Function of Proton Channels in Lung Epithelia

Abstract

The properties of the voltage-dependent H⁺ channel have been studied in lung epithelial cells for many years, and recently HVCN1 mRNA expression has been linked directly to H⁺ channel function in lung epithelium. The H⁺ channel is activated by strong membrane depolarization, intracellular acidity, or extracellular alkalinity. Early on it was noted that these are surprising physiological channel characteristics when considering that lung epithelial cells have rather stable membrane potentials and a well pH-buffered intracellular milieu. This raised the question under which conditions the H⁺ channel is active in lung epithelium and what is its physiological function there. Current understanding of the HVCN1 H⁺ channel in lung epithelial acid secretion, its activation by an alkaline mucosal extracellular pH, and its role in the regulation of the mucosal pH of the lung has resulted in a model of mucosal pH regulation based on the parallel function of the HVCN1 H⁺ channel and the CFTR HCO₃^- channel, which suggests that HVCN1 is a critical factor that maintains a neutral surface pH in the lung.

INTRODUCTION

Lung epithelial cells were the first intact mammalian cell system in which voltage-dependent H⁺ channels were identified. Using primary cultures of rat alveolar type II cells, DeCoursey (1) originally described a plasma membrane H⁺ conductance in whole cell patch clamp recordings. The H⁺ conductance was shown to be active during membrane depolarization and extracellular alkalinity supported current activation. The activation by extracellular alkalinity was later found to be governed by the inside-to-outside H⁺ gradient such that the H⁺ conductance could be expected to activate physiologically by intracellular acidification, extracellular alkalinization, or membrane depolarization (2). In type II cells it was empirically determined that the threshold potential (V_threshold) at which the H⁺ conductance activates is linearly related to the transmembrane inside-to-outside H⁺ gradient (ΔpH) and the membrane potential by V_threshold = -40 mV ΔpH + 20 mV (ref. (2)). For epithelial cells that do not depolarize significantly this relation represents a substantial hurdle for the activation of the H⁺ conductance. For example, in the absence of a transmembrane pH gradient, activation of the H⁺ conductance would require a depolarization of the membrane potential above 20 mV, which would be unusual for any epithelial cell. Also, when considering a normal apical membrane potential of airway cells of V_a=-24 mV, a pH-gradient of 1.1 pH units would be required for H⁺ channel activation.

Similar voltage and pH-dependent characteristics have been found for the H⁺ conductance that is present in phagocytes (3-6). These cells, however, greatly depolarize their plasma membrane potential during the respiratory burst and measurably acidify their cytosol, and the H⁺ conductance functions to recover the intracellular pH from an acid load and to repolarize the membrane potential (7, 8). In contrast, lung epithelial cells have a rather stable membrane potential, and the intracellular pH changes little (2). The discrepancy between the functional properties of the H⁺ conductance and the conditions present in lung epithelial cells has been noted as soon as the H⁺ conductance was identified in type II cells (9), and the physiological function of the H⁺ conductance in these cells was initially rather unclear (9).

The purpose of this review is to shed some light into the physiological function of the voltage-dependent H⁺ conductance in lung epithelia by particularly considering the context of an epithelial setting. I will first review epithelial characteristics of the lung, cellular pH regulation, the pH of the lung lining fluid, and transepithelial and membrane potentials in polarized epithelial cells as they would affect the activity of the H⁺ conductance. Then, the involvement of the H⁺ conductance in acid secretion and the regulation of the pH of the airway surface liquid will be considered. In contrast to the role of the H⁺ conductance in phagocytes, I make the argument here that its function in the lung epithelium is the reacidification of an alkaline lung lining fluid.

