Cholangiocyte cilia express TRPV4 and detect changes in luminal tonicity inducing bicarbonate secretion

Abstract

Cholangiocytes, epithelial cells lining the biliary tree, have primary cilia extending from their apical membrane into the ductal lumen. Although important in disease, cilia also play a vital role in normal cellular functions. We reported that cholangiocyte cilia are sensory organelles responding to mechanical stimuli (i.e., luminal fluid flow) by alterations in intracellular Ca²⁺ and cAMP. Because cholangiocyte cilia are also ideally positioned to detect changes in composition and tonicity of bile, we hypothesized that cilia also function as osmosensors. TRPV4, a Ca²⁺-permeable ion channel, has been implicated in signal transduction of osmotic stimuli. Using purified rat cholangiocytes and perfused intrahepatic bile duct units (IBDUs), we found that TRPV4 is expressed on cholangiocyte cilia, and that hypotonicity induces an increase in intracellular Ca²⁺ in a TRPV4-, ciliary-, and extracellular calcium-dependent manner. The osmosensation of luminal tonicity by ciliary TRPV4 induces bicarbonate secretion, the main determinant of ductal bile formation, by a mechanism involving apical ATP release. Furthermore, the activation of TRPV4 in vivo, by its specific agonist, 4αPDD, induces an increase in bile flow as well as ATP release and bicarbonate secretion. Our results suggest that cholangiocyte primary cilia play an important role in ductal bile formation by acting as osmosensors.

Bile initially secreted by hepatocytes is modified by cholangiocytes (epithelial cells lining the bile ducts) via absorptive and secretory processes (1). Although cholangiocytes account for only 3–5% of the liver cell population, they generate up to 40% of total bile volume (1, 2). Hepatocytes secrete osmotically active solutes, primarily bile salts, and glutathione, establishing a gradient that generates water movement via aquaporin water channels (1–4). Primary bile then flows through the lumen of intrahepatic bile ducts where cholangiocytes modify it by absorption of solutes, mainly bile salts and glucose, and secretion of ions like Cl⁻ and HCO₃⁻. Ultimately, these processes lead to water secretion (i.e., ductal bile production) by cholangiocytes, also involving aquaporin water channels (2, 5).

Cholangiocytes, but not hepatocytes, express primary cilia, which are microtubule-based organelles (6), extending from their apical plasma membrane into the ductal lumen. Thus, cholangiocyte cilia are ideally positioned to detect changes in the flow, composition, and tonicity of luminal bile (7). Until recently, primary cilia were considered vestigial organelles with no physiologically relevant functions (8). Triggered by observations that mutations in ciliary-associated proteins cause disease (e.g., situs inversus, hydrocephalus, obesity, and polycystic kidney and liver disease), primary cilia have become the subject of intense investigation (9–11).

Primary cilia are also important to normal cell function. In mammals, primary cilia detect different stimuli, including light (by photoreceptors) and urine flow (by primary cilia on renal tubular epithelia) (9, 12). We reported that primary cilia on cholangiocytes act as mechanosensors, responding to luminal fluid flow by alterations in intracellular Ca²⁺ and cAMP. The ciliary proteins involved in this transduction of mechanical stimuli included polycystin-1 (PC1), a cell surface receptor; polycystin-2 (PC2), a Ca²⁺ channel; and adenylyl cyclase isoform 6 (7). Given that cholangiocyte cilia are exposed to various osmotic stresses induced by the movement of osmotically active molecules into and out of bile flowing through the lumen of intrahepatic bile ducts, we hypothesized that these organelles might also function as osmosensors.

TRPV4 is a member of the transient receptor potential (TRP) superfamily of cation channels that function as integrators of physical and chemical stimuli. This Ca²⁺-permeable channel functions as an osmosensor, being activated by anisosmotic conditions (13–20). In Trpv4-transfected CHO cells, TRPV4 participates in regulatory volume decrease (RVD), sensing a hypotonic stimulus and mediating influx of extracellular calcium that activates signaling cascades resulting in secretion of K⁺ and Cl⁻, followed by loss of intracellular water (13, 21). The worm, Caenorhabditis elegans, expresses Osm-9, a gene of the TRP family and an ortholog of Trpv4, the protein product of which is found in its ciliated sensory neurons. Worms with a mutation in Osm-9 have defects in their response to osmotic, mechanical, and odorant stimuli, defects that are rescued by neuron-directed transgene expression of TRPV4 (22). In mammals, TRPV4 is expressed in multiple organs (e.g., kidney, trachea, lung, and brain) in cells exposed to changes in tonicity of fluid (14, 23, 24).

