Ezrin knockdown reduces procaterol-stimulated ciliary beating without morphological changes in mouse airway cilia

Mucociliary clearance is the first line of defense in the airway epithelium. Its function is the result of beating cilia on the surface of airway ciliary cells cooperating with the protective mucous layer. The surface mucous layer entraps inhaled small particles and pathogens (viruses, bacteria and fungi), and the beating cilia transport the mucous layer to the oropharynx. Thus, the beating cilia are the key apparatus involved in mucociliary clearance (Wanner et al., 1996; Knowles and Boucher, 2002; Salathe, 2007). Defects in beating cilia in humans cause primary ciliary dyskinesia (PCD), whose clinical symptoms include chronic rhinosinusitis, chronic bronchitis, infertility and hydrocephalus (Afzelius, 2004).

Airway ciliary cells are multiciliated cells (MCCs). Their ciliogenesis proceeds in several steps: (1) MCC insertion into the epithelium, and basal body (BB) biogenesis and duplication from centrioles; (2) apical expansion of the MCC, and BB migration and docking to the apical surface; and (3) BB distribution with planar cell polarity followed by elongation of cilia (Kulkarni et al., 2018). F-actin plays central roles in these steps, especially expansion of the apical surface of MCCs and distribution of BBs. Ezrin, an F-actin binding protein, is associated with BBs throughout the apical region of MCCs and is possibly required for anchoring of BBs to the apical cytoskeleton (Gomperts et al., 2004). Ezrin is also associated with the microvilli of epithelial cells and participates in the anchoring of a variety of proteins to the apical cytoskeleton. Furthermore, ezrin is indispensable for intact microvilli formation in small intestine and retinal pigment epithelia, as reported previously using ezrin-knockout (Vil2^−/−) mice (Saotome et al., 2004; Bonilha et al., 2006; Casaletto et al., 2011). Ezrin functions as a general crosslinker between apical membrane proteins and the actin cytoskeleton, either directly or indirectly via scaffold proteins such as Na⁺/H⁺ exchanger regulatory factors (NHERF) 1 or 2 (also known as SLC9A3R1 and SLC9A3R2, respectively), and is involved in the functional expression of apical membrane proteins and the regulation of endocytosis (Tsukita and Yonemura, 1999).

Previously, Tamura et al. have created ezrin-knockdown (Vil2^kd/kd) mice by introducing a mutational cassette between exons 2 and 3 of the Vil2 gene, which encodes ezrin (Tamura et al., 2005). Consequently, in the Vil2^kd/kd mice, the expression level of ezrin is reduced to 5% compared with the expression level in wild-type (WT) mice, and the Vil2^kd/kd mice have been used to study the physiological roles of ezrin. Vil2^kd/kd mice show achlorhydria, due to defects in membrane fusion, without any effects on the total expression of gastric H⁺/K⁺-ATPase in parietal cells (Tamura et al., 2005). The mice also show hypophosphatemia, which is caused by urinary loss of phosphate via the impairment of apical localization of sodium-dependent phosphate transporters, Npt2a and Npt2c (also known as SLC34A1 and SLC34A3, respectively), in the renal proximal tubule (Hatano et al., 2013). The Vil2^kd/kd mice also show severe intrahepatic cholestasis characterized by intrahepatic bile acid accumulation, which is caused by reduced cell surface expression of cystic fibrosis transmembrane conductance regulator (CFTR), anion exchanger 2 (AE-2, also known as SLC4A2), and aquaporin 1 (AQP1) in cholangiocytes (Hatano et al., 2015).

In bronchiolar ciliary cells, procaterol – a β₂ adrenergic receptor (β₂AR) agonist – activates ciliary beating via increasing the intracellular cyclic AMP (cAMP) level (Komatani-Tamiya et al., 2012). It has been reported that ezrin indirectly binds to β₂AR and the CFTR chloride channel via NHERF1 to form a multiprotein complex in vitro and in cultured cells (Naren et al., 2003). In this complex, ezrin recruits protein kinase A (PKA), acting as a PKA-anchoring protein (AKAP) to promote phosphorylation and activate CFTR (Dransfield et al., 1997). Therefore, ezrin may regulate cell surface expression of β₂AR and ciliary beating stimulated by adrenergic agonists. However, the physiological roles of ezrin in airway motile cilia in vivo remain unclear. In the present study, we examined the morphology of trachea of Vil2^kd/kd mice, including ciliary length and number, as well as BB orientation, using scanning electron microscopy (SEM), thin-section transmission electron microscopy (TEM) and immunofluorescence analyses. We also studied the expression of β₂AR at the cell surface, using immunofluorescence and cell surface biotinylation, and ciliary beating in the presence and absence of procaterol.

