Cholangiocyte primary cilia are chemosensory organelles that detect biliary nucleotides via P2Y₁₂ purinergic receptors

Animals and models.

Male Sprague-Dawley rats (225–250 g) were obtained from Harlan Sprague Dawley (Indianapolis, IN), housed in a temperature-controlled room (22°C) with 12-h light/dark cycles, and maintained on a standard diet with free access to water. All experimental procedures were approved by the Animal Use and Care Committee of the Mayo Foundation. Rats were anesthetized with pentobarbital sodium (50 mg/kg body wt ip) for in vivo procedures. Livers were harvested, fixed in 10% formaldehyde, and embedded in paraffin. IBDs ranging in luminal diameter from 100 to 125 μm and in length from 1.0 to 1.5 mm were isolated from rat liver and microperfused as previously described (26, 28). Normal rat cholangiocytes (NRCs) were isolated from rat liver and cultured as described (3). Cilia were isolated from cultured normal mouse cholangiocytes (NMCs) by the slide-pull technique (21).

Solutions.

The composition of standard Ringer-HCO₃⁻ buffer (KRB) (in mM) was: 120.0 NaCl, 5.9 KCl, 1.2 Na₂HPO₄, 1.0 MgSO₄, 25.0 NaHCO₃, 1.25 CaCl₂, and 5.0 d-glucose, pH 7.4.

RNA isolation and RT-PCR.

Total RNA was obtained from freshly isolated rat IBDs, cholangiocytes, hepatocytes, brain, NRCs, and NMCs with the use of TRIZOL reagent (Invitrogen, Carlsbad, CA). The expression of P2Y₁₂ and P2Y₁₃ were detected by RT-PCR using specific primers. The PCR products were sequenced, and the identities of the amplicons were verified by data base homology searches (BLAST; NCBI, National Institutes of Health).

Western blotting.

Freshly isolated rat cholangiocytes, hepatocytes, brain and NMCs were resuspended in RIPA Sample Buffer (Santa Cruz Biotechnology, Santa Cruz, CA). Samples (10 to 40 μg) of protein extracts were run in 7.5% SDS-PAGE and electrotransferred to a nitrocellulose membrane (Bio-Rad, Hercules, CA). After being blocked with 5% nonfat dried milk in PBS-0.5% Tween, membranes were incubated overnight at 4°C with antibodies against P2Y₁₂ and P2Y₁₃ (Alomone Laboratories, Jerusalem, Israel; dilution 1:200). After three washes with PBS-0.5% Tween, the samples were incubated with peroxidase-conjugated secondary antibodies (dilution 1:5,000) for 45 min at room temperature. Bands were visualized with ECL Plus Western Blotting Detection Kit (BD Biosciences, Franklin Lakes, NJ).

Immunogold scanning electron microscopy.

Isolated IBDs were immersed in 4% phosphate-buffered glutaraldehyde (pH 7.4) for 1 h, treated with 0.1% Triton X-100 for 5 min, and then rinsed with phosphate buffer three times. The samples were incubated overnight at 4°C with anti-P2Y₁₂ (Alomone, dilution 1:100) and then incubated for 2 h at room temperature with anti-rabbit IgG conjugated with 1.4 nm of nanogold. The samples were fixed in 1% glutaraldehyde for 10 min, and gold enhancement was performed with a gold enhancement kit (Nanoprobes, Yaphank, NY). The samples were dehydrated, critical-point dried, and carbon coated. Images were generated at 8 kV by a Hitachi S-4700 microscope (Hitachi, Pleasanton, CA).

Immunofluorescence confocal microscopy.

Immunofluorescence microscopy was performed with a confocal microscope Zeiss LSM 510 (Carl Zeiss, Thornwood, NY) with a 100× Plan-Apochromat 1.4 NA oil objective. Isolated split-open IBDs, and rat liver samples were incubated with two antibodies, one against a ciliary marker, acetylated α-tubulin (dilution 1:200, Sigma), and the other against protein of interest, i.e., P2Y₁₂, P2Y₁₃ (Alomone, dilution 1:100), AC4, AC5, AC6, AC7 (FabGennix International, Frisco, TX; dilution 1:100), AC8 (Santa Cruz Biotechnology, dilution 1:100), PKA RI-α, PKA RII-α, and PKA RII-β (Santa Cruz Biotechnology, dilution 1:50), PKA RI-β (Calbiochem, San Diego, CA; dilution 1:50), exchange protein directly activated by cAMP (EPAC)1 and EPAC2 (Santa Cruz Biotechnology, dilution 1:50), followed by incubation for 1 h at room temperature with fluorescent secondary antibodies (Molecular Probes, Eugene, OR; dilution 1:100). Nuclei were stained in blue with 4′,6-diamidino-2-phenylindole (Sigma). Corresponding negative controls (i.e., no primary antibodies) were performed in all experiments. Images were taken with magnification ×100 and afterwards cropped using Adobe Photoshop to show the ciliary location of each protein.