In the work described here using rat and mouse cholangiocytes, perfused intrahepatic bile duct units (IBDUs), and in vivo rat models, we show that TRPV4 is expressed on cholangiocyte cilia, and that hypotonicity induces an increase in intracellular Ca²⁺ in a TRPV4-, ciliary-, and extracellular calcium- dependent manner. The osmosensation of luminal tonicity detected by ciliary TRPV4 induces bicarbonate secretion, the main determinant of ductal bile formation, through a mechanism involving ATP release into the ductal lumen. Our results provide evidence of TRPV4 expression in primary cilia and suggest that cholangiocyte primary cilia play an important role in the regulation of ductal bile formation by acting as osmosensors.

Results

Cholangiocytes Express TRPV4 Message and Protein.

We performed RT-PCR on total RNA from isolated mouse and rat cholangiocytes (25) and cultured mouse (NMC) and rat (NRC) cholangiocytes (26) with cDNA from kidney as positive control. As shown in Fig. 1A, both mouse and rat cholangiocytes express Trpv4 mRNA. We sequenced the PCR products and verified the identities of the amplicons by database homology searches (BLAST; National Center for Biotechnology Information, National Institutes of Health).

Fig. 1.

TRPV4 message and protein are expressed in cholangiocytes. (A) RT-PCR for Trpv4 on normal mouse cholangiocytes cell line (NMC), freshly isolated mouse cholangiocytes (FIMC), isolated rat bile ducts, normal rat cholangiocytes in primary culture (NRC), and NMC-overexpressing TRPV4 (NMC-Trpv4). Kidney RNA was used as positive control. (B) Thirty micrograms of total homogenate of rat bile ducts and NRCs was loaded onto 7.5% SDS–polyacrylamide gels. Kidney and CHO cells were used as positive and negative controls, respectively. (C) Ten micrograms of NMC, NMC-Trpv4, and NMC-Trpv4-shRNA total homogenates was immunoblotted with anti-TRPV4. Anti-actin immunoblots were used as loading controls. Anti-TRPV4 immunoblot of NMC-Trpv4 homogenates before (−) or after (+) digestion with peptide/N-glycosidase (PNGaseF).

As shown in Fig. 1B, immunoblotting for TRPV4 detected a double band between 75 and 100 kDa in rat kidney and IBDUs, a pattern similar to that reported in mouse kidney (27); only the high molecular-weight isoform was seen in NRCs. Bands were absent when TRPV4 antibody was preabsorbed with a molar excess of immunizing peptide. To further evaluate the identity of the TRPV4 band, we obtained total protein from NMCs transfected to overexpress TRPV4 (NMC-Trpv4) and NMC-Trpv4 cotransfected with a shRNA against TRPV4 (NMC-Trpv4-shRNA) cell lines. Western blots showed a similar double band pattern in all samples (Fig. 1C); as expected, protein expression was higher in the overexpressing NMC-Trpv4 cell line than in NMC and NMC-Trpv4-shRNA. After incubation with N-glycosidase, the highest molecular-mass band shifted to the 75-kDa band (Fig. 1C), indicating the presence of N-glycosylation in the TRPV4 molecule, as reported (28), and further confirming antibody specificity.

TRPV4 Protein Is Expressed on Cholangiocyte Primary Cilia.

We performed immunofluorescent confocal microscopy on NMC and NMC-Trpv4 cultured for 7 days after confluence, a manipulation that results in the development of mature cilia on cholangiocytes in culture (29). Using antibodies to TRPV4 and to acetylated α-tubulin, a known component of the ciliary axoneme, we found that TRPV4 is expressed in cholangiocyte cilia (Fig. 2A). Immunofluorescent confocal microscopy on rat liver tissue sections confirmed the ciliary localization of TRPV4 in vivo (Fig. 2B). The presence of TRPV4 in cilia was also confirmed by immunogold electron microscopy on IBDUs (Fig. 2C). Finally, a highly enriched fraction of cilia isolated from NRCs (29) stained positively for TRPV4 (Fig. 2D).

