INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

BEFORE BIRTH, THE FETAL LUNG is filled with fluid. This fluid is critical for normal development of the lung and is secreted by the fetal lung epithelium through a process driven by active secretion of Cl from the interstitium of the lung into the lumen of the air space, with subsequent movement of water down its osmotic gradient. After birth, however, reversal of this phenomenon occurs: fluid is cleared from the lumen of alveoli through a process driven by absorption of Na⁺ from the apical side of alveoli into the interstitium of the lung and subsequent vectorial movement of water down its osmotic gradient.

Discovery of the epithelial Na⁺ channel (ENaC) (5, 6, 16) has shed new light on the mechanism of Na⁺ transport. Although the stoichiometry of the ENaC is still controversial (7, 9), it is generally believed to consist of the three subunits, , , and  (6). The -subunit is considered essential for the proper function of the channel in the lung because the mouse lacking this subunit fails to clear its fetal lung liquid and dies within the first 40 h after birth (14). Although mice lacking the - and -subunits of ENaC have near-normal lung clearance, they die shortly after birth because of an abnormal electrolyte balance (3, 18, 22), likely due to the inability of the kidneys to maintain electrolyte homeostasis.

Although there is some information available, the actual mechanism involved in the dramatic conversion of the lung from a fluid-secreting to a fluid-absorbing organ is poorly understood. For example, it is known that an increase in the circulating levels of adrenaline in the fetal lamb (4) can rapidly and reversibly switch the fetal lamb from fluid secretion to fluid absorption. However, this is a reversible phenomenon, and because the increase in the level of this hormone declines shortly after birth, it is uncertain whether this signal is sufficient to permanently convert the lung to Na⁺ absorption. Fetal (20, 23) and adult (17, 21) distal lung epithelia have enhanced Na⁺ transport and higher mRNA levels of all three subunits of ENaC in a 21% O₂ relative to a 3% O₂ environment, a phenomenon that is reversible and is likely mediated, at least in part, by reactive oxygen species (23). Therefore, O₂ may be an important molecular switch to "turn on" lung epithelial Na⁺ absorption at birth and maintain it at sufficiently high levels thereafter.

The process of O₂ induction of ENaC mRNA expression in fetal lung distal epithelial (FDLE) cells would require an O₂-sensing cellular repertoire. A putative heme O₂ sensor has been implicated in mediating the induction of erythropoietin (Epo) synthesis and vascular endothelial growth factor mRNA expression by hypoxia in a liver cell line (11, 12). The function of this sensor is dependent on extracellular iron levels and can be modulated by iron chelators, selected divalent transition metals such as Co²⁺ and Ni²⁺, and carbon monoxide (CO). Although it is an increase rather than a decrease in O₂ concentration that stimulates ENaC mRNA expression and Na⁺ transport in FDLE cells, we speculated that we may obtain insight into the mechanism of O₂ induction using these agents and approaches.

In this study, we show that iron, but not the iron binding proteins contained within serum, is required for the induction of amiloride-sensitive Na⁺ transport and ENaC mRNA expression when FDLE cells are exposed to 21% O₂. Three series of experiments suggested that this process is mediated through or dependent on heme-containing protein(s). First, when FDLE cells were cultured in the presence of Co²⁺, a transition metal that displaces iron from the protoporphyrin moiety of heme proteins and locks the heme molecule in its deoxygenated conformation (30), 21% O₂ could no longer induce ENaC mRNA expression and amiloride-sensitive Na⁺ transport in these monolayers. Second, when FDLE cells were incubated in the presence of CO, a gas that binds to the heme molecule and locks it in its oxygenated conformation (11), there was an induction of ENaC mRNA expression and Na⁺ transport, even though we maintained the FDLE cells in a 3% O₂ environment. Third, incubation of FDLE cells with dioxoheptanoic acid (DHA), an inhibitor of heme synthesis (29), attenuated the induction of FDLE cell Na⁺ transport by a 21% O₂ environment.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Primary cell culture. FDLE cells were isolated and cultured as previously described (19). In brief, 20-day-gestation Wistar rats (breeding day = day 0, term 22 days; Charles River, St. Constant, PQ) were killed with an ether overdose. Fetal lung tissue was dispersed with 0.125% trypsin, and the resulting cell pellet was further incubated with 0.1% collagenase to separate associated fibroblasts from epithelial cells. Dissociated fibroblasts were separated from epithelial cells with a differential adherence and centrifugation procedure. FDLE cells were then seeded at 1 × 10⁶ cells/cm² onto 0.4-µm pore size Snapwell cell culture inserts (Corning Costar, Cambridge, MA) for Ussing chamber studies and at 0.5 × 10⁶ cells/cm² onto 75-mm-diameter, 0.4-µm pore size Transwell cell culture inserts (Corning Costar) for subsequent RNA isolation. All cells were submersion cultured for a total of 3 days after being seeded in Dulbecco's modified Eagle's medium (4.5 g/l of glucose with 2 mM L-glutamine and 110 mg/l of sodium pyruvate) supplemented with 10% fetal bovine serum (Cansera, Rexdale, ON), 100 U/ml of penicillin G sodium, and 100 µg/ml of streptomycin sulfate. All cell culture reagents were purchased from GIBCO BRL (Life Technologies, Burlington, ON).

O₂ environment and interventions. After being seeded, the FDLE cells were returned to an incubator containing 3% O₂-5% CO₂-balance N₂. After 24 h, the medium was replaced with fresh medium that contained various agents. Immediately thereafter, the monolayers were either transferred to 21% O₂ or kept in 3% O₂ for the next 48 h. For the CO experiments, the FDLE cells were placed in a hypoxic chamber (Fisher Scientific, Unionville, ON), purged with a gas mixture containing 3% O₂, 5% CO₂, 10% CO, and balance N₂ and then tightly sealed. The chamber was then placed within the 3% O₂ incubator. The concentration of these gases was analyzed by the manufacturer (Praxair, Oshawa, ON). The O₂ concentration in the hypoxic chamber was monitored during the 48-h experimental period with a Miniox I O₂ analyzer (MSA Medical Products, Pittsburgh, PA). After the 48-h induction period, the monolayers were used for either Ussing chamber studies or Northern analyses. Deferoxamine, CoCl₂, NiCl₂, ZnCl₂, ferrous ammonium sulfate, and DHA were purchased from Sigma-Aldrich Canada (Mississauga, ON).

Measurement of FDLE monolayer bioelectric properties. The bioelectric properties of the FDLE monolayers were determined as previously described (19) with modified Ussing chambers (World Precