Surface fluid absorption and secretion in small airways

Abstract

Native small airways must remain wet enough to be pliable and support ciliary clearance, but dry enough to remain patent for gas flow. The airway epithelial lining must both absorb and secrete ions to maintain a critical level of fluid on its surface. Despite frequent involvement in lung diseases, the minuscule size has limited studies of peripheral airways. To meet this challenge, we used a capillary to construct an Ussing chamber (area <1 mm²) to measure electrolyte transport across small native airways (∼1 mm ø) from pig lung. Transepithelial potentials (V_t) were recorded in open circuit conditions while applying constant current pulses across the luminal surface of dissected airways to calculate transepithelial electrical conductance (G_t) and equivalent short circuit current () in the presence and absence of selected Na⁺ and Cl⁻ transport inhibitors (amiloride, GlyH-101, Niflumic acid) and agonists (Forskolin + IBMX, UTP). Considered together the responses suggest an organ composed of both secreting and absorbing epithelia that constitutively and concurrently transport fluids into and out of the airway, i.e. in opposite directions. Since the epithelial lining of small airways is arranged in long, accordion-like rows of pleats and folds that run axially down the lumen, we surmise that cells within the pleats are mainly secretory while the cells of the folds are principally absorptive. This structural arrangement could provide local fluid transport from within the pleats toward the luminal folds that may autonomously regulate the local surface fluid volume for homeostasis while permitting acute responses to maintain clearance.

Key points

• Using a small glass capillary, an Ussing chamber was designed and used to measure electrical properties across very small pieces (<1 mm²) of freshly dissected epithelia.

• The system was applied to small airways of the lung to show that fluids on the airway surfaces are constantly being simultaneously secreted and absorbed.

• A new model proposes that the accordion-like structure of folds and pleats in the epithelial lining of the airways serves to secrete fluid within the pleats and to absorb fluid along the folds so that fluid levels on the airway surfaces are maintained automatically and airways do not flood or become too dry.

• These results help us understand the first line of lung defence against infections from bacteria and viruses and may be used to treat or prevent lung disease.

Introduction

Failed maintenance of airway surface fluid (ASF) threatens either asphyxia from drowning or lung disease from contamination and infection. Consequently, mechanisms to secrete as well as absorb surface fluid must be constantly engaged to acutely adjust and maintain a critical volume of ASF. Although it is well accepted that the epithelia lining the small airways are capable of fluid secretion and absorption, understanding how these processes occur and are coordinated in maintaining strict control of the ASF volume has been a significant challenge due to size and accessibility of these tissues.

At present, it is widely held that the ASF is precisely controlled by an epithelium that, as a whole, oscillates between absorptive and secretory activities (Kunzelmann, 1999; Boucher, 2003). The cells of such an epithelium must be capable of bidirectional transport, and every transporting cell should possess all the requisite components for absorption as well as secretion. To oscillate between secretory and absorptive transport, a number of cellular functions and properties must acutely change in unison. That is, as a minimum, to change efficiently from absorbing to secreting, the cells should at least (a) significantly down-regulate the apical Na⁺ and the basolateral Cl⁻ membrane conductances, because a cation conductance in the apical membrane would depolarize and undermine the electrical driving force required for anion secretion and a basolateral anion conductance would allow anion transport in the wrong direction; (b) upregulate the K⁺ conductance and the NKCC cotransporter activity in the basolateral membrane, the former to hyperpolarize the cell and provide the electrical driving force, and the latter to supply Cl⁻ anions to the cell for apical secretion; and (c) render the paracellular conductance Na⁺ selective, to selectively provide co-ions for the anions secreted into the lumen. In general, epithelial cells do not effect this complex set of changes because they are designed to transport in only one direction; that is, epithelial cells either absorb or secrete, but not both. For example, in the intestine, secretory crypt cells secrete fluid while absorptive villar cells simultaneously absorb fluid. More generally, in exocrine gland structures, fluid and electrolytes are secreted by proximal secretory units and absorbed or modified by more distal duct cells, e.g. sweat glands, salivary glands, pancreas, etc. Indeed, examples of the presence of both functions in cells that alternate between fluid secretion and absorption are not well characterized.

These considerations strain the present, widely accepted notion that the volume of the ASF is controlled by the relative activity and duration of the direction of fluid transport of an epithelium that alternately cycles from secretion to absorption. Alternatively, although not at present identified, if separate systems of secretory and absorptive epithelial cells also exist in the airway epithelial architecture, the crucial volume of ASF could be maintained by simply varying the relative activity of either function.

The epithelial characteristics underlying the mechanisms of fluid secretion and absorption that would address these issues are not defined for smaller peripheral bronchi and bronchioles. Consequently, the mechanisms for maintaining the surface fluids of small airway have been largely deduced from properties of cultured airway epithelial cells, large bronchii, and tracheal epithelia (Welsh et al. 1982; Joris & Quinton, 1991; Kondo et al. 1993; Boucher, 2003). The paucity of such information seems puzzling if abnormalities of electrolyte transport are inherent to such critical airway pathologies as cystic fibrosis, pseudohypoaldosteronism, and possibly asthma, bronchitis, and chronic obstructive pulmonary disease. At the same time, the complexity of the anastomosing, ever-diminishing size of each branch of the bronchial tree is, no doubt, responsible for the limited transepithelial measurements of electrolyte transport in these tissues (Ballard et al. 1992; Al-Bazzaz, 1994; Blouquit et al. 2002; Wang et al. 2005). In fact, at this level, branching occurs at lengths equivalent to about three to four luminal diameters of each segment, and the diameter of each segment decreases to about 80–85% of the diameter of the preceding segment at each sequential branch (Weibel, 1963).

At the level of small airways, retrieval of intact epithelial specimens large enough for studies in conventional Ussing chambers is virtually impossible. Consequently, we took a new approach to help overcome these limitations. We used the end of a small glass capillary pressed against the luminal surface of an opened airway to form an electrically isolated compartment of the capillary lumen. This arrangement allowed transepithelial electrophysiological measurements from segments of small airways of about 1 mm φ (Fig. 1). With this system, we examined the ion secretory and absorptive properties of native small airways by determining the spontaneous electrical properties as well as responses to specific inhibitors and agonists of these functions. Although subject to interpretation, overall the results indicate the presence of an epithelium that consists of separate secretory and absorptive units that control ASF volume by the relative activities of concurrent, constitutive fluid transport in opposite directions.

Figure 1. Dissected small airways and Capillary-Ussing chamber.