ANATOMICAL CONSIDERATIONS

The lung is made of 23 generations of branching airway tubes where in most cases each airway divides to form two smaller airways. Starting at the trachea (generation 0) and terminating at the alveolar sac (generation 23) the function and cellular composition of the epithelium that lines the lung changes gradually. The terms “airway epithelium” or “tracheobronchial epithelium” usually refer to the ciliated epithelium as it is found in the nasal cavity, trachea, and bronchi until approximately generation 10 of the airways (Fig. 1A). This part of the lung is lined with a moderately-tight epithelium with a transepithelial electrical resistance of 300-400 Ω·cm² (10). Epithelial tightness is a functional feature that determines the ability of the epithelium to maintain different apical and basolateral membrane potentials (V_a and V_b, respectively); epithelial tightness is also critical to maintain concentration gradients, including pH gradients, between the airway surface liquid (ASL, the thin layer of fluid that lines the epithelium) and the serosa. Structurally, epithelial tightness results from the organization of the tight junctional cell contacts which serve to form an epithelial barrier and define the two distinct membranes, the apical (facing the lumen) and the basolateral (facing the serosa; Fig. 1). The transepithelial resistance is governed by the parallel transcellular and paracellular conductive pathways. The two principal cell types that form the airway epithelial barrier are ciliated cells and goblet cells. Ciliated cells comprise 50% to 70% of the epithelium; they are ~50 μm tall and continuously move the ASL with their cilia. Goblet cells release mucus that forms a normally quite thin layer that floats on top the periciliary fluid. ASL usually refers to the periciliary fluid plus the overlying mucus blanket. In normal airways, the ASL is ~10-20 μm high resulting in a volume of ~1-2 μl per square centimeter epithelial area (11). The small volume of the ASL is a determinant of the concentrations of factors secreted by the epithelium into the ASL including pH buffers and protons. The ASL contains a number of antibacterial factors and bicarbonate and its normal pH is regulated around neutral, although large variations and extremes between pH~4.5 and 8.5 have been reported for a number of airway conditions ((12) and see below). The major function of the airway epithelium is to form an innate defense barrier, filter air, and kill microbes.

Fig. 1. The two major epithelia of the lung.

A. The proximal airways (nasal cavity, trachea, bronchi, conducting bronchioles) are lined with a ciliated, columnar epithelium. The major two cell types that from the epithelial barrier are ciliated and mucus-producing goblet cells. The luminal surface is lined by the airway surface liquid (ASL) that is composed of the periciliary fluid (in which the cilia beat) and overlayed by a mucus blanket. The major function of the airway epithelium is innate defense. B. The alveoli in the distal lung are lined by the very thin alveolar epithelium composed of large and squamous type I cells and of small, surfactant-producing type II cells. The alveolar epithelium is lined by an extremely thin alveolar lining fluid. The major functions of the alveolar epithelium are gas exchange and innate defense.

Starting at generation 17, the airways become respiratory bronchioles with an increasing number of alveoli budding off, until generation 23, where the airways terminate in an alveolar sac. Each alveolus (diameter roughly 0.25 mm) is lined by the extremely thin alveolar epithelium. It is composed of very large but thin type I cells (13) and small, cuboidal type II cells, which are characterized by lamellar bodies as the storage organelle of surfactant. Alveoli are the sites of gas exchange. Alveoli are lined with an extremely thin fluid film with a volume of ~10-20 nl/cm² and a corresponding height of 0.1-0.2 μm, which is maintained by the surface properties of the surfactant generated by type II cells. This is by far the thinnest fluid film on any epithelium and as a result epithelial secretions result in correspondingly high luminal concentration changes. In contrast to the epithelium of the upper airways, the alveolar epithelium is extremely tight with transepithelial resistances of >2000 Ω·cm² (14-16). A high transepithelial resistance helps to depolarize the apical membrane potential (see below) and to maintain large concentration gradients between the alveolar lining fluid and the serosa.

THE H⁺ CONDUCTANCE IN LUNG EPITHELIA

A voltage-dependent H⁺ conductance has been found in type II cells and every other type of lung epithelial cell line that has been investigated. The epithelial cell types investigated are of tracheobronchial origin or alveolar type II cells, and a few commonly used cell lines (Table 1) which expressed H⁺ conductances with biophysical characteristics similar to the initial reports in rat type II cells, which are 1) strong voltage-dependence, slow time constants of current activation, block by Zn²⁺, and dependence on the pH gradient (17-19). Interestingly, investigations performed in confluent, polarized airway epithelia allowed to localize the H⁺ conductance to the apical (that is, luminal) membrane and to establish its involvement in airway epithelial acid secretion (Table 1). This has been confirmed further in freshly excised human nasal tissue samples (20) supporting the physiological role of the H⁺ conductance in epithelial acid secretion. Thus, it became clear that a H⁺ conductance is present and active in the epithelia of the lung to support acid secretion into the luminal compartment.

Table 1.

