An autocrine ATP release mechanism regulates basal ciliary activity in airway epithelium

Abstract

Key points

• Extracellular ATP, in association with [Ca²⁺]_i regulation, is required to maintain basal ciliary beat frequency.

• Increasing extracellular ATP levels increases ciliary beating in airway epithelial cells, maintaining a sustained response by inducing the release of additional ATP.

• Extracellular ATP levels in the millimolar range, previously associated with pathophysiological conditions of the airway epithelium, produce a transient arrest of ciliary activity.

• The regulation of ciliary beat frequency is dependent on ATP release by hemichannels (connexin/pannexin) and P2X receptor activation, the blockage of which may even stop ciliary movement.

• The force exerted by cilia, measured by atomic force microscopy, is reduced following extracellular ATP hydrolysis. This result complements the current understanding of the ciliary beating regulatory mechanism, with special relevance to inflammatory diseases of the airway epithelium that affect mucociliary clearance.

Abstract

Extracellular nucleotides, including ATP, are locally released by the airway epithelium and stimulate ciliary activity in a [Ca²⁺]_i‐dependent manner after mechanical stimulation of ciliated cells. However, it is unclear whether the ATP released is involved in regulating basal ciliary activity and mediating changes in ciliary activity in response to chemical stimulation. In the present study, we evaluated ciliary beat frequency (CBF) and ciliary beating forces in primary cultures from mouse tracheal epithelium, using videomicroscopy and atomic force microscopy (AFM), respectively. Extracellular ATP levels and [Ca²⁺]_i were measured by luminometric and fluorimetric assays, respectively. Uptake of ethidium bromide was measured to evaluate hemichannel functionality. We show that hydrolysis of constitutive extracellular ATP levels with apyrase (50 U ml⁻¹) reduced basal CBF by 45% and ciliary force by 67%. The apyrase effect on CBF was potentiated by carbenoxolone, a hemichannel inhibitor, and oxidized ATP, an antagonist used to block P2X7 receptors, which reduced basal CBF by 85%. Additionally, increasing extracellular ATP levels (0.1–100 μm) increased CBF, maintaining a sustained response that was suppressed in the presence of carbenoxolone. We also show that high levels of ATP (1 mm), associated with inflammatory conditions, lowered basal CBF by reducing [Ca²⁺]_i and hemichannel functionality. In summary, we provide evidence indicating that airway epithelium ATP release is the molecular autocrine mechanism regulating basal ciliary activity and is also the mediator of the ciliary response to chemical stimulation.

Key points

• Extracellular ATP, in association with [Ca²⁺]_i regulation, is required to maintain basal ciliary beat frequency.

• Increasing extracellular ATP levels increases ciliary beating in airway epithelial cells, maintaining a sustained response by inducing the release of additional ATP.

• Extracellular ATP levels in the millimolar range, previously associated with pathophysiological conditions of the airway epithelium, produce a transient arrest of ciliary activity.

• The regulation of ciliary beat frequency is dependent on ATP release by hemichannels (connexin/pannexin) and P2X receptor activation, the blockage of which may even stop ciliary movement.

• The force exerted by cilia, measured by atomic force microscopy, is reduced following extracellular ATP hydrolysis. This result complements the current understanding of the ciliary beating regulatory mechanism, with special relevance to inflammatory diseases of the airway epithelium that affect mucociliary clearance.

Abbreviation

8‐SPT

8‐(p‐sulphophenyl) theophylline

ADO

adenosine

AFM

atomic force microscopy

APY

apyrase

CBF

ciliary beat frequency

CBX

carbenoxolone

Cx

connexin

HBSS

Hank's balanced salt solution

HC

hemichannel

MCC

mucociliary clearance

MTECs

mouse tracheal epithelial cells

oATP

oxidized ATP

P2X7R

P2X7 receptor

P2Y2R

P2Y2 receptor

Panx

pannexin

Introduction

The continuous influx of air towards the lungs is accompanied by the entry of microorganisms and contaminants that can cause damage or respiratory infections. Ciliated and secretory cells are structural components of the airway epithelium and are functionally responsible for mucociliary clearance (MCC); the basic defence mechanism of the airway. MCC displaces the mucus layer that traps harmful molecules by spontaneous synchronic ciliary movement (Seybold et al. 1990). This process requires a finely‐tuned ciliary beat frequency (CBF), which in turn is translated into a ciliary beating force. Ciliary forces have been directly measured in the frog oesophagus using atomic force microscopy (AFM) (Teff et al. 2007, 2008). However, no studies currently exist concerning the overall effect of ciliary forces in mammalian airway cells under physiological conditions. This information would be valuable for understanding the mechanical effect of CBF on MCC.

Ciliary movement is produced by the sliding of microtubules inside the cilium. The energy required for this process to occur is provided via the hydrolysis of ATP by dynein ATPase (Satir, 1980). In isolated cilia from cell membranes extracted with detergent, spontaneous movement is partially recovered in the presence of ATP and Mg²⁺ (Naitoh & Kaneko, 1972; Navarrette et al. 2012); however, this does not account for all the CBF regulatory mechanisms, considering that basal CBF varies from 6 to 15 Hz depending on the species and the anatomic location of the epithelium (Nakahari, 2007).