P2Y₁₂ siRNAs gene silencing.

P2Y₁₂ siRNAs were purchased from Ambion (Austin, TX). Isolated IBDs were cultured for 24 h at 37°C with 20 nM of P2Y₁₂ siRNAs or corresponded scrambled siRNAs by a procedure previously described (53). P2Y₁₂ mRNA was detected by real-time quantitative RT-PCR (Light Cycler Apparatus; Roche Molecular Biochemicals, Indianapolis, IN). 18S RNA was determined with the mRNA universal 18S standard (Ambion) and used as a normalizing control. The expression of P2Y₁₂ protein was detected by Western blotting with actin as a loading control.

Measurement of cAMP.

A single IBD was equilibrated for 30 min in the perfusion chamber; then forskolin (FSK, 100 μM) was added into the bathing solution (KRB) containing 0.5 mM IBMX, an inhibitor of phosphodiesterases. Afterwards, an IBD was perfused through its lumen for 15 min with standard KRB (control, no ADP) or with KRB containing 100 μM ADP; during perfusion IBDs remained in the bathing KRB containing FSK and IBMX. After perfusion, an IBD was removed from the microperfusion system, and cAMP levels were measured by using the Bridge-It cAMP designer cAMP assay (Mediomics, St. Louis, MO). Results were expressed in pmol of cAMP per μg DNA. The amount of DNA in a single IBD was determined by using the DNeasy Tissue Kit (Qiagen, Valencia, CA).

NRCs were stimulated for 15 min at 37°C with FSK (100 μM) in the presence of 0.5 mM IBMX and simultaneously treated with ADP (100 μM), ATP-γS (10 μM), or UTP (100 μM). cAMP levels were measured by the Bridge-It cAMP designer cAMP assay (Mediomics) and expressed in pmol of cAMP per well.

Statistical analysis.

All values are expressed as mean ± SE. Statistical analysis was performed by the Student's t-test, and results were considered statistically different at P < 0.05.

Expression of P2Y₁₂ and P2Y₁₃ in rat cholangiocytes and cilia.

RT-PCR and Western blotting show that both P2Y₁₂ and P2Y₁₃ are expressed in rat cholangiocytes (Fig. 1, A and B). P2Y₁₂ and P2Y₁₃ are also expressed in rat hepatocytes and brain, which we used as a positive control (Fig. 1, A and B).

To test the expression of P2Y₁₂ and P2Y₁₃ in cholangiocyte cilia, paraffin sections of rat liver were stained with antibodies to a ciliary marker, acetylated α-tubulin, and to either P2Y₁₂ or P2Y₁₃. Figure 1C shows the IBD with cilia (stained in red) extending from the cholangiocyte apical plasma membrane into the ductal lumen. Images in green and overlay images in yellow suggest that P2Y₁₂ is expressed in the cholangiocyte apical plasma membrane and cilia. In contrast, P2Y₁₃ is not detected in cholangiocyte cilia (data not shown).

Expression of P2Y₁₂ in cholangiocyte cilia was confirmed by immunogold scanning electron microscopy (Fig. 1D) (51). The scanning electron microscopy image of the lumen of isolated split-open IBD (Fig. 1D, a) shows the cholangiocyte apical surface with a single cilium; the image of the same field generated by backscattered electrons identified the position of the gold-labeled secondary antibodies, which are seen as bright white dots localized to the ciliary membrane.

Expression of components of the cAMP signaling cascade in cholangiocyte cilia.

Given that P2Y₁₂ is a G protein-coupled receptor associated with the cAMP signaling pathway, ACs should also be localized to the same cellular compartment, i.e., to the ciliary membrane. Figure 2 shows that at least three ACs (i.e., AC6, AC4, and AC8) are expressed in cilia. AC5 and AC7 are not present in cilia. AC9 was not tested in this study because high quality antibodies against this protein were not available.