In adult mouse tracheal epithelial cells (MTECs), ezrin is expressed in the apical domain of ciliated cells but not in non-ciliated cells (Huang et al., 2003). When we examined the localization of ezrin in isolated airway ciliary cells and trachea from WT mice using immunofluorescence analysis, we observed that ezrin was localized with F-actin at the BB compartment, which was shown by anti-outer dense fiber 2 (Odf2) antibody staining (Nakagawa et al., 2001), near the base of motile cilia (Fig. 1A,B). To perform detailed analysis of the subcellular localization of ezrin in multiciliated airway cells in trachea, we used super-resolution fluorescence microscopy. In mouse trachea, ezrin was localized to microvilli and around the BB, which is an Odf2-positive region, in airway ciliary cells (Fig. 1C,D).

It has been reported that, in addition to ezrin, moesin is also expressed in airway ciliary cells, although the function of the two proteins in the cells is apparently different (Huang et al., 2003). Ezrin–radixin–moesin (ERM) proteins may infrequently compensate for their respective functions (Doi et al., 1999). However, there was no compensatory upregulation of moesin in Vil2^kd/kd mice (Fig. S1A,B).

To visualize the morphology of motile cilia in multiciliated airway cells, we stained isolated airway ciliary cells and trachea of Vil2^kd/kd mice with an antibody against acetylated α-tubulin. The morphology of motile cilia in Vil2^kd/kd mice was apparently similar to that in WT mice (Fig. 2A). To analyze the surface structure of airway cilia in greater detail, we performed SEM analysis. In mouse trachea, no apparent morphological differences, including in the length of cilia and the percentage of ciliated cells, were observed between WT and Vil2^kd/kd mice (Fig. 2B–D). The orientation of the BB basal foot (BF) regulates the bending direction of cilia and directly influences the efficiency of mucociliary transport. Thus, we examined the BF orientation in trachea from WT and Vil2^kd/kd mice using TEM. We evaluated the orientation by calculating the index of orientation (I_O), which indicates the degree of uniformity of the BB BF orientation (Herawati et al., 2016; Nakayama et al., 2021). In the results, no morphological changes, including in BF orientation, were observed for the cilia of trachea from Vil2^kd/kd mice (Fig. 2E,F).

To evaluate the whole-cell orientation of BBs [I_O(cell)] in mouse trachea, we performed super-resolution immunofluorescence microscopy and analyzed the positions of Odf2 and centriolin (a BF marker; Gromley et al., 2003). I_O(cell) showed no significant differences between WT (0.85±0.03; mean±s.e.m.) and Vil2^kd/kd (0.86±0.02) mouse tracheas (Fig. 2G–I). These results suggest that ezrin knockdown did not affect the functional morphology and planar cell polarity of airway motile cilia.