Studies in unicellular organisms such as paramecium showed that membrane‐regulated [Ca²⁺]_i changes control ciliary activity. The removal of extracellular Ca²⁺ with EGTA stops ciliary activity and subsequent injection of Ca²⁺ reactivates cilia movement and paramecium locomotion. In addition, increased [Ca²⁺]_i changes the direction of ciliary beating and increases ciliary frequency. In the absence of stimulation, Ca²⁺ that slowly leaks into the cell is pumped out at a similar steady‐state rate (Eckert & Murakami, 1972), maintaining ciliary activity. In mammalian cells, where the evolution has led to unidirectional movement, it is unknown whether [Ca²⁺]_i regulation establishes the basal CBF.

ATP, as a chemical signalling molecule, is characterized as a powerful activator of CBF (Villalón et al. 1989; Weiss et al. 1992; Gheber & Priel, 1994; Morales et al. 2000; Morse et al. 2001; Zhang & Sanderson, 2003; Barrera et al. 2005; Barrera et al. 2007; Zhao et al. 2012; González et al. 2013), whose effect is initially mediated by the P2Y2 receptor (P2Y2R) and Ca²⁺ mobilization from intracellular Ca²⁺ stores (Gheber et al. 1995; Morales et al. 2000; Morse et al. 2001) to induce a fast initial increase of CBF followed by a sustained activity that gradually decays. Amplification or maintenance of the response is dependent on Ca²⁺ influx (Barrera et al. 2004) via ionotropic P2X cilia receptors (Ma et al. 1999), whose ATP‐activated currents have properties unique to P2X7 receptor (P2X7R) channels (Kawakami et al. 2004; Ma et al. 2006; Surprenant & North, 2009). Additionally, the sustained component depends on adenosine (ADO), provided by extracellular ATP and hydrolysis of its metabolites by epithelial ecto‐5′‐nucleotidases (Lazarowski et al. 2004; Zuo et al. 2008), which increases intracellular cAMP levels by metabotropic A2b receptor activation (Morse et al. 2001).

A broad range of extracellular ATP concentrations were measured in airway epithelia under several conditions. Under resting conditions, human nasal and bronchial epithelial cultures have low extracellular ATP levels (1–5 nm) (Lazarowski et al. 2004), which increase by up to 1500 nm after mechanical stress (Yegutkin et al. 2000; Zhao et al. 2012). In addtion, extracellular ATP accumulates in cultures or bronchoalveolar lavage fluid of patients with chronic inflammation, such as cystic fibrosis or asthma, therefore increasing its concentration (Donaldson et al. 2000; Idzko et al. 2007; Esther et al. 2011).

ATP, stored at millimolar concentrations within cells, is released by different mechanisms, such as disrupting plasma membrane integrity (Rock & Kono, 2008) and under non‐cytolytic conditions. In the airway epithelium, ATP can be released by exocytosis of mucin granules from goblet cells (Okada et al. 2011; Sesma et al. 2013) and by a conductive mechanism mediated by pannexin (Panx) or connexin (Cx) hemichannels (HCs) (Lazarowski, 2012). Pharmacological studies have shown that Panx1 and Cx43 are the most relevant channels for ATP release. ATP is released via Panx1 in astrocytes (Orellana et al. 2013), erythrocytes (Montalbetti et al. 2011) and skeletal muscle fibre (Riquelme et al. 2013). Cx43 HCs release ATP in melanoma cells (Ohshima et al. 2012) and astrocyte cultures (Stehberg et al. 2012). Panx1 mediates hypotonic stress‐activated ATP release from airway epithelial cells of mouse trachea (Seminario‐Vidal et al. 2011). However, the effect of ATP released via HCs on basal ciliary activity and increased CBF induced by ATP is unknown.

In the present study, we show that the ecto‐hydrolysis of constitutive extracellular ATP reduced basal CBF, an effect that was potentiated by blocking HCs and P2XR. The maintenance of ATP‐dependent CBF increase was reduced by blocking ATP release via HCs, suggesting an autocrine ATP release mechanism for regulating basal and stimulating CBF.

Methods

Ethical approval

All procedures were performed in accordance with the ethical principles of The Journal of Physiology and were approved by the Bioethics Committee at Pontificia Universidad Católica de Chile. BALB/c mice (male, weighing between 25–30 g, 6–8 weeks old), supplied by the Research Animal Facility at Pontificia Universidad Católica de Chile, were maintained on a standard chow diet with water available ad libitum and under a controlled environment (12 : 12 h light/dark cycle at 23–25°C). Animals were killed by cervical dislocation and tracheas were removed to prepare cell cultures.

