Effects of cholesterol on CCK-1 receptors and caveolin-3 proteins recycling in human gallbladder muscle

Abstract

The contraction of gallbladders (GBs) with cholesterol stones is impaired due to high cholesterol concentrations in caveolae compared with GBs with pigment stones. The reduced contraction is caused by a lower cholecystokinin (CCK)-8 binding to CCK-1 receptors (CCK-1R) due to caveolar sequestration of receptors. We aimed to examine the mechanism of cholesterol-induced sequestration of receptors. Muscle cells from human and guinea pig GBs were studied. Antibodies were used to examine CCK-1R, antigens of early and recycling endosomes, and total (CAV-3) and phosphorylated caveolar-3 protein (pCAV-3) by Western blots. Contraction was measured in muscle cells transfected with CAV3 mRNA or clathrin heavy-chain small-interfering RNA (siRNA). CCK-1R returned back to the bulk plasma membrane (PM) 30 min after CCK-8 recycled by endosomes, peaking at 5 min in early endosomes and at 20 min in recycling endosomes. Pretreatment with cholesterol-rich liposomes inhibited the transfer of CCK-1R and of CAV-3 in the endosomes by blocking CAV-3 phosphorylation. 4-Amino-5-(4-chloro-phenyl)-7-(t-butyl)pyrazolo[3,4-d]pyrimidine (inhibitor of tyrosine kinase) reproduced these effects by blocking pCAV-3 formation, increasing CAV-3 and CCK-1R sequestration in the caveolae and impairing CCK-8-induced contraction. CAV-3 siRNA reduced CAV-3 protein expression, decreased CCK-8-induced contraction, and accumulated CCK-1R in the caveolae. Abnormal concentrations of caveolar cholesterol had no effect on met-enkephalin that stimulates a δ-opioid receptor that internalizes through clathrin. We found that impaired muscle contraction in GBs with cholesterol stones is due to high caveolar levels of cholesterol that inhibits pCAV-3 generation. Caveolar cholesterol increases the caveolar sequestration of CAV-3 and CCK-1R caused by their reduced recycling to the PM.

human gallbladers (GBs) with ChS exhibit an impaired contraction induced by cholecystokinin (CCK)-8. GBs with cholesterol stones (ChS) respond with a weak contraction to intravenous infusions of CCK-8 compared with the normal contraction of GBs with pigment stones (PS) (1). The impaired contraction has also been demonstrated in muscle strips and cells from GBs with ChS compared with muscle cells from GBs with PS (2, 5, 30). However, muscle cells from GBs with ChS contract normally in response to CCK-8 after treatment with cholesterol-free liposomes that removes the excess cholesterol from the plasma membrane (PM) (30). Muscle cells from prairie dog GBs develop a weak contraction when fed a high-cholesterol diet or when muscle cells from human GB with PS are treated with cholesterol-rich liposomes that increases the caveolar cholesterol concentration (5, 28). Furthermore, the contraction of other types of muscle cells is also impaired by higher concentrations of PM cholesterol (24).

These defective cells contract normally when stimulated with agonists such as guanosine 5′-O-(3-thiotriphosphate) (a G protein activator) or with second messengers such as inositol triphosphate and diacylglyreol that bypass PM receptors, suggesting that defect resides at the receptor level (30).

Most of the cholesterol in the PM is concentrated in caveolae, where cholesterol is transported in and out of the cell (14, 15). Caveolae also are compartments through which a large number of receptors is internalized (15). Activated receptors couple to G proteins, then internalize into caveolae binding to caveolins that modulate their functions. They are then transferred to recycling endosome organelles that recycle most of these receptors back to the bulk PM where they again become available for activation (27).

High caveolar concentrations of cholesterol impair the muscle contraction by sequestering receptors in this domain, resulting in their reduction in the bulk PM (PM − caveolae) (28). These findings were further supported by the lower binding of radiolabeled CCK-8 and of its antagonist l-364,718, consistent with the presence of a fewer number of CCK-1 receptor (CCK-1R) in the bulk PM (27). The reduced agonist and antagonist binding is normalized after these muscle cells are incubated with cholesterol-free liposomes for 90 min, resulting in removal of the excess cholesterol from the caveolae. Thus receptors are present in muscle cells from GBs with ChS but are unavailable for binding and activation at the bulk PM because a higher percentage of receptors remain sequestered in caveolae containing high levels of cholesterol (28).