Fig. 2.

TRPV4 is expressed on cholangiocyte cilia. (A) Confocal microscopy on normal and Trpv4-transfected NMCs. Cells were stained with anti-acetylated α-tubulin (α -tub) in red and anti-TRPV4 in green. (B) Normal rat liver tissue showing a longitudinal cut of an intrahepatic bile duct and the expression of TRPV4 on cholangiocyte cilia (yellow). (C) Immunogold electron microscopy indicates the presence of TRPV4 on cholangiocyte cilia in IBDUs isolated from rat liver (arrows). (D) Confocal immunofluorescence showing rat cholangiocytes “Peel off” isolated cilia costained with anti-acetylated α-tubulin (red) and anti-TRPV4 (green).

Extracellular Hypotonicity Induces a Ca²⁺ Response in Cholangiocytes in a TRPV4- and Extracellular Ca²⁺-Dependent Manner.

It was previously reported that Mz-Cha-1 cells, derived from a human cholangiocarcinoma, respond to hypotonicity by an increase in [Ca²⁺]_i (30, 31). To test the possibility that TRPV4 is involved in this response, we exposed NMC, NMC-TrpV4, and NMC-Trpv4-shRNA cells in suspension and preloaded with fura-2 to a hypotonic buffer (200 mOsm/liter) and measured [Ca²⁺]_i. NMCs maintain a stable [Ca²⁺]_i when in isotonic buffer but show a 2-fold increase in [Ca²⁺]_i when in hypotonic buffer [supporting information (SI) Fig. 8A, P < 0.05, n = 3]. TRPV4-overexpressing cells show a 6-fold increase in [Ca²⁺]_i when exposed to hypotonicity, a response that was impaired when NMC-Trpv4-shRNA cells were used (SI Fig. 8A, P < 0.05, n = 3). The hypotonicity-induced increases in [Ca²⁺]_i depended on the presence of Ca²⁺ in the extracellular media (SI Fig. 8B). As a positive control, we measured the effect of the TRPV4 agonist, 4αPDD, on intracellular Ca²⁺, and we found an increase in [Ca²⁺]_i that was dose-dependent (SI Fig. 8C). These data suggest that cholangiocyte TRPV4 is involved in a hypotonicity-induced Ca²⁺ influx from the extracellular media.

Luminal Hypotonicity Induces TRPV4- and Ciliary-Dependent Changes in [Ca²⁺]_i Levels in Cholangiocytes of Microperfused IBDUs.

To test the hypothesis that ciliary TRPV4 could sense luminal changes in osmolarity, we used a previously developed physiologically relevant experimental model: microperfused IBDUs prepared from rat liver by microdissection (7, 32). To avoid a potential confounding mechanical response by cholangiocyte cilia, luminal perfusion was kept at a rate that would not stimulate a mechanical ciliary response [<20 nl/min (7)]. Because we had previously observed that osmotically induced water movement into or out of the luminal space of IBDUs occurs very rapidly depending on the osmotic gradient between the lumen and the bathing buffer (32), we induced intraluminal hypotonicity by bathing the exterior of IBDUs in hypotonic KRB (100 mOsm/liter). This maneuver resulted in a rapid decrease in fluorescence of a nonabsorbable fluorescent marker in the luminal perfusate, consistent with a drop in luminal osmolarity (Fig. 3A) and validating this experimental approach to modify the luminal tonicity of IBDUs. When cholangiocytes of IBDUs were preloaded with a Ca²⁺-sensitive fluorescent dye (fluo-4), we found that a decrease in intraluminal tonicity caused an increase in dye fluorescence (i.e., an increase in [Ca²⁺]_i) (Fig. 3 A and B). The hypotonicity-induced increase in [Ca²⁺]_i levels was 45 ± 9.9% compared with basal isotonic conditions. To address whether cholangiocyte cilia were involved in this hypotonicity-induced [Ca²⁺]_i signaling response, we deciliated IBDUs with chloral hydrate (ClHy), a technique described and validated by us and others (7, 9, 33). Functional studies showed that in IBDUs treated with ClHy, the hypotonicity-induced rise in [Ca²⁺]_i was reduced by 93.8 ± 10.6% (Fig. 3D), suggesting that cholangiocyte cilia are involved in the transmission of hypotonic stimuli into an [Ca²⁺]_i response.