A, dissected small airway. B, opened airway. C, opened airway tissue pinned and mounted over supporting polycarbonate filter or nylon mesh. (Scale bar = 1 mm for A/C.) D, capillary-Ussing chamber system for measuring V_t and passing constant current. The system for guiding the polished glass capillary onto the centre of the tissue for electrical isolation and sealing is assembled on a microtubule perfusion device (Burg et al. 1968) (not shown) in order to precisely control the advance of the capillary end onto the apical surface of the epithelium. E, micrograph of example of small airway epithelium mounted as described and used for analysis (scale bar = 200 μm).

Methods

Ethical approval

The Institutional Animal Care and Use Committee of the University of California San Diego approved the procedures used in this study. All studies here comply with the policies and regulations of The Journal of Physiology.

Tissue collection, transport and preservation

Intact lungs were harvested from a local slaughterhouse from farm pigs (30–80 kg) of either sex immediately after sacrifice by captive bolt stunning and knockdown followed immediately by exsanguination to reduce haemorrhage. Excised lungs were sealed in plastic bags placed under crushed ice (<10°C) until used. Tissues so maintained usually retained ion transport properties for at least 6 h, but the freshest tissues usually gave the best electrical responses. All specimens reported were examined within 6 h of sacrifice.

Dissection and mounting in capillary-Ussing chamber

We cut small blocks of about 1–3 cm³ from the peripheral lung parenchyma, usually near the costal diaphragmatic ridge of the lower lobes that were then placed in a chilled (<10°C) bath of NaCl-Ringer solution. Under a dissecting microscope, we identified the openings of small airways (∼ 1 mm ø) on the surfaces of the wedges of tissue and isolated a length of 2–4 mm of the airway from surrounding parenchyma (Fig. 1A) with sharpened tweezers and iridectomy scissors. We slit the lumen longitudinally to expose about ∼4 mm² of luminal surface (Fig. 1B), which was then held open with modified 00 insect pins (Ward's Scientific, Rochester, NY, USA) on a ‘trampoline’ of a polycarbonate filter (20 μm pores; SPI Supplies, West Chester, PA, USA) or a close weave nylon mesh (salvaged from common tea bags) (Fig. 1C and D). The trampoline was previously glued over the end of a custom fabricated plastic tube (I.D. 3 mm ø; O.D. 6 mm ø), the interior of which formed the serosal compartment of the capillary-Ussing chamber (Fig. 1C and D). We used a V-track concentric pipette carrier designed originally for renal microtubule perfusion (Burg et al. 1968) to precisely advance the end of a carefully squared, fire polished glass capillary (inside area ≍ 0.66 mm²) against the luminal surface of the airway pinned flat over the trampoline support (Fig. 1D). Prepared in this manner, the pleats of the airway epithelial folds were spread and exposed to apical perfusing solutions in the capillary lumen (Fig. 1E). In the presence of a 150 mm Cl⁻ serosal to mucosal gradient, the capillary was gradually advanced to achieve an apparent maximum transepithelial potential (V_t). Thus, the capillary lumen formed the luminal compartment of the Ussing chamber for transepithelial electrical measurements (Fig. 1D). Both the apical and serosal sides of the tissue were perfused with defined solutions at ∼37°C. Preparations that did not spontaneously exhibit a V_t of at least −25 mV lumen negative in the presence of a 150 mm Cl⁻ transepithelial gradient were discarded.

Transepithelial electrical measurements

We measured open circuit V_t continuously via free-flow solution bridges of 3 m KCl from Ag–AgCl electrodes that formed rinseable junctions with the stream of perfusates leaving the tissue. This arrangement allowed the perfusate–KCl bridge junction potentials to be minimized and stabilized by rinsing the junction with fresh KCl after each change of perfusate. Electrode asymmetries were measured before and after mounting tissues, which were usually stable and less than ±0.5 mV throughout the protocol. All reported values are based on measured values with asymmetries subtracted. G_t was calculated from Ohm's law using the deflection in V_t that resulted from constant 1.0 μA current pulses passed across the tissue for 1.5 s at 10 s intervals. was calculated from G_t and the open circuit V_t using Ohm's law. The equivalent circuit was taken as the current that is ‘equivalent’ () to active ion transport (current source) passing through the leak (shunt) resistance of the epithelium (1/G_t) in a current loop that gives the open circuit voltage drop across the epithelium (V_t). The resistance pathway, G_t, does not require selectivity, but mostly likely conducts mainly Na⁺ and Cl⁻.

Solutions

The Ringer solution contained (in mm): Na⁺ (150), K⁺ (4.6), Mg²⁺ (1.0), Ca²⁺ (1.0), PO₄³⁻ (2.5), Cl⁻ (150), SO₄²⁻ (1.0), acetate (2.0), glucose (10), buffered to pH 7.4 with HCl. To minimize endogenously generated prostaglandins during dissection, indomethacin (1 μm) was present in the dissecting Ringer solution. For anion diffusion studies, 150 mm NaCl was replaced with equimolar sodium gluconate (150 mm Cl⁻ gradients) or with 75 mm NaCl plus 150 mm mannitol for osmolar balance (1:2 dilution diffusion gradients).

3-Isobutyl-1-methylxanthine (IBMX; 100 μm), forskolin (10 μm), uridine 5′-triphosphate (UTP; 100 μm), amiloride (10 μm), NFA (100 μm), indomethacin (1 μm) and bumetanide (100 μm) were all obtained from Sigma-Aldrich (St Louis, MO, USA). N-(2-Naphthalenyl)-[(3,5-dibromo-2,4-dihydroxyphenyl) methylene]glycine hydrazide (GlyH-101; 50 μm), used as a cystic fibrosis transmembrane conductance regulator (CFTR) inhibitor of Cl⁻ conductance, was a generous gift from Drs A. Verkman and N. Sonawane.

Luminal and basolateral solutions were changed rapidly via manifolds that distributed stores of the above solutions as required.

Statistical analysis

The data are presented as means ± standard error of the mean (SEM); n is the number of tissues examined. Statistical significance was determined on the basis of Student's paired t test with P < 0.05 taken as significantly different.

Results

System validation

Our previous results using microperfused intact small airways (∼1 mm φ)(Wang et al. 2005) showed that the airway epithelium is highly Cl⁻ selective. In these microperfused preparations, the mean V_t was about −3 mV in bilateral NaCl-Ringer solution, but V_t significantly hyperpolarized to −58 mV when the luminal perfusate was changed to 150 mm sodium gluconate (an impermeant anion) to create a steep 150 mm bath-to-lumen Cl⁻ gradient (Wang et al. 2005). Using these data as an index of tissue integrity for our dissected specimens mounted in the capillary-Ussing chamber, we first tested the response of V_t to a 150 mm Cl⁻ bath to lumen gradient. When the impermeant gluconate anion replaced luminal Cl⁻, V_t hyperpolarized to nearly −40 mV (Fig. 2 and Table 1), and at times to near −70 mV. Adding amiloride to the luminal perfusate depolarized V_t about 5 mV (not shown), demonstrating that most of the potential is due to Cl⁻ diffusion. When a 1:2 dilution of luminal NaCl-Ringer solution perfused the lumen, the mean V_t hyperpolarized from −2 to −9 mV (Fig. 2).