Proton conductance and acid secretion of lung epithelial cells

Lung Cell Type	Species	Characteristics	Whole cell conductance, (pS/pF)	Epithelial acid Secretion nmole·h-1·cm-2	reference
Ciliated	Human	Primary tracheal	20	410	(17, 18)
				460
Mixed	Human	Primary sinonasal	--	1070 2)	(20)
Calu-3	Human	Serous-like gland cell line	12	22	(18)
JME	Human	Nasal epithelial cell line, CF genotype	9	125	(17, 19)
Type II	Human	Primary fetal alveolar	10 3)	80	(14)
Type II	Rat	Primary alveolar	150 1)	--	(2)
A549	Human	Type II-like cell line	11 1)	--	(76)

Whole cell conductance is given at 0 mV for recordings at room temperature; epithelial acid secretion is the Zn²⁺-sensitive fraction of total at a mucosal pH of 7.3.

¹⁾original value adjusted to one pH unit gradient;

²⁾Original value adjusted to mucosal pH of 7.3 using the pH-dependence in (19);

³⁾H. Fischer, unpublished;

^--, not determined.

When considering the biophysical characteristics of the H⁺ conductance, cellular pH, and membrane potentials, the observation that a H⁺ conductance contributes to acid secretion raises the question of 1) how the H⁺ conductance is physiologically activated in lung epithelium, and 2) what is its role in epithelial function. For comparison, the best described function of the H⁺ conductance is in innate defense in phagocytes (21) where it is activated in parallel with the NOX2 NADPH oxidase to kill phagocytosed microbes during the respiratory burst (Fig. 2A). NADPH oxidases are plasma membrane proteins that transfer electrons from intracellular NADPH across the membrane to form extracellular superoxide and kill microbes. The phagosomal membrane is strongly depolarized by the electron transport by NOX2. During this process, large amounts of intracellular acid are generated by the oxidation of NADPH. Activated neutrophils acidify their cytosol at a rate of 0.1 pH/min and depolarize their membrane potential from a resting value of -60 mV to +50 mV (7, 21). As a result, the plasma membrane H⁺ conductance activates (by both membrane depolarization and intracellular acidification) leading to a massive release of acid from the cell and a repolarization of the plasma membrane (5, 22). In phagocytes, the role of the H⁺ conductance is closely linked to the activity of NOX2 where it serves as an acid relief valve to recover from intracellular acid loading by NOX2 activity and to repolarize the membrane potential (21).

Fig. 2. Comparison of HVCN1 in phagocytes compared to Lung epithelial cells.

A. In phagocytes the HVCN1 H⁺ channel operates to support NOX2 activity by maintaining intracellular pH and membrane potential. NOX2 transfers electrons (e^-) from intracellular NADPH to extracellular O₂ to form O₂^-. This process generates intracellular H⁺. B. Initial proposal of H⁺ channel function in lung epithelial cells was modeled after phagocytes as a similar arrangement of HVCN1 with the DUOX NADPH oxidase. There, DUOX transfers e^- to form extracellular H₂O₂ and similarly generates intracellular H⁺; for proper stoichiometry see (77). NOX2, NADPH oxidase 2; HVCN1, voltage-dependent H⁺ channel 1; DUOX, dual NADPH oxidase.

Interestingly, epithelial cells of the lung express their own NADPH oxidase based on the DUOX isoforms (14, 18) and, similar to phagocytes, lung epithelial cells kill bacteria and are part of the innate defense system. When both the DUOX NADPH oxidase and the H⁺ conductance were identified in lung epithelial cells (18) it appeared that the H⁺ conductance in lung epithelium serves a similar function as in phagocytes and is required to release intracellular H⁺ during DUOX activity (23). Accordingly, an initial model by us for the function of the H⁺ channel in lung epithelium was closely modeled after phagocytes, as shown in Fig. 2B. However, later it became clear that DUOX has only very small effects on pH_i (14, 18), which is well buffered by the pH-regulating membrane transporters present in lung epithelial cells. Nevertheless, local acidic pH_i compartments near the apical membrane have been discussed (23, 24) and might affect this conclusion. Acidic pH_i compartments at the apical membrane could be caused by apical DUOX or the tightly packed apical mitochondria in airways (25); however, a local acidic pH_i at the apical membrane has not been observed experimentally.

DUOX also shows relatively low rates of H₂O₂ production in comparison to NOX2 (26) and, by extension, small transmembrane electron currents, which in the presence of the dominating Na⁺ and Cl^- conductances in lung epithelia are expected to show little effects on the membrane potential. These findings suggested that, in contrast to NOX2, DUOX does not necessarily require a H⁺ channel for its function because lung epithelial cells express mechanisms to stably regulate the intracellular pH and the membrane potentials. This suggests that the role of the H⁺ channel in the lung is quite different from the one in phagocytes. To support this argument, below I describe in detail the mechanisms that determine the membrane potentials and the intracellular pH in lung epithelial cells.

