INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

AT THE CELLULAR LEVEL, loss of cystic fibrosis (CF) transmembrane conductance regulator (CFTR) function results in two primary electrophysiological abnormalities in the airway epithelia of CF patients. First is the loss of chloride permeability in response to cAMP, consistent with the role of the CFTR as a cAMP-regulated chloride channel (21, 24). Second, increased absorption of sodium across the airway epithelium through amiloride-sensitive sodium channels has been demonstrated in CF airways (19). The role that increased absorption of sodium plays in disease progression, if any, is not known at this time. It has been postulated that increased sodium absorption coupled with decreased chloride secretion serves to desiccate mucus secretions, which prevent proper clearance and enhance bacterial growth and infection. The activity of the epithelial sodium channel (ENaC) has been shown to be altered by the presence or absence of CFTR in transfected MDCK cells (29). ENaC-dependent sodium absorption is increased in the absence of the CFTR and is further stimulated by cAMP-increasing agents compared with cells cotransfected with both CFTR and ENaC. How the loss of CFTR function or expression results in sodium hyperabsorption is not yet clearly understood. The expression of the ENaC is reportedly unchanged between non-CF and CF airways, indicating that regulation of ENaC activity is in some way modulated by the CFTR, perhaps via a disruption of some physical interaction between CFTR and ENaC (2, 29).

Although CF mice do not develop detectable airway disease, the nasal epithelia of CFTR(F508/F508) and CFTR(/) mice serve as good models for the ion transport defect found in the human CF airway; they not only exhibit the chloride transport abnormality associated with reduced or lost CFTR function but also exhibit the hyperabsorption of sodium characteristic of human CF airways (8). Kelley et al. (18) have previously demonstrated that the combination of forskolin and milrinone is effective in restoring some CFTR function in the nasal epithelium of CFTR(F508/F508) mice but not in that of CFTR(/) mice. Kelley and colleagues (14, 15) have shown that activation results from blocking the activity of cGMP-inhibited phosphodiesterase (cGI-PDE), thereby raising the local concentration of cAMP near the apical plasma membrane of epithelial cells to a level sufficient to activate F508 CFTR. This signaling pathway would indicate that cGMP could play a significant role in the regulation of CFTR activity. We have been able to substantiate this mechanism by demonstrating nearly identical activation of chloride permeability in CFTR(F508/F508) mice with the combination of forskolin and C-type natriuretic peptide (CNP), an agonist for the membrane-associated guanylate cyclase (GC) B receptor (17). Natriuretic peptides have been shown to regulate chloride transport in shark rectal gland cells (28) and in porcine colon (1). Heat-stable enterotoxin (Sta) and guanylin, peptides that stimulate membrane-bound GCs, have been shown to activate the CFTR through protein kinase (PK) A-dependent pathways. These peptides are structurally and functionally related to the natriuretic peptides and have been shown to stimulate CFTR-mediated chloride transport in the human colonic cell lines T84 and Caco-2 (3, 5) through apparent cross-activation of PKA by cGMP.