Fig. 3.

Luminal hypotonicity induces an [Ca²⁺]_i increase in a TRPV4- and cilia-dependent manner. (A) Representative traces of IBDUs microperfused with KRB containing 1 mmol/liter of FS (gray) or IBDUs loaded with fluo-4 (black). The tonicity of the bath perfusate was changed and fluorescence intensity followed over the time. (B) Fluorescence tracing of fluo-4 loaded IBDUs perfused with isotonic mannitol-KRB buffer (gray) or hypotonic KRB buffer (black) over time. (C) Trpv4 qRT-PCR normalized to 18s RNA and Western blot analysis on IBDUs incubated overnight with scrambled or Trpv4-siRNA. (D) Fluo-4 fluorescence changes induced by hypotonicity on control, Trpv4-siRNA and ClHy treated IBDUs. (*, P < 0.01, n = 6.)

To explore whether TRPV4 is involved in the hypotonicity-induced [Ca²⁺]_i response, we inhibited endogenous Trpv4 in IBDUs by a specific siRNA. As shown in Fig. 3C, both Trpv4 mRNA and protein decreased by >70% with this reagent. The hypotonicity-induced increase in [Ca²⁺]_i of IBDUs treated with a scrambled siRNA was not different from that in normal IBDUs. In IBDUs treated with Trpv4-siRNA, the [Ca²⁺]_i response was reduced by 90.1 ± 4.7% (Fig. 3D), supporting TRPV4 involvement in the hypotonicity-induced increase in [Ca²⁺]_i level.

Hypotonicity Induces TRPV4- and Ciliary-Dependent Luminal Alkalinization and ATP Release.

To determine a possible physiological role of luminal tonicity sensation by ciliary TRPV4, we measured changes in luminal pH as a reflection of bicarbonate secretion using the cell-impermeable pH-sensitive dye, BCECF-dextran (32, 34, 35). As shown in Fig. 4, hypotonicity, as expected, induced a decrease in the luminal isosbestic fluorescence intensity because of water movement into the ductal lumen. As a consequence of this stimulus, control IBDUs showed an increase in luminal pH of 0.23 ± 0.04 unit/min (Fig. 4 B and C). Treatment of IBDUs with the Trpv4-siRNA or with ClHy decreased the hypotonicity-induced luminal pH increase by 71 ± 17% and 89 ± 3%, respectively. These experiments demonstrate that luminal hypotonicity sensed by ciliary TRPV4 contributes to regulation of bicarbonate secretion by cholangiocytes.

Fig. 4.

Luminal tonicity changes in IBDUs induce cholangiocyte bicarbonate secretion in a TRPV4- and cilia-dependent manner. Changes in ductular pH (ΔpH) induced by hypotonicity were examined by using luminal perfusion with BCECF-dextran. (A) When IBDUs were exposed to hypotonicity, water moved into the ductal lumen inducing a decrease in the isosbestic fluorescence of the dye. (B) Control IBDUs (●) showed a net increase in luminal pH over time that was prevented in the siRNA- (■) or ClHy- (▾) treated IBDUs. (C) Slopes of the first portion of the curves after the addition of hypotonic buffer were calculated and expressed as ΔpH units/min. (*, P < 0.05; #, P < 0.01 compared with control group, n = 4.)