Figure 2. NaCl dilution diffusion potentials across small airways.

A, a representative trace of V_t with constant current pulses (1 μA) illustrating effects of (1) diluting luminal NaCl-Ringer solution 1:2 on V_t hyperpolarization, which indicates that Cl⁻ permeability significantly exceeds Na⁺ permeability (P_Cl/P_Na∼3), and (2) completely replacing luminal Cl⁻ with gluconate. B, summary of the results of spontaneous V_t in bilateral isotonic NaCl-Ringer solution (open bar), 1:2 NaCl-Ringer solution luminal dilution diffusion V_t (filled bar), and a 150 mm Cl⁻ gradient diffusion potential (hatched bar). The diffusion gradients were established by replacing the luminal solution with 75 mm NaCl or Cl⁻-free Ringer solution (150 mm NaGlu) as described in Methods (*P < 0.002, **P < 0.0002, n = 6).

Table 1.

Basic parameters: spontaneous V_t, G_t and

	Vt (mV)	Gt (mS cm−2)	(μA cm−2)
NaCl (L)/NaCl (B) (n = 37)	−2.1 ± 0.2	21.2 ± 0.9	44.4 ± 3.9
NaGlu (L)/NaGlu (B) (n = 13)	−8.5 ± 1.0	5.1 ± 0.3	43.6 ± 5.5
NaGlu (L)/NaCl (B) (n = 37)	−37.7 ± 1.6	9.4 ± 0.6	–

Luminal (L) and basolateral (B) sides of the tissue were perfused with either 150 mm NaCl or 150 mm NaGlu as indicated. Open circuit transepithelial potentials (V_t) were measured continuously, transepithelial conductances (G_t) were calculated from Ohm's law from the deflection in V_t during constant current pulses (∼1 μA) passed across the tissue, and equivalent short circuit currents () were calculated from transepithelial conductance (G_t) and spontaneous V_t. (n = number of measurements in as many tissues; means ± SEM.)

Since the maximum free solution Cl⁻ diffusion potential for a 150 mm Cl⁻ gradient is less than ∼−10 mV, the fact that these preparations exhibited such large Cl⁻ diffusion potentials indicates that the isolated small airways (1) remain intact with electroconductive properties that are well preserved when mounted in the capillary Ussing chamber (Fig. 1 D,E) and (2) are constitutively highly anion selective, comparing favourably in this property with the salt absorbing human sweat duct (Quinton, 1986).

Basic electrophysiological parameters: spontaneous V_t, G_t and

In the presence of bilateral NaCl-Ringer solution, the mean values of spontaneous V_t, G_t and were −2 mV (lumen negative), 21 mS cm⁻², and 44 μA cm⁻², respectively. When the impermeant anion gluconate replaced Cl⁻ bilaterally (150 mm sodium gluconate (NaGlu)/150 mm NaGlu), V_t hyperpolarized 4-fold, and G_t decreased 4-fold, leaving essentially unchanged (Table 1).

Absorptive and secretory transport inhibitors

We applied amiloride (10 μm) to the luminal surface to block Na⁺ conductance and inhibit electroconductive Na⁺ absorption through ENaC to indicate the magnitude of the spontaneous absorptive transport current in symmetric 150 mm NaCl-Ringer solutions. Amiloride consistently and significantly depolarized V_t about 32%, decreased G_t slightly (10%), and inhibited 35% of the spontaneous (Tables 2 and 3; Fig. 3). When the Ca²⁺ activated Cl⁻ channel (CaCC) inhibitor, niflumic acid (NFA), was applied to the luminal surface, V_t likewise depolarized by 36%, G_t decreased by 20% and fell by 49% (Table 2 and Fig. 4A). However, when the lumen was perfused with GlyH-101 to block the CFTR Cl⁻ channel, V_t hyperpolarized by 50%, G_t decreased by 22% and increased by 19% (Table 2 and Fig. 4B). Alternatively, when the basolateral side of the tissue was treated with bumetanide to inhibit the Na⁺–K⁺–2Cl⁻ cotransporter in the basolateral membrane and block Cl⁻-dependent secretion, V_t depolarized 58%, G_t was not affected, but 58% of the spontaneous was inhibited (Table 2).

Table 2.

Effect of ion transport inhibitors in bilateral NaCl

		Vt (mV)	Gt (mS cm−2)	(μA cm−2)
Amiloride	Pre	−2.2 ± 0.2	18.9 ± 1.9	42.3 ± 7.0
(n = 10)	Post	−1.5 ± 0.3*	17.0 ± 1.4*	27.4 ± 6.3*
NFA	Pre	−2.2 ± 0.1	20.5 ± 0.8	45.0 ± 3.7
(n = 5)	Post	−1.4 ± 0.2*	16.6 ± 1.2*	23.1 ± 2.9*
GlyH-101	Pre	−1.6 ± 0.5	26.0 ± 8.7	38.3 ± 15.4
(n = 16)	Post	−2.4 ± 1.1*	20.3 ± 7.3*	45.5 ± 21.1*
Bumetanide	Pre	−1.9 ± 0.3	25.0 ± 2.4	45.7 ± 8.3
(n = 6)	Post	−0.8 ± 0.2*	24.9 ± 1.9	19.0 ± 3.9*

In bilateral NaCl-Ringer solution, after stabilizing (Pre) either the Na⁺ transport inhibitor amiloride (10 μm) or the anion transport inhibitor GlyH-101 (50 μm) or niflumic acid (NFA, 100 μm) was added to the luminal solution or bumetanide (100 μm) was added to the contraluminal solution (Post). The effects of each inhibitor on transepithelial voltage (V_t), conductance (G_t), and equivalent short circuit current () were compared to its paired control condition (Pre).

^*Significantly different (P < 0.03) from paired value before (Pre) applying drug. (n = number of measurements in as many tissues; means ± SEM.)

Table 3.