In addition to their role as Cl secretagogues, natriuretic peptides within the renal and vascular systems have been shown to effectively inhibit sodium transport through cGMP-dependent mechanisms (reviewed in Ref. 25). For example, atrial natriuretic peptide stimulates cGMP formation via activation of GC-A and causes a decrease in amiloride-sensitive sodium absorption in renal inner medullary collecting duct cells (23). Similarly, decreasing cGMP production through the inhibition of GC activity by the compound LY-83583 increases amiloride-sensitive sodium absorption (23). Clearly, the regulation of electrolyte and fluid levels in the airways is an important process. An intriguing aspect of CNP as a natural modulator of ion transport in airway epithelium is the possibility that this peptide hormone may regulate sodium, as well as chloride, transport. Coupled with the ability of CNP to increase ciliary beat frequency (6), modulating both chloride secretion and sodium absorption through the production of cGMP would make CNP an excellent candidate for a primary regulator of airway clearance. It has been demonstrated that apical application of CNP can effectively increase cGMP levels and regulate ciliary beat frequency in airway epithelial cells (6). Production of CNP in the lumen has been shown to likely take place in stimulated macrophages and to be secreted in response to lipopolysaccharide or other challenges (30). In this paper, we explore whether CNP is capable of influencing sodium absorption in murine airways and examine the signaling pathways involved in sodium regulation compared with those involved in the cGMP-mediated stimulation of chloride secretion.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Measurement of mouse nasal transepithelial potential difference values. Mouse nasal transepithelial potential difference (TEPD) was measured with the protocols of Grubb et al. (8) and Kelley et al. (16). Briefly, mice were anesthetized with 10 µl/g body weight of 0.4 mg/ml of acepromazine, 11 mg/ml of ketamine, and 2 mg/ml of xylazine in PBS. PE-10 tubing drawn out to approximately one-half of its original diameter was inserted 2-3 mm into the nostril of the mouse. The solutions were perfused at room temperature by use of a Razel A-99 (Razel Scientific Instruments, Stamford, CT) syringe pump at a rate of ~7 µl/min. A series of valves was used to change solutions, with a delay time of ~45 s between solution change and solution contact with the nasal epithelium. Ringer solutions consisted of chloride-replete HEPES-buffered Ringer (HBR; 10 mM HEPES, pH 7.4, 138 mM NaCl, 5 mM KCl, 2.5 mM Na₂HPO₄, 1.8 mM CaCl₂, and 1.0 mM MgSO₄) and chloride-free HBR (10 mM HEPES, pH 7.4, 138 mM sodium gluconate, 5 mM potassium gluconate, 2.5 mM Na₂HPO₄, 3.6 mM hemicalcium gluconate, and 1.0 mM MgSO₄; all chemicals from Sigma, St. Louis, MO).

Measurement of mouse tracheal potential difference. Excised tracheae were mounted on holding pipettes, and the luminal and bath compartments were perfused independently. All experiments were performed at 37°C by placing all solutions and the mounted trachea in an Isolette infant incubator (Narco, Hatboro, PA). HBR was perfused to both the luminal and basolateral sides by gravity. Luminal perfusion rates were between 3 and 5 ml/min. TEPD was measured via 4% agar bridges in HBR placed on both the luminal and basal sides and connected through calomel electrodes to a DVC 1000 voltage-current clamp (WPI). Data were collected on a MacLab/4e from Advanced Instruments.

Mice. Mice were genotyped from tail clip DNA. F508 mice (Cftr^tm1Kth) were a generous gift from Kirk Thomas (University of Utah School of Medicine) and were genotyped by the procedures previously described (31). Cftr^m1Unc mice (27) were obtained from Jackson Laboratories and were genotyped as described by Koller et al. (22). To increase survival of CF animals, the mice were fed a liquid diet as described by Eckman et al. (4). The mice were cared for in accordance with Case Western Reserve University Institutional Animal Care and Use Committee guidelines.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Effect of CNP on amiloride-sensitive sodium absorption in mouse trachea. The effects of CNP and 8-bromo-cGMP (8-BrcGMP) on amiloride-sensitive sodium absorption were studied by measuring the TEPD in excised tracheae of normal mice (Fig. 1). The GC agonist CNP was tested for its ability to change sodium absorption by measuring depolarization of TEPD in both the presence and absence of amiloride. The results were compared with the effects of the direct addition of 8-BrcGMP. The baseline TEPD of these tissues was 3.7 ± 1.1 mV (n = 7 mice). CNP (1 µM) and 8-BrcGMP (100 µM) added to the luminal perfusate reversibly depolarized TEPD by 21.6 ± 8.1 and 23.3 ± 8.4%, respectively. Amiloride (100 µM) induced a 65.4 ± 2.9% depolarization of baseline TEPD. When tracheae were exposed to amiloride before and during CNP addition, CNP no longer had any effect on TEPD. These results indicate that CNP is able to reduce amiloride-sensitive sodium absorption across the epithelium of mouse tracheae, likely through a cGMP-dependent mechanism.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Effect of C-type natriuretic peptide (CNP) and 8-bromo-cGMP (8-BrcGMP) on tracheal transepithelial potential difference (TEPD). A: raw trace of TEPD assay with excised murine trachea in presence of HEPES-buffered Ringer (HBR) solution, CNP (1 µM), amiloride (Amil; 100 µM), and 8-BrcGMP (100 µM). B: average depolarization mediated by each compound given as percentage of baseline TEPD. Amil-pretreated CNP was measured relative to depolarization caused by Amil alone. Values are means ± SE and represent an average of 4 experiments.

Differential effects of the phosphodiesterase inhibitor milrinone and CNP on nasal TEPD of CFTR(/) mice. Other groups (13, 29) have demonstrated that agonists of PKA activity increased amiloride-sensitive sodium absorption in cells lacking CFTR but expressing the amiloride-sensitive ENaC. In the nasal epithelium of CFTR(/) mice, milrinone (100 µM) hyperpolarized the lumen negative TEPD by 2.0 ± 1.2 mV (n = 4 experiments) when measured in chloride-replete HBR in the absence of amiloride (Fig. 2). This result is consistent with our hypothesis that milrinone is stimulating PKA activity through the inhibition of cGI-PDE. Kelley et al. (16) have previously shown that CNP is capable of stimulating chloride secretion through a mechanism indistinguishable from that of milrinone in wild-type and F508 homozygous mice. In CFTR(/) mice, CNP caused a depolarization in the lumen negative potential of 6.3 ± 0.5 mV (31.5 ± 7.7% of baseline TEPD; n = 4 experiments; Fig. 2). Time-controlled traces with HBR alone varied only 0.3 ± 0.4 mV (n = 3 experiments) in the direction of depolarization. Treatment of the nasal epithelium of CFTR(/) mice with amiloride before and during CNP addition reduced any further depolarization of TEPD by CNP to ~1.5 ± 0.6 mV (Fig. 3). Although milrinone and CNP both act through cGI-PDE inhibition to stimulate chloride secretion, they differ in their ability to regulate sodium absorption. The results obtained with milrinone are consistent with previous reports that show that PKA activation increases amiloride-sensitive sodium absorption in CF cells (13, 29). However, the effects of CNP suggest that cGMP may have dual roles in sodium and chloride transport regulation by acting through different signaling pathways.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Effects of milrinone and CNP on nasal TEPD in CFTR(/) mice. A: averaged nasal TEPD traces showing points every 15 s. Data are change in TEPD (TEPD), with zero TEPD and zero time being defined as time either CNP (1 µM) or milrinone (100 µM) was added. Control, HBR time control. Positive values, hyperpolarization of lumen negative TEPD; negative values, depolarization. Values are means ± SE; nos. in parentheses, no. of mice. B: effects of conditions shown in A 150 s after drug addition. Values are means ± SE; nos. in parentheses, number of experiments. Significant difference compared with control: * P = 0.05; ** P < 0.001 by Duncan's multiple range test.

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Effect of CNP on nasal TEPD in CFTR(+/+,) (n = 5) and CFTR(/) (n = 4) mice in presence of Amil. Averaged traces show points every 15 s after addition of Amil (100 µM). Amil and CNP were perfused in presence of HBR. Zero time, time of CNP addition. Positive values, hyperpolarization of lumen negative TEPD; negative values, depolarization. Values are means ± SE. * P < 0.05 by t-test compared with equivalent time points in traces from CFTR(/) mice.

Effects of GC inhibition on mouse nasal TEPD. It has previously been shown that the GC inhibitor LY-83583 increases amiloride-sensitive sodium transport in rabbit medullary collecting duct cells (23). With the use of the mouse nasal TEPD assay, the GC inhibitors methylene blue (100 µM) and LY-83583 (50 µM) were perfused onto the nasal epithelium of normal mice. Lumen negative TEPD was hyperpolarized 3.3 ± 0.7 (n = 5 experiments) and 2.5 ± 0.6 mV (n = 8 experiments) with LY-83583 and methylene blue, respectively (Fig. 4). The addition of amiloride to the perfusing solutions depolarized the TEPD in the presence of either GC inhibitor. GC inhibitors added after the addition of amiloride had no effect on the TEPD. The hyperpolarization induced by LY-83583 also showed sensitivity to the addition of 8-BrcGMP, suggesting that cGMP is involved in the basal regulation of sodium absorption. The addition of milrinone, an agent to increase cAMP levels, had no effect other than a slight hyperpolarization of nasal TEPD. This result is consistent with the hyperpolarizing effect milrinone exhibited on the nasal TEPD of CFTR(/) mice as shown in Differential effects of the phosphodiesterase inhibitor milrinone and CNP on nasal TEPD of CFTR(/) mice.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Effect of guanylate cyclase (GC) inhibition on nasal TEPD in CFTR(+/+) mice before (A) and after (B) Amil addition. C: effect of addition (open bar) of milrinone (100 µM) and 8-BrcGMP (100 µM) on LY-83583-induced hyperpolarization of lumen negative TEPD. In A and C, TEPD was allowed to reach steady state before addition of drugs. In B, steady state with Amil was reached before addition of GC inhibitors. Amil (100 µM) and GC inhibitors [methylene blue (MethBlue; 100 µM) and LY-83583 (50 µM)] were perfused in presence of HBR. Zero time is time of GC inhibitor addition. Positive values, hyperpolarization of lumen negative TEPD; negative values, depolarization. Values are means ± SE; n, no. of experiments.

The effects of the GC inhibitor LY-83583 on mouse nasal TEPD were compared between normal mice and CFTR(F508/F508) mice. The nasal epithelium of the CFTR(F508/F508) mice exhibits the CF airway characteristic of hyperabsorption of sodium compared with that in non-CF mice. Baseline nasal TEPD values of CFTR(F508/F508) and non-CF mice were 19.3 ± 2.0 (n = 10) and 8.7 ± 1.3 mV (n = 10), respectively. Unlike the normal mice, the CFTR(F508/F508) mice exhibited no further hyperpolarization of lumen negative TEPD with the addition of LY-83583 (Fig. 5). After 2 min of exposure to LY-83583 (50 µM), the nasal TEPD of CFTR(F508/F508) mice depolarized an average of 0.3 ± 1.1 mV (n = 4). These data suggest that abnormal regulation of the cGMP signaling pathway may be at least partially responsible for the increased absorption of sodium across the airway epithelium, which is characteristic of CF.

View larger version (29K): [in this window] [in a new window]   Fig. 5.   Effect of GC inhibition on nasal TEPD in CFTR(+/+) and CFTR(F508/F508) mice. LY-83583 (50 µM) in HBR was perfused onto nasal epithelium of CFTR(+/+) and CFTR(F508/F508) mice followed by perfusion with Amil (100 µM). Zero time is time of GC inhibitor addition. Positive values, hyperpolarization of lumen negative TEPD; negative values, depolarization. Values are means ± SE; n, no. of experiments. *P < 0.05 by t-test compared with equivalent time points in trace from CFTR(F508/F508) mice.

Differential regulation of ion transport by PKA and PKG. GC control of sodium absorption implies a role for cGMP in this regulatory pathway. To assess the extent to which cGMP-dependent protein kinase is involved in the regulation of sodium absorption, the PKA- and PKG-specific inhibitors Rp diastereomers of adenosine 3',5'-cyclic monophosphothioate (Rp-cAMPS) and 8-(p-chlorophenylthio)guanosine 3',5'-cyclic monophosphothioate (Rp-8-pCPT-cGMPS), respectively, were tested for their effect on the nasal TEPD in wild-type mice (Fig. 6). Measured 3.5 min after Rp-8-pCPT-cGMPS application, PKG inhibition resulted in a 2.4 ± 0.9-mV hyperpolarization (n = 4 experiments) of lumen negative TEPD that was amiloride sensitive. PKA inhibition resulted in a further depolarization of TEPD. Nasal TEPD continued to depolarize an average of 1.8 ± 1.2 mV (n = 3 experiments) after 3.5 min and responded much less to amiloride in the presence of Rp-cAMPS.

