Fluid-shear-stress-induced translocation of aquaporin-2 and reorganization of actin cytoskeleton in renal tubular epithelial cells

Abstract

In vivo, renal tubular epithelial cells are exposed to luminal fluid shear stress (FSS) and a transepithelial osmotic gradient. In this study, we used a simple collecting-duct-on-a-chip to investigate the role of an altered luminal microenvironment in the translocation of aquaporin-2 (AQP2) and the reorganization of actin cytoskeleton (F-actin) in primary cultured inner medullary collecting duct (IMCD) cells of rat kidney. Immunocytochemistry demonstrated that 3 h of exposure to luminal FSS at 1 dyn cm⁻² was sufficient to induce depolymerization of F-actin in those cells. We observed full actin depolymerization after 5 h exposure and substantial re-polymerization within 2 h of removing the luminal FSS, suggesting that the process is reversible and the fluidic environment regulates the reorganization of intracellular F-actin. We demonstrate that several factors (i.e., luminal FSS, hormonal stimulation, transepithelial osmotic gradient) collectively exert a profound effect on the AQP2 trafficking in the collecting ducts, which is associated with actin cytoskeletal reorganization.

---

Insight, innovation, integration

Renal tubular epithelial cells are exposed to large transepithelial osmotic gradients, a luminal fluid shear stress (FSS) and a number of hormones. In this work, we studied the role of altered renal environments such in fluidic shear stress, hormonal stimulation, and transepithelial osmotic gradients on renal tubular epithelial cells using a simple collecting-duct-on-a-chip. We discovered that various combinations of environmental factors are directly involved with depolymerization of actin cytoskeleton and trafficking of aquaporin-2 to the plasma membrane. This provides an insight into how a simple microfluidic device in a multi-layer format can generate tubular dynamics in vitro and be used to find optimum fluidic conditions for renal tubular cells.

Introduction

Renal tubular epithelial cells in vivo are exposed to a mechanical force in the form of fluid shear stress (FSS) in the tubular lumen as well as large transepithelial osmotic gradients between the apical and basolateral plasma membranes. They are also under the control of a number of hormones and cytokines acting on different sides of the cells for renal reabsorption and excretion of water and electrolytes. The collecting duct is the vasopressin-regulated water reabsorption site which is important in maintaining body water homeostasis. Aquaporin-2 (AQP2) is a vasopressin-regulated apical water channel protein expressed in the collecting duct principal cells and constitutes a major route of transcellular water transport across the apical plasma membrane. Numerous studies have shown that actin microfilaments and motor proteins are interactively involved in the translocation of AQP2 to the plasma membrane.^1–4

Despite the well-known role of phosphorylation in the regulation of AQP2 trafficking,^5,6 the effects of changes in the luminal microenvironment of the collecting ducts on the actin reorganization and AQP2 trafficking are poorly understood. Two-dimensional substrate-based approaches using a culture dish or transwell system have been widely employed to investigate such changes in the cellular morphology and function in vitro. However, these conventional techniques can only poorly mimic the cellular microenvironments consisting of a combination of precise biochemical and mechanical cues. Thus, researchers have developed various microfabrication- and microfluidics-based techniques to better mimic the complexity of in vivo microenvironment.^7–14 For the renal tubular cells, we recently developed a simple collecting-duct-on-a-chip in the form of a multi-layer microfluidic device (MMD), which allows for in vivo-like fluidic environments and efficient culture and analysis of the cells, as verified by enhanced cell polarization and cytoskeletal reorganization.¹⁵ Following the introduction of the chip, here we present the effects of changes in the luminal or basolateral microenvironments such as differential FSS, hormonal stimulation, and transepithelial osmotic gradients on the translocation of AQP2 and the reorganization of actin cytoskeleton in primary cultured inner medullary collecting duct (IMCD) cells of rat kidney (Fig. 1). A range of parameters were tested for FSS, transepithelial osmotic gradients, and stimulation with arginine vasopressin (AVP), with the aim of developing a simple tool for drug screening for the regulation of AQP2 and AVP V2 receptor and a model system for studying renal physiology and pathophysiology for body water balance.