To examine the influence on mucociliary clearance of ezrin knockdown, we analyzed the fluid flow for mucociliary transport using live imaging of fluorescent beads in WT and Vil2^kd/kd mouse tracheas. We evaluated the mucociliary transport by quantifying the directionality and speed of fluid flow. The directionality of the flow of fluorescent beads showed no significant difference between WT and Vil2^kd/kd tracheas, which is consistent with the finding that there was no difference in the degree of uniformity of the BB BF orientation between WT and Vil2^kd/kd tracheas. However, a significant difference was observed in the speed of fluid flow between these tracheas after stimulation with procaterol, which is a selective β₂AR agonist. The speed of fluid flow in Vil2^kd/kd trachea did not significantly change upon stimulation with 1 nM procaterol, whereas that in WT trachea significantly increased following stimulation with 1 nM procaterol (Fig. 3A,B; Movie 1). To evaluate ciliary function in detail, we analyzed ciliary beat frequency (CBF) and ciliary bend angle (CBA) in isolated airway ciliary cells of mice. The CBF and CBA values in basal conditions showed no differences between WT and Vil2^kd/kd mice (data not shown). Furthermore, we quantitatively evaluated the activation of ciliary beating by procaterol treatment. To normalize the data across experiments, the CBF and CBA measurements shown in Figs 4 and 5 are presented as the ratio of the value obtained at a given stimulation timepoint and the average value just before stimulation (CBF_t/CBF₀ and CBA_t/CBA₀, respectively; see Materials and Methods for further details). The ratios of the CBF and CBA values were 1.81±0.04 and 1.76±0.02 (mean±s.e.m.), respectively, at 8 min after stimulation with 1 nM procaterol in WT ciliary cells (Fig. 4A). However, in Vil2^kd/kd mice, the ratios of CBF and CBA values were only 1.40±0.03 and 1.39±0.02, respectively, at 8 min after stimulation with 1 nM procaterol (Fig. 4B), which is consistent with the finding that the speed of fluid flow in Vil2^kd/kd trachea did not change upon treatment with procaterol. Conversely, there was no significant difference in the increase of CBF and CBA values following stimulation with 100 μM IBMX, which is a phosphodiesterase inhibitor with broad-spectrum activity, between WT and Vil2^kd/kd mice (Fig. 4C,D). These results suggest that increased intracellular cAMP levels result in activation of ciliary beating not only in WT mice but also in Vil2^kd/kd mice.

To evaluate whether the difference in procaterol-induced ciliary beating activation between WT and Vil2^kd/kd mice was caused by intracellular cAMP levels, we measured intracellular cAMP levels in isolated airway ciliary cells of these mice. Procaterol significantly increased the intracellular cAMP levels in WT and Vil2^kd/kd cells. However, the cAMP level in the WT cells was significantly higher than that in the Vil2^kd/kd cells (Fig. 5A). We also measured the intracellular cAMP levels in WT and Vil2^kd/kd cells after treatment with IBMX. IBMX significantly increased the intracellular cAMP levels in both WT and Vil2^kd/kd cells, and the intracellular cAMP level in the WT cells was similar to that found in the Vil2^kd/kd cells (Fig. 5B). These results, that intracellular cAMP levels are increased upon stimulation with procaterol or IBMX in WT cells, are similar to those reported previously (Kogiso et al., 2017). These results suggest that the reduced responsiveness to procaterol found in the ciliary beating of isolated airway ciliary cells from Vil2^kd/kd mice can be attributed to the reduced β₂AR-dependent generation of intracellular cAMP in response to procaterol.

We performed immunoblotting and immunofluorescence analyses of β₂AR and NHERF1, which is a member of the PDZ scaffold protein family, in isolated airway cells from WT and Vil2^kd/kd mice, since the formation of a β₂AR–NHERF1–ezrin–actin multiprotein complex at the apical surface of epithelial cells has been reported (Naren et al., 2003). The immunoblotting analysis revealed that total expression levels of NHERF1 and β₂AR do not differ between airway ciliary cells isolated from WT and Vil2^kd/kd mice (Fig. 6A,B). To further investigate the cell surface expression of β₂AR, we performed a surface biotinylation assay using isolated airway ciliary cells from WT and Vil2^kd/kd mice. We confirmed the expression of GAPDH to ensure exclusion of intracellular proteins. In the biotinylated sample, GAPDH was not detected (data not shown). The results of the surface biotinylation assay revealed that surface expression of β₂AR in Vil2^kd/kd mice was significantly lower than that in WT mice (Fig. 6C,D). To determine whether the low cell surface expression of β₂AR in Vil2^kd/kd mice was due to reduced apical targeting of newly synthesized β₂AR or to reduced retention of β₂AR at the apical surface, we performed an internalization assay using cell surface biotinylation in combination with chase experiments. We observed that 80% of the biotinylated β₂AR of both WT and Vil2^kd/kd mice was internalized following 1 h incubation at 37°C (Fig. 6E,F). Conversely, internalized and biotinylated β₂AR was scarcely detected in WT and Vil2^kd/kd mice following 1 h incubation at 4°C, indicating that the internalization found in the experiment represents temperature-dependent endocytosis. These results suggest that the internalization step, which is related to retention of the β₂AR at the apical surface, was not impaired in the Vil2^kd/kd mice.