The caveolae of muscle cells from GBs with ChS also contain higher concentrations of caveolin-3 (CAV-3), the main caveolin protein in smooth muscle cells (13, 14), than muscle cells from GB with PS (28). The mechanism responsible for the cholesterol-induced sequestration of receptors and caveolin proteins in caveolae of muscle cells from GBs with ChS is not known (27). It has been suggested that increased caveolin mRNA and transfer from the Golgi to caveolae is induced by cholesterol influx (13). However, it is also possible that the increase in caveolin protein synthesis is the result of greater sequestration of these proteins in the caveolae by the effects of higher cholesterol levels.

The aim of these studies was to examine the mechanism whereby abnormal concentrations of PM cholesterol cause the accumulation of receptors, G proteins, and CAV-3 proteins in the caveolae.

METHODS

Patients.

Human GBs were obtained by elective laparoscopic cholecystectomy performed for gallstone disease. None of the patients had a history or clinical evidence of acute cholecystitis. Gallstones were classified as cholesterol and pigment according to their gross appearance and chemical analysis (2). The use of these human GB tissues was approved by the Rhode Island Hospital Human Protection Committee.

Animals.

Male guinea pigs (weight 450–500 g) were purchased from Elm Hill Breeding Laboratories (Chelmsford, MA). The Animal Welfare Committee of Rhode Island Hospital approved their use. Animals were housed in thermoregulated rooms with free access to food and water. After an overnight fast, the guinea pigs were killed with isofluorane.

GBs were kept in ice-cold oxygenated Krebs solution (in mM: 116.6 NaCl, 3.4 KCl, 21.9 NaHCO₃, 1.2 NaH₂PO₄, 2.5 CaCl₂, 1.2 MgCl₂, and 5.4 glucose). After removal of the serosa and mucosa, the muscle layer was carefully cleaned by removing any remaining connective tissue and small blood vessels and then cut into strips for further use.

Isolation of GB muscle cells.

Muscle cells were isolated using methods described previously (1, 3, 5, 29, 30). The muscle layer was cut into 2-mm-wide strips and digested in HEPES buffer containing 0.5 mg/ml of type F collagenase and 2 mg/ml papain (activity of 13.9 U/mg protein) for 20 min at 35°C in a shaking water bath. The buffer was gently gassed with 100% O₂ during the tissue digestion. At the end of the digestive process, the tissue was filtered through Nitex mesh 200 mm (Tetko, Elmsford, NY) and rinsed with 20 ml HEPES. The tissue remaining on the filter was collected and incubated in HEPES buffer at 35°C for 15 min to allow free dispersion of cells. GB muscle cells were permeabilized as previously described to perform studies with antibodies (27, 30).

Muscle cell contraction.

It was measured as previously described (3, 4). Muscle cells were pretreated with buffer (control) or with maximal concentrations of CCK-8 (10⁻⁸ M) and then fixed in 1% acrolein. The cell length was measured with a phase contrast microscope (Carl Zeiss, Jena, Germany) and a closed circuit television camera (Panasonic, Seacaucus, NJ) connected to a Macintosh Computer with NIH Image software. The average length of 30 cells, measured in the absence of agonists, was taken as “control” length and compared with the length measured after addition of agonists. Shortening was defined as the percent length decrease of 30 cells after treatment with agonists compared with control length.

Preparation of cholesterol-free and/or -rich liposomes.

Cholesterol-free liposomes were prepared by using egg phosphatidylcholine (3, 28, 29). Three milliliters of phosphatidylcholine (20 mg/ml in chloroform) in a glass test tube were dried under a stream of nitrogen. The dried lipids were suspended in 3 ml of normal saline and sonicated for 30 min with a Branson 2200 sonicator (Branson Ultrasonics, Danbury, CT). The suspension was then centrifuged at 10,000 g for 30 min to sediment the undispersed lipids. Two milliliters of the supernatant and 8 ml of 0.2% BSA-HEPES buffer were mixed to make cholesterol-free liposomes (∼1.2 mg/ml). Cholesterol-rich liposomes were made using cholesterol-free liposomes plus cholesterol (cholesterol-to-phosphatidylcholine ratio = 3:1 mg/mg).

Preparation of PMs.

PM were prepared and purified by sucrose gradient centrifugation as previously described (27). GB muscle cells preincubated with buffer, cholesterol-free liposomes, or cholesterol-rich liposomes were homogenized separately for 90 min by using a tissue tearer (Biospec Products, Racine, WI) in 10 vol by weight of a sucrose-HEPES buffer. The homogenates were centrifuged at 600 g for 5 min. The supernatant was collected in a clean centrifuge tube (Beckman Instruments) and centrifuged at 150,000 g for 45 min. The pellet was resuspended in sucrose-HEPES, layered over a linear 9–60% sucrose gradient, and centrifuged at 90,000 g for 3 h. The PMs were collected at ∼24% sucrose. They were then diluted and pelleted by centrifugation at 150,000 g for 30 min. The membrane pellet was stored at −70°C.