Effects of combining amiloride with bumetanide and niflumic acid

	Vt (mV)	Gt (mS cm−2)	(μA cm−2)
Control (n = 9)	−1.5 ± 0.3	28.6 ± 4.2	42.2 ± 8.7
Amil	−1.0 ± 0.2*	26.1 ± 3.9*	27.0 ± 7.1*
Amil + Bumet	−0.3 ± 0.1*	26.1 ± 3.6	7.9 ± 2.0*
Control (n = 6)	−1.7 ± 0.2	28.3 ± 1.3	47.0 ± 4.4
Amil	−1.1 ± 0.2*	24.9 ± 1.1*	28.1 ± 3.7*
Amil + NFA	−0.2 ± 0.0*	22.1 ± 0.5*	3.5 ± 0.8*
Control (n = 5)	−1.9 ± 0.6	19.9 ± 3.6	32.1 ± 6.4
Amil	−1.4 ± 0.4*	16.5 ± 3.7	22.4 ± 4.4*
Amil + GlyH101	−1.1 ± 0.3	17.0 ± 3.4	14.9 ± 2.0

In bilateral NaCl, first, the ENaC inhibitor amiloride (Amil 10 μm) was added to the luminal solution to block Na⁺ absorption, and then the Na⁺–K⁺–2Cl⁻ cotransporter inhibitor bumetanide (Bumet, 100 μm) was added to the contraluminal solution to block Cl⁻ secretion. Similarly, after applying amiloride, the Ca²⁺ activated Cl⁻ channel inhibitor niflumic acid (NFA, 100 μm) was applied to the luminal side. Also as above, after amiloride, the lumen was perfused with the CFTR Cl⁻ channel inhibitor GlyH-101 (50 μm).

^*Significantly different from preceding value (P < 0.03). (n = number of measurements in as many tissues; means ± SEM.)

Figure 3. Effects of inhibiting of absorption and secretion.

A representative trace of V_t with constant current pulses shows the depolarizing effects of adding the ENaC inhibitor amiloride (10 μm) to the luminal surface followed by adding the Na⁺–K⁺–2Cl⁻ cotransporter inhibitor bumetanide (100 μm) to the serosal bath on V_t. Bilateral 150 mm NaCl-Ringer solution. See Tables 2 and 3 for quantitative summaries of the effects on V_t, G_t and .

Figure 4. Effects of anion conductance inhibitors on spontaneous transport.

Traces of spontaneous electrical properties illustrating effects of inhibitors in bilateral 150 mm NaCl-Ringer solution. A, Ca²⁺ activated Cl⁻ channel inhibitor niflumic acid (100 μm) depolarized V_t slowly. B, in contrast, CFTR inhibitor GlyH-101 (50 μm) acutely hyperpolarized V_t. See Tables 2 and 3 for quantitative summaries of the effects on V_t, G_t and .

To determine whether the residual current after blocking Na⁺ absorption with amiloride was due to Cl⁻ secretion, we added basolateral bumetanide (100 μm), which further depolarized V_t by 70% and inhibited most of the remaining without affecting G_t (Table 3 and Fig. 3). In a few experiments, we reversed the order of adding bumetanide and amiloride with similar results. Likewise, when amiloride and NFA (see below) were applied together, V_t depolarized by 88%, G_t fell 22% and the was almost completely blocked, indicating that as with bumetanide and amiloride, both absorptive and secretory transport were almost completely inhibited (Table 3). On the other hand, although blocking CFTR with GlyH-101 in the presence of amiloride had no significant effect on G_t, it significantly depolarized V_t by 21%, but only further decreased by 32% to about half of the spontaneous, uninhibited (Table 3).

Specificity of GlyH-101 and niflumic acid for Cl⁻ conductances

We then sought to better delineate the components of the Cl⁻ conductance. To confirm that GlyH-101 and NFA were acting on Cl⁻ conductances, we applied each inhibitor to the luminal surface in the presence of (1) amiloride to remove the Na⁺-dependent transport potential and ENaC conductance, (2) IBMX/forskolin (Fsk) to maximally activate CFTR-dependent Cl⁻ conductance, and (3) a hyperpolarizing 150 mm Cl⁻ gradient (150 mm Glu (L)/150 mm Cl⁻ (B)) to reveal effects on the Cl⁻ diffusion potential. Under these conditions, both NFA and GlyH-101 depolarized the diffusion V_t, although NFA acted more slowly (Fig. 5A and B). When GlyH-101 and NFA were added sequentially, G_t decreased from 31.7 mS cm⁻² to 24.4 mS cm⁻² after adding GlyH-101 and to 19.6 mS cm⁻² after adding NFA, indicating that the effects were additive and consistent with blocking separate conductances (data not shown; n = 6; all effects differed significantly from controls). These combined results indicate that both inhibitors block anion (Cl⁻) conductances in the apical membrane.

Figure 5. Effects of Cl⁻ conductance inhibitors on V_t in the presence of 150 mm Cl⁻ gradient.

A and B, representative traces of the effects of Cl⁻ conductance inhibitory drugs on V_t: NFA (100 μm) (A) and GlyH-101 (50 μm) (B). C and D, summary of V_t responses after luminal Cl⁻ replacement to niflumic acid (*P < 0.05, n = 5) (C) and GlyH-101 (P < 0.04, n = 5) (D). All measurements made in the presence of amiloride, IBMX+Fsk, and a lumen-to-serosa 150 mm Cl⁻ gradient (150 mm Glu (L)/150 mm Cl⁻ (B)). Both inhibitors significantly reduced the transepithelial bi-ionic Cl⁻ diffusion potentials, which indicates the likely presence of both CFTR and a CaCC in the small airway. Traces as described in Fig. 2. *Significantly different (P < 0.05).

Transport agonists UTP and Fsk/IBMX

To further examine whether separate anion conductances might be associated with absorption and secretion, we assayed for the effects of UTP, a Ca²⁺ mediated purinergic agonist, on a CaCC, and for the effects of the cAMP mediated agonist, Fsk/IBMX, on CFTR. In bilateral NaCl, as expected for a secretory response, application of UTP (100 μm) after an acute peak response (Fig. 6) maintained V_t significantly hyperpolarized by 45%, increased G_t by 12% and increased by 67% compared to spontaneous values (Table 4 and Fig. 6). On the other hand, when the adenylyl cyclase activator, Fsk (10 μm), and the phosphodiesterase inhibitor, IBMX (100 μm), were applied together to increase intracellular cAMP and activate CFTR, V_t significantly depolarized by 37% as G_t significantly increased by 68% (Table 4 and Fig. 6), although mean decreased slightly without statistical significance.

Figure 6. Effects of agonists on transepithelial ion transport.

A representative trace of V_t with constant current pulses (1 μA) showing effects of applying apical purinergic agonist UTP (100 μm) or cAMP mediated agonist Fsk (10 μm) + IBMX (100 μm) in bilateral 150 mm NaCl-Ringer solutions. Left trace shows the effects of purinergic agonist UTP (100 μm) on V_t, which acutely hyperpolarized before returning to a level significantly hyperpolarized compared to pre-stimulation levels. Right trace shows the effects of CFTR agonists Fsk (10 μm) plus IBMX (100 μm) on V_t, which depolarized slightly.

Table 4.