We performed immunofluorescence analysis in isolated airway ciliary cells to examine the localization patterns of NHERF1 and β₂AR in WT and Vil2^kd/kd mice. NHERF1 colocalized with F-actin in the subapical region, which contains the root of motile cilia, in WT mice (Fig. 7A). This localization pattern was similar to that of ezrin, as shown in Fig. 1A. In addition, β₂AR colocalized with F-actin in WT mice (Fig. 7B), suggesting that β₂AR is anchored to the apical membrane via NHERF1 and ezrin in airway ciliary cells. However, in Vil2^kd/kd mice, the subapical localization of NHERF1 and the apical localization of β₂AR were disturbed (Fig. 7C,D). These results suggest that ezrin plays an important role in formation of the multiprotein complex consisting of β₂AR and NHERF1 in the subapical region and in the functional expression of β₂AR on the apical membrane.

Ezrin is a multifunctional protein. First, it is a general crosslinker between plasma membrane proteins or phospholipids (bound to the amino-terminal domain) and the actin cytoskeleton (bound to the carboxy-terminal domain), and it is involved in the functional expression of membrane proteins on the cell surface (Kawaguchi et al., 2017). As such, ezrin is important for apical membrane localization of the sodium-dependent phosphate transporters Npt2a and Npt2c in renal proximal tubules (Hatano et al., 2013), and of CFTR, AE-2 and AQP1 in cholangiocytes (Hatano et al., 2015). Second, ezrin also functions as an AKAP, acting as a scaffold protein that tethers target proteins with PKA to promote their phosphorylation. In fact, it has been reported that ezrin recruits PKA, CFTR and Na⁺/H⁺ exchanger 3 (NHE3, also known as SLC9A3) on the same platform to promote phosphorylation and modulate function of CFTR and NHE3 (Bagorda et al., 2002; Naren et al., 2003). Third, ezrin also acts as a regulator of small G proteins (Ivetic and Ridley, 2004). It regulates cytoskeletal organization by sequestering Rho-related proteins (such as GDP/GTP exchange protein and Rho GDI) and functions as both an upstream and downstream effector of Rho GTPases (Takahashi et al., 1997).

In mouse and human pulmonary epithelia, ezrin is associated with the BBs of cilia (Gomperts et al., 2004; Epting et al., 2015), with its expression being under the control of Foxj1, which is a master transcriptional regulator of ciliated cells. Foxj1 expression is required for correct apical localization of ezrin (Huang et al., 2003). Conversely, the expression of ezrin is decreased in the pulmonary epithelium of Foxj1-knockout mice due to increased calpain proteolysis (Gomperts et al., 2004). Therefore, ezrin seems to be required for anchoring of BBs to the apical cytoskeleton.

In zebrafish and Xenopus, ezrin has been reported to be required for apical docking of BBs and for ciliogenesis (Epting et al., 2015). In the zebrafish embryo, ezrin expression is detected in ciliated cells, the olfactory placode and pit, brain ventricles, and pronephric tubules. Ezrin knockdown by treatment with a translation-blocking morpholino (TB-MO) leads to reduced microvilli formation, defects in BB docking, shortened cilia and pronephric cyst formation (Epting et al., 2015). In Xenopus epidermal MCCs, ezrin knockdown using a TB-MO leads to impaired actin remodeling and intracellular ciliary axoneme formation, indicating that ezrin is required for apical docking of BBs as well as ciliogenesis. However, very few studies have addressed the physiological role of ezrin in mammalian airway cilia. Here, we examined the effects of ezrin knockdown on ciliary morphology by using immunofluorescent staining of acetylated α-tubulin, SEM and TEM to analyse isolated airway ciliary cells and tracheas prepared from WT and Vil2^kd/kd mice. Furthermore, we quantitatively measured the values of CBF and CBA to evaluate beating of airway cilia by using video microscopy equipped with a high-speed camera. Surprisingly, we did not observe any abnormal morphology of cilia in both isolated airway ciliary cells and trachea from Vil2^kd/kd mice. Moreover, there was no significant difference in the values of I_O, which indicates the degree of uniformity of the BB BF orientation, between WT and Vil2^kd/kd mice. These results suggest that ezrin is not directly involved in the development and planar cell polarity of mouse cilia. It cannot be completely excluded that the very low level of ezrin expressed in Vil2^kd/kd mice (less than 5% of the expression in WT mice) may allow normal ciliogenesis. It should be noted that injection of zebrafish embryos with a TB-MO targeting ezrin2 (also known as ezrb) led to a clear reduction of ezrin2 expression but not a complete knockout (Link et al., 2006). Rather, the expression level of ezrin2 relative to the WT expression level after treatment with the TB-MO, as observed by western blotting, looks higher than that found for ezrin in our Vil2^kd/kd mice. The effects of ezrin knockdown on BB docking as well as ciliogenesis seem to be quite different between Xenopus, zebrafish and mouse.

It has been reported that ezrin interacts β₂AR via a scaffold protein, NHERF1, and tethers a multiprotein complex consisting of ezrin, β₂AR and NHERF1 to the actin cytoskeleton. In the complex, a quadruplet amino acid motif at the C terminus of β₂AR, DSLL (Asp-Ser-Leu-Leu), binds to the PDZ domain of NHERF1. Therefore, it seems that functional expression of β₂AR at the cell surface is regulated by ezrin. Here, we discovered that cell surface expression of β₂AR and NHERF1 was disturbed in Vil2^kd/kd mice, as shown in Fig. 8. Cell surface biotinylation in combination with chase experiments revealed that disturbed cell surface expression of β₂AR in the Vil2^kd/kd mice was not due to endocytosis-induced internalization. Consequently, airway cilia beating stimulated by procaterol was impaired in the Vil2^kd/kd mice. Conversely, ezrin did not act as an AKAP in airway ciliary cells, because there was no significant difference between Vil2^kd/kd and WT mouse airway ciliary cells in the increase of CBF and CBA values upon treatment with IBMX, which increases intracellular cAMP without going through β₂AR. Therefore, ezrin seems to affect ciliary beating, with its point of action being located upstream of cAMP production.

In conclusion, our results suggest that morphogenesis of motile cilia is not disturbed in airway cilia in Vil2^kd/kd mice. Rather, ezrin plays a key role in functional expression of β₂AR at the cell surface and ciliary beating stimulated by procaterol via β₂AR.

All procedures involving animals were approved by the Committees for Animal Research of Ritsumeikan University Institutional Animal Care and Use Committee (BKC2020-027). Vil2^kd/kd mice were the kind gifts of Professor Sachiko Tsukita of the Graduate School of Frontier Biosciences, Osaka University, Japan. Eight- to eleven-week-old male WT and Vil2^kd/kd mice were used in this study.

For isolation of airway ciliary cells, the mice were first anesthetized by inhalational isoflurane (3%) and were then further anesthetized and heparinized by an intraperitoneal injection (ip) of pentobarbital sodium (70 mg/kg) plus heparin (1000 units/kg) for 15 min. Finally, the mice were euthanized by a high dose of pentobarbital sodium (100 mg/kg, ip).

The following antibodies were used in this study: rabbit anti-acetylated α-tubulin antibody (ab179484; Abcam, Cambridge, UK; 1:500), mouse anti-acetylated tubulin antibody (6-11B-1 T6793; Sigma-Aldrich, St Louis, MO; 1:100), mouse anti-ezrin antibody (3C12; Abcam; 1:1000), rabbit anti-Odf2 antibody (ab43840; Abcam; 1:400), rat anti-centriolin antibody generated previously (Ishikawa et al., 2005; 1:2), mouse anti-β₂AR antibody (E-3; Santa Cruz Biotechnology, Dallas, TX; 1:100 for immunofluoresence, 1:500 for immnoblotting), rabbit anti-GAPDH antibody (14C10 #2118; Cell Signaling Technology, Beverly, MA; 1:1000), Alexa Fluor 488-conjugated anti-mouse antibody and Alexa Fluor 633-conjugated anti-rabbit antibody (Invitrogen, Waltham, MA; 1:200).

We used Rhodamine phalloidin (R415; Thermo Fisher Scientific, Waltham, MA) to stain F-actin. Procaterol was a generous gift from Otsuka Pharmaceutical Co., Ltd. (Tokyo, Japan). Heparin, elastase, IBMX and bovine serum albumin (BSA) were purchased from FUJIFILM Wako Pure Chemical Industries, Ltd. (Osaka, Japan).

Lung epithelial cells, including airway ciliary cells, were isolated from the lungs as previously described (Ikeuchi et al., 2018). After mice were euthanized, the lungs were perfused with a nominally Ca²⁺-free solution (121 mM NaCl, 4.5 mM KCl, 25 mM NaHCO₃, 1 mM MgCl₂, 5 mM glucose and 5 mM HEPES, pH 7.4) containing heparin (20 units/ml) to clear blood by perfusion via the pulmonary artery, and the lungs with trachea and heart were removed from the mice. After removal, the lung cavity was washed with the nominally Ca²⁺-free solution and then with the control solution (121 mM NaCl, 4.5 mM KCl, 25 mM NaHCO₃, 1 mM MgCl₂, 1.5 mM CaCl₂, 5 mM glucose and 5 mM HEPES, pH 7.4). After washing, the lung cavity was digested with elastase (0.2 mg/ml) and DNase I (0.02 mg/ml; Sigma-Aldrich) for 40 min at 37°C. After elastase digestion, lungs were cut into small pieces using fine forceps in the control solution containing DNase I (0.02 mg/ml) and BSA (4%), and the control solution containing small lung pieces was filtered through 300 μm nylon mesh to separate cells from undigested tissues. The isolated lung cells were washed three times with centrifugation at 160 g for 5 min and were then resuspended in the control solution at 4°C.

The isolated lung cells containing airway ciliary cells attached to the coverslip were air-dried. The cells on the coverslip were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) for 30 min at room temperature and washed with ice-cold PBS containing 10 mM glycine (PBSG). Then, the cells were permeabilized with PBS containing 0.1% Triton X-100 for 15 min at room temperature and washed with PBSG, followed by incubation with 3% BSA in PBS for 60 min at room temperature. For immunostaining, the coverslip was incubated with primary antibody diluted with the Can Get Signal A solution (TOYOBO, Osaka, Japan) at 4°C overnight. Cells were washed with PBS containing 0.1% BSA to remove unbound antibodies. Finally, cells were stained with secondary antibody for 1 h at room temperature. After washing with PBS containing 0.1% BSA three times, cells were observed using confocal laser microscopy (FV10i; Olympus, Tokyo, Japan).

The mouse distal tracheas were fixed in cold methanol for 10 min at −20°C, and in HEPES-buffered saline (HBS, pH 7.5) for 7 min at room temperature, and then permeabilized with 0.2% Triton X-100 in HBS for 5 min at room temperature, followed by incubation in 1% BSA in PBS for 30 min at room temperature. Incubations with primary and secondary antibodies were performed at room temperature for 1 h each, as previously described (Herawati et al., 2016; Nakayama et al., 2021). Samples were observed using spinning-disk Olympus super-resolution microscopy (Olympus) and disk scanning unit (DSU) microscopy (Olympus). Super-resolution immunofluorescence images were obtained using a 1.6× conversion lens with either a scientific complementary metal–oxide–semiconductor (sCMOS) camera (ORCA-Flash 4.0 v2; Hamamatsu Photonics, Hamamatsu, Japan) or a charge-coupled device (CCD) camera (ORCA-R2; Hamamatsu Photonics).

Isolated lung cells including airway ciliary cells were collected as described above. The cells were lysed in RIPA buffer containing 150 mM NaCl, 1% Nonidet P40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris-HCl, pH 7.4 and protease inhibitor cocktail (cOmplete; Roche, Basel, Switzerland). After centrifugation at 3000 g for 10 min at 4°C to remove cell debris, the supernatants were collected. Protein (5–20 µg) was loaded into each lane for Laemmli SDS-polyacrylamide gel electrophoresis (10%) and then transferred to a polyvinylidene difluoride membrane. The membrane was blocked for 1 h using 5% milk powder in TBS-T (150 mM NaCl, 10 mM Tris-HCl, pH 8.5 and 0.1% Tween 20) and exposed to primary antibody diluted with Solution 1 (Can Get Signal; TOYOBO) at 4°C overnight. After a TBS-T rinse, the secondary antibody diluted with Solution 2 (Can Get Signal; TOYOBO) was applied to the membrane for 1 h at room temperature. After rinsing with TBS-T, antigen–antibody complexes were visualized with a chemiluminescence system (Immobilon; Merck Millipore, Burlington, MA).