Isolation of caveolae.

Muscle cells (∼3 × 10⁷ cells from one human GB) were suspended in 10 ml of buffer A (0.25 M sucrose, 1 mM EDTA, and 20 mM Tricine, pH 7.8) and homogenized with a Dounce tissue grinder (28). The supernatant was collected and laid on the top of 30% Percoll in buffer A and centrifuged at 84,000 g for 30 min. The PM band was collected and sonicated with a Vibra Cell Sonicator for six sonication bursts. An aliquot of suspension was saved as total PM fraction. The remainder was mixed with 1.84 ml of buffer C (0.25 M sucrose, 6 mM EDTA, and 120 mM Tricine, pH 7.8, 50% OptiPrep) and 0.16 ml of buffer A (final OptiPrep concentration is 23%) in the bottom of a TH641 tube. A linear 20% to 10% OptiPrep gradient was poured on top of the sample and centrifuged at 52,000 g for 90 min. The top 5 ml of the gradient were collected and placed in a fresh TH641 centrifuge tube. It was mixed with 4 ml of buffer B (0.25 M sucrose, 6 mM EDTA, 120 mM Tricine, pH 7.8). The sample was overlaid with 1 ml of 15% OptiPrep and 0.5 ml of 5% OptiPrep and centrifuged at 52,000 g for 90 min. The band in the 5% interface was collected and designated caveolae membranes.

Recycling of CCK-1R and CAV-3 proteins in human GB muscle cells.

CCK-1R internalization and postendocytic trafficking was studied in the whole PM, caveolae, in early and recycling endosomes. Based on previous studies, most of the CCK-1R are localized in the bulk PM (whole PM − caveolae) in the basal state (28). After activation with CCK-8, the CCK-1R undergoes desensitization and rapid internalization into caveolae. The receptor-caveolin complex is rapidly and transiently shuttled to early endosomes and then slowly returned to the cell surface (4).

CCK-1R and CAV-3 were colocalized in caveolae and endosomal membranes that were separated in a sucrose equilibrium gradient (11). A postnuclear supernatant prepared from GB muscle cells showed a stable coexpression of CAV-3 and CCK-1R after they were fractionated by 5–50% sucrose gradient centrifugation (48,000 rpm for 20 h in a sw50.1 rotor). The gradient was fractionated from the bottom into 12 aliquots, and protein levels were measured in each fraction. Marker antibodies were used against CCK-1R, against EEA1 (for early endosomes), Rab11 (for recycling endosomes), CAV-3, and CAV-1, and caveolin phosphospecific proteins were used to determine protein expression determined by Western blot.

Immunoprecipitation and immunoblotting.

Normal or treated GB muscle cells were washed one time with cold PBS and homogenized with lysis buffer (150 mM NaCl, 50 mM Tris-Cl, 10.01% NaN₃, 2 mM EDTA, 1 mM sodium orthovanadate, 10 μg/ml leupeptin, and 25 μg/ml aprotinin) (11). Samples were centrifuged at 2,800 rpm for 10 min. Postnuclear supernatant was obtained and fractioned by 10–50% sucrose gradient with early endosomes and recycling late endosomes present in fractions 8 to 10 (12, 33). Immunoprecipitation was performed by standard methodologies (12, 32). Briefly, each fraction was precleared with 25 μl of packed protein A cross-linked with 4% beaded agarose at 4°C for 30 min. The beads were removed by centrifugation (13,000 rpm for 10 min), and the supernatant was collected. Primary antibody (two endosome marker antibodies) was added to the cleared supernatant and incubated at 4°C with constant mixing overnight. The immune complex was captured by addition of packed protein A-agarose beads (50 μl) overnight at 4°C. Beads were washed with immunoprecipitation buffer, and their separation was achieved by centrifugation (13,000 rpm for 10 min) followed by the removal of the supernatant. Specificity of the immunoprecipitation was confirmed by negative control reactions performed with beads alone. Subsequently, samples were subjected to Western blot analysis with equal amounts of samples prepared in SDS sample buffer.

Western blot.