Fig. 1 Schematic of the collecting-duct-on-a-chip. The multi-layer microfluidic device (MMD) is a sandwiched assembly of a PDMS channel, polyester porous membrane, and PDMS reservoir. Renal cells are cultured inside the channel. The system allows for precise manipulation of environmental parameters such as luminal fluid shear stress, hormonal stimulation, and transepithelial osmotic gradient.

Experimental

Multi-layer microfluidic device (MMD)

We recently developed a simple collecting-duct-on-a-chip by integrating a polydimethylsiloxane (PDMS) channel and a porous polyester membrane¹⁵ for analyzing the actin cytoskeleton rearrangement and AQP2 translocation in the IMCD cells. Briefly, a sandwiched assembly of PDMS (Sylgard 184 Silicone elastomer kit, Dow Corning, Midland, MI) microfluidic molds (1 mm in width, 1 cm in length, and 100 μm in height), PDMS wells (∼2 mm in width and ∼0.6 cm in length) and polyester membrane (∼6 mm in width and ∼1.6 cm in length, pore size: 0.4 μm, 10 μm thickness, transwell insert, Corning Inc, Acton, MA) were irreversibly bonded after oxygen plasma treatment (60 W, PDC-32G, Harrick Scientific, Ossining, NY) for 1 min. Details of the fabrication procedure can be found in our previous studies.^15,16

Primary culture of rat kidney IMCD cells and flow experiments

Primary culture of IMCD cells was performed, as previously described.¹⁷ IMCD cells were seeded on MMD with hypertonic culture medium (DMEM/F-12, 80 mM urea, 130 mM NaCl, 10 mM HEPES, 2 mM L-glutamine, penicillin–streptomycin (10 000 U ml⁻¹) and 10% FBS (pH 7.4, 640 mOsm per kg H₂O) and the PDMS well was filled with medium. The MMD was plated on a Petri dish (30 mm) with medium. IMCD cells were cultured for 3 days in hypertonic culture medium at 37 °C in a humidified 5% CO₂ incubator. At day 4, after confluency, the cells were exposed to a FSS of 0.2–5 dyn cm⁻² for 0–12 h with hypertonic flow medium (DMEM/F12 with 1% FBS, 640 mOsm per kg H₂O) using a syringe pump (KD Scientific, Holliston, MA) at 37 °C. This FSS was calculated using the equation τ = 6μQ/bh², where μ is medium viscosity (g cm⁻¹ s⁻¹), Q is the volumetric flow rate (cm³ s⁻¹), b is channel width, and h is channel height.¹⁸ All experiments were completed in triplicate.

Immunocytochemistry

After stimulation, the cell-attached membrane substrates were taken apart from the MMD and immunocytochemistry was performed, as previously described.¹⁵ Cells were labeled with tetramethylrhodamine isothiocyanate (TRITC)–phalloidin (1 : 200 in PBS, Sigma, St. Louis, MO) for filamentous actins (F-actin) or anti-AQP2 primary antibodies (1 : 100 in PBS, Santa Cruz Biotechnology, Inc., CA). The secondary antibodies were donkey anti-goat IgG-FITC (1 : 100 in PBS, Jackson ImmunoResearch, West Grove, PA). A confocal laser microscope (LSM510, Carl Zeiss, Jena, Germany) was used to capture the images.