After isolating the mouse trachea, the samples were fixed with 0.1% tannic acid, 2% formaldehyde and 2.5% glutaraldehyde in 100 mM HEPES buffer (pH 7.5) at 37°C for 1 h, followed by post fixation with 1% OsO₄ in 100 mM HEPES buffer (pH 7.5) on ice for 2 h, as previously described (Tsukita et al., 1980). The samples were dehydrated and embedded in Poly/Bed 812 (08791-500; Polysciences, Warrington, PA).

For SEM, observations were performed on an S-4800 scanning electron microscope (Hitachi, Tokyo, Japan) at an accelerating voltage of 5.0 kV. The lengths of the motile cilia in isolated tracheas were measured using ImageJ software (NIH, Bethesda, MD). The averages of percentage of ciliated cells were calculated from independent eight visual fields in tracheas from WT and Vil2^kd/kd mice.

For TEM, the samples were serially sectioned at either 50 nm or 70 nm and analyzed using a JEM-1400 plus microscope (JEOL, Tokyo, Japan). Colloidal gold particles (20 nm diameter) were deposited on each section, and the samples were observed at an acceleration voltage of 1 MeV (H-3000; Hitachi). Images were taken at 25,000× magnification from −60° to +66° at 2° intervals around a single-tilt axis and acquired with a slow scan CCD camera (model F415S; TVIPS GmbH, Bayern, Germany). The image calibration and 3D reconstructions of each series were performed using IMOD software (Kremer et al., 1996).

To analyze mucociliary transport, adult mouse tracheal samples were isolated and observed using a fluorescence microscope and a DSU microscope (BX53-DSU; Olympus). The flow of fluorescent beads (a 500-fold dilution of Fluoresbrite, 0.5 µm; Polysciences) in the mouse trachea was recorded at 25 ms/frame using a water immersion objective lens (LUMPlan FLN 60×; NA 1.00; WD, 2.0 mm; Olympus), an sCMOS camera (ORCA-Flash 4.0 v2; Hamamatsu Photonics), and a thermoplate (37°C; Tokai Hit, Fujinomiya, Japan). The mouse trachea was incubated with HBS (pH 7.5) containing 1 nM procaterol and fluorescent beads after preincubation with HBS without procaterol. Hardware was controlled using MetaMorph software (Molecular Devices, San Jose, CA). After subtracting images processed by Gaussian Blur from all acquired images in ImageJ, each bead was tracked and analyzed using the TrackMate plugin for Fiji (https://fiji.sc/). Mucociliary transport of the fluorescent beads was quantified using two indicators: speed and directionality. ‘Speed’ indicates the velocity of the flow of fluorescent beads (μm/s), and ‘directionality’ indicates the ratios of end-to-end distances against distances along the trajectory of the fluorescent beads in arbitrarily selected regions of isolated tracheal surfaces from WT and Vil2^kd/kd mice before and after the stimulation with procaterol.

The cells were set on a coverslip precoated with Cell-Tak (Becton Dickinson Labware, Bedford, MA) in a perfusion chamber (20 μl) mounted on an inverted light microscope (T-2000; NIKON, Tokyo, Japan) connected to a high-speed camera (IDP-Express R2000; Photron Ltd., Tokyo, Japan). The experiments were carried out at 37°C, because the CBF is temperature dependent (Delmotte and Sanderson, 2006; Lorenzo et al., 2008). The chamber was perfused with the control solution aerated with a gas mixture (95% O₂ and 5% CO₂ at 37°C) at a constant rate (200 μl/min). The ciliary cells were distinguished from other lung epithelial cells by their beating cilia. For measurements of CBF and CBA, video images were recorded for 2 s at 500 frames per s. Previous reports have already described the method used to measure CBF and CBA in detail (Komatani-Tamiya et al., 2012). Before the start of experiments, cells were perfused with the control solution for 5 min and then with test solution containing 1 nM procaterol or 100 μM IBMX. After the experiments, CBF and CBA were measured using an image analysis program (DippMotion V2D; DITECT, Tokyo, Japan). The ratios of CBF and CBA (CBF_t/CBF₀ and CBA_t/CBA₀, where the subscripts ‘0’ and ‘t’ indicate measurements made just before and at the time from the start of experiments, respectively; normalized CBF and CBA) were calculated to make comparisons across the experiments. Five CBF or CBA measurements made every 1 min during 5 min of control perfusion were averaged, and the average value was used as CBF₀ or CBA₀, respectively. For each coverslip, we selected a visual field with 1–2 cells or a cell block and measured their CBFs or CBAs. The ratios of CBF and CBA were calculated by averaging values obtained from 3–11 cells (n values indicate the number of cells).