To address mechanisms linking the hypotonicity-induced activation of ciliary TRPV4 with bicarbonate secretion, we tested the hypothesis that ATP release mediates this phenomenon. As shown in Fig. 5A, hypotonicity induced a 5-fold increase in ATP release in NMCs compared with isotonic conditions. This result is consistent with data previously reported on Mz-ChA-1 cells (36). Cells overexpressing TRPV4 showed a 7-fold increase in ATP release when exposed to hypotonicity, an effect that was reduced by 70% when cells were cotransfected with a Trpv4-shRNA (Fig. 5A). TRPV4 activation by 4αPDD induced a 3-fold increase in extracellular ATP (Fig. 5B). Furthermore, the hypotonicity-induced increase in luminal pH was prevented when IBDUs were perfused with suramin, a purinergic receptor antagonist (Fig. 5C). Both the hypotonicity- and 4aPDD-induced ATP release depended on the presence of cilia (Fig. 6 A, B, and D) and on the presence of extracellular calcium (Fig. 6E). To exclude a possible direct inhibition of TRPV4 by ClHy, we measured the intracellular calcium response induced by 4αPDD on TRPV4-overexpressing nonciliated cells preincubated with ClHy; we found no significant differences between cells treated with ClHy and those exposed to vehicle (Fig. 6C). Taken together, these data support the notion that the hypotonicity-induced activation of ciliary TRPV4 generates apical release of ATP that in turn induce luminal bicarbonate secretion.

Fig. 5.

Hypotonicity induces ATP release in a TRPV4-dependent manner, and P2Y receptors mediate the hypotonicity-induced bicarbonate secretion. (A) ATP release by NMC, NMC-Trpv4, and NMC-Trpv4-shRNA after exposure to isotonic or hypotonic conditions. (*, P < 0.001 compared with isotonic conditions; **, P < 0.001 and P < 0.05 compared with isotonic conditions and NMC Hypo, respectively; *#, P < 0.001 compared with isotonic conditions and NMC Hypo; n = 10.) (B) ATP release induced by 4αPDD. (*, P < 0.01, n = 6.) (C) Changes in ductular pH induced by hypotonicity were examined by using luminal perfusion with BCECF–dextran in control and suramin-treated IBDUs show the previously undescribed observation. (#, P < 0.001, n = 6.)

Fig. 6.

Hypotonicity-induced ATP release depends on cilia integrity and extracellular calcium. (A) SEM images from control NMCs cultured 7 days after confluence showed normal cilia (Upper) but, when treated 24 h with ClHy cilia, were absent (Lower). (B) Hypotonicity-induced ATP release on control and ClHy-treated NMC and NMC-Trpv4 cell lines. (C) 4αPDD-induced changes in fluorescence of fluo-4 loaded TRPV4-overexpressing nonciliated cells ± ClHy. (*, P < 0.01, n = 12.) (D) 4αPDD-induced ATP release on control and ClHy-treated NMC-Trpv4 cell line. (*, P < 0.01, n = 6.) (E) Hypotonicity-induced ATP release in NMCs exposed to media in which calcium was omitted. (*, P < 0.001 compared with controls, n = 10.)

In Vivo Activation of TRPV4 Induces an Increase in Bile Flow, Bicarbonate Secretion, and ATP Release.

To analyze the potential contribution of TRPV4 to bile flow in vivo, rats were given a retrograde intrabiliary infusion of 0.2 ml of saline solution containing 30 μM 4αPDD or its vehicle, DMSO. This experimental method has been validated (34). Exposure of the biliary tract to this TRPV4 agonist for 10 min led to an increase in bile flow compared with animals retroinjected with the vehicle (Fig. 7A). The increase in bile flow was reduced by 84% when a TRP channel inhibitor, Gd³, was coadministered with 4αPDD (Fig. 7B). The biliary excretion of bicarbonate and ATP was also significantly increased by retrograde intrabiliary administration of 4αPDD (Fig. 7 C and D), an effect that was also prevented by Gd³⁺. Taken together, these in vivo data are consistent with our in vitro results and further support the hypothesis that TRPV4 activation induces ATP release and bicarbonate secretion, generating an increase in bile flow.

Fig. 7.

In vivo activation of TRPV4 induced an increased bile flow and bicarbonate and ATP secretion. (A) Effect of intrabiliary administration of 30 μM 4αPDD or its vehicle DMSO on bile flow. (B) Effect of 100 μM Gd³⁺ on the 4αPDD-induced increase in bile flow. (C and D) Biliary bicarbonate and ATP excretion on control vs. 4αPDD-treated rats. (*, P < 0.01, n = 4.)

