Hyperoxic inhalation (PO₂ higher than that in the ambient air) is used in a variety of clinical practices including anaesthesia and post-operative recovery. The duration of hyperoxic ventilation can be prolonged depending on the patient conditions such as peripheral O₂ saturation. Also, in patients with respiratory disorders which cause hypoxemia, long-term oxygen therapy is often used. Although the benefits of O₂ supplementation are obvious in clinical situations of fatal hypoxemia, there also are harmful effects of O₂. For example, the side effects of increasing concentrations of O₂ supplementation are frequently observed as pulmonary injury (1). In neonatal neural tissues such as the retina, the harmful influence of hyperoxic ventilation has also been observed (2). The cough reflex, a defensive respiratory reflex, is also impaired in hyperoxia, and the inhibition of cough reflex is prevented by dietary antioxidants (3).

It is generally thought that tissue injury caused by O₂ is mediated by the formation of reactive oxygen species (ROS), which can react with and damage essential biomolecules via lipid peroxidation, protein sulfhydryl oxidation, and DNA damage (4). Because airway epithelium is constantly and frequently exposed to oxidative stress, it is highly likely that ROS-mediated oxidative stress affects the functions of airway epithelium (5-7). In this context, there is increasing evidence for the protective effects of antioxidant supplementation in respiratory diseases (3, 4, 8).

Because humans cannot synthesize ascorbic acid, dietary uptake of vitamin C is essential to cope with oxidative stress and to preserve physiological homeostasis. It has been reported that vitamin C is present in airway surface liquid (ASL), a thin (10-30 µm) layer of fluid covering the luminal surface of the airway epithelium (9-11). The physiological role of vitamin C in ASL is to stimulate Cl^- secretion via cAMP-activated Cl^- channels known as cystic fibrosis transmembrane transport regulator (CFTR) in the luminal membrane of the airway epithelium (12). A balanced level of ASL is critical for the protection of the airway epithelium. For transepithelial fluid secretion, an electrogenic Cl^- secretion model is regarded as the ionic mechanism in which the cAMP-dependent activations of the luminal Cl^- channel (CFTR) and the basolateral K⁺ channel (potassium voltage-gated channel, KQT-like subfamily, member 1, KCNQ1) are critical steps (11, 13).

While there have been numerous studies on the structural and biochemical changes in respiratory epithelial cells in response to oxidative stress, a direct investigation on the physiological function (i.e., electrolyte secretion) has been rarely conducted. The studies by Cowley and Linsdell showed that exogenous hydrogen peroxide (H₂O₂, 0.5-2 mM) directly activates the electrogenic Cl^- secretion of Calu-3, a cell line model of serous airway epithelial cells (14). In contrast, the oxidative stress caused by pyocyanin, a redox-active phenazine compound, impairs CFTR-dependent Cl^- secretion in the bronchial epithelium (15). Consistent with this, vitamin C activates CFTR in primary cultured human airway epithelial cells (12).

Apart from the acute effects, chronic effects of oxidative stress and of vitamin C deprivation on the airway electrolyte transport are very important. To the best of our knowledge, there has been no investigation on the effects of sustained hyperoxic ventilation on the electrolyte secretion of airway epithelium in vivo. Also, the functional changes in the airway epithelium in the vitamin C-deficient animal model have not yet been investigated.

Unlike humans, rodents synthesize vitamin C (ascorbic acid) from glucose in situ. Recently, a mouse line has been generated with a targeted deletion of the gene coding for L-gulono-c-lactone oxidase (Gulo), which catalyzes the final step of ascorbic acid biosynthesis (16). Mice null for Gulo (gulo[-/-] mice) and that are not provided with dietary vitamin C supplements become scorbutic, lose weight, and eventually die. The ambient level of oxygen might induce oxidative stress when the intrinsic antioxidant levels are insufficient. In this respect, it was tempting to investigate the airway epithelial secretory function in gulo(-/-) mice exposed to normal atmospheric conditions.

Based on the results outlined above, we investigated electrolyte secretion and the absorption functions of mouse airway epithelium using the Ussing chamber apparatus to measure short-circuit current (I_sc). The changes in I_sc (ΔI_sc) reflecting the Cl^- secretion and Na⁺ absorption functions of airway epithelia were compared to mice exposed to hyperoxic conditions (80%-98% PO₂) and to a normoxic environment. Also, we compared the I_sc values in the airway epithelium of gulo(-/-) mice according to the duration of vitamin C deficiency.