Results

The FSS within the renal tubular system was estimated in the range of 0.2–20 dyn cm⁻², which is ∼10% of that of the endothelial cell.¹⁹ Here, a FSS of 1 dyn cm⁻², which is assumed to be a representative value in the collecting ducts, was applied for 0–12 h to the luminal membrane of primary cultured IMCD cells in the absence of transepithelial osmotic gradients and AVP stimulation (osmolality: 640 mOsm per kg H₂O). By doing so, we were able to see the onset of actin depolymerization as a function of time. Fig. 2A shows the fluorescent images of F-actin in response to a luminal FSS of different durations. Up to 1 h, F-actin was assembled to form thick fibers abundantly localized throughout the cytoplasm. At 3 h, F-actin fibers became thinner in a depolymerized form. A FSS of longer duration caused a partial redistribution of F-actin with a reduction of actin bundles and an appearance of a dot-like depolymerized form. At 5 h, F-actin was fully depolymerized and dot-like patterns of F-actin were seen in the cytoplasm, whereas thick actin fibers were only observed at the periphery of the cells (i.e., lateral cell borders). This change was also seen in the x–z reconstruction images of the F-actin staining (lower panels in Fig. 2A) and magnified images of apical and basal regions (right panels denoted as a and b in Fig. 2A). To examine the F-actin remodeling, we performed semi-quantitative analysis of F-actin intensity in randomly selected areas of the cytoplasm in the apical and basal regions of the cells, based on the z-sectioned confocal images at five different durations (Fig. 2B). As shown, the fluorescent intensity was reduced by ∼60% in the apical region and ∼50% in the basal region with respect to the original intensity and remained almost unchanged after 5 h. Moreover, the F-actin intensity ratio of membrane to cell body after 0 h and 5 h of FSS in four different regions (with images from the apical to the basal region) revealed that F-actin was fully depolymerized in the cytoplasm, whereas thick actin fibers were observed at the periphery of the cells (Fig. 2C and D). These results suggest that luminal FSS per se can induce spontaneous depolymerization of F-actin in the IMCD cells and decrease F-actin intensity in both apical and basal regions of the cells. These effects initially become distinguishable after 3 h of exposure at 1 dyn cm⁻² of FSS.

Fig. 2 Effects of luminal fluid shear stress on IMCD F-actin reorganization. Labeling of F-actin in the IMCD cells cultured on the device in response to a FSS of 1 dyn cm⁻². (A) Representative confocal microscope images of F-actin staining, along with x–z reconstruction, for different durations of FSS application. For each 10 μm × 10 μm panel, magnified views of the (a) apical and (b) basal regions are shown. F-actin figures are depolymerized by 3 h. (B) Quantitative analysis of F-actin intensity in the randomly selected areas of the cytoplasm with five different durations in both apical and basal regions. Twelve regions of interest (each 100 μm²) at each time point were randomly selected and compared by using ImageJ. (C) Four optical sectioned images from the apical to the basal region at 0 h and 5 h of exposure to the FSS. The plot profile shows the F-actin intensity for the white dotted line. Blue lines indicate the cell membrane. (D) The F-actin intensity ratio of membrane to cell body with 0 h and 5 h of FSS in four different regions. The increase in ratio indicates loss of F-actin in the cytoplasm, with relative preservation of F-actin associated with the membranes. Twenty regions of interest (4 μm²) at each optical section were randomly selected and compared by using ImageJ. *, P < 0.01 versus 0 h; **, P < 0.001 versus 0 h. Error bars indicate SEM.

Based on the above observation, F-actin appeared to be completely depolymerized after 1 dyn cm⁻² of FSS for 5 h. In order to test whether F-actin can be polymerized again after withdrawal of the luminal FSS, the cells were left undisturbed for a period of time (<2 h) under the static condition (0 dyn cm⁻²) and the fluorescent images of actin fibers were taken for each time point. During the initial duration up to 1 h, F-actin seemed to remain depolymerized with a less intensity in the cytoplasm (Fig. 3A). After withdrawl of FSS for 2 h, repolymerization was observed in the cytoplasm, as judged by the appearance of thick and abundant fibers in several cells (Fig. 3B), suggesting that the actin polymerization–depolymerization is reversible and strictly regulated by FSS. In Fig. 3C, the F-actin fibers prior to the withdrawal of FSS (5 h FSS) and the application of FSS (0 h FSS) are shown for comparison. The fluorescence intensity of F-actin in the cytoplasm showed a significant increase during withdrawal of FSS (∼2-fold increase within 2 h rest), corresponding to the intensity of F-actin prior to the application of FSS (Fig. 3C and D). This finding indicates that the luminal FSS alone regulates the reorganization of intracellular F-actin in the IMCD cells.