Primary cultures from mouse tracheal epithelial cells (MTECs)

Cell cultures were performed as descibed by Gonzalez et al. (2013). Mouse tracheas, placed in DMEM/F12 medium with antibiotics (10 μg ml⁻¹ streptomycin, 100 U ml⁻¹ penicillin G, 0.125 μg⁻¹ml⁻¹ amphotericin B) were washed and longitudinally cut to expose the epithelium. Then, tracheas were incubated in pronase 0.05% w v⁻¹ for 30 min in a CO₂ chamber at 37°C. Tissue was then transferred into a Petri dish with DMEM‐F12 and 5% foetal bovine serum. The epithelium was mechanically removed with forceps and scissors, cut into 2–4 mm² pieces and soaked in NHS (137 mm NaCl; 5.09 mm KCl; 1.14 mm Na₂HPO₄ × 2H₂O; 0.18 mm KH₂PO₄; 0.923 mm MgCl₂ × 6H₂O; 0.91 mm CaCl₂ × 2H₂O; 4.07 mm NaHCO₃; 21.5 mm glucose, pH 7.4), supplemented with 1% vitamins, 1% essential amino acids, 1% non‐essential amino acids, 1% pyruvate, and antibiotics: 0.2 mg ml⁻¹ neomycin and 0.12 mg ml⁻¹ penicillin. The pieces of mouse trachea epithelia were placed on a coverslip, pre‐treated with 0.1% gelatin, in Rose chambers. Explants were covered with a sterile dialysis membrane (Spectra/Por*2, MWCO 12–14 000; #25218‐468; VWR Scientific, Radnor, PA, USA) pre‐hydrated with NHS medium. Rose chambers were filled with 2 ml of NHS medium; containing 10% heat inactivated horse serum (pH 7.2–7.4) and maintained in an incubator at 37°C. The culture medium was replaced every 48 h. Experiments were performed on the fifth day of culture, after a monolayer of ciliated cells and spontaneous ciliary activity was observed.

CBF measurements

CBF was recorded in the cultures at 35°C, with a high‐speed video camera (scA640; Basler AG, Ahrensburg, Germany) (100 frames s⁻¹) mounted on a TE200 (40×) microscope (Nikon, Tokyo, Japan). The video (2.26 s long) was captured and files were reloaded and analysed with the Sisson‐Ammons Video Analysis (SAVA) software (Ammons Engineering, Clio, MI, USA).

The software performs a power spectrum analysis (using a fast Fourier transform) and displays the number of readings vs. frequencies to determine the dominant frequency in the whole field or in a region of interest (Sisson et al. 2003). Whole field analysis was used, except when debris interfered with the measurement. The SAVA system is unable to determine frequencies lower than 2 Hz; therefore, the lack of cilia movement was determined visually. Recordings with less than 1000 points of movement and higher than 6000 points of movements (corresponding to < 3 and > 20 cells) were excluded.

Prior to CBF measurements, primary cultures were washed three times with warmed Hank's balanced salt solution (HBSS). Cultures were maintained for 15 min at 35°C to stabilize basal CBF, which was then recorded for 5 min to determine the baseline (0% CBF). Finally, different treatments were added to the culture medium and CBF was measured for the next 10 or 20 min.

AFM

AFM MFP‐3D (Asylum Research, Goleta, CA, USA) with an iDrive (BL‐TR400PB, k = 0.01–0.05 N m⁻¹) module was used to acquire topological images and obtain single force curves, using the AFM tip as a motion sensor. Tip calibration and thermal tuning was performed in air and liquid medium prior to the experiment. Images were obtained in AC mode, with a scan rate of 0.22 Hz. Single force was performed in contact mode, with a trigger point of 0.5 V. The AFM probe was directed into the sample and stopped automatically by the system. The probe was then retracted away from the cells. CBF was calculated by fast Fourier transform power spectrum of the deflection recorded as a function of time. Ciliary force was calculated as:

$$F=kA$$

where the spring constant of the tip (k) was multiplied by the amplitude of tip deflection (A). The AFM experiments were performed at room temperature (Cartagena & Raman, 2014).

Quantification of extracellular ATP levels

Using a luminometric technique, cultures were washed with HBSS medium three times. Then, cultures were maintained for 20 min at 35°C. A 100 μl extracellular medium was used directly or diluted in HBSS medium. Immediately, the sample was boiled for 1 min to prevent ATP degradation by ecto‐nucleotidases (Lazarowski et al. 2004). Samples were maintained at −70°C prior to measurements. A calibration standard curve (between 0.05 and 20 nm) was performed in each experiment. Luciferin/luciferase (100 μl) was added to each well using a microinjector and luminescence was recorded for 10 s (Luminoskan Ascent 2.5; Thermo Labsystems, Philadelphia, PA, USA). Because primary cultures are not confluent, the area occupied by the cells on the glass was calculated and ATP levels were expressed as pmol cm^–2.

Measurements of Ca²⁺ _i

Changes in Ca²⁺ _i (intracellular calcium) levels were evaluated using a fluorometric technique as described previously (Barrera et al. 2004). MTECs were loaded with 1.5 μm Fura‐2AM for 1 h at 37°C. Cultures were washed three times and stabilized for 30 min. Experiments were performed at 25°C. Fluorescence images were acquired at an excitation wavelength of 340 and 380 nm and detected at 510 nm with an Olympus fluorescence microscope coupled to an image acquisition system (MetaFluor, version 6.1; Universal Imaging Corporation, Bedford Hills, CA, USA).

Dye uptake rate

MTECs were loaded in HBSS containing 5 μm ethidium bromide and fluorescence intensity was recorded in selected regions of interest. Images were captured every 15 s using an Olympus fluorescence microscope coupled to an image acquisition system (MetaFluor, version 6.1) that was used for offline image analysis and fluorescence quantification. Slopes of dye uptake were calculated and expressed as AU min^–1.