Discussion

The key findings reported here relate to an osmosensory function of cholangiocyte primary cilia. The data suggest that, in addition to being mechanosensors, primary cilia on cholangiocytes, and perhaps on other epithelia, are osmosensors, and that TRPV4 is a key protein in this osmosensory process. Moreover, the TRPV4-dependent detection of hypotonicity by cholangiocyte primary cilia triggers an increase in [Ca²⁺]_i, a key second messenger in cholangiocytes, and induces ATP release. This chain of events culminates in the secretion of bicarbonate, the ion that accounts for alkalinization and generation of ductal bile. Our results suggest the existence of a new ciliary-dependent regulatory pathway for ductal bile secretion.

In addition to the ability of primary cilia in the kidney and biliary tract to act as mechanosensors (7, 37), primary cilia may function as chemosensors, as evidenced by the fact that receptor proteins are often expressed on cilia [e.g., somastostatin receptor 3 and 5-HT6 serotonin receptor on brain neuronal cilia and the progesterone receptor on motile cilia of mouse oviduct epithelial cells (38, 39)]. However, to our knowledge, our data show the previously undescribed observation that an osmosensor, TRPV4, is expressed on the primary cilia of an epithelial cell. Importantly, the osmosensory ability of cholangiocyte cilia via TRPV4 has downstream consequences (i.e., alterations in [Ca²⁺]_i, ATP release, and HCO₃⁻ secretion) that could profoundly influence the key physiologic role of cholangiocytes, that is, the production of ductal bile.

The cellular topography of TRPV4 is variable and tissue-specific, reported to be on both the apical and basolateral membranes of renal tubule epithelia (40–43), on the basolateral membrane of the mouse mammary cell line, HC11 (23), and on the apical domain of salivary epithelial cells (44). Importantly, TRPV4 is also expressed on motile cilia of cells in the female reproductive tract (24), and preliminary data suggest the localization of TRPV4 on primary cilia of kidney epithelia (45, 46). The TRPV4 ortholog in C. elegans, Osm-9, is located on cilia of the sensory neurons of the worm (47). Because luminal tonicity in our studies was changed by modification of the bath perfusate, activation of TRPV4 on the basolateral domain cannot be excluded. However, this scenario seems unlikely, because the hypotonicity-induced calcium response was reduced in IBDUs by deciliation, and cilia are expressed only on the apical domain of cholangiocytes. Furthermore, the TRPV4 agonist, 4αPDD, induces an intracellular Ca²⁺ response only when it is perfused into the lumen of IBDUs, but not when it is applied to the basolateral side (data not shown).

The dependence of TRPV4 activation on extracellular Ca²⁺ is not surprising, because the intracellular Ca²⁺ response due to a mechanical stimulus, in both kidney epithelia and cholangiocytes, and the odorant receptors in olfactory epithelial, all depend on extracellular Ca²⁺ (7, 48, 49).

Studies carried out on Trpv4-null mice show that TRPV4 is necessary for the maintenance of osmotic equilibrium, and it is conceivable that TRPV4 acts as an osmotic sensor in the central nervous system (16). TRPV4 also plays a role in maintenance of cellular osmotic homeostasis (50). Here we show the function of TRPV4 as osmosensor in a different context, i.e., not in the regulation of global body osmolarity but in the specific sensation of bile tonicity, leading to the secretion of ions such as bicarbonate in response to a decrease in the tonicity of bile. Although activation of TRPV4 has been reported to induce an increase in paracellular permeability for small solutes in an epithelial cell line, this effect occurs over hours and, thus, is not likely relevant to our findings, which occur much more rapidly (23). The mechanism that links the hypotonicity-induced TRPV4-mediated [Ca²⁺]_i increase with the stimulation of bicarbonate secretion appears to involve ciliary- and TRPV4-dependent hypotonicity-induced ATP release. ATP is present in bile in physiological concentrations (51). Moreover, cholangiocytes express a number of P2Y and P2X nucleotide receptors on their apical domain, and their stimulation by ATP induces apical Cl⁻ secretion and provokes alkalinization of bile via Cl⁻/HCO₃⁻ exchange activity mediated by AE2 (35, 36, 52–54). Interestingly, an intracellular P2X receptor required for osmoregulation in Dictyostelium discoideum was recently reported (55). Based on this information, we speculated that ATP released by the hypotonic activation of ciliary TRPV4 would activate purinergic receptors on the apical domain of cholangiocytes, which might then influence cholangiocyte transport processes and ultimately ductal bile secretion. Our data support this chain of events.