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Effect of protein kinase (PK) A and PKG inhibition on nasal TEPD in CFTR(+/+) mice. Rp diastereomers of 8-(p-chlorophenylthio)guanosine 3',5'-cyclic monophosphothioate (Rp-cGMPS; 100 µM; n = 4; A) and adenosine 3'.5'-cyclic monophosphothioate (Rp-cAMPS; 100 µM; n = 3; B) in HBR were perfused onto nasal epithelium of CFTR(+/+) mice followed by perfusion with Amil (100 µM). Averaged traces show points every 15 s. Positive values, hyperpolarization of lumen negative TEPD; negative values, depolarization. Values are means ± SE. * P < 0.05 by t-test compared with equivalent time points in trace from Rp-cAMPS-treated mice.

Conversely, chloride secretion measured in response to forskolin and CNP was sensitive to PKA inhibition (Fig. 7). The addition of Rp-cAMPS reduced forskolin- and CNP-induced hyperpolarization from ~3 to 0.4 ± 0.6 mV (n = 4 experiments). Rp-8-pCPT-cGMPS did not significantly reduce hyperpolarization mediated by the combination of forskolin and CNP (2.3 ± 0.7 mV; n = 3 experiments). These data are consistent with previous findings (16) that CNP-induced chloride secretion in Calu-3 cells and in mouse nasal epithelia is dependent on PKA activity, whereas PKG activity has no significant effect on this pathway. Both sodium and chloride secretion in murine airway epithelium can be regulated through cGMP, although the specific signaling pathways appear to diverge (Fig. 8).

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Effect of PKA and PKG inhibition on chloride secretion in CFTR(F508/F508) mice. Either Rp-cGMPS (100 µM) or Rp-cAMPS (100 µM) (Rp-cN) was perfused in presence of Amil (100 µM), forskolin (Forsk; 10 µM), and CNP (1 µM) in chloride-free HBR onto nasal epithelium of CFTR(F508/F508) mice. No Agonist, time-controlled trace with chloride-free HBR and Amil (100 µM). A: averaged traces showing points every 15 s. Values are means ± SE; n, no. of experiments. Zero time and zero TEPD are point at which Forsk and CNP were added. B: TEPD 2 min after drug treatment. Positive values, hyperpolarization of lumen negative TEPD; negative values, depolarization. Significant difference compared with Forsk+CNP in absence of inhibitors: * P = 0.01; ** P = 0.002 (both by Duncan's multiple range test).

View larger version (21K): [in this window] [in a new window]   Fig. 8.   Schematic diagram showing dual regulation of chloride and sodium transport by cGMP at apical membrane of an airway epithelial cell. (+), Stimulation of activity; (), inhibition. PDE, phosphodiesterase; PDE III, cGMP-inhibited, cAMP-specific PDE; PDE V, cGMP-specific PDE.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Kelley and colleagues (14, 15, 18) have previously shown that CNP is capable of regulating CFTR-mediated chloride secretion through the inhibition of cGI-PDE by increasing PKA activity, presumably through localized elevations of cAMP concentration. CNP can substitute effectively for the cGI-PDE inhibitor milrinone in stimulating nasal epithelial chloride secretion in CFTR(F508/F508) mice when used in combination with forskolin. This observation raises the possibility that CNP is a natural regulator for airway epithelial chloride secretion (16).