MEMBRANE POTENTIALS

A significant determinant of the activity of the H⁺ conductance is the membrane potential. The lung epithelium is a salt transporting tissue and the membrane potential of the apical membrane (V_a) is governed by its major conductances, which are the apical Na⁺ and Cl^- conductances (Fig. 3). The basolateral membrane potential (V_b) is largely governed by its K⁺ conductance. The primary ion transport function of lung epithelia is Na⁺ absorption across apical Na⁺ channels driven by the basolaterally located Na⁺-K⁺-ATPase. Recycling of K⁺ across the basolateral membrane results in a hyperpolarization of the cell, and Cl^- distributes passively across the apical membrane according to its electrochemical driving forces. Although there are a number of other ion channels expressed in lung epithelium, including the apical H⁺ channel, their contributions to the membrane potentials can be safely neglected owing to the high Na⁺ and Cl^- conductances. For comparison, the apical Na⁺ and Cl^- conductance in rabbit airways was found to be 1.5 and 2 mS/cm², respectively, and the basolateral conductance about 3 mS/cm² (27). These can be physiologically upregulated, and in Calu-3 cells, an astounding forskolin-stimulated apical Cl^- conductance of 200 mS/cm² has been reported (28). For comparison, the apical H⁺ conductance in human airway epithelium was found to be 9 μS/cm² (17) indicating that its effect on V_a can be safely ignored.

Fig. 3. Major ion transport and equivalent circuit model for lung epithelium.

Left, major ion conductances are shown and lumped together. Lung epithelia absorb Na⁺ across apical Na⁺ channels driven by a basolateral Na⁺-K⁺-ATPase; K⁺ is recycled across basolateral K⁺ channels. Cl^- distributes passively over the apical membrane. Right, arrangement and activity of transport mechanisms results in an equivalent circuit model where R_a and R_b represent the Thévenin equivalents of the apical and basolateral membrane resistances and V_a and V_b the corresponding membrane potentials. Note that V_a is reported inside-vs-mucosa, V_b inside-vs-serosa, and V_t mucosa-vs-serosa, which results in the relation V_t = V_b - V_a for practical purposes. R_j represent the Thévenin equivalent of the paracellular, junctional resistance over which V_t drops.

Membrane potentials in primary tracheobronchial human epithelia have been found to be comparably stable during normal ion transport activity. Under control conditions, during basal Na⁺ absorption, the apical membrane potential is approximately V_a = -24 mV and the basolateral membrane potential V_b = -34 mV (29-32). Regulation of conductances in either membrane result in voltage changes in both membranes owing to their arrangement in a 3-resistor electrical circuit formed by the apical, basolateral, and junctional resistors (R_a, R_b, and R_j; see Fig. 3). The differential expression of the major ion conductances in the two membranes and the selective ion gradients towards the mucosal and the serosal compartment result in distinct electromotive forces across the two membrane. However, the difference in the membrane potentials of the apical and basolateral membrane is solely caused by the junctional resistance, R_j, where the voltage difference between V_a and V_b drops (Fig. 3). Thus, the tightness of the epithelial monolayer (that is, the value of R_j relative to R_a+R_b) determines the depolarization of V_a with respect to V_b. In leaky epithelia, where R_j is small compared to R_a+R_b, there is little difference between V_a and V_b, and as a result the transepithelial potential, V_t, is small. The difference V_b-V_a is readily measured experimentally as V_t. In primary tracheal cultures V_t is approximately -10 mV. Interestingly, in measurements of the nasal potential difference in human subjects, an average V_t of -19 mV is reported as the normal value (33) suggesting that in vivo V_a is significantly more depolarized than V_b compared to in vitro epithelial cultures.