Fig. 3 Repolymerization following removal of luminal FSS. Full depolymerization of actin fibers was induced in F-actin-labeled IMCD cells by 5 h of luminal FSS at 1 dyn cm⁻². Confocal micrographs, with x–z reconstructions, are shown for different durations following the removal of luminal FSS. (A) 1 h after withdrawal, F-actin fibers are still depolymerized with lower intensity in the cytoplasm. (B) F-actin fibers are repolymerized in the cytoplasm at 2 h of withdrawal after depolymerization. (C) F-actin fibers prior to withdrawal of FSS (5 h FSS) and application of FSS (0 h FSS) are shown for comparison. (D) Quantitative analysis of F-actin intensity in the randomly selected areas of the cytoplasm. Twelve regions of interest (each 100 μm²) at each time point are randomly selected and compared by using ImageJ. Scale bar, 10 μm. *, P < 0.01; **, P < 0.001 versus 5 h FSS. Error bars indicate SEM.

Next, to examine the role of altered tubular FSS on F-actin depolymerization, the cells were exposed to a range of FSS (0.2–5 dyn cm⁻²) for the same duration of 5 h. As shown in Fig. 4A, a low FSS of 0.2 dyn cm⁻² was insufficient for inducing the full depolymerization of F-actin, as can be seen from randomly organized F-actin along the cytoplasm (Fig. 4A). In contrast, a FSS larger than 1 dyn cm⁻² induced the full depolymerization of F-actin (Fig. 4B–D), in that thick actin fibers were mostly observed at the periphery of the cells with inner dot-like patterns. This change was further demonstrated by an x–z reconstruction of the F-actin staining, showing that depolymerization mainly occurred in both apical and basal regions of the cells. These results suggest that F-actin depolymerization is regulated by both the magnitude and duration of the luminal FSS.

Fig. 4 Effects of flow velocity on F-actin depolymerization. Confocal micrographs, with x–z reconstructions, of F-actin-labeled IMCD cells cultured on the device for 5 h with different flow velocities. (A) 0.2 dyn cm⁻². (B) 1 dyn cm⁻². (C) 2 dyn cm⁻². (D) 5 dyn cm⁻². F-actin exposed to a low flow velocity is less depolymerized (A). For flow velocities larger than 1 dyn cm⁻², full depolymerization of F-actin was observed (B–D). Scale bars: 10 μm.

In the absence of FSS and AVP stimulation, we saw largely dispersed AQP2 and well-organized fibers of F-actin along the cytoplasm (Fig. 5A). When triggered by AVP hormonal stimulation, dramatic changes were observed. For example, a simple treatment with 10⁻¹¹ M of AVP for 30 min gave rise to an increase in AQP2 labeling of the plasma membrane and a decrease in labeling of the cytoplasm in some cells (Fig. 5B). Importantly, for the cells exhibiting AQP2 trafficking, the depolymerization of F-actin fibers was clearly observed, indicating that the AQP2 trafficking to the plasma membrane is associated with an AVP-induced actin depolymerization. A separate treatment with 10⁻⁹ M (Fig. 5C) or 10⁻⁷ M AVP (Fig. 5D) for 30 min also led to an intense AQP2 labeling of the plasma membrane, which was also accompanied by complete F-actin depolymerization.

Fig. 5 Effects of arginine vasopressin (AVP) stimulation on F-actin and AQP2 in a static environment. Confocal micrographs of TRITC-labeled F-actin (left) and FITC-labeled AQP2 (center) in IMCD cells cultured on the glass substrate without luminal fluid shear stress. Merged channels shown at right. (A) No AVP control. F-actin is polymerized and AQP2 is located in the cytoplasm. (B–D) Representative micrographs following 30 min of AVP treatment. With increasing concentrations of AVP, F-actin depolymerization and AQP2 translocation to the plasma membrane were observed. Scale bars: 10 μm.