Effects of Agonists in bilateral NaCl

	Vt (mV)	Gt (mS cm−2)	(μA cm−2)
Control (n = 6)	−2.0 ± 0.5	24.4 ± 5.3	43.3 ± 8.4
UTP (peak)	−4.0 ± 0.6*	26.6 ± 5.7*	94.2 ± 11.7*
UTP (stable)	−2.9 ± 0.5*	27.4 ± 5.4*	72.6 ± 12.0*
Control (n = 8)	−1.9 ± 0.5	19.1 ± 1.6	38.9 ± 12.4
FSK/IBMX	−1.2 ± 0.5*	32.1 ± 4.1*	34.6 ± 14.0

In bilateral NaCl, the purinergic agonist UTP (100 μm) was added to the luminal solution. Adenylyl cyclase activator Fsk (10 μm) combined with phosphodiesterase inhibitor IBMX (100 μm) was applied to the luminal side after recording the spontaneous condition (Control). No inhibitors were present with the agonists.

^*Significantly different from control value (P < 0.04). (n = number of measurements in as many tissues; means ± SEM.)

To confirm that these agonists were acting on Cl⁻ conductances, we again applied each in the presence of a 150 mm hyperpolarizing Cl⁻ gradient (150 mm Glu⁻ (L)/150 mm Cl⁻ (B)). In separate experiments (data not shown), Fsk/IBMX acutely and significantly hyperpolarized V_t by −12.8 ± 3.7 mV (n = 8; P < 0.01), and UTP acting somewhat more slowly also significantly hyperpolarized V_t by −4.6 ± 0.8 mV (n = 7; P < 0.001). Both responses are consistent with activating apical Cl⁻ conductances. Additionally, we also inhibited the UTP response with luminal NFA in a few experiments, which not only inhibited the response, but also reduced the to 50% of the control spontaneous indicating the presence of spontaneous CaCC mediated secretion.

Stimulated secretion and Na⁺ conductance (G_Na)

If the epithelium consists of cells capable of bidirectional transport that spontaneously absorb and secrete NaCl in unison, converting the epithelium to secretion should require inhibiting or significantly down-regulating apical membrane G_Na and would thereby decrease G_t (see Introduction). To test whether the epithelium performs as a single uniform group of transporting cells that oscillate between secretion and absorption, we assayed for changes in ENaC-dependent, amiloride sensitive, Na⁺ conductance before and after stimulating secretion (150 mm Cl⁻ (L)/150 mm Cl⁻ (B)). We measured G_t before and after adding amiloride to the luminal solution and repeated the two measures again after stimulating with a secretagogue, Fsk/IBMX or UTP. In bilateral NaCl, stimulating with either agonist had no significant effect on G_Na, defined here as the change in G_t due to amiloride. That is, there was no difference in the amiloride sensitive component of G_t after stimulating the tissue with either secretagogue (Table 5). Similarly, no difference in the amiloride sensitive G_Na occurred when these manoeuvres were carried out in the presence of a Cl⁻ gradient (150 mm Glu⁻ (L)/150 mm Cl⁻ (B)); data not shown).

Table 5.

Effect of stimulating secretion on Na⁺ conductance

	ΔGNa (mS cm−2)
	Pre-agonist	Post-agonist
UTP (n = 10)	1.4 ± 0.4	1.3 ± 0.4	P > 0.8
Fsk/IBMX (n = 7)	2.6 ± 0.8	2.8 ± 1.0	P > 0.6

ENaC inhibitor amiloride (10 μm) was applied to the luminal surface before (Pre) and after (Post) adding UTP or Fsk/IBMX. ΔG_Na is the decrease in conductance that amiloride induced under each condition. (n = number of measurements in as many tissues; values are means ± SEM.)

Discussion

While large airway fluid balance might be achieved by matching the volume of fluid secreted by submucosal glands with the volume of fluid absorbed by the airway surface epithelium, it is generally accepted that the airway surface epithelium per se is capable of both processes independent of glands (Al-Bazzaz, 1994; Ballard et al. 1995; Boucher, 2003; Chambers et al. 2007). However, since CFTR is highly expressed in the airways (Kreda et al. 2005; Wang et al. 2005) and defective CFTR in cystic fibrosis affects both absorption and secretion in general (Quinton, 2007), a better understanding of the normal processes of maintaining ASL balance may help to explain and possibly to control some forms of airway pathogenesis. Toward that end, we modified a microperfusion device (Burg et al. 1968) to enable measurements of V_t, G_t and of small airways on the order of 1 mm φ (Fig. 1).

With this preparation, we asked whether the fluid transporting cells of the airway surface epithelium are all inherently capable of bidirectional transport, that is of both secreting and absorbing, or whether the epithelium consists of more than one type of cell, one that only secretes and another that only absorbs. Since fluid cannot be transported in opposite directions simultaneously by the same cell, if the epithelium were composed of similar cells that can transport in either direction, it would be obligated to alternate directions of transport to maintain ASF volume (Fig. 7A). On the other hand, concurrent transport in both directions could occur if the epithelium consists of intrinsically secretory and intrinsically absorptive epithelial units that transport independently (Fig. 7B).

Figure 7. Schematic model and structure of small airways.

A, schematic model of an epithelium composed of cells that maintain the ASF layer by secreting (blue) fluid until the volume becomes excessive, whereupon the same cells reverse the direction of transport to absorb (red) fluid. When the volume depletes, the process reverts to secretion again. Appropriate fluid levels in this model would be maintained by continuous oscillations between absorption and secretion along the airway. (Blue represents secretory capacity; red represents absorptive capacity.) B, schematic model of separate groups of absorptive and secretory cells in airway epithelia that maintain the fluid layer covering the small airways. Cells located within the pleats of the epithelia secrete (blue) fluid while cells located around the folds in the epithelium concurrently absorb (red) secreted fluid. The volume of ASF is maintained by the relative activities of the two processes. C, cross section of small airway epithelium illustrating its arrangement into folds and pleats that run parallel to the longitudinal axis of the airway (perpendicular to the plane of the page). Mucous cells are generally found in clusters at the contraluminal bases of the pleats (intensely stained cells; inset scale bar = 50 μm). This arrangement suggests a distribution of secretory cells toward the bottom of the pleats proximal to the mucus cells with absorptive cells located more distally toward and along the luminal folds in the epithelium. (Periodic acid Schiff staining, diameter ∼1 mm; scale bar = 100 μm.)

Spontaneous transepithelial electrolyte transport

In bilateral NaCl-Ringer solutions, the unstimulated small airway epithelia exhibited a small, lumen negative electrical potential of about −2 mV, consistent with values reported earlier for small airways (Welsh et al. 1983; Al-Bazzaz et al. 1991; Ballard et al. 1992; Al-Bazzaz, 1994; Wang et al. 2005; Blouquit et al. 2006). This voltage derives from a G_t (21 mS cm⁻²) and (44 μA cm⁻²) (Table 1; Figs 2, 3 and 4). It seems unlikely that the small V_t is the result of tissue trauma, leaks, or edge damage, since bi-ionic Cl⁻:gluconate diffusion potentials (150 mm Glu⁻ (L)/150 mm Cl⁻ (B)) hyperpolarized V_t more than 20-fold (Table 1; Fig. 2). Similarly, 1:2 NaCl dilution potentials also hyperpolarized V_t 4- to 5-fold (Fig. 2), showing that the epithelium is highly and constitutively selective for Cl⁻ with selectivity for Cl⁻ over Na⁺ of more than 3:1. These values characterize this tissue as an epithelium with a moderately low resistance due to a highly selective anion conductance that spontaneously and constitutively transports electrolytes, consistent with previous findings that attributed the high anion selectivity to CFTR (Wang et al. 2005).

Concurrent absorption and secretion

Since amiloride specifically blocks ENaC-dependent Na⁺ absorption (Benos, 1982), the fact that amiloride alone blocked less than 40% of the (Table 2 and 3; Fig. 3) not only denotes constitutive Na⁺ absorption activity, but also the presence of additional, concurrently active ion transporting components. The fact that basolateral bumetanide, apical NFA or GlyH-101 all largely inhibited the remaining amiloride insensitive current (Table 3; Figs 3 and 4) seems to indicate concurrent secretory and absorptive transport. The fact that GlyH-101 combined with amiloride was less effective than either bumetanide or NFA may suggest that CFTR participates in both absorptive and secretory units, but that its role in secretion is not as predominant as in absorption as observed in human sweat secretion (Quinton & Reddy, 1991; Reddy & Quinton, 1992b).

On the other hand, since secretion in general depends on Cl⁻, the fact that was virtually unchanged by removing Cl⁻ bilaterally (Table 1) argues that the spontaneous ion transport of the unstimulated airway is essentially absorptive, with little, if any, contribution from simultaneous, spontaneous secretion. This result seems counter to the notion of concurrent spontaneous secretion and absorption in Cl⁻-replete solutions, but if no spontaneous secretion is present, it is even more puzzling that amiloride inhibited less than half of the current in bilateral Cl⁻. If these effects are on an epithelium with bidirectional transport capacities, inhibiting ENaC with amiloride may convert absorption to an NKCC-dependent Cl⁻ secretion, while inhibiting NKCC with bumetanide might enhance ENaC-dependent Na⁺ absorption by lowering intracellular Cl⁻ (Xie & Schafer, 2004; Adam et al. 2005b). Reduced intracellular Cl⁻ may play a role in the increased ENaC activity and Na⁺ absorption reported for CF airways (Boucher, 2003, 2007) and CF colon (Mall et al. 1999) that lack CFTR-dependent Cl⁻ conductance. If this is the case, it seems very serendipitous that the increase in Na⁺ absorption in Cl⁻ free media almost exactly matches the amount of Cl⁻ secretion lost upon completely removing Cl⁻ whereupon the epithelial cells must be able to switch almost instantly from one state to the other as the absorptive or secretory function is inhibited.

Still, if this epithelium is strictly absorbing Na⁺ in bilateral NaCl, it is peculiar that the spontaneous V_t is so very small (∼− 2 mV) compared to the V_t values of other epithelia characterized by predominately ENaC-dependent Na⁺ absorption. That is, in bilateral isotonic NaCl, the V_t of the salivary duct is about −30 mV (Bijman et al. 1983); the sweat duct is about −15 mV (Reddy et al. 2005); renal collecting duct is about −20 mV (Barratt et al. 1975); distal colon is about −20 mV (Schultz et al. 1977); and urinary bladder and frog skin are from −60 to −70 mV (Erlij, 1976; Leaf, 1982). This comparison begs the question of why the V_t of small airway is uniquely so low. The fact that secretory epithelia are low resistance due to highly conductive paracellular Na⁺ shunts argues that in small airways, secretory units are arranged in parallel with absorptive units (Fig. 7B) and may thereby shunt the V_t of the Na⁺ absorbing epithelia. At this point, we have no clear explanation for this apparent paradox in Cl⁻ free media.

Distinct Cl⁻ conductance inhibitor effects

If secretion accounts for a significant fraction of the spontaneous current in isotonic bilateral NaCl-Ringer solution, Cl⁻ channel blockers should decrease the spontaneous . The depolarization of V_t and inhibition of about half of the spontaneous current after luminal application of the CaCC inhibitor niflumic acid (NFA) is consistent with inhibiting a spontaneous Cl⁻ dependent secretory current (Table 2 and Fig. 4). That is, during secretion, decreasing Cl⁻ diffusion through an apical anion conductance acts to hyperpolarize the apical membrane and thereby to depolarize V_t and reduce (V_t = V_b–V_a, where V_a is the apical membrane potential and V_b is the basal membrane potential). However, the fact that the CFTR inhibitor GlyH-101 decreased G_t and simultaneously hyperpolarized V_t without decreasing seems most consistent with predominantly inhibiting Cl⁻ conducting channels in an absorptive system. That is, during electrogenic Na⁺ (salt) absorption, decreasing Cl⁻ movement through an apical anion conductance shifts the apical membrane potential toward the Na⁺ membrane potential and depolarizes the apical membrane, which hyperpolarizes V_t. To wit, the loss of the apical CFTR Cl⁻ conductance in cystic fibrosis results in marked hyperpolarization of airways and sweat ducts – in the latter case by about an order of magnitude (Quinton, 1983). The slight increase in seems paradoxical, but might be due to a small stimulatory effect of lowering intracellular Cl⁻ on Na⁺ absorption as mentioned above (Dinudom et al. 1995; Xie & Schafer, 2004; Adam et al. 2005a).

Barring secondary effects on basolateral K⁺ conductance, these results imply that NFA prevents Cl⁻ exit from, while GlyH-101 seems to predominantly prevent Cl⁻ entry into, the cell, which is impossible in the same cell since the electrochemical driving force cannot be both inward and outward at the same time in the same cell for the same ion. These results indicate that each drug acts on separate groups of cells, one constitutively secretory possibly dominated by a CaCC and the other constitutively absorptive possibly dominated by a CFTR Cl⁻ conductance. This is not to conclude that CFTR is expressed only in absorptive cells. Certainly, CFTR clearly plays both roles elsewhere (Quinton & Reddy, 1991; Reddy et al. 1992).