Statistical analysis

Data are expressed as the mean ± SEM or mean ± SD, as indicated; n represents the number of cell cultures tested from at least three different animals. Statistical analysis was performed with Prism, version 6 (GraphPad Software Inc., San Diego, CA, USA) or OriginPro, version 9.2 (OriginLab Corp., Northampton, MA, USA). A t test or one‐way ANOVA followed by a Tukey's post hoc test was performed. Changes in CBF and dye uptake rate were analysed after arcsine transformation. P < 0.05 was considered statistically significant.

Results

Effect of modifying extracellular ATP levels on CBF

To determine the dependence of CBF on a broad range of extracellular ATP levels, we used MTECs that have a basal CBF between 10.4 and 16.2 Hz (mean ± SD: 13.3 ± 2.9 Hz) (n = 858 whole fields; 85 cultures; 43 tracheas).

To reduce basal extracelullar ATP levels, MTECs were treated with apyrase (APY), an ectonucleotidase that hydrolyzes extracellular ATP (Zimmermann & Braun, 1996). APY reduced CBF in a concentration‐dependent manner, reaching a maximum reduction of 53% from baseline with 100 U ml⁻¹ of APY (baseline of 12.3 ± 0.3 Hz was reduced to 5.7 ± 0.3 Hz with apyrase) (Fig. 1 A and B) supporting the involvement of basal extracellular ATP levels on the regulation of basal ciliary activity. Similar observations were obtained in primary cultures from paediatric human adenoid tissue biopsies, where CBF reductions of 6.1 ± 3.2%, 19.8 ± 3.7%, 68.8 ± 10.6% and 71.8 ± 4.0% were observed after application of 5, 10, 20 and 50 U ml⁻¹ APY, respectively (n = 4, data not shown).

Figure 1. Effect of modifying extracellular ATP levels on CBF.

A, time course of the CBF change in ciliated cells treated with APY. Basal CBF was measured over 5 min followed by APY application. B, maximum CBF decrease induced by APY followed by a concentration‐dependent effect (^* P < 0.05 vs. 0; #P < 0.05 vs. 20 U ml⁻¹). C, quantification of extracellular ATP levels in culture medium samples after APY (50 U ml⁻¹) was added to the culture. Extracellular ATP levels were significantly reduced at both 1 and 20 min after APY application (^* P < 0.05 vs. Vehicle). D, maximum percentage CBF change was measured following the addition of ATP at different concentrations. A concentration‐dependent effect was observed between 1 nm and 10 μm ATP, whereas an increase of ATP concentration of up to 0.5 and 1 mm reduced basal CBF. Insert: time courses of the CBF response to ATP and the non‐hydrolysable ATP analogue, ATP γS (n = representative curves of three to five experiments). Numbers in bars indicate the number of experiments performed.

Basal extracellular ATP levels measured in undisturbed cultures were not significantly different after 1 and 20 min; however, APY (50 U ml⁻¹) significantly lowered extracellular ATP levels after 1 and 20 min of treatment (Fig. 1 C). These data confirm that APY reduced extracellular ATP levels by at least 50% after 1 min of treatment.

To increase extracellular ATP levels, cell cultures were exposed to exogenous ATP application (1–100 μm). As expected, this induced a rapid increase in CBF, followed by sustained response that did not decay during the recorded time (Fig. 1 D, insert). Non‐linear fit analysis of CBF changes induced by ATP (0.01–10 μm) showed an EC₅₀ of 0.04 μm (Fig. 1 D). High levels of ATP (0.5 and 1 mm) and ATP‐γS (1 mm), a non‐hydrolyzed ATP analogue, reduced CBF by 45.0% (n = 6) and 78.5 (n = 4) compared to baseline, respectively (Fig. 1 D). The transient effect of high ATP compared to the sustained effect of ATP‐γS on CBF demonstrated ATP metabolism by ectonucleotidase enzymes present in the airway epithelium. These results show a dual ATP response on CBF. The increase in extracellular ATP levels produced an increased CBF, whereas high levels of ATP (> 0.5 mm) triggered decreased CBF.

Effect of extracellular ATP hydrolysis on CBF and ciliary forces, measured by AFM

Using the AFM tip as a motion sensor, we obtained an oscillatory deflection pattern that correspond to a ciliated cell with a calculated CBF of 11.5 ± 4.3 Hz (mean ± SD, n = 30) (Fig. 2 A, insert), which is similar to the frequency range detected by the videomicroscopy technique. The calculated ciliary force increased as the tip approached the cell surface (Fig. 2 A). APY (100 U ml⁻¹) reduced both CBF by 41% and ciliary forces by 67% (Fig. 2 B), which strongly suggests that cilia would be less active for MCC.

Figure 2. CBF and ciliary forces measured by AFM.