Muscle cells were homogenized in Triton X-100 lyse buffer containing 50 mM Tris·HCl (pH 7.5), 100 mM NaCl, 50 mM NaF, 5 mM EDTA, 1% (vol/vol) Triton X-100, 40 mM glycerol phosphate, 40 mM p-nitrophenylphosphate, 200 μM sodium orthovanadate, 100 μM phenylmethylsulfonyl fluoride, 1 μg/ml leupeptin, 1 μg/ml pepstatin A, and 1 μg/ml aprotinin (8). The suspension was centrifuged at 15,000 g for 5 min, and the protein concentration in the supernatant was determined. The supernatants underwent SDS-PAGE, and the separated proteins were electrophoretically transferred to a nitrocellulose membrane at 30 volts overnight. The nitrocellulose membranes were blocked in 5% nonfat dry milk and then incubated with antibody against CCK-1R and a monoclonal antibody against total CAV-3 and polyclonal antibody that detected both pCAV-1 and pCAV-3 protein bands (1:2,000) for 1 h, followed by 60 min incubation in horseradish peroxidase-conjugated secondary antibody (Amersham Biosciences, Piscataway, NJ). Detection was achieved with an enhanced chemiluminescence agent (Amersham Biosciences). The results were expressed as optical density (optical density/mg of protein, optical density/β-actin) or as percent of pCAV-3 of total CAV-3 (pCAV-3-to-total CAV-3 ratios).

Small-interference RNA transfection of human muscle strips in organ culture.

Muscle strips were cultured and transfected with small-interference RNA (siRNA) using the method described before (8, 23). Briefly, human GB muscle strips devoid of mucosa and serosa were placed in serum and antibiotic-free RPMI 1640 medium containing CAV-3 protein siRNA (100 pmol) [CAV-3: sense (5′ → 3′) 5′-CGAAGGAACAAGGUCUAAUtt-3′ (catalog no. 145685; Ambion, Austin, TX); antisense AUUAGACCUUUUCGtc (length 21); clathrin heavy-chain siRNA sense AACCUGCGGUCUGGAGUCAAC and antisense AAUGGAUCUCUUUGAAUACGG (catalog no. 35067; Santa Cruz Biotechnology, Santa Cruz, CA)] or scrambled siRNA (catalog no. 1022076; Qiagen Laboratories, Valencia, CA) with lipofectamine 2000 (Invitrogen, Carlsbad, CA) as instructed by the manufacturer.

After a 24-h transfection with sense and anti-sense siRNA, muscle strips were rinsed in warm PBS and examined for the expression of the proteins by Western blotting. Transfection experiments were performed in triplicate. For studies of muscle contraction and measurements of CCK-1R and G_i-3 proteins, the strips were subjected to the previously described protocol to obtain enzymatically dissociated muscle cells.

Transfection efficiency studies.

The efficiency of the siRNA transfection was verified with a fluorescent-labeled oligo (Invitrogen) transfected to muscle strips at the same concentration of the transfection mixture (100 pmol/ml) and vector lipofectamine (7, 8). After 24 h incubation, muscle cells were dissociated and then frozen for biochemical determinations. In addition, tissue sections (10 μM) were placed on slides to determine the optimal transfection efficiency. The transfection efficiency was 65% using an Olympus IX50/FIA fluorescent microscope equipped with a nap-fix camera (Olympus Optical, Melville, NY).

Chemicals and antibodies: goat polyclonal antibodies against CCK-1R (SC-16173) and G_i-3 proteins.

Monoclonal antibody that only detected total CAV-3 and polyclonal antibody that detected both phosphorylated CAV-1 and CAV-3 (p-CAV-Y14, catalog no. G10406; BD Bioscience, San Jose, CA), clathrin HC antibody (sc-12734; San Cruz laboratories), and EEA1 and Rab11 antigens, 4-amino-5-(4-chloro-phenyl)-7-(t-butyl)pyrazolo[3,4-d]pyrimidine (PP2) (Calbiochem, San Diego, CA) were used. Egg phosphatidylcholine, chlorpromazine, and other reagents were purchased from Sigma Chemical (St. Louis, MO).

Protein determination.

Protein content in each tissue sample was measured according to the methods of Bradford using the Bio-Rad protein assay kit (Bio-Rad Laboratories).

Statistics.

The differences between controls and experimental results were determined by one- or two-factorial repeated analysis by ANOVA. P < 0.05 was considered to be statistically significant.

RESULTS

The traffic of CCK-1R was identified in whole plasma membrane (bulk PM + caveolae), early endosomes, and recycling endosomes by Western blots using CCK-1R antibodies; EEA1 antibody, an antigen present in early endosomes; and Rab11 antibody, an antigen present in recycling endosomes. With the use of these antibodies, receptors and endosome antigens were colocalized in aliquots of the same cell samples at 0 (basal), 5, 20, and 30 min after treatment with CCK-8 for 15 s (Fig. 1). After treatment with CCK-8, CCK-1R were promptly transferred to endosomes with their concentration peaking within 5 min in the early endosomes and peaking within 20 min in the recycling endosomes. CCK-1R returned to the whole PM within 30 min, since the receptor bands were not different between before (0 min) and 30 min after CCK-8 stimulation. This finding is consistent with the response of GB muscle cells to a second dose of CCK-8 that after 30 min caused a contraction similar to the one caused by the first dose (27).