To test the effects of FSS on AQP2 trafficking to the plasma membrane, changes of intracellular AQP2 and F-actin distribution were examined in response to a FSS with different magnitudes as shown in Fig. 6. Immunofluorescent images of AQP2 and F-actin demonstrated that both were initially dispersed in the cytoplasm when the cells were grown in a static condition in the absence of AVP (Fig. 6A). In response to a FSS of 0.2 dyn cm⁻² to the luminal side for 5 h in the absence of AVP, AQP2 labeling was still largely dispersed in the cytoplasm but F-actin started to depolymerize in many cells (Fig. 6B). In contrast, a FSS of 1 dyn cm⁻² (Fig. 6C) to the luminal side for 5 h in the absence of AVP was accompanied by significant AQP2 trafficking to the apical plasma membrane and the full depolymerization of F-actin. These observations indicate that the luminal FSS alone induces AQP2 translocation to the apical plasma membrane in the collecting duct cells, which is associated with actin cytoskeletal reorganization. In addition, in response to an FSS of 1 dyn cm⁻² to the luminal side for 5 h combined with subsequent basolateral AVP stimulation for 30 min at 10⁻⁹ M, more intense AQP2 targeting to the apical plasma membrane and F-actin depolymerization were observed (Fig. 6D). Consistent with this, semi-quantitative analysis of AQP2 intensity demonstrated that luminal FSS induces redistribution of AQP2 to the apical region of the collecting duct cells (Fig. 6A–D). FSS of 1 dyn cm⁻² to the luminal side for 5 h in the absence or the presence of basolateral AVP stimulation induced more AQP2 redistribution to the apical region of the cells than that of a static condition and FSS of 0.2 dyn cm⁻² in the absence of AVP.

Fig. 6 Effects of luminal fluid shear stress (FSS) on F-actin and AQP2. Confocal micrographs of IMCD cells with TRITC-labeled F-actin (top) and FITC-labeled AQP2 (middle). Merged channels with x–z reconstructions are shown under the images. Quantitative analysis of AQP2 intensity is shown at bottom. Ten regions of interest (each 100 μm²) at five different sections are randomly selected and compared by using ImageJ. (A) Static control with no AVP stimulation. (B–D) Representative micrographs following 5 h of exposure to fluidic condition. F-actin depolymerization and AQP2 translocation increase with greater fluid velocity (B, C). Combining FSS with AVP (basolateral treatment for 30 min at 10⁻⁹ M concentration) further heightens the effect (D). Scale bars: 10 μm.

Finally, we tested the culture environment with different osmolalities by introducing a solution containing different amount of urea and sodium between microfluidic channel (luminal flow) and PDMS well (basolateral flow). We then investigated whether a transepithelial osmotic gradient across the luminal and basolateral side of the IMCD cells affects the intracellular AQP2 and F-actin distribution. For this experiment, a FSS of 1 dyn cm⁻² was applied for 5 h to all experimental groups (Fig. 7). Increased AQP2 trafficking to the plasma membrane was observed in the presence of a gradient of 300 mOsm per kg H₂O on the luminal side vs. 600 mOsm per kg H₂O on the basolateral side (Fig. 7B), than with 600 mOsm per kg H₂O on the luminal side vs. 600 mOsm per kg H₂O on the basolateral side (Fig. 7A). Full F-actin depolymerization was observed in both conditions. This finding suggests that transepithelial osmotic gradient across the luminal and basolateral plasma membranes, at least in part, plays a role in AQP2 trafficking.

Fig. 7 Effect of a transepithelial osmotic gradient on F-actin and AQP2. Confocal micrographs of TRITC-labeled F-actin and FITC-labeled AQP2 in IMCD cells cultured in the absence of AVP stimulation. (A) No-gradient control: 600 mOsm per kg H₂O on the luminal side and 600 mOsm per kg H₂O on the basolateral side. (B) Gradient condition: 300 mOsm per kg H₂O on the luminal side and 600 mOsm per kg H₂O on the basolateral side. F-actin depolymerization and AQP2 translocation are observed. Scale bars: 10 μm.