Recent developments suggest that cAMP signaling requires tight regulation via a flexible downstream machinery to ensure an appropriate cellular response (30). The components of such machinery are PKA and EPAC, which are recruited to a given subcellular compartment by A-kinase anchoring proteins (AKAPs) (6, 52). According to our recent observation, rat cholangiocytes express all four regulatory subunits of PKA (i.e., PKA RI-α, PKA RI-β, PKA RII-α, and PKA RII-β) and both isoforms of EPAC (i.e., EPAC1 and EPAC2) (4). Dual labeling of rat liver sections with antibodies against acetylated α-tubulin and proteins of interest (i.e., four regulatory subunits of PKA and both EPAC isoforms) shows that PKA (i.e., two regulatory subunits, PKA RII-α and PKA RI-β) and EPAC (isoform 2) are expressed in cholangiocyte cilia (Figs. 3 and 4). AKAP150, which is known to form a functional complex with AC6, PKA, and EPAC (6, 52), is also expressed in cholangiocyte cilia (Fig. 4). All tested components of the cAMP signaling pathway are not exclusively localized to cholangiocyte cilia; they are also localized intracellularly and on the cholangiocyte plasma membrane. However, their expression in the ciliary membrane suggests that cilia possess key elements of the cAMP signaling cascade.

ADP decreases cAMP levels in cholangiocytes in a P2Y₁₂-dependent manner.

P2Y₁₂ is a G protein-coupled receptor linked to ACs via G_i, and its activation by the most potent endogenous agonist, ADP, leads to inhibition of ACs and to a decrease in cAMP levels (1, 56). To determine whether ADP affects cAMP signaling in cholangiocytes via P2Y₁₂ receptors, experiments were performed on microperfused IBDs and on ciliated NRCs in which cAMP levels were increased by FSK (100 μM). In nontreated IBDs and NRCs, the basal levels of cAMP were 0.99 ± 0.26 pmol/μg DNA and 0.10 ± 0.06 pmol/well, respectively. ADP (100 μM) alone did not affect the basal cAMP levels in either microperfused IBDs or in NRCs, whereas FSK increased cAMP concentrations to 9.78 ± 1.35 pmol/μg DNA and to 14.26 ± 0.06 pmol/well, respectively (Fig. 5, A and B). Perfusion of IBDs and incubation of NRCs for 15 min at 37°C with ADP (100 μM) resulted in a decrease of FSK-induced cAMP levels from 9.78 ± 1.35 to 5.08 ± 1.06 pmol/μg DNA (i.e., by 1.9-fold; P < 0.05) and from 14.26 ± 0.06 to 9.30 ± 1.12 pmol/well (i.e., 1.5-fold; P < 0.05), respectively (Fig. 5, A and B). ATP-γS, a nonhydrolyzed analog of ATP, which is considered as another agonist for P2Y₁₂, did not affect basal levels of cAMP but lowered the FSK-stimulated cAMP increase by 1.3-fold (P < 0.05). In contrast, UTP, a known agonist for P2Y₂ and P2Y₄ but not for P2Y₁₂ (1), did not affect either basal cAMP levels or FSK-induced cAMP increase (Fig. 5C), suggesting that P2Y₁₂ is likely involved in effects of ADP and ATP-γS on FSK-induced cAMP levels in rat cholangiocytes.

To directly address the involvement of P2Y₁₂ receptors in the ADP-induced changes in cAMP levels, experiments were performed on NRCs pretreated with a nonspecific inhibitor of P2Y receptors, suramin (50 μM), and on microperfused IBDs in which P2Y₁₂ receptors were silenced by siRNAs. Data in Fig. 6 show that suramin completely abolished the ADP-induced cAMP decrease in NRCs. Likewise, P2Y₁₂ siRNAs, which inhibit the expression of P2Y₁₂ on both mRNA and protein levels by 86 ± 5% and 53 ± 8%, respectively (Fig. 7 A and B), abolished the ADP-induced cAMP decrease in microperfused IBDs (Fig. 7C). Thus the data suggest that ADP affects cAMP signaling in rat cholangiocytes via P2Y₁₂ purinergic receptors.

ADP decreases cAMP levels in cholangiocytes in a ciliary-dependent manner.

To further address whether cholangiocyte cilia are involved in the ADP-induced cAMP signaling response, experiments were performed on IBDs and NRCs after pharmacological removal of cilia by chloral hydrate. Chloral hydrate deciliates different cell types, including cholangiocytes, via an unknown mechanism but do not cause other significant structural and/or functional alterations in the cells (16, 28, 42).

Functional studies revealed that ADP does not affect cAMP levels in deciliated IBDs (i.e., 5.55 ± 1.69 compared with 6.88 ± 2.12 pmol/μg DNA in IBDs stimulated with FSK alone; Fig. 8A) and in deciliated NRCs (i.e., 13.29 ± 1.23 compared with 13.87 ± 0.66 pmol/well in NRCs stimulated with FSK alone; Fig. 8B). Thus cholangiocyte cilia play an important role in the ADP-induced intracellular cAMP signaling response.