Fig. 1.

Cholecystokinin (CCK)-1 receptor (R) in the plasma membrane (PM), early endosomes (EEA1), and recycling endosomes (RAB11) in muscle cells from gallbladders (GBs) with pigment stones (PS) over a 30-min period. Antibodies against EEA1 proteins identified early endosomes and antibodies against Rab11 proteins identified recycling endosomes. Recycling of receptors peaked at 5 min in early endosomes (*P < 0.001 by ANOVA) at 20 min in recycling endosomes (**P < 0.01 by ANOVA), returning to the PM after 30 min. P values compare peak optical densities (ODs) of CCK-1R of early endosomes at 5 min with all other times and of recycling endosomes at 20 min with all other times. There were no differences between CCK-1R in the PM between 0 and 30 min. Values are means ± SE of n = 3 experiments.

We then examined the contraction induced by CCK-8 (10⁻⁸ M) to determine the effect of cholesterol-rich and cholesterol-free liposome treatment for 90 min on muscle cells from GB with PS and ChS. Figure 2A shows that treatment with cholesterol-rich liposomes impaired the contraction of muscle cells from GB with PS and cholesterol-free liposomes increased the contraction of muscle cells from GB with ChS. Likewise cholesterol-rich liposomes increase the cholesterol content in the PM of muscle cells from GB with PS, whereas cholesterol-free liposomes decreased the cholesterol content of the PM of muscle cells from GB with ChS (Fig. 2B). Figure 2C shows that cholesterol-rich liposome treatment reduced the CCK-1R recycling in the basal state and after CCK-8 stimulation. The optical density of CCK-1R was measured at 0, 5, and 20 min in the fractions of whole PM (that includes caveolae), early (EEA1), and recycling (Rab11) endosomes. CCK-8 increased the optical density of CCK-1R bands in both endosomes in normal muscle cells when incubated with buffer. Cholesterol-rich liposome treatment increased these receptors in the whole PM in the basal state and after CCK-8 and markedly decreased them in the early and recycling endosomes in the basal state and more clearly after CCK-8 stimulation. These findings suggest that high caveolar concentrations of cholesterol inhibit the transfer of CCK-1R from the whole PM (caveolae) to the endosomes.

Fig. 2.

Effect of treatment with cholesterol-rich liposomes (Ch-rich L) and Ch-free L on muscle contraction, PM Ch, and CCK-1R traffic in muscle cells from GBs with PS compared with buffer treated GBs (controls). A and B: treatment with Ch-rich L decreased the contraction induced by CCK-8 and increased the PM Ch content of muscle cells from GB with PS. Treatment with Ch-free L increased the contraction induced by CCK-8 and decreased the Ch content in the PM of muscle cells from GBs with cholesterol stones (ChS) (&, &&, ^, and ^^P < 0.001). C: treatment with Ch-rich L blocked the transfer of receptors from the whole PM to early (EEA1) and recycling (Rab11) endosomes determined over a 20-min period. Ch-rich L increased the OD of CCK-1R in the whole PM (that includes caveolae) in the basal state and after CCK-8 (* and #P < 0.01) and decreased their OD levels in early and recycling endosomes before (basal levels) and after (** and ##P < 0.001 by ANOVA) CCK-8. Values are means ± SE of n = 3.

We next examined the effect of high cholesterol concentrations on pCAV-3 in normal muscle cells with a polyclonal antibody that detected the total and pCAV-3 and pCAV-1 pretreated with buffer or with cholesterol-rich liposomes for 90 min in the basal state and after CCK-8 (Fig. 3A). CCK-8 increased pCAV-3 and to a lesser extent pCAV-1. Cholesterol-rich liposomes blocked the phosphorylation of both caveolins in the basal state and after CCK-8. The inhibition of pCAV by cholesterol-rich liposomes was confirmed by treating muscle cells with PP2, a tyrosine kinase inhibitor. PP2 (10⁻⁵ M) reduced the ratio pCAV-3-to-total CAV-3 in the basal state and after CCK-8 (Fig. 3B).

Fig. 3.