Discussion

Recently, some fluidic studies have been carried out to clarify the effects of FSS on the changes of morphology and function of renal tubular cells. For example, Duan et al.¹⁸ demonstrated that FSS caused cytoskeletal reorganization and junctional reformation in renal proximal tubular epithelial cells, which could provide the underlying signaling mechanisms for FSS-induced proximal tubule transport for sodium, bicarbonate, and water reabsorption. Moreover, this finding could suggest the FSS-induced cytoskeletal reorganization as a potential mechanism for the regulation of cytoskeleton-related protein trafficking in tubular epithelial cells (e.g., AQP2). In this study, we utilized a simple multi-layer fluidic device and demonstrated that the luminal FSS increased AQP2 translocation to the apical plasma membrane in the IMCD cells, which was associated with F-actin depolymerization in both apical and basal regions of the cells despite the absence of AVP stimulation (Fig. 2–4 and 6).

The cytoskeleton is known to be involved in the AQP2 trafficking in the kidney collecting duct. In particular, the microtubular network has been implicated in this process, since chemical disruption of microtubules inhibits the increase in osmotic water permeability both in the toad bladder and in the mammalian collecting duct.^20,21 Actin filaments are also involved in the hydrosmotic response.^4,22–26 A number of previous studies demonstrated that vasopressin or forskolin stimulation (i.e., protein kinase A (PKA) activator) increased the apical sorting of water channels associated with depolymerization of actin cytoskeleton in mammalian collecting duct and amphibian bladder.^4,27–29 Moreover, okadaic acid, a phosphatase inhibitor, also induced actin depolymerization leading to AQP2 translocation in renal collecting duct CD8 cells, which was similar to induction by forskolin treatment.²⁹ Interestingly, okadaic-acid-induced actin depolymerization was still visible in CD8 cells pre-treated with H89 (a selective PKA-inhibitor) where AQP2 translocation was apparent.²⁹ In contrast, forskolin did not induce actin network organization in H89-pre-treated cells, in which AQP2 did not translocate to the plasma membrane. This suggests that the reorganization of the actin network is essential for promoting AQP2 translocation to the membrane.

In conjunction with the role of cytoskeleton in the AQP2 trafficking, it has been reported that small GTPase Rho affects AQP2 trafficking through regulation of actin cytoskeleton.³⁰ Inactivation of RhoA by phosphorylation and increased formation of RhoA–RhoGDI complex seem to control the dissociation of actin fibers seen after vasopressin stimulation. Consistent with this, forskolin-induced inhibition of RhoA leads to a partial depolymerization of actin cytoskeleton, promoting AQP2 translocation. Moreover, the FSS modulates the activity of small GTPases in the endothelial cells,^31–33 for which further studies are needed to fully understand the mechanistic details of how small GTPases regulate the response of collecting duct cells to FSS. In addition, it was demonstrated that the FSS also modulates nitric oxide production and increases intracellular concentration of Ca²⁺ in the IMCD cells, both of which play a role in AQP2 trafficking.¹⁹ Finally, it should be noted that the primary cilium functions as a flow sensor in the cultured renal epithelial cells, thus mediating an increase in intracellular Ca²⁺ concentration.^34,35 Further studies would be needed to fully understand the mechanistic details of how small GTPases, nitric oxide production, and primary cilium regulate the response of the IMCD cells to FSS.

Regarding the relatively short (3 day) incubation period of the primary cultured IMCD cells within the microfluidic channel, our preliminary data and other previous studies³⁶ revealed that AQP2 expression in the primary cultured IMCD cells could be significantly reduced after day 2 when the cells were grown under standard culture conditions without treatment of AVP, dDAVP, or cAMP. Moreover, the cells were not treated by vasopressin or cAMP analogue to maintain the AQP2 expression, since the FSS-induced effects on F-actin depolymerization and AQP2 translocation could be complicated by the co-treatment. Thus, we allowed the IMCD cells to grow to confluency for 3 days after an isolation of inner medulla and performed the subsequent experiments at day 4 to obtain an acceptable amount of AQP2 expression for immunocytochemistry. However, due to the relatively short duration of culture, it could be possible that cells were not fully polarized, and the actin fibers that were being lost looked more like stress fibers than the classical “terminal web” seen in native epithelial cells. Thus, a longer incubation could increase the degree of polarization, which might be a more useful way to study the actin cytoskeleton reorganization.

