throughout the cytosol under baseline, non-stimulated conditions (left panel) while AQP2 accumulated on the apical membrane after VP stimulation (right panel), but not on the basolateral membrane (S2 Fig). Bars = 500 nm
Fig 4.
Constitutive recycling of AQP2 is detected in the basolateral domain of the MDCK-AQP2 cyst.
In the MDCK-AQP2 cyst, blockade of endocytosis by addition of mβCD for 20 minutes led to predominant enrichment of AQP2 in the basolateral domain. (A) Immunofluorescence staining reveals basolateral accumulation of total AQP2 with mβCD treatment. (B) In transmission electron micrographs, some 15nm gold AQP2-labeled particles are found in the apical/subapical domain (left panel) but they are more predominantly presented on the basolateral membranes of two neighboring cells. Bar = 500 nm
Fig 5.
Subcellular distribution of phosphorylated AQP2 in MDCK cells grown in 3D culture and in Brattleboro rat kidney.
Antibodies recognizing AQP2 phosphorylated at S256, S261, S264 or S269 highlighted the subcellular localization of phosphorylated AQP2 in MDCK-AQP2 cysts (A-H). The staining pattern of differentially phosphorylated AQP2 in the cyst is reminiscent of that observed in Brattleboro rat kidneys (I-P). Prior to stimulation, cysts were pre-incubated in serum free medium for 120 min. Comparing the distribution of phosphorylated AQP2 in cyst versus kidney tissue, pS256 is found in the cytosol and apical membrane in both the cyst and kidney tissue without VP treatment (A, I) Total AQP2 staining of the tissue in panel I can be seen in Fig 3B. After stimulation with VP (E,M), the redistribution of pS256 AQP2 to the apical membrane is pronounced in the kidney tissue, but is not significantly different in the cyst tissue (quantification in panel R). VP stimulation resulted in a reduction of both apical and cytosolic pS261 fluorescence intensity in cyst and kidney tissue (B vs. F, J vs. N). Under stimulated conditions, some non-specific labeling of the nuclei by the p261 antibody can be observed (F). pS264 staining signal translocated from the cytosolic compartment (C,K) to the apical membrane (G, O) after VP stimulation in both cyst and kidney tissue (H, P). In both the cyst, and the rat tissue pS269 staining was only observed following stimulation, and solely located on the apical membrane (H,P). Bar = 10 μm. (Q) Differential phosphorylation of AQP2 at various serine residues after FK stimulation was detected by western blot in AQP2 expressing MDCK cells grown as a monolayer culture. The bar graph represents quantification of western blots (means ± SE, N = 3 experiments): total AQP2 and phosphorylation antibody results are presented as relative to loading control. No significant difference was observed between non-stimulated and VP/FK simulated conditions for AQP2 pS256 or pS264. On the other hand, stimulation resulted in a significant (asterisk) decrease in pS261 (p = .002) and a significant increase in pS269 (p = .04). (R) Quantitative assessment of cyst immunofluorescence images reveal significant (P < = 0.05) increases in apical, relative to intracellular, pS264 (N = 4 cysts) and pS269 (N = 8 cysts) AQP2 staining following AVP stimulation. No significant difference in apical, relative to intracellular pS261 (N = 5 cysts) fluorescence is observed due to concomitant decreases (relative to treatment) in both apical and intracellular pS261 staining.