An original trace of the Ussing chamber experiments with an ICR mouse is shown in Fig. 1A, in which the V_te and ΔV_te are directly indicated. After the equilibrium period in NT solution, a stable level of V_te was confirmed. The initial spontaneous V_te was reduced by the luminal application of amiloride (10 µM), which blocks the epithelial Na⁺ channels (ENaC). In the presence of amiloride, cAMP-dependent Cl^- secretion was induced via application of forskolin (2 µM) and isobutylmethyl xanthine (IBMX, 100 µM) to the basolateral side. The addition of forskolin/IBMX induced a negative shift in V_te and increased the lumen negative I_sc. It is well known that luminal Cl^- channels (CFTR) and basolateral K⁺ channels (KCNQ1) are activated by cAMP-dependent signalling pathways (17, 18). Consistent with this model, the application of the KCNQ1 channel blocker (100 µM of chromanol 293B) to the basolateral side suppressed the cAMP-induced I_sc. Then, ATP (50 µM) was applied to the luminal side of the tracheal epithelium, which is known to induce a transient increase in cytosolic Ca²⁺ concentration ([Ca²⁺]_c) via phospholipase C (PLC)/inositol-trisphosphate (InsP₃)-coupled signalling pathways inducing the release of stored Ca²⁺ from the endoplasmic reticulum (ER) (19). This luminal ATP-induced increase in [Ca²⁺]_c (Δ[Ca²⁺]_c) is known to induce Cl^- secretion via Ca²⁺-activated Cl^- channels (Cl_Ca) in the luminal membrane. Consistently, transient increases in V_te and I_sc were observed in response to the luminal ATP application (19, 20). The I_sc values measured in each phase of the above protocol showed a decreasing tendency in the mice exposed to strong hyperoxic conditions (PO₂, 98%) for 36 hr (Fig. 1B).

The average values of I_sc change (ΔI_sc) induced by the above pharmacological treatments are summarized in Fig. 1C. In mice that were in hyperoxia conditions (80%, 90%, and 98% of PO₂ for 36 hr), the peak amplitude of ΔI_sc induced by forskolin/IBMX (ΔI_sc,forsk) decreased at the highest PO₂ condition compared to the responses of the mice exposed to 80% PO₂ or to the normoxic condition (Fig. 1C). The peak increase in I_sc due to the luminal ATP (ΔI_sc,ATP) of 98% PO₂ was smaller than that of 80% PO₂ and not significantly different from that of the control (Fig. 1D). In contrast, the amiloride-sensitive I_sc (ΔI_sc,amil) was not affected by chronic hyperoxia (Fig. 1E).

Next, we investigated the I_sc responses of airway epithelia obtained from wild-type C57BL/6 mice (WT), gulo(-/-) mice with vitamin C-supplementation for three weeks (K/O+Vit-C), and gulo(-/-) mice reared with no vitamin C-supplementation for 5-7 days (K/O-1wk), 14-15 days (K/O-2wk), 20-22 days (K/O-3wk), and 27-29 days (K/O-4wk). In Fig. 2A, the representative traces of V_te and ΔV_te are shown for WT and K/O-3wk mice. Also, the averaged I_sc values measured in each phase of the above protocol (control, amiloride, forskolin/IBMX, 293B, and ATP) are summarized for the WT and K/O-3wk mice (Fig. 2B). In general, the responses of airway epithelial I_sc to the above pharmacological agents were smaller in WT mice than those in the ICR mice described above (Fig. 1B). Nevertheless, it was notable that the amplitudes of I_sc were markedly suppressed in K/O-3wk mice (Fig. 2B).

The amplitudes of ΔI_sc,forsk were compared between the groups of different duration of vitamin C deprivation. Overall, ΔI_sc,forsk showed a decreasing tendency beginning with K/O-2wk mice, then became significant in K/O-3wk mice, and was completely abolished in K/O-4wk mice (Fig. 2C). ΔI_sc,ATP was also decreased in the K/O-3wk and K/O-4wk mice (Fig. 2D). In contrast, ΔI_sc,amil was abolished only in the K/O-4wk group (Fig. 1E). We also noted that the tissue resistance (R_te) showed a decreasing tendency in K/O-3wk mice and was significantly decreased in K/O-4wk mice (Fig. 2F).

The above results suggest that the Cl^- secretory function of the mouse airway epithelium is susceptible to ambient levels of oxidative stress when the endogenous antioxidant is deficient. Because the cAMP-activated CFTR is the major pathway of Cl^- secretion (11, 13), we compared the expression of CFTR in airway epithelia between WT and gulo(-/-) mice. Strong expression of CFTR in the luminal membrane was commonly observed in the control and K/O+Vit-C mice. However, consistent with the decrease in ΔI_sc,forsk according to the extent of vitamin C deficiency, the expression of CFTR was decreased in K/O-2wk mice and was nearly absent in K/O-3wk and K/O-4wk mice (Fig. 3).