Conclusions

We have presented that a simple collecting-duct-on-a-chip in the form of a multi-layer microfluidic device can be used as an experimental platform for emulating luminal fluidic microenvironments and for studying dynamic changes in the translocation of AQP2 and reorganization of actin cytoskeleton for renal tubular epithelial cells. Various combinations of environmental factors have been systematically investigated so as to generate tubular dynamics in vitro and to find optimum fluidic conditions for kidney tubular cells. These factors include FSS with various durations (1–12 h) and magnitudes (0.2–5 dyn cm⁻²), AVP hormonal stimulation with various concentrations (10⁻¹¹–10⁻⁷ M), and two transepithelial osmotic gradient conditions (300–600 mOsm per kg H₂O).

Our results demonstrated that the F-actin polymerization–depolymerization in both apical and basal regions of the cells is reversible and triggered by FSS, which can be enhanced by AVP hormonal stimulation and a transepithelial osmotic gradient. The underlying mechanisms for the effects of FSS on the F-actin polymerization–depolymerization and AQP2 trafficking need to be elucidated in future studies. The work we have presented here should be useful for developing a simple tool for drug screening for the regulation of AQP2 and a model system for studying renal physiology and pathophysiology for water balance in the body.

Acknowledgements

This study was supported by Nano Systems Institute-National Core Research Center (NSI-NCRC), World Class University (WCU) Program (R31-2008-000-10083-0), Korea Research Foundation Grant (MOEHRD) (Grant KRF-J03003), National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology, Korea (R01-2007-000-20441-0 and 2010-0008225) and the Korea Healthcare Technology R&D Project, Ministry of Health & Welfare (A080143).