A: effect of Ch-rich L on caveolin (CAV)-1 and CAV-3 phosphorylation (pCAV-3) in the basal state and after CCK-8. A polyclonal antibody detected CAV-1 and CAV-3 proteins. OD only was determined on total and pCAV-3 proteins since CCK-8 stimulated mainly the phosphorylation of CAV-3. CCK-8 increased the levels of pCAV-3 in muscle cells treated with buffer (*P < 0.001 by ANOVA). Ch-rich L reduced the phosphorylation of CAV-3 compared with Ch-free L treatment (#P < 0.01 by ANOVA). Ch-rich L treatment reduced CAV-3 phosphorylation after CCK-8 compared with CCK-8 alone (**P < 0.001 by ANOVA). B: likewise, 4-amino-5-(4-chloro-phenyl)-7-(t-butyl)pyrazolo[3,4-d]pyrimidine (PP2) (a tyrosine kinase inhibitor) blocked the phosphorylation of CAV-3 in the basal state (&P < 0.01 by ANOVA) and after CCK-8 stimulation (&&P < 0.001 by ANOVA). Values are means ± SE of n = 3.

Moreover, in buffer-treated muscle cells, total CAV-3 and pCAV-3 were transferred to the endosomes during basal conditions (Fig. 4A). The transfer was further increased after CCK-8 stimulation since both fractions were present in early (EEA1) and recycling (Rab11) endosomes. Treatment with cholesterol-rich liposomes blocked the transfer of pCAV-3 and total CAV-3 to the early and recycling endosomes, resulting in the accumulation of CAV-3 in the caveolae. Figure 4B presents the same data of Fig. 4A to show more clearly the statistical analysis between control and cholesterol-rich liposome treatment in the basal state and after CCK-8 stimulation in the bulk PM and early and recycling endosomes.

Fig. 4.

Western blot and OD analysis illustrates the transfer of total CAV-3 and pCAV-3 from the whole PM to EEA1 and Rab11 endosomes in muscle cells from GBs with PS at 0, 5, and 20 min. CCK-8 increased the percent pCAV-3 in the EEA1 (P < 0.01 by ANOVA)- and in Rab11 (P < 0.001 by ANOVA)-labeled endosomes. Ch-rich L treatment reduced pCAV-3 levels in the basal state in the PM (*P < 0.01 by ANOVA) and in the early endosomes (P < 0.01 by ANOVA). Ch-rich L also lowered pCAV-3 levels after CCK-8 stimulation in the whole PM and in the endosomes. It reduced pCAV-3 in the PM (**P < 0.05 by ANOVA) and in EEA1 (#P < 0.001 by ANOVA) and Rab11 endosomes (##P < 0.005 by ANOVA). Values are means ± SE of n = 4.

We then examine whether PP2, like high caveolar cholesterol concentrations, inhibits the muscle contraction by sequestering CCK-1R in the caveolae, thus preventing their return back to the bulk PM. Treatment with PP2 (10⁻⁵ M) for 30 min blocked the contraction induced by CCK-8 in muscle cells from GB with PS and ChS (Fig. 5A). It also increased the sequestration of CCK-1R in the caveolae in the basal state and after CCK-8 stimulation (Fig. 5B). PP2 reduced the optical density of the CCK-1R bands in the bulk PM, causing a corresponding increase in the caveolar bands and reproducing the effects induced by high concentrations of caveolar cholesterol. The bulk PM-to-caveolae ratio of the CCK-1R/β-actin was 2.3 ± 0.4 in controls. After PP2 treatment, it decreased to 0.9 ± 0.1 (P < 0.01 by ANOVA) and was further reduced after CCK-8 stimulation to 0.5 ± 0.06 (P < 0.05 by ANOVA). The accumulation of CCK-1R in the caveolae suggests that PP2 blocked the recycling of receptors back to the bulk PM. These results suggest that phosphorylation of CAV-3 proteins is a required step for the recycling of CAV-3 proteins and CCK-1R.

Fig. 5.

Effect of PP2 (a tyrosine kinase inhibitor) on muscle contraction and CCK-1R distribution within the PM. A: treatment with PP2 decreased the CCK-8-induced contraction of muscle cells from PS and ChS (#P < 0.001 and ##P < 0.01 by ANOVA). B: PP2 lowered the concentration of CCK-1R in the bulk PM, but it increased it in the caveolae in the basal state (*P < 0.05 by ANOVA) and after CCK-8 (&P < 0.001) (**P < 0.005 by ANOVA). The receptors were not different in the whole plasma membrane (WPM). Values are means ± SE of n = 3.