As their name implies, the natriuretic peptides are most well known for their ability to regulate sodium transport. Along with a loss of cAMP-stimulated chloride secretion in CF airway epithelia, an increase in the rate of sodium absorption is also associated with the CF phenotype. A role for CNP in the regulation of sodium absorption in the airways would be consistent with the previously established role for natriuretic peptides. We have been able to show that CNP, via a cGMP-dependent pathway, decreases amiloride-sensitive ion transport. The larger CNP-induced decrease of TEPD in CFTR(/) mice compared with that in CFTR(+/+,) mice suggests that CF-related sodium hyperabsorption can be modulated by cGMP. Consistent with this hypothesis, we have shown that inhibition of either GC or PKG activity leads to an increase in amiloride-sensitive sodium transport in the nasal epithelia of non-CF mice. However, GC inhibition has no effect on sodium transport when tested in CFTR(F508/F508) mice, thus providing evidence that this pathway may already be involved in CF-associated sodium hyperabsorption (Fig. 5). Our data suggest that abnormal regulation of cGMP-dependent pathways may be at least partially involved in elevated sodium absorption found in CF airway epithelia. Although these data are consistent with other reports (23, 25) showing the regulation of sodium absorption by natriuretic peptides, there are data that indicate that cGMP has other effects on airway epithelial ion transport. Geary et al. (6) reported that CNP has no effect on either chloride or sodium transport in primary nasal epithelial cells obtained from scrapings placed in culture. It is possible that culture conditions altered the electrical response to CNP, although cGMP-mediated effects on ciliary beat frequency were preserved. Also, a recent paper by Schwiebert et al. (26) demonstrated that cGMP could stimulate an increase in sodium and chloride currents in rat tracheal epithelial cells. Our differing results may reflect the different techniques employed in each of the studies or differences between cGMP-mediated regulation of sodium transport in rat tracheal and mouse nasal and tracheal epithelial cells. Rat tracheal epithelial sodium transport appears to be regulated primarily through cyclic nucleotide-gated nonselective cation channels, whereas this study suggests that amiloride-sensitive channels play a larger role in mouse airways.

These findings may have implications for the use of pharmacological therapies for CF. The effects on sodium absorption should be considered when various pharmacological approaches are used to stimulate mutant CFTR activity. Our data are consistent with other reports that show that stimulation of PKA activity by agents that raise cAMP levels increases sodium absorption in CF airway epithelia (Fig. 2). Each of the compounds commonly tested, milrinone (18), genistein (11, 12), NS004 (7), and CPX (9, 10), requires adenylate cyclase activation to optimize the stimulatory effects of these compounds on CFTR activity. Although each of these compounds may be effective at stimulating chloride permeability through some mutant form of CFTR, a concomitant increase in already elevated sodium absorption will likely occur. It is not currently known what effect sodium absorption plays in disease progression, and further increasing the rate of absorption may be detrimental to the health of the patient. Most experimental protocols used for in vivo measurements of chloride permeability, such as nasal TEPD, utilize high concentrations of amiloride- and chloride-free Ringer solutions to optimize both electrical and chemical gradients for chloride secretion and generation of TEPD. Such conditions that promote net secretion would not be present in a therapeutic setting. As discussed by Knowles et al. (20), an increase in chloride permeability in a CF airway might be expected to result in a net absorption of fluid rather than secretion. This concept has led to the notion that chloride secretion, coupled with the inhibition of sodium absorption by amiloride, is likely to provide the most effective mode for driving the net balance toward fluid secretion. Finding agents such as CNP that may help restore a balance to chloride and sodium transport, as opposed to merely increasing CFTR-mediated chloride permeability, may have a more therapeutic benefit.

In summary, our data identify a crucial role for cGMP in the coregulation of ion transport in the airway epithelia. A previous report (6) demonstrated that CNP increases ciliary beat frequency in primary human airway epithelia. These observations suggest that CNP may play an important role in acute fluid balance and mucociliary clearance in mammalian cells. CNP is reportedly produced by murine peritoneal and bone marrow macrophages in response to lipopolysaccharide (30). Thus, if CNP is also produced by airway macrophages, CNP may represent an excellent candidate for a natural regulator of airway clearance. Perhaps modified forms of CNP and other possible airway clearance regulators that inversely regulate both chloride secretion and sodium absorption should be examined for therapeutic potential.