A, measurement of ciliary force as the tip approaches the cell surface. Ciliary force increased as the tip approached the surface below the contact point (0 μm) (mean ± SE: 182 ± 34.5 pN at 1 μm; 307 ± 109.4 pN at 1.5 μm; 400.0 ± 66.7 pN at 2 μm; n = 3). Insert: CBF was calculated from the cantilever deflection recording of different cells using FFT power spectra. Calculated CBF was 11.5 ± 4.3 Hz (mean ± SD, n = 30). B, ciliary force and CBF were calculated before and after application of APY (100 U ml⁻¹). APY reduced CBF by 41% and ciliary forces by 67% (n = 4). Time course of deflection recording in a ciliated cell, prior to (Ba) and following (Bb) APY application.

Reduced CBF correlates with reduced Ca²⁺ _i

To determine whether a reduction in basal CBF correlates with changes in Ca²⁺ _i levels, this second messenger was measured in the presence of APY and ATP 1 mm. Figure 3 A shows that 10 U ml⁻¹ APY, as well as 1 mm ATP, reduced Ca²⁺ _ilevels during the recording time (Fig. 3 A). These results indicate that reduced basal CBF is associated with a reduced Ca²⁺ _ilevels.

Figure 3. HCs and P2XR inhibition decreases basal CBF.

A, time course of Ca²⁺ _i decrease induced by APY (10 U ml⁻¹) (black line) or ATP 1 mm (grey line). B, reduction of dye uptake rate following ATP 1 mm or CBX treatment (^*, P < 0.05). Insert: representative time course of dye uptake in cultures treated with CBX. C, CBF decrease was measured after treatment with oATP, CBX and APY. These inhibitors were added to the culture alone or together. Coapplication of oATP and CBX significantly reduced basal CBF compared to oATP and CBX applied separately. Simultaneous application of APY, oATP and CBX further reduced basal CBF (^* P < 0.05 vs. Vehicle; ^# P < 0.05 vs. CBX; °P < 0.05 vs. oATP; &P < 0.05 vs. APY). Insert: time course of the Ca²⁺ _i level decrease induced by CBX plus oATP treatment. D, quantification of extracellular ATP levels 5 min after the addition of vehicle, oATP (100 μm), CBX (100 μm) and APY (APY) (50 U ml⁻¹). No significant differences were observed with oATP, CBX and oATP plus CBX. However, concomitant treatment with APY, oATP and CBX significantly decreased extracellular ATP levels (^* P < 0.05 vs. Vehicle). E, maximum percentage CBF decrease was measured after treatment with the purinergic receptor antagonist (SUR) and HC inhibitors (Gd³⁺, La³⁺ and PROB). Basal CBF was recorded over 5 min, followed by the addition of blockers to the cultures and CBF was then measured for 10 min. Gd³⁺ and La³⁺ transiently reduced CBF (^* P < 0.05 vs. Vehicle). F, basal CBF was measured in cultures incubated with 8‐SPT (100 μm), an adenosine receptor antagonist, and APY (50 U ml⁻¹). Once applied together, 8‐SPT was added to the culture 5 min before APY (^* P < 0.05 vs. 8‐SPT, ^# P < 0.05 vs. Vehicle/APY, n = 4). Although 8‐SPT did not modify basal CBF, this potentiates the CBF decrease induced by APY. Numbers in the bars indicate the number of experiments performed.

Contribution of HCs and P2X on basal CBF regulation

To corroborate the functionality of HCs in cell cultures, we evaluated the uptake of ethidium bromide from the extracellular milieu in control solution, which has been used as an indicator of HC opening (Contreras et al. 2003). At resting conditions, the rate of ethidium bromide uptake was reduced after carbenoxolone (CBX) (100 μm) treatment, an inhibitor of Cx and Panx1 HCs (Fig. 3 B, insert). ATP (1 mm) also reduced the dye uptake rate (Fig. 3 B), suggesting that the CBF reduction after high ATP stimulation could be explained by closing HCs.

To check the participation of P2X7R and HCs in the regulation of basal CBF, oxidized ATP (oATP; a P2X7R antagonist) and CBX were added to the culture. oATP (100 μm) had no effect on basal CBF (Fig. 3 C), whereas CBX (100 μm) decreased transiently basal CBF by 24.9 ± 6.8% over 15 s (Fig. 3 C). oATP and CBX, added simultaneously to MTECs, decreased CBF by 57.5 ± 3.0% at 5 min, with a return to baseline after 20 min (Fig. 3 C). Concomitant treatment with CBX, oATP and APY further reduced basal CBF by 85.2 ± 4.8% (n = 10), which is a significant decrease compared to oATP, CBX and APY added separately (Fig. 3 C). It is important to highlight that, in some cultures, when oATP, CBX and APY were added together, ciliary activity was reduced completely (100%) for ∼2 min. Furthermore, the blocking of HCs with CBX did not modify Ca²⁺ _i levels (Fig. 3 C, insert), and treatment with oATP produced a transient fast increase in Ca²⁺ _i levels (data not shown); however, concomitant treatment with oATP and CBX to block HCs and P2XR together resulted in a reduction in Ca²⁺ _i levels (Fig. 3 C, insert) in accordance with the CBF reduction. These results suggest that extracellular ATP, HCs and/or P2XR are necessary to maintain basal ciliary activity in ciliated epithelia and indicate a relationship between P2XR and HCs associated with Ca²⁺ _i that regulates basal CBF.