As seen from the inhibition of ΔI_sc,forsk by 293B, KCNQ1 activity is also critical to the maintenance of Cl^- secretion through CFTR (18). Therefore, we investigated whether KCNQ1 expression is altered in gulo(-/-) mice. However, the expression of KCNQ1 was persistently observed in K/O-3wk and K/O-4wk mice (Fig. 4). While not rigorously analyzed here, morphological changes were also observed in the airway epithelia of gulo (-/-) mice; the typical ciliated columnar epithelium became cuboid or flattened in response to increased duration of vitamin C deficiency (Figs. 3, 4, K/O-3wk and K/O-4wk).

In the present study, we found that chronic vitamin C deficiency suppressed Cl^- secretion and downregulated CFTR expression in mouse airway epithelium. The cAMP-dependent Cl^- secretion (ΔI_sc,forsk) as well as the luminal ATP-induced Cl^- secretion (ΔI_sc,ATP) was decreased in gulo (-/-) mice. Since the CFTR was fully stimulated by forskolin/IBMX upon ATP application in the present experimental procedure, ΔI_sc,ATP was believed to be due to Cl^- secretion via CFTR as well as Ca²⁺-activated Cl^- channels (19). In this respect, the decreased ΔI_sc,ATP might also reflect the downregulation of CFTR. However, we could not conclude that the ion channels other than CFTR were actually affected by hyperoxia or vitamin C deficiency in mice. The inhibitory tendency of chronic hyperoxia on ΔI_sc,Fsk also suggested that an imbalance between oxidizing influence and antioxidant capacity could impair the cAMP-dependent Cl^- secretion mechanisms (Fig. 1C).

While CFTR-mediated secretion is more susceptible to vitamin C deficiency than it is to other parameters, the sustained deprivation of vitamin C generally suppresses the transepithelial transport functions. This functional impairment seems to precede the morphological changes in the epithelia of gulo(-/-) mice. Notably, the parameter reflecting the integrity of tissue (R_te) is severely lowered in K/O-4wk mice, indicating that it may be one of the scorbutic symptoms. It is well known that chronic deficiency of vitamin C weakens connective tissue due to impaired collagen synthesis (16). In addition to the decreased R_te, the histological findings of the K/O-3 and -4wk (flattened airway epithelium) mice indicate that a transformation from a ciliated/columnar epithelium occurs during the sustained loss of vitamin C in vivo.

The human respiratory tract is constantly exposed to transient instances of oxidative stress resulting from the inhalation of a variety of foreign materials including atmospheric pollutants and microorganisms. Furthermore, the production of ROS during episodes of infection and inflammation has been implicated in the pathogenesis of a number of pulmonary diseases such as asthma, adult respiratory distress syndrome, chromic obstructive pulmonary disease, and cystic fibrosis (21-24). Since inflammation and oxidative stress are closely related, inflammation also plays a role in the development of chronic lung disease (21, 22). Clinically, oxygen therapy is frequently applied to treat systemic hypoxemia. In this process, harmful effects of hyperoxia are often reported. The risk of hyperoxic stress and its sequel (e.g., bronchopulmonary dysplasia) are more prominently observed in neonates (2, 25).

The concomitant suppressions of ΔI_sc,forsk and CFTR expression in vitamin C-deprived strongly suggest that chronic oxidative stress negatively regulates CFTR protein expression. Recently, it has been reported that oxidative stress induced by the pharmacological agent tert-butylhydroquinone (BHQ) suppresses CFTR expression in T84, a colonic epithelial cell line (6). The same group also reported that the functional expression of CFTR was suppressed in the human nasal mucosa of cigarette smokers (5). Based on the results of these previous studies, it is highly likely that the suppressions of ΔI_sc,Fsk and CFTR in gulo(-/-) mice reflect a transcriptional regulation of CFTR genes under chronic oxidative stress (26, 27). Interestingly, CF-like symptoms such as thickened airway secretions are often seen in chronic inflammatory airway diseases that are not associated with mutations in the CFTR gene, and there is emerging evidence that posttranslational damage to CFTR by reactive oxygen and nitrogen species decreases CFTR function (27).

Antioxidants such as vitamin C may have beneficial effects via restoration of the balanced levels of ROS in tissues, as has been suggested in a study of hepatic ischemic-reperfusion injury model (28). Also, the lack of vitamin C in the ASL of asthmatics has been reported (9). However, careful interpretation is still required for the role of dietary supplementation of vitamin C in human, since there are controversies in the relation between plasma levels of vitamin C and cancer incidence (29). Furthermore, because we have not directly measured the oxidative stress in this study, the downregulation of CFTR by vitamin C deprivation might owe to the lack of vitamin C per se in addition to the putative oxidative stress.

In summary, this study confirms the inhibitory effects of vitamin C deprivation on the Cl^- secretion function in murine airway epithelium in vivo. Among the ion channels associated with epithelial Cl^- secretion, CFTR appears more to be vulnerable to oxidative stress than are other types of channels.