References

1. Y. Noda, S. Horikawa, E. Kanda, M. Yamashita, H. Meng, K. Eto, Y. Li, M. Kuwahara, K. Hirai, C. Pack, M. Kinjo, S. Okabe and S. Sasaki, J. Cell Biol., 2008, 182, 587–601.
2. G. Tamma, E. Klussmann, J. Oehlke, E. Krause, W. Rosenthal, M. Svelto and G. Valenti, J. Cell Sci., 2005, 118, 3623–3630.
3. Y. Tajika, T. Matsuzaki, T. Suzuki, A. Ablimit, T. Aoki, H. Hagiwara, M. Kuwahara, S. Sasaki and K. Takata, Histochem. Cell Biol., 2005, 124, 1–12.
4. G. Ding, N. Franki, J. Condeelis and R. M. Hays, Am. J. Physiol., 1991, 260, C9–C16.
5. S. Nielsen, J. Frokiaer, D. Marples, T. H. Kwon, P. Agre and M. A. Knepper, Physiol. Rev., 2002, 82, 205–244.
6. R. A. Fenton, H. B. Moeller, J. D. Hoffert, M. J. Yu, S. Nielsen and M. A. Knepper, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 3134–3139.
7. A. Khademhosseini, R. Langer, J. Borenstein and J. P. Vacanti, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 2480–2487.
8. S. Nagrath, L. V. Sequist, S. Maheswaran, D. W. Bell, D. Irimia, L. Ulkus, M. R. Smith, E. L. Kwak, S. Digumarthy, A. Muzikansky, P. Ryan, U. J. Balis, R. G. Tompkins, D. A. Haber and M. Toner, Nature, 2007, 450, 1235–U10.
9. A. Groisman, C. Lobo, H. J. Cho, J. K. Campbell, Y. S. Dufour, A. M. Stevens and A. Levchenko, Nat. Methods, 2005, 2, 685–689.
10. D. Huh, H. Fujioka, Y. C. Tung, N. Futai, R. Paine, J. B. Grotberg and S. Takayama, Proc. Natl. Acad. Sci. U. S. A., 2007, 104, 18886–18891.
11. R. Derda, A. Laromaine, A. Mammoto, S. K. Y. Tang, T. Mammoto, D. E. Ingber and G. M. Whitesides, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 18457–18462.
12. D. E. Ingber, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 11571–11572.
13. T. Fuji, T. Anada, Y. Honda, Y. Shiwaku, H. Koike, S. Kamakura, K. Sasaki and O. Suzuki, Tissue Eng. A, 2009, 15, 3525–3535.
14. S. R. Khetani and S. N. Bhatia, Nat. Biotechnol., 2008, 26, 120–126.
15. K. J. Jang and K. Y. Suh, Lab Chip, 2010, 10, 36–42.
16. K. W. Kwon, S. S. Choi, S. H. Lee, B. Kim, S. N. Lee, M. C. Park, P. Kim, S. Y. Hwang and K. Y. Suh, Lab Chip, 2007, 7, 1461–1468.
17. Y. J. Lee, I. K. Song, K. J. Jang, J. Nielsen, J. Frokiaer, S. Nielsen and T. H. Kwon, Am. J. Physiol.: Renal Physiol., 2007, 292, F340–F350.
18. Y. Duan, N. Gotoh, Q. S. Yan, Z. P. Du, A. M. Weinstein, T. Wang and S. Weinbaum, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 11418–11423.
19. Z. Q. Cai, J. D. Xin, D. M. Pollock and J. S. Pollock, Am. J. Physiol.: Renal Physiol., 2000, 279, F270–F274.
20. M. E. Phillips and A. Taylor, J. Physiol. (London), 1989, 411, 529–544.
21. M. E. Phillips and A. Taylor, J. Physiol. (London), 1992, 456, 591–608.
22. D. R. Dibona, Am. J. Physiol., 1983, 245, C297–C307.
23. W. A. Kachadorian, S. J. Ellis and J. Muller, Am. J. Physiol., 1979, 236, F14–F20.
24. J. Muller and W. A. Kachadorian, Am. J. Physiol., 1984, 247, C90–C98.
25. M. Pearl and A. Taylor, Am. J. Physiol., 1983, 245, C28–C39.
26. J. B. Wade and W. A. Kachadorian, Am. J. Physiol., 1988, 255, C526–C530.
27. R. M. Hays, J. Condeelis, Y. Gao, H. Simon, G. H. Ding and N. Franki, Pediatr. Nephrol., 1993, 7, 672–679.
28. H. Simon, Y. Gao, N. Franki and R. M. Hays, Am. J. Physiol., 1993, 265, C757–C762.
29. G. Valenti, G. Procino, M. Carmosino, A. Frigeri, R. Mannucci, I. Nicoletti and M. Svelto, J. Cell Sci., 2000, 113, 1985–1992.
30. G. Tamma, E. Klussmann, G. Procino, M. Svelto, W. Rosenthal and G. Valenti, J. Cell Sci., 2003, 116, 1519–1525.
31. S. Li, B. P. C. Chen, N. Azuma, Y. L. Hu, S. Z. Wu, B. E. Sumpio, J. Y. J. Shyy and S. Chien, J. Clin. Invest., 1999, 103, 1141–1150.
32. E. Tzima, M. A. del Pozo, S. J. Shattil, S. Chien and M. A. Schwartz, EMBO J., 2001, 20, 4639–4647.
33. E. Tzima, Circ. Res., 2006, 98, 176–185.
34. H. A. Praetorius and K. R. Spring, J. Membr. Biol., 2003, 191, 69–76.
35. H. A. Praetorius and K. R. Spring, Annu. Rev. Physiol., 2005, 67, 515–529.
36. K. Maric, A. Oksche and W. Rosenthal, Am. J. Physiol., 1998, 275, F796–F801.

---