Extracellular ATP levels in cell culture meadium treated with CBX and oATP were measured to determine the involvement of HCs and P2XR. Extracellular ATP levels measured 5 min following the addition of oATP, CBX (Fig. 3 D) or oATP plus CBX (data not shown) were not significantly different compared to vehicle; however, a slight reduction with CBX and high variability with oATP plus CBX were observed (19.9 ± 9.5 pmol cm⁻², n = 11, for vehicle; 24.7 ± 4.5 pmol cm⁻², n = 4, for oATP; 8.4 ± 3.6 pmol cm⁻², n = 4, for CBX; and 95.5 ± 56.5 pmol cm⁻², n = 6, for oATP plus CBX). When APY, oATP and CBX were added simultaneously, extracellular ATP levels were significantly reduced (Fig. 3 D). Extracellular ATP levels were also measured 10 and 20 min after the addition of oATP, CBX and APY, although no differences were observed compared to the 5 min treatments (data not shown).

Suramin (200 μm), a non‐specific antagonist of purinergic receptors, had no effect on basal CBF (Fig. 3 E). A similar response was observed with KN‐62 (10 μm), which inhibits P2X7R (data not shown). Gd³⁺ and La³⁺ (200 μm), blockers of Ca²⁺ permeable channels including Cx HCs (Garré et al. 2010), decreased significantly CBF, reaching a maximum decay at 1 min (Fig. 3 E). Neither CBX, at a concentration that only inhibits panx1 (1 and 10 μm, data not shown), nor probenecid (PROB) (1 mm), a panx1 inhibitor (Hayoz et al. 2012) (Fig. 3 E), had any effect on basal CBF. These results suggest that Cx HCs participate in the mechanism controlling ATP release required to maintain basal CBF.

To determine the contribution of ADO, a metabolite derived from extracellular ATP hydrolysis, on basal CBF, 8‐(p‐sulphophenyl) theophylline (8‐SPT) (100 μm), an ADO receptor antagonist (Garcia et al. 2013) was added to the cultures. 8‐SPT did not modify basal CBF; however, pre‐incubation with 8‐SPT (100 μm) for 5 min followed by APY (50 U ml⁻¹) showed a larger basal CBF reduction than cultures incubated with APY alone (Fig. 3 F). This result suggests that ATP metabolites, similar to ADO, might also contribute to maintaining basal CBF.

ATP release mediates the CBF increase induced by an exogenous ATP stimulus

Cultures were pre‐incubated with HCs and P2X7R blockers to evaluate the CBF response to ATP. Because some of these blockers effected basal CBF, ATP was added to the culture medium only when the basal CBF was re‐established and remained stable for 2 min. oATP (10 and 100 μm) and CBX (10 μm) did not modify the CBF increase induced by ATP (10 μm); however, CBX (100 μm) inhibited the CBF increase induced by ATP (Fig. 4 A). This inhibition was also observed after the application of PROB (1 mm). Pre‐incubation with oATP (10 μm) plus CBX (10 μm) did not modify the CBF response induced by ATP; however, increasing each blocker concentration up to 100 μm in the presence of ATP triggered a reduction of basal CBF levels (Fig. 4 A). These results suggest that Cx and Panx1 HCs are required for the CBF response induced by ATP.

Figure 4. The HC blocker, CBX, affects the autocrine release mechanism and reduces the CBF increase and extracellular ATP levels following ATP addition.

A, CBF change was measured following the addition of ATP (10 μm) to cultures pre‐incubated with oATP, CBX and PROB. Increased CBF induced by ATP was prevented in cultures pre‐incubated with 100 μm CBX. In addition, ATP reduced CBF basal levels in cultures pre‐incubated with 100 μm oATP and 100 μm CBX (^* P < 0.05 vs. Vehicle; #P < 0.05 vs. Vehicle/ATP). B, time course of the ATP (1 μm) effect on CBF in the presence of oATP (100 μm). oATP significantly reduced the sustained component of the ATP effect on CBF (^* P < 0.05). C, time course of the ATP (1 μm) effect on CBF in the presence of CBX (100 μm). CBX transiently reduced the ATP‐triggered CBF increase at 4 min (^* P > 0.05). D, extracellular ATP levels measured at 4 or 5 min in the absence (left axis) or after adding ATP (right axis) in cultures treated with CBX. CBX decreased extracellular ATP levels following the addition of ATP (^* P < 0.05). ATP levels following the addition of vehicle or CBX alone were 0.0199 ± 0.0095 nmol cm⁻² and 0.0084 ± 0.0036 nmol cm⁻² at 5 min, respectively. Numbers in or under the bars indicate the number of experiments performed.