To confirm the role of CAV-3 proteins and receptor accumulation in caveolae in the contraction induced by CCK-8, human permeable muscle cells from GB with PS were treated with CAV-3 antibody (titer 1:800) or in cultured muscle strips transfected with scrambled or with CAV-3 siRNAs. The CAV-3 antibody reduced the contraction induced by CCK-8 compared with buffer-treated cells (Fig. 6A). CAV-3 protein knockdown also impaired the contraction induced by CCK-8 compared with controls (treated with culture medium or scrambled siRNA) (Fig. 6B). The magnitude of the contraction of cultured control human and guinea pig muscle cells was not significantly different from that of freshly dissociated muscle cells. These observations were confirmed in cultured muscle cells from guinea pig GB transfected with scrambled or CAV-3 siRNA (Fig. 6B). CAV-3 siRNA reduced the protein expression of CAV-3 determined by Western blot by 58.4% from the levels determined in muscle cells treated with scrambled siRNA (Fig. 6C).

Fig. 6.

A: effect of CAV-3 antibody (Ab) on the contraction of muscle cells from human GBs with PS induced by CCK-8. The antibody (titer 1:800) reduced the contraction (%shortening) induced by CCK-8 (&P < 0.001 by ANOVA). B: effect of CAV-3 siRNA on the contraction induced by CCK-8 in cultured muscle cells from human GBs with PS and from guinea pig GBs. CAV-3 small-interfering RNA (siRNA) reduced the contraction compared with controls (culture media) and with scrambled siRNA in human (* and **P < 0.01 by ANOVA) and guinea pig (# and ##P < 0.01 by ANOVA) GB muscle cells. C: CAV-3 protein expression in muscle cells from human GBs with PS. CAV-3 was reduced in muscle cells transfected with CAV-3 siRNA compared with scrambled siRNA (&&P < 0.005 by ANOVA). Values are means ± SE of n = 3.

CAV-3 siRNA also increased CCK-1R (Fig. 7A) and G_i-3 proteins (Fig. 7B) in the caveolae and a concomitant decrease in the bulk PM. These results are similar to the findings observed in muscle cells treated with cholesterol-rich liposomes or with PP2.

Fig. 7.

CAV-3 siRNA altered the distribution of CCK-1R (A) and G_i-3 (B) proteins in the WPM, bulk PM, and caveolae 30 min after CCK-8. CAV-3 siRNA reduced the levels of CCK-1R (*P < 0.001) and G_i-3 proteins (**P < 0.001) in the bulk PM and increased their levels in the caveolae compared with controls (culture media or scrambled siRNA, #P < 0.005 and ##P < 0.01 by ANOVA). Values are means ± SE of n = 3.

We then examine whether abnormal cholesterol concentrations in the whole PM affect agonists that bind to receptors that are internalized through clathrin-coated pits, a low-cholesterol membrane compartment. Met-enkephalin was selected as an agonist because it stimulates δ-opioid receptors that are internalized through clathrin (9, 10, 16). Chlorpromazine (a specific clathrin antagonist) had no effect on the contraction induced by CCK-8 but blocked the contraction induced by met-enkephalin (10⁻⁵ M) in muscle cells from GB with PS (Fig. 8A). The contraction of muscle cells from GBs with ChS induced by met-enkephalin was normal (Fig. 8B). Similar results were obtained in cells transfected with the clathrin heavy-chain siRNA (Fig. 8, C and D). Clathrin heavy-chain siRNA blocked the contraction elicited by met-enkephalin and had no effect on that induced by CCK-8.

Fig. 8.

A: chlorpromazine (10⁻⁶ M) had no effect on CCK-8-induced contraction in muscle cells from GBs with PS. However, it blocked the contraction elicited by met-enkephalin (m-enkephalin) (*P < 0.01 by ANOVA). B: the contraction induced by met-enkephalin was not impaired in muscle cells from GBs with ChS treated with buffer or with Ch-free liposomes. It was blocked by chlorpromazine (#P < 0.001 by ANOVA). C: the clathrin HC siRNA had no effect on CCK-8-induced contraction but blocked by met-enkephalin-induced contraction (**P < 0.001 by ANOVA). Values are means ± SE of n = 3.

DISCUSSION

The data show that high caveolar concentrations of cholesterol in human GB muscle cells sequester CAV-3 proteins, CCK-1R, and G_i-3 proteins in the caveolae in the basal state and after stimulation with CCK-8. The sequestration of receptors decreases CCK-1R turnover inhibiting their transfer to the endosomes resulting in their accumulation in the caveolae and concomitant reduction in the bulk PM. Fewer CCK-1R in the bulk PM contributes to the reduced CCK-8 binding to receptors and impaired muscle contraction in muscle cells from GBs with ChS.