V_a and V_b hyperpolarize in presence of the Ca²⁺-agonist histamine by ΔV_a = -5 mV and ΔV_b = -6 mV (29), which is likely governed by the activation of basolateral Ca²⁺-activated K⁺ channels. V_a and V_b are slightly depolarized when stimulated by the cAMP agonist isoproterenol by ΔV_a = 1.5 mV and ΔV_b = 2 mV (34), which is caused by the activation of the apical Cl^- conductance. For comparison, the serous cell-like Calu-3 airway cell line expresses very high amounts of apical CFTR Cl^- channels. When Calu-3 epithelia were stimulated with the cAMP-agonist forskolin to activate the CFTR Cl^- conductance, V_a depolarized from -52 mV to -21 mV, and, correspondingly, V_b depolarized from -60 mV to -44 mV (28). These membrane depolarizations are the largest reported for any lung epithelium. Nevertheless, in comparison to the depolarizations seen in phagocytes and the requirements for the activation of the H⁺ conductance, the depolarizations in airway cells appear small.

In alveolar cells, in comparison, membrane potentials have been little studied owing to their complicated anatomy in vivo and the lack of good cell culture models for alveolar epithelia. In vivo, the alveolar epithelium is composed of type I and type II cells (Fig. 1B); however, type I cells are difficult to harvest and culture, and type II cells dedifferentiate quickly into a type I-like phenotype in culture. As a result, there is little information about membrane potentials in these cells or in an intact epithelium. In two early reports on freshly isolated single type II cells, a resting membrane potential of -27 mV (rat lung, using a membrane potential dye (35)) or -63 mV (rabbit lung, using the Rb⁺ distribution potential (36)) has been found, and Matalon recently estimated -40 mV(37). None of these studies allowed estimates of apical or basolateral potentials because single cells were used. However, the transepithelial potential of confluent human primary type II cell monolayers is -9 mV (14), or in rat type II monolayers, -10 mV (38). There is no transepithelial information about type I cells but given a similar expression of channel types in type I and type II cells (13), these values are likely found in intact alveoli as well. Thus, although V_a and V_b have not been measured in alveolar cells, V_a is expected to be ~10 mV depolarized compared to V_b. It should be noted that some studies found relatively small transepithelial potentials (39), including a study of microimpaled intact alveoli in the rabbit (V_t = 3.5 mV; (40)).

Stimulation of type II monolayers with a cAMP agonist activates both apical CFTR Cl^- channels and ENaC Na⁺ channels and results in a transepithelial hyperpolarization by -6 mV (41) suggesting a depolarization of V_a by 6 mV. For comparison, selective membrane potentials have been measured in H441 cells, which are Clara-like, small-airway cells but share some transport characteristics with type II cells. H441 cells had a baseline V_a of -43 mV and during stimulation with forskolin V_a depolarized to -18 mV by way of activation of the apical ENaC Na⁺ conductance (42).

These studies of membrane potentials in both airway and alveolar epithelia suggest that the apical membrane is approximately 10 mV more depolarized than the basolateral membrane. The absolute membrane potentials change only little during epithelial transport regulation when compared to the requirements of the voltage-dependent regulation of the H⁺ conductance. The reported depolarizations in primary cells are in the order of <10 mV and only cell lines with high CFTR or ENaC expression depolarize V_a somewhat more (i.e., Calu-3 and H441 cells). However, the relatively small changes of V_a suggest that the apical H⁺ conductance is only little regulated by the membrane voltage in lung epithelial cells.

pH GRADIENTS

Because the airway pH, and in particular the pH of the luminal surface varies widely (12), and the threshold potential for H⁺ channel activation is sensitively regulated by the pH gradient, the pH gradient across the apical membrane is a candidate as a regulator of H⁺ channel activity in lung epithelium.

The pH of the airways has been measured over many years using pH electrodes placed on the luminal epithelial surface and, more recently, using pH sensitive fluorescent dyes (12). The earliest reliable measurement of normal ASL pH by Kyle et al. (43) reported a pH of 6.85 in ferret trachea using a pH-sensitive microelectrode. This value under resting conditions has been confirmed using the fluorescent, pH-sensitive dye BCECF (as dextran to maintain extracellular localization) in bovine tracheal cultures (pH 6.81; (44)) and in basal human airway gland secretions (pH 6.90; (45)). When determining the ASL pH indirectly by measuring the pH of an unbuffered solution that is perfused over the airway mucosa, a pH of 6.93 and 6.85 was found in pig and human airway epithelium, respectively (17, 46). A similar value (pH 6.92) was found in vivo in rabbit alveoli using pH sensitive microelectrodes (47). These studies, using different systems and techniques, confirmed the initial measurements by Kyle et al. and suggest that the normal pH of the ASL is pH 6.85.