The physiological implications of an osmosensory function of cholangiocyte cilia are unclear. Bile is considered isotonic, but no technology permits measurements of variations in bile tonicity during its long passage through the ductal lumen, from the hepatocyte canaliculus to the common bile duct; indeed, measurements of bile tonicity can be made only on bile collected at the most distal end of the biliary tree (i.e., the common bile duct). It is known that cholangiocytes can absorb glucose and bile salts from primary bile generated by hepatocytes (1, 32). Thus, it seems plausible that these absorptive processes can generate transient decreases in ductal bile tonicity that could be sensed by cholangiocyte cilia. As shown by our in vivo and in vitro results, the activation of ciliary TRPV4, in turn, induces a bicarbonate secretory response, mediated by the apical release of ATP and purinergic receptor activation. This secretion would restore the tonicity of ductal bile and also generate water secretion as indicated by the increased bile flow.

In summary, our study shows that primary cilia on cholangiocytes may function as osmosensors, and that ciliary osmosensation, mediated by a mechanism that involves an increase in [Ca²⁺]_i and apical ATP release, induces luminal bicarbonate secretion. The physiological, pathophysiological, and therapeutic implications of this phenomenon remain to be fully elucidated.

Materials and Methods

Animals.

Male Sprague–Dawley rats (225–250 g) were maintained on a standard diet. In vivo experiments using a validated retrograde intrabiliary infusion model were performed as described (34). All experimental procedures were approved by the Animal Use and Care Committee of the Mayo Foundation.

Solutions.

The composition of Ringer-HCO₃⁻ buffer (KRB) was as follows: 120.0 mmol/liter of NaCl, 5.9 mmol/liter of KCl, 1.2 mmol/liter of Na₂HPO₄, 1.0 mmol/liter of MgSO₄, 25.0 mmol/liter of NaHCO₃, 1.25 mmol/liter of CaCl₂, and 5.0 mmol/liter of d-glucose, pH 7.4. For tonicity-change studies, isotonic KRB was prepared with 40 mmol/liter of NaCl and completed to isotonicity with mannitol. For hypotonic KRB, addition of mannitol was omitted. In calcium-free KRB, CaCl₂ was removed.

Cell Culture.

NMCs were grown in Eagle's modified essential medium supplemented with 10% (vol/vol) FBS and penicillin (100 units/ml)/streptomycin (100 μg/ml). NRCs were grown as described (56).

Isolated IBDUs.

IBDUs were isolated from normal rat liver as described (32).

RT-PCR.

Specific oligonucleotides were synthesized based on the published sequence for rat Trpv4 5′-TACCACGGTGGACTACCTGAG-3′ (forward) and 5′-CATGATGCTGTAGGTCCCTGT-3′ (reverse).

Trpv4 Cloning, shRNA Constructs, and Stable Transfection.

Trpv4 was cloned from NMC total RNA. RT-PCR amplification was achieved with Trpv4 specific primers (5′-CCAAGCTTGCCACCATGGCAGATCCTGGTGATGGT-3′ and 5′-CCGGATCCAGTGGG G C A TCGTCCGT-3′), and cloned in frame into the pcDNA6/myc-His mammalian expression vector. After selection, cells were cotransfected with the Trpv4-shRNA expressing vector. Top- and bottom-strand oligonucleotides corresponding to the nucleotide sequence of the target site (AACATGAAGGTCTGTGACGAG) were cloned into pSilencer3.1-H1 puro shRNA vector.

Trpv4 Silencing by siRNA and Chemical Deciliation on IBDUs.