Distinct agonist effects

The possibility of two distinct Cl⁻ conductances distributed between separate groups of absorptive and secretory cells may be supported further by the disparate responses of V_t to the purinergic Ca²⁺ mediated agonist UTP and to the cAMP mediated agonist forskolin combined with IBMX (Fsk/IBMX). Apical UTP caused a sharp hyperpolarizing peak of more than double the spontaneous V_t, which stabilized at about 140% of the spontaneous V_t. The hyperpolarized V_t was associated with a correspondingly large increase in and a slight increase in G_t (Table 4 and Fig. 6). In distinct contrast, Fsk/IBMX markedly increased G_t with a consistent depolarization of V_t and a slight increase in (Table 4 and Fig. 6). Again, the agonists, like the inhibitors above, appear to be acting on separate anion conductances in separate types of epithelia each with a Cl⁻ driving force directed in opposite directions – secretion or absorption. That is, if UTP predominantly stimulated cells with an apical CaCC in the presence of an outwardly directed Cl⁻ electrochemical potential favouring Cl⁻ secretion, V_t would hyperpolarize as seen, whereas if Fsk/IBMX predominantly stimulated cells with CFTR Cl⁻ channels in the apical membrane in the presence of an inwardly directed Cl⁻ electrochemical potential favouring absorption, V_t would depolarize as seen.

We expected UTP to activate a CaCC-dependent Cl⁻ conductance and Fsk/IBMX to activate a CFTR-dependent Cl⁻ conductance in the apical membrane of secretory cells in order to hyperpolarize V_t. However, the fact that Fsk/IBMX did not mimic UTP, but instead elicited a significant depolarization of V_t with a clear increase in G_t without changing suggests that it activated an anion conductance that is not rate limiting for spontaneous Cl⁻ absorption. This response resembles that of Fsk/IBMX stimulation of increased CFTR Cl⁻ conductance in the purely absorptive sweat duct (Quinton & Reddy, 1991; Reddy & Quinton, 1992a). Again, we do not exclude the possibility that Fsk/IBMX stimulates both secretory and absorptive cells.

Stimulation does not alter ENaC conductance

If salt absorption occurs via ENaC (Na⁺) and CFTR (Cl⁻) in the apical membrane of cells that also secrete, the apical Na⁺ conductance must be inactivated when bidirectional cells shift from absorbing to secreting. Inactivation is necessary to prevent Na⁺ from depolarizing the apical membrane and undermining the hyperpolarization essential to drive Cl⁻ secretion. Moreover, for epithelial fluid secretion to occur, Na⁺ must pass paracellularly through the tight junctions (Silva et al. 1977). If ENaC remained open, Na⁺ would simply recycle through the cell and luminal NaCl accumulation in the lumen would be subverted (Silva et al. 1977).

In short, the amiloride sensitive ENaC-dependent portion of G_t should become significantly lower, if not negligible, during secretion when an apical membrane Na⁺ conductance would be a liability. In symmetrical NaCl-Ringer solutions (150 mm NaCl (L)/150 mm NaCl (B)), when either UTP or Fsk/IBMX was applied to stimulate secretion in the small airway, there was no detectable change in G_t in response to amiloride (Table 5) or in the presence of a 150 mm Cl⁻ gradient (150 mm Glu (L)/150 mm Cl⁻ (B)). These results indicate that the secretory epithelium of the airway is composed of secretory cells that respond to UTP (and Fsk/IBMX), but do not have an apical Na⁺ conductance that deactivates upon stimulation to secrete and therefore do not have the ability to absorb. Moreover, this result also indicates that ENaC in absorptive cells of the small airway is insensitive to secretagogue and that in turn these cells are therefore not likely to have a coexisting secretory capacity.

Organization of concurrent absorption and secretion in airway epithelia

If the transport functions of the airway are performed by separate, independently secreting and absorbing epithelial units, the microanatomy of the airway surface by analogy with other organ systems suggests a parallel or tandem structure–function arrangement (Fig. 7B). It is common for groups of secretory cells to be structurally proximal to, but separated from, groups of absorptive cells in the same organ (e.g. secretory acini and absorptive ducts in salivary glands and pancreas; secretory tubules and absorptive ducts in sweat glands; secretory crypts and absorptive villi in intestines). Given that airway surfaces consist of multiple parallel folds of epithelial cells along the longitudinal axis of the airway tube (Fig. 7C), absorptive cells should be located along the more luminal surfaces of the folds while the secretory cells should be located along the more contraluminal sides of the pleats between the folds (Fig. 7B). The ends of the pleats are frequently characterized by the presence of multiple mucous cells (Fig. 7C) whose mucin products are likely to require fluid secretion by neighbouring cells for optimal discharge. Such a structural arrangement for absorptive and secretory epithelia in the airway would support constitutive and/or regulated absorption as well as concurrent constitutive and/or regulated secretion that in effect would ‘recycle’ fluid secreted from within the pleats back through the absorptive cells along the more luminal surfaces of the folds (Fig. 7B). Intuitively, this coupling of function with structure would inherently modulate or avert excessive luminal fluid accumulation.

Since active fluid movements in either direction are driven by differences in water activity created by the net transport (accumulation) of osmolytes across a semipermeable barrier, water always flows passively according to its activity gradient across water permeable epithelia (Tormey & Diamond, 1967; Sackin & Boulpaep, 1975; Quinton, 1979), including airways (Verkman et al. 2000). In this specific case, mainly net secretion of NaCl into the pleated space would lower the activity of the water in the adjacent ASF causing water from the basolateral extracellular fluid space to move down its chemical concentration gradient into the pleated spaces of the lumen to achieve equilibrium (isotonicity) (Fig. 7B). Fluid reabsorption would be the reverse of this process driven by active net absorption of mainly NaCl from ASL on the surfaces around the luminal folds. Water follows NaCl transported into the basolateral intercellular spaces of the epithelial cells of the folds and thence into the contraluminal extracellular space within the folds to complete a cycle of fluid movement (Fig. 7B).

The observation that fluid is virtually always, if not always, secreted into confined compartments (tubules, crypts, acini) as opposed to unconfined, open flat surfaces may support the need for separate secretory structures in the airways (Fig. 7B and C). A confined compartment provides for the accumulation of solutes essential for iso-osmotic fluid flow (Diamond & Tormey, 1966; Sackin & Boulpaep, 1975). While overt osmotic gradients are probably rapidly dissipated by high membrane water permeability (Sackin & Boulpaep, 1975), the pleats between the folds in the airway surface provide a ‘confined’ compartment to accommodate solute/water coupling for efficient isotonic fluid secretion. During breathing, the pleats in the small airways tend to open and collapse with each breath cycle so that the confined compartments in the pleats must transiently shorten or open with each inspiration. A slight extension of this concept suggests that normal ASF volumes may be maintained almost autonomously if the fluid absorptive capacity of the folds normally exceeds the secretory rate of the pleats. That is, as the airway collapses and pleats close, secreted fluid would be forced out of the pleats over the folds where excesses would be removed by constitutive fluid absorption. Thus, the mechanics of the respiratory cycle could nearly autonomously maintain homeostasis of ASF levels, ensuring that the fluid covers the entire airway surface without inundating or drying the lumen.