Very large variations of ASL pH have been reported in measurements in human subjects for a number of disease conditions indicating that the apical membrane can be exposed to large pH changes. Measurements in vivo in human subjects are currently mainly done by using the exhaled breath condensate, tracheal aspirates, or bronchoalveolar lavage fluid as reporters of ASL pH. Comparison of pH values between different studies and techniques is somewhat limited by the different sample processing techniques, and, in particular, the maintenance of the CO₂ content in the samples (48), however, the inclusion of normal control values in each study allows for a relative comparison. In a study of exhaled breath condensate, normal average pH was 7.65, which dropped in asthmatics to an average of pH 5.23, and extreme values of pH 4.5 to pH 8.5 were reported (49). In another study, the pH of the tracheal aspirate from neonates was found to be alkaline, ranging between 5.8 and 8.8 (50) and with average pH of 7.8. Bronchoalveolar lavage fluid sampled from neonates is on average acidic with a pH of 5.5 but ranges up to pH 7.5 during respiratory distress (51). In cystic fibrosis, exhaled breath condensate was found to be ~1 pH unit more acidic than normal (52, 53), and in pulmonary tuberculosis the exhaled breath was 0.5 pH units more acidic than normal (range pH 4 to 8.2 (54)). The origin of the pH changes in the lung are not well investigated but likely relate to acid and base secretion by the epithelium (12). These observations indicate that the lung epithelium is exposed to sizeable luminal pH changes of several pH units.

Intracellular pH (pH_i) has been studied in both airway and alveolar epithelia. The most reliable measurement of pH_i in airways is a study by Paradiso et al. (55) who used confluent, polarized human cultures in presence of HCO₃^-/CO₂ buffered solutions. Use of these conditions appears critical when considering the effect on membrane potentials, transporter activities, and buffering systems on pH_i. Paradiso et al. (55) reported an average pH_i of 6.95. For comparison, other studies in single cells or HCO₃^- free solutions reported a more alkaline pH_i of 7.1 – 7.4 (14, 55-58). Under physiological conditions, pH_i is stably buffered by basolateral mechanisms (Fig. 4A) including the Na⁺-H⁺ exchanger (57, 59), the Na⁺-HCO₃^- co-transporter (60), and the Cl^--HCO₃^- anion exchanger (55, 61). The activity of these transporters is tightly regulated by pH_i so that any intracellular acid or base excess is removed across the basolateral membrane. Any observed changes of pH_i are small and are only observed experimentally when these major pH regulating membrane transporters are inhibited.

Fig. 4. Airway epithelial H⁺ and HCO₃^- transport determines ASL pH.

A. Distribution of H⁺ and HCO₃^- transporters in lung epithelium. The HVCN1 H⁺ conductance and the CFTR HCO₃^- conductance are expressed apically. HVCN1 conducts H⁺ only outwardly while CFTR conducts HCO₃^- both ways. The Na⁺-H⁺ exchanger NHE1, the Cl^--HCO₃^- anion exchanger AE2, and the Na⁺-HCO₃^- cotransporter NBC1 are present in the basolateral membrane. Resting pH values for the airway surface liquid (pH_ASL), intracellularly (pH_i), and serosally (pH_ser) are given. For stoichiometry of transporters see (28, 78). B. ASL pH determines airway H⁺ and HCO₃^- secretion. HCO₃^- secretion (open symbols) is modeled for pH_i 6.95; V_a -24 mV; apical, linear, voltage-independent HCO₃^- conductance 0.4 mS/cm²; average pCO₂ 3 kPa. H⁺ secretion data (filled symbols) are mean±SE compiled from previous measurements (17, 19). C. Predicted change of ASL pH by H⁺ and HCO₃^- secretion. Calculated from rates in B using buffer capacities for the ASL from ref. (65) (range 6.5 to 15 mM/pH) and an ASL volume of 1.5 μl/cm². CFTR, cystic fibrosis transmembrane conductance regulator.

A stably buffered pH_i in combination with a variable extracellular pH is likely to affect the activity of the apical H⁺ conductance. This also suggests that the regulation of the H⁺ conductance by the extracellular pH is involved in the acidification of the ASL. This contrasts from its role in phagocytes, where its role is acid release and alkalinization of the cytosol.

PROTON CHANNELS GOVERN PROTON SECRETION

H⁺ channels are present in cell types all along the respiratory tract epithelium (Table 1). When investigating intact, polarized epithelia in Ussing chambers using the pH stat technique, these cell types also secrete substantial amounts of acid into the mucosal (luminal) medium. However, the relation between H⁺ channel activity in whole cell patch clamp measurements and epithelial acid secretion was initially difficult to make.