We measured the intracellular cAMP level in isolated airway ciliary cells using a modified method described previously (Kogiso et al., 2017). The cells were incubated in the control solution containing 1 nM procaterol or 100 μM IBMX or not for 10 min at 37°C. The intracellular cAMP level was measured using a cAMP EIA kit (Cayman Chemical, Ann Arbor, MI). Protein concentrations were measured by BCA protein assay (Thermo Fisher Scientific, Waltham, MA). The measurements of cAMP levels were normalized to protein concentrations (μmol/g protein).

To measure the cell surface β₂AR expressions in isolated airway ciliary cells, a surface biotinylation assay was performed as described previously (Kawaguchi et al., 2018). The cells were collected as described above. Surface proteins were labeled with 2 mg/ml NHS-SS-biotin in HEPES-buffered solution (130 mM NaCl, 2.5 mM NaH₂PO₄, 4 mM KCl, 1.2 mM MgSO₄, 5.5 mM glucose, 2 mM CaCl₂ and 10 mM HEPES, pH 7.4) for 30 min at 4°C. Excess NHS-SS-biotin was quenched with 100 mM glycine in the HEPES-buffered solution containing 6 mM L-alanine and 1 mM Na citrate (HEPES++ solution). After washing with HEPES++ solution, cell suspensions were pelleted at 120 g for 10 min and lysed with RIPA buffer. Protein concentrations were measured by BCA protein assay. A total of 100 μg protein was incubated with streptavidin-coated agarose beads (Sigma-Aldrich) at 4°C overnight. Then, beads were washed twice with wash buffer [150 mM NaCl, 5 mM EDTA, 50 mM HEPES (pH 7.5) and 0.1% Triton X-100]. Then, proteins were extracted from the beads by heating for 15 min at 65°C in 60 μl of SDS loading buffer containing 50 mM DTT and 10% β-mercaptoethanol to cleave the disulfide bridge in sulfo-NHS-SS-biotin. These samples were analyzed by immunoblotting.

Endocytosis of β₂AR was evaluated by a surface biotinylation-based internalization assay. Biotinylation of surface proteins was performed as described above. After biotinylation, the cells were incubated at 4°C or 37°C for 1 h to allow endocytosis of β₂AR. Then, the cells were immediately cooled by ice-cold PBS, and treated with 50 mM sodium 2-mercaptoethanesulfonate (MesNa), which is a membrane-impermeable reducing agent, in a reducing buffer (125 mM NaCl, 10 mM HEPES, 1 mM MgCl₂, and 50 mM Tris-HCl, pH 8.0) to strip biotin from surface proteins. Excess MesNa was quenched with 25 mM iodoacetamide for 15 min at 4°C. The cells were lysed, and protein amounts were measured by BCA protein assay. Equal amounts of protein (200 μg) were incubated with streptavidin-agarose beads at 4°C overnight. Proteins were extracted with SDS loading buffer containing 10% β-mercaptoethanol, and analyzed by immunoblotting.

Data are expressed as the mean±s.e.m. Statistical significance between the means was assessed by one-way ANOVA, or two-tailed Student’s paired or unpaired t-test, as appropriate. For post hoc multiple comparison tests, a Tukey–Kramer test was used after one-way ANOVA. Differences were considered significant at P<0.05. The statistical significance is shown in the figures. The data in boxplots are presented as medians, 25th–75th percentiles, ranges and outliers.

We thank Professor Sachiko Tsukita and Dr Atsushi Tamura for Vil2^kd/kd mice and helpful discussions. We also thank Fumiko Takenaga for her technical assistance.

The peer review history is available online at https://journals.biologists.com/jcs/article-lookup/doi/10.1242/jcs.259201.