Isolated IBDUs were cultured for 24 h with 20 nmol/liter of functional or scrambled Trpv4-siRNA (target sequence, 5′-AAGAACTCAGGCACAGATGAA-3′) or with 4 mmol/liter of ClHy (57).

Measurement of [Ca²⁺]_i Level in Suspended Cholangiocytes.

Cells (1.0 × 10⁶) were washed and loaded with 3 μM Fura-2/AM for 30 min at 37°C. After washing, cells were resuspended in isotonic or hypotonic KRB and fluorescence-quantified. Excitation was performed at both 340 and 380 nm, with emission determined at 520 nm. [Ca²⁺]_i was calculated from the ratios of these values as described (58). Alternatively, cells were seeded on 96-well plates, loaded with fluo-4, and, after washing, stimulated with KRB containing 4αPDD or its vehicle DMSO. In related experiments, cells were preincubated with ClHy before Fluo-4 loading.

Microperfusion of IBDUs.

Individual isolated IBDUs were mounted in a custom-designed temperature-controlled chamber on the stage of an inverted fluorescence microscope, as described (32). The lumen was perfused at very low rate (20 nl/min). The external surface of the perfused IBDU was simultaneously bathed with isotonic or hypotonic KRB.

Measurement of Water Movement into the Lumen of IBDUs.

The lumen of IBDUs was perfused with KRB containing 1 mmol/liter of fluorescein-5-(6)-sulfonic acid, trisodium salt (FS), a fluorescent volume marker. To decrease luminal tonicity, water movement into the lumen was induced by changing the tonicity of the bath solution and measured as described (32).

Measurement of [Ca²⁺]_i Level in Cholangiocytes of Microperfused IBDUs.

Changes in [Ca²⁺]_i levels of cholangiocytes of KRB-perfused IBDUs were detected by using the cell-permeant Ca²⁺-sensitive fluorescent probe Fluo-4 AM, as we described in detail (35).

Measurement of pH in the IBDU Lumen.

The luminal pH of perfused IBDUs was measured by using the cell-impermeant pH-sensitive dye, BCECF-dextran, as described (32). IBDUs were perfused with HCO₃⁻-free buffer containing BCECF-dextran (pH 7.2) and simultaneously bathed with isotonic or hypotonic KRB. After excitation at 495 nm, luminal pH was measured as the ratio of emission intensities at 530 nm (pH- and concentration-sensitive) and 440 nm (only concentration-sensitive), respectively. In situ calibration curves were generated at the end of each experiment to convert fluorescence-excitation ratios (F495/F440) to pH values (32, 35).

ATP Release.

NMC, NMC-trpv4, and NMC-Trpv4-shRNA were cultured 7 days after confluence in 24-well plates. After washing three times with KRB, cells were incubated for 2 min in 200 μl of isotonic KRB. Two hundred microliters of isotonic KRB or water was added and incubated for 4 min. The viability of the cells was not affected. Alternatively, cells were incubated with KRB containing 4αPDD or its vehicle DMSO. Media were then collected, centrifuged for 10 min × 500 × g, and passed to clean tubes. ATP concentration was measured with a luciferin–luciferase bioluminescence assay kit.

Immunoblot Analysis of TRPV4.

Protein fractions were subjected to 7.5% SDS–polyacrylamide gel and transferred to nitrocellulose membranes. After blocking, blots were incubated overnight at 4°C with rabbit affinity-purified polyclonal TRPV4 antibody (1:5,000, Alomone Labs), washed, and incubated 1 h at room temperature with HRP-conjugated goat anti-rabbit secondary antibody (1:5,000 dilution). For protein detection, ECL system was used. Selected samples were digested with 10 μg/ml peptide/N-glycosidase F.

Immunofluorescence Confocal Microscopy.

Samples were incubated with antibodies against acetylated α-tubulin (1:500, Sigma–Aldrich) and TRPV4 (1:100) overnight at 4°C followed by incubation for 1 h with fluorescent secondary antibodies (1:100). Nuclei were stained with DAPI.

Immunogold and Scanning Electron Microscopy.

Immunogold electron microscopy of IBDUs and SEM of NMCs were performed as described (29, 59).

Cilia Isolation.

Cilia were isolated from cultured cells by using the peeloff technique as described (29).