The identification of the role of H⁺ channels in epithelial acid secretion was complicated by the fact that airway epithelia express at least three functional mechanisms for H⁺ secretion in their apical membrane including the vacuolar-type H⁺-ATPase, the non-gastric (colonic) H⁺-K⁺-ATPase, and H⁺ channels (for details see (12)). Initially it was unclear why three parallel mechanisms for acid secretion are present in the apical membrane; however, when investigating airway epithelia in presence of an acidic luminal pH of around 6, low rates of acid secretion are observed (~5 nmole·h^-1·cm^-2, (19), compare to rates in Table 1). These are likely mediated by ATP-dependent processes based on the observation that H⁺ secretion occurs against an H⁺ gradient. In contrast, when the mucosal medium is alkalinized, acid secretion increases in parallel with the pH gradient, and a substantial fraction of the acid secretion is Zn²⁺ sensitive (Table 1, and (19)). This suggests that the ATP-dependent mechanisms operate at a low rate and provide basal H⁺ secretion even against an H⁺ gradient, while H⁺ channels allow for large rates of acid secretion when the ASL pH is alkaline.

Additional evidence for a role of H⁺ channels in epithelial acid secretion came from the similar sensitivity to Zn²⁺ and the similar dependency on pH gradients of whole cell H⁺ currents and epithelial acid secretion. Even more convincing evidence is based on functional characteristics of a misfunctional mutation of HVCN1 (19). In a study of epithelial acid secretion and whole cell H⁺ currents in primary airway cultures we found a tracheal isolate that expressed the M91T HVCN1 mutant. Acid secretion measured in these cultures was characterized by a reduced sensitivity to an alkaline pH such that approximately a 0.5 pH-unit larger mucosal pH was required to activate acid secretion compared to normal. A similar behavior was found when the M91T HVCN1 mutant was recombinantly expressed and investigated with whole cell patch clamping, where again larger bath-to-pipette pH gradients were required for H⁺ current activation compared to wild type HVCN1. Thus, the functionally distinct characteristics of the M91T HVCN1 mutant were found in acid secretion by epithelia and in whole cell patch clamp measurements and directly related H⁺ channel activity to acid secretion.

PROTON CHANNELS REGULATE ASL pH

Since H⁺ channels are evidently present and active in the lung epithelium, we wondered under which conditions the electrochemical H⁺ gradient would be outward and large enough to activate the H⁺ conductance. Because both the intracellular pH and the membrane potential are relatively stable, we focused in a recent investigation on the role of the ASL pH in the activation of apical H⁺ channels and H⁺ secretion (19). The pH-dependence of Zn²⁺-sensitive, H⁺ channel-mediated acid secretion is shown in Fig. 4B, filled circles. At a mucosal pH of 7.3 and above, acid secretion is active while at a mucosal pH of 7 and below no acid is secreted. Thus, when considering the resting ASL pH of 6.85 (Fig. 4A), the H⁺ channel is likely closed and opens only when the ASL pH turns alkaline. In general, the pH gradient-driven activation of H⁺ currents is consistent with the observed pH dependence of H⁺ currents in phagocytes; however, the H⁺ gradient in phagocytes is generated by intracellular acidity, while in lung epithelium it appears to be governed by extracellular alkalinity.

This has two major consequence for the function of H⁺ channels in lung epithelium. First, in contrast to phagocytes, where H⁺ channels are involved in the regulation of pH_i, H⁺ channels in lung epithelial cells appear little involved in the regulation of pH_i. Second, the observed activation of H⁺ channels by extracellular alkalinity will re-acidify an alkaline ASL pH, which indicates the H⁺ channel in the regulation of the extracellular pH.

Alkalinization of the ASL is likely caused by epithelial HCO₃^- secretion. Figure 4B displays, in comparison, the predicted effect of ASL pH on both H⁺ and HCO₃^- secretion by the epithelium. Because both H⁺ and HCO₃^- move passively across the apical membrane, both H⁺ and HCO₃^- secretion are driven by V_a and the apical transmembrane pH gradient. Alkalinization of the airways is thought to be mainly related to the activity of the cAMP-activated CFTR Cl^- channel¹⁾ in the apical membrane of lung epithelia. CFTR shows a voltage-independent and linear conductance for HCO₃^- that is ~20% compared to Cl^- (62, 63) and CFTR mediates HCO₃^- secretion by airway epithelia (60, 64). When modeling airway epithelial HCO₃^- secretion for an apical HCO₃^- conductance of 0.4 mS/cm² with V_a = -24 mV and pH_i = 6.95 then gradient-driven HCO₃^- secretion across the apical membrane linearly increases with decreasing pH for all ASL pH values <7.35 (Fig. 4B). Accordingly, at high ASL pH values CFTR may mediate HCO₃^- absorption.