To determine whether P2X7‐R modified the sustained component of the CBF increase induced by ATP, cultures were stimulated with ATP (1 μm) for 1 min and then treated with oATP (100 μm) or vehicle. oATP induced a faster fall of the CBF increase triggered by ATP compared to the control and significant differences were observed 2 min after addition of ATP (Fig. 4 B). CBX (100 μm) also triggered a faster decay to the ATP induced CBF increase compared to controls, with maximum differences being observed 4 min after addition of ATP (Fig. 4 C). CBX reduced extracellular ATP levels at 4 min (Fig. 4 D) and 20 min after ATP stimulation (7.7 ± 0.9 nmol cm⁻² for vehicle vs. 4.9 ± 0.6 nmol cm⁻² for CBX, P < 0.05, n = 6) (data not shown), suggesting that intracellular ATP is released by HCs following an exogenous ATP stimulus. Furthermore, the reduction percentage of ATP levels by CBX under the basal condition was similar under ATP‐stimulated conditions (58% and 63%, respectively) (Fig. 4 D), suggesting that HCs participate in ATP release to control basal and chemically stimulated CBF. Interestingly, extracellular ATP levels in cultures following the addition of ATP (1 μm) were 1.01 ± 0,09 μm at 4 min and 1.12 ± 0.057 μm at 20 min (n = 6), indicating that extracellular ATP levels are stable over the time of the experiment. These results suggest that exogenous ATP induces P2XR activation and ATP release via HCs to participate in maintaining the CBF increase.

Discussion

In the present study, with the aim of establishing a relationship between extracellular ATP levels and basal CBF in MTECs, we modified extracellular ATP levels either using a commercial ecto‐nucleotidase APY or by adding exogenous ATP to the culture.

APY decreased basal CBF in a concentration‐dependent manner, which was corroborated by measuring ATP levels in the culture medium, observing a reduction of basal extracellular ATP levels by ∼50% at 1 min. Extracellular ATP levels measured under resting conditions (∼20 pmol cm⁻² or 30 nm) were similar to those previously reported in primary cultures from human respiratory epithelia, suggesting that ATP is released constitutively (Lazarowski et al. 2004). Furthermore, the reduction of basal CBF following APY treatment may have important implications on MCC velocity because this can reach up to 56% after an increase in CBF of only 16% (Seybold et al. 1990). In addition, Kur & Newman (2014) showed that APY dilated retinal arterioles under control conditions, demonstrating the importance of ATP to the control of basal vascular tone. Using AFM, we measured CBF in a similar range by videomicroscopy. Furthermore, we simultaneously measured ciliary forces corresponding to ∼750 pN under resting conditions, a value of similar magnitude to that reported in frog oesophagus epithelia via AFM (210 pN) (Teff et al. 2007). Our results showed that APY reduced both airway epithelium CBF and ciliary forces, which would probably produce a decrease in the number of active cilia or an overall reduction of the cilia stroke momentum, which, as a consequence, would affect the MCC defence mechanism. Altogether, these results indicate that constitutive levels of extracellular ATP are required to maintain basal ciliary activity under physiological conditions.

We observed that the addition of ATP produced a dual effect in the airway epithelium. Although concentrations between 10 nm and 100 μm of ATP increased CBF, higher ATP concentrations (0.5 and 1 mm) reduced it. Under airway inflammatory conditions, extracellular ATP levels are elevated, which, in conjunction with the reduced basal CBF observed in our protocol, could explain the impaired MCC (Donaldson et al. 2000). Furthermore, the prolonged effect of ATP γS (compared to ATP) on CBF suggests that ecto‐nucleotidases act as regulators of extracellular nucleotide levels.

An increased [Ca²⁺]_i is recurrently associated with increased CBF induced by extracellular ATP (Villalón et al. 1989), an observation that supports our results showing that [Ca²⁺]_i is required to maintain basal CBF.

To establish the contribution of HCs in the regulation of basal CBF, we measured dye uptake rate in cultured cells. Dye uptake increased under basal conditions, indicating that resting ciliated cells have functional HCs. Furthermore, dye uptake rate was reduced with ATP 1 mm and also after blocking HCs with CBX, suggesting that high ATP levels induce HC closure. A similar interpretation was previously suggested by Dubyak (2009), who proposed negative feedback closure of Panx1 channels when extracellular ATP levels became too high in oocytes expressing Panx1. PROB, which inhibits Panx1 but not Cx HCs (Ransford et al. 2009; Hayoz et al. 2012), had no effect on basal CBF; however, CBX reduced CBF by 25%. Gd³⁺ and La³⁺, when used as inhibitors of Ca²⁺ channels, including Cx HCs (Lipski et al. 2006; Gkoumassi et al. 2009; Garré et al. 2010), caused a decrease in CBF by ∼50%, suggesting the participation of Cx HCs in the regulation of basal CBF. Cx23, Cx43 and Cx46 are expressed in tracheal epithelial cultures (Yeh et al. 2003; Isakson et al. 2006); however, because Cx43 is the main Cx related to ATP release (Ohshima et al. 2012; Stehberg et al. 2012), we propose that it is involved the regulation of basal CBF. The transient CBF response following CBX treatment suggests that a compensatory mechanism regulates CBF, where Cx partially controls CBF by ATP release or Ca²⁺ entry. When Cx is blocked, other channels could be activated to regulate CBF, such as TRPV4, which contributes to an ATP‐induced increase in CBF via Ca²⁺ entry (Lorenzo et al. 2008).