The turnover of CCK-1R in normal muscle cells after stimulation with maximal CCK-8 concentrations is ∼30 min, since there are no differences in the optical density of the CCK-1R bands between basal (0 min) and 30 min after CCK-8. These results are consistent with our previous findings that there was no difference in the magnitude of muscle cell contraction between the first dose at time 0 and second dose of CCK-8 after 30 min (25, 27).

CCK-8 activates CCK-1R that couples to G_i-3 proteins in GB muscle cells. This complex is rapidly internalized into caveolae where it binds to caveolin proteins. These caveolar proteins negatively regulate the functions of G proteins and recycling of CCK-1R (18, 26). High caveolar concentrations of cholesterol decreased the transfer of CCK-1R from the caveolae to early and recycling endosomes, resulting in their accumulation in caveolar domains; therefore, fewer receptors return to the bulk PM where they can become accessible to their respective ligands.

Moreover, high cholesterol concentrations cause a rise in the levels of caveolin proteins in the caveolae. It has been suggested that the cellular influx of cholesterol increases the levels of caveolin proteins by stimulating their synthesis and their transfer from the Golgi or from the endoplasmic reticulum to the caveolae (13, 19). Our results suggest that greater influx of cholesterol increases the concentration of CAV-3 protein by blocking the transfer of these proteins from the caveolae to early and recycling endosomes. Cholesterol inhibits the phosphorylation of CAV-3 proteins that is normally induced by the internalization of receptor-G protein complexes into the caveolae (20). High caveolar concentrations of cholesterol lower the pCAV-3-to-total CAV-3 ratios in the caveolae in the basal and stimulated states, reducing their transfer to endosomes. It is therefore conceivable that increases in CAV-3 mRNA induced by cholesterol influx are stimulated in part by the accumulation of these proteins in the caveolae.

These conclusions are supported by the effects of PP2, an inhibitor of Src-tyrosine phosphorylation of caveolins, on normal GB muscle cells. Receptor-G protein complexes stimulate the phosphorylation of caveolins by tyrosine kinases. PP2, by blocking CAV-3 phosphorylation, reproduced the effects of high levels of cholesterol in GB muscle cells. PP2 impaired the muscle contraction induced by CCK-8 and increased the caveolar accumulation of CAV-3 proteins and CCK-1R. These actions of PP2 are mostly due to an almost complete inhibition of the transfer of CAV-3 and CCK-1R to the endosomes in the basal state and after CCK-8. These results suggest that CAV-3 phosphorylation is a necessary step for the transfer of caveolin proteins into endosomes that may facilitate the recycling of receptors.

The role of CAV-3 proteins in the recycling of CCK-1R and G_i-3 proteins was further examined by transfecting normal muscle cells with CAV-3 siRNA. Knockdown of CAV-3 proteins in control GB muscle cells reduced the muscle contraction induced by CCK-8 compared with muscle cells transfected with scrambled siRNA. It also increased the accumulation of CCK-1R and G_i-3 proteins in the caveolae and their concomitant reduction in the bulk PM. These findings suggest that normal levels of CAV-3 proteins are also required for normal turnover of receptor-G protein complexes and muscle contraction induced by agonists that activate receptors that internalize through caveolae.

High concentrations of cholesterol, however, do not seem to alter the functions of receptors that internalize through clathrin-coated vesicles although a normal concentration of cholesterol seems to be required for the normal function of clathrin-dependent endocytosis (21). Cholesterol did not affect the contraction induced by met-enkephalin, a ligand of δ-opioid receptors that internalize through clathrin pits (9, 10, 16). Met-enkephalin-induced contraction was blocked by chlorpromazine, an inhibitor of clathrin, and by knock down of the clathrin heavy chain in muscle cells transfected with clathrin HC siRNA. However, chlorpromazine or clathrin HC siRNA had no effect on the contraction induced by CCK-8. In contrast to ileal muscle cells, these studies suggest the CCK-1R in GB muscle cells internalize mostly through caveolae (17, 18). Furthermore, the finding of a normal contraction mediated by δ-opioid receptors suggests that high levels of cholesterol appear to be limited to caveolae-dependent endocytosis since they do not seem to affect the recycling of receptors once they have been transferred to the endosomes since the same organelles recycle receptors that internalize through caveolae and through clathrin-coated pits (16).

These results also suggest that inhibitory effects of high caveolar concentrations of cholesterol are likely to impair the recycling of other receptors that are internalized through these cholesterol-rich domains affecting receptors that are activated by anti-inflammatory and cytoprotective agonists such as PGE₂ (29), creating an environment in the muscle and in the GB wall as a whole that may facilitate the development of acute and chronic cholecystitis.

GRANTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant RO-1 DK-027389-20A.

DISCLOSURES

The authors declare no financial conflicts of interest.