Because the volume of the airway surface liquid is very small, these rates of H⁺ and HCO₃^- secretion are expected to affect the ASL pH. The estimated effects of H⁺ and HCO₃^- secretion on changes of the ASL pH are shown in Fig. 4C. Rates of ASL pH changes were calculated from ΔASL pH = J / (β · v), where J is the H⁺ or HCO₃^- flux, β is the pH-dependent buffer capacity of the ASL (from Holma (65) for respiratory secretions), and v is the ASL volume (1.5 μl/cm²). The resulting rates are quite fast (Fig. 4C). For example, an alkaline ASL pH of 8 is predicted to re-acidify within a few minutes back to pH 7 based on H⁺ channel activity (with an initial rate of -0.25 pH/min); in a cAMP-stimulated epithelium with active CFTR this is predicted to re-acidify even faster. These calculations suggest a role for the H⁺ channel in the re-acidification of the ASL to values near pH 7, where the H⁺ channel is closed. Because the H⁺ channel is expressed in parallel to CFTR, the activities of both channels are predicted to mutually complement each other during the regulation of the ASL pH. We speculate that the activities of CFTR and HVCN1 in airways provide a feedback control system for the regulation of pH ASL as depicted in Fig. 5. The normal set-point under resting conditions is pH~7 (activation of H⁺ conductance; Fig. 4B) for this system, although this set-point is dependent on V_a. A disturbance of ASL pH results in a H⁺ or HCO₃^- secretion as a response and adjustment of ASL pH. At an acidic ASL pH, the H⁺ channel is closed but HCO₃^- is secreted along its electrochemical gradient through CFTR and alkalinizes the ASL. At an alkaline pH ASL, H⁺ channels are open and secrete acid to acidify the ASL. HCO₃^- absorption might support the effect of acid secretion on ASL pH, however, the H⁺ channel appears better positioned to perform this function because it is selectively activated by the alkaline pH, while CFTR is not pH-activated, but instead is regulated through the cAMP/PKA system (e.g., requires activation by β agonists).

Fig. 5. Proposed role for HVCN1 in ASL pH regulation.

The CFTR HCO₃^- channel and the HVCN1 H⁺ channel provide a system for the regulation of the pH of the ASL. The ASL pH is the control variable to that governs H⁺ or HCO₃^- secretion into the ASL. An alkaline ASL pH activates HVCN1 and provides a driving force for H⁺ secretion. An acidic ASL pH provides a driving force for HCO₃^- secretion, however, requires additional activation of CFTR by the cAMP/PKA system. Arrow with circled + indicates activation. PKA, protein kinase A.

This model can be used to explain the observed acidic or alkaline ASL pH values in disease conditions. In inflammatory airway diseases an alkaline ASL pH is generally observed (12). Inflamed airways are characterized by increased Ca²⁺ signaling (66) and break-down of the junctional resistance (67, 68), both of which result in a hyperpolarization of the apical membrane. As a result, HCO₃^- secretion is favored and the H⁺ channel is inactivated, which shifts the set-point of the regulatory feedback to an alkaline value. On the other hand, an acidic ASL pH is observed in cystic fibrosis (12). Cystic fibrosis airways are characterized by a depolarized apical membrane (10) and a misfunctional CFTR, which results in a shift of the set-point to an acidic ASL and an inhibition of CFTR as a control element in this model.

Further information might be generated from HVCN1 knockout mouse models that have been introduced recently (69-72) in comparison or in combination with CFTR knockout mice. In mice, ASL pH is regulated by CFTR as shown by an acidification of the ASL when CFTR is blocked (73) or in CFTR knockout mice (74). As of the writing of this review, HVCN1 H⁺ channel activity has not been studied in mouse lung, however, investigations of mouse lung epithelium derived from HVCN1 knockout mice may shed additional light into the function of HVCN1 is lung epithelial biology.