For protection of the airway surface, secretory cell activity may be enhanced by local stimulation. That is, local irritation from extraneous debris would increase fluid secretion in excess of local absorption and engage the ciliary escalator in clearing that site. Once the stimulus is relieved, secretion would diminish and allow constitutive absorption to restore and maintain ASF volume to normal levels.

Proof of a structure–function arrangement for secretion and absorption in the airways will require more direct evidence of separate transport activities in the cells of the pleats and cells of the folds of the airway epithelial lining. If established, it will be noteworthy that the pleated and folded airway surfaces may have a crucial purpose well beyond simply accommodating the distension and contraction of airway diameters during the volume changes that accompany breathing.

Caveats

Edge damage

While the Ussing chamber provides the essential advantage of controlling solutions on both surfaces for transepithelial assays, the pressure on the edge of the tissue held between the two compartments unavoidably damages the epithelial barrier along the lines of contact, creating electrical shunts which yield underestimates of V_t and and overestimates of G_t (Erlij, 1976). We surmise that even though V_t in symmetrical isotonic NaCl was small, edge damage was not a major handicap because V_t compared well with other studies (Ballard et al. 1992; Al-Bazzaz, 1994; Blouquit et al. 2002; Wang et al. 2005). More critically, our mean bi-ionic Cl⁻ diffusion V_t was robust, −38 mV, amounting to two-thirds of the mean bi-ionic potential reported from undissected, native perfused airways of similar diameter (Wang et al. 2005) and markedly, very much more negative than any other reported values of this measure. Furthermore, our rule of discarding tissues that did not hyperpolarize to at least 25 mV should have avoided including significantly compromised preparations or those with leaks due to undetected airway branches in the assay field. That is, damaged tissue could not have mounted Cl⁻ gradient diffusion potentials much more negative than free solution (or maximally about −10 mV). Nonetheless, some edge damage is likely to be present causing underestimates of V_t and overestimates of G_t.

Small potentials

The most likely errors in the present data are in the accuracy of measurements of very small potentials, which potentially are corrupted by even very small drifts, asymmetries, or extraneous liquid junctions in the system. Although absolute values may contain such errors, most of the results used for interpretation are based on relative changes that occurred in the same preparation so that values were paired and most systematic errors should have cancelled or been greatly minimized. The pitfall here is that variations in small measurements lead to larger relative errors and reduce the ability to detect significant differences with a feasible number of observations. Thus, small effects might be missed erroneously or left uncertain.

Small airways and submucosal glands

We avoid referring to the small airways in our preparations as ‘bronchioles’ since, when strictly defined, ‘bronchioles’ are airways without cartilage or submucosal glands in the surrounding parenchyma. In pigs of the size used here, airways on the order of ∼1 mm diameter generally contain a sparse population of small submucosal glands. When we examined our specimens for gland density histologically, we found about 1–2 glands mm⁻² at this level, which is much less than reported earlier (Ballard et al. 1995) in pigs much smaller and younger than used here. While it may be possible to reduce the diameter of the capillary and examine smaller airways free of these structures, we have not done so. The area of airway epithelium enclosed by the capillary orifice was about 0.66 mm² suggesting that, on average, perhaps one gland may have been present in each tissue assayed.

It is doubtful that these glands contribute significantly to transepithelial electrophysiological properties since (1) electrical currents generated within the glands are likely to be dissipated by a cable effect before reaching the airway surface, and (2) apical changes in fluid composition and drugs seem unlikely to reach submucosal glands rapidly. Likewise, gland openings might have provided transepithelial shunts, but if these were consistent and significant, it seems unlikely that such large Cl⁻ diffusion potentials could have been supported across the tissue.

Drug specificity

GlyH-101 and NFA were intended to specifically block CFTR and CaCC, respectively. While there is evidence that niflumic acid also inhibits single CFTR channel activity (Scott-Ward et al. 2004), it has been used successfully to discriminate between these channels in native glands (Lee and Foskett, 2010; Ousingsawat et al. 2011) and airway cells (Ousingsawat et al. 2011). Likewise, forskolin with IBMX and UTP were intended to stimulate CFTR and CaCC, respectively, even though there is growing evidence that forskolin and IBMX may raise intracellular Ca²⁺ concentrations at least in some types of cultured cells (Namkung et al. 2011). Nonetheless, the dramatic difference in the responses in this native tissue between UTP and forskolin plus IBMX strongly suggests they are acting on distinct pathways. Although we cannot be certain that these were the only actions of these drugs, the responses seemed consistent with the intended actions and yet distinct enough in their elicited responses to warrant reasonable confidence in the presumed actions.

Summary

We have introduced a novel device and method for examining the transepithelial electrophysiological properties of small specimens of epithelia. Applying this system to isolated native small airways shows that this epithelium spontaneously exhibits a low transepithelial potential (V_t) and a moderately high, predominantly anion selective conductance that generates a stable equivalent short circuit current (). To block the current, both a Na⁺ absorption inhibitor, amiloride, and a Cl⁻ secretion inhibitor, bumetanide, were required, suggesting that both absorption and secretion occur concurrently. The contrasting effects of Cl⁻ conductance inhibitors and the contrasting effects of cAMP and Ca²⁺ mediated agonists also suggest that separate groups of secretory and absorptive cells compose the small airway epithelium that are concurrently and constitutively active. A new model places secretory cells at the base of pleats and absorptive cells toward the tips of the folds inherent to the plicated epithelia of distal airways. This arrangement of secretory and absorptive cells may provide for autonomous regulation of homeostatic ASF levels during normal conditions, while also providing for rapid fluid transport responses to counter threats to airway hygiene.

Glossary

ASF

airway surface fluid

CaCC

Ca²⁺ activated Cl⁻ channels

cAMP

cyclic adenosine monophosphate

CF

cystic fibrosis

CFTR

cystic fibrosis transmembrane conductance regulator

ENaC

epithelial sodium channel

Fsk

forskolin

G_Cl

chloride conductance

G_Na

amiloride sensitive sodium conductance

G_t

transepithelial conductance

GlyH-101

(N-(2-naphthalenyl)-[(3,5-dibromo-2,4-dihydroxyphenyl)methylene]glycine hydrazide

Glu

gluconate

IBMX

3-isobutyl-1-methylxanthine

equivalent short circuit current

NFA

niflumic acid

NKCC

Na⁺–K⁺–2Cl⁻ cotransporter

UTP

uridine 5′-triphosphate

V_t

transepithelial potential

Author contributions

P.Q. designed the study; A.S. performed the experiments and compiled the data; A.S. and P.Q. wrote and edited the manuscript. The work was conducted at the University of California, San Diego.