Basal extracellular ATP levels did not change following the addition of oATP and did not show a significant reduction with CBX. Nevertheless, extracellular ATP levels were significantly decreased only when APY, oATP and CBX were added to the culture at the same time, in concordance with the maximum reduction of basal CBF. It is possible that the ATP released via HCs and Ca²⁺ entry via P2XR represent a local mechanism in cilium proximity that controls basal CBF. Previously, studies indicated that P2X7R activation is required for the sustained component of ATP‐induced CBF increase (Ma et al. 1999; Surprenant & North, 2009, 2006; Li et al. 2010) and that P2X7R could be physicaly associated with Panx1 and Cx43, whereas P2X7R activation by extracellular ATP promotes the gradual opening of HCs, leading to ATP release (Faria et al. 2005; Suadicani et al. 2006). The results of the present study showed that concomitant blocking of HCs and P2XR reduces Ca²⁺ _i levels, in accordance with the possible mechanisms described for the interaction between P2X7R and HCs that involves Ca²⁺ _i movements (D´hondt et al. 2011). Further studies are necessary to establish a potential structural relationship between P2X7R and HCs on Ca²⁺ _i and its contribution to basal ciliary activity.

Extracellular ADO, derived from ATP hydrolysis by ecto‐nucleotidases expressed in the respiratory epithelium such as ecto‐5′‐nucleotidases (NTPDase3, CD39L3‐, E‐NPP) and non‐specific alkaline phosphatases (Lazarowski et al. 2004; Zuo et al. 2008), increased CBF (Morse et al. 2001; Zhang et al. 2004), thus activating A2bR (Morales et al. 2000; Morse et al. 2001). Adding APY to cultures preincubated with 8‐SPT reduced basal CBF to a greater extent than APY alone. Previously, we showed that the concomitant addition of ATP and ADO to oviductal epithelial cultures increases CBF in concentrations that do not have a significant effect by themselves, which suggests the existence of a synergistic mechanism (Morales et al. 2000), where cross‐talk between transduction pathways, associated with inositol trisphosphate, Ca²⁺, protein kinase C, cAMP and protein kinase A, produce an increased CBF response (Barrera et al. 2004, 200, 2007). Furthermore, in respiratory epithelial cells, stimulation with ATP and adenosine resulted in the sustained stimulation of CBF (Morse et al. 2001). The results of the present study suggest that extracellular ATP participates in the control mechanism of basal CBF both directly and after its hydrolysis to ADO.

The CBF increase induced by an exogenous ATP stimulus was significantly reduced with CBX or PROB, suggesting that Cx and Panx HCs participate in the regulation of this response; in particular, Panx1, which has already been detected in respiratory epithelial cells (Bruzzone et al. 2003; Ohbuchi et al. 2013). We have previously demonstrated that an increase in CBF induced by ATP is dependent on Ca²⁺ movements from intracellular stores and extracellular compartments (Morales et al. 2000; Barrera et al. 2004). Preincubation with oATP had no effect on the CBF increase induced by ATP, although concomitant treatment with oATP and CBX not only blocked, but also reduced CBF in response to ATP (∼60%). The addition of CBX 1 min after adding ATP, reduced the sustained component of CBF, an effect that was associated with a decrease in extracellular ATP levels in cultures exposed to CBX, suggesting that exogenous ATP induces ATP release via HCs to maintain the CBF response. Coincidently, in neonatal mouse olfactory epithelium, activation of purinergic receptors by ATP evokes ATP release via HCs (Hayoz et al. 2012). We also observed that the addition of oATP reduced the sustained component of CBF, suggesting the participation of P2X7R.

In summary, the results of the present study contribute to the identification of the broad role of extracellular ATP with respect to regulating ciliary beating. This occurs by a mechanism associated with P2XR activation and ATP release by Cx/Panx HCs, where ATP induces its own release that could be a control mechanism of basal ciliary activity and also explain the increase in CBF induced by extracellular signals (Fig. 5).

Figure 5. Model of CBF regulation by extracellular ATP.

It is proposed that basal CBF is regulated by the release of lower levels of ATP to the extracellular space via Cx HCs, that stimulates P2X7R to maintain Ca²⁺ _i levels. When extracellular ATP increases, following mechanical stimulation (Zhao et al. 2012), P2Y2Rs are activated, increasing CBF by increasing Ca²⁺ _i levels. To maintain this response, ATP induces ATP release via Cx or Panx HCs, as well as by P2X7R activation. Higher extracellular ATP levels, associated with inflammatory respiratory diseases, reduce CBF by closing HCs and decreasing Ca²⁺ _i levels. Dashed arrows represent the presence of transduction steps downstream from A2b and P2Y2 receptor activation.

Additional information

Competing interests

The authors declare that they have no competing interests.

Author contributions

Experiments were performed in the laboratories of MV and NPB. KD participated in the conception and design of the work; data acquisition, analysis and interpretation; and the writing of the manuscript. MR participated in CBF and dye uptake data acquisition and analysis. DVC participated in CBF data interpretation. CN and CF participated in AFM data acquisition, analysis and interpretation. MV and NPB participated in the conception and design of the work, as well as data analysis and interpretation. All authors participated in drafting the work and revised it critically for intellectual content. All authors approved the final version of the manuscript; they agree to be accountable for all aspects of the work ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved; and all people named as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

Fondecyt grants 3150652 (KD) and 1120169 (NPB), DPI‐Conicyt 20140080 (NPB), Anillo ACT‐1108 (NPB) and ICM P10‐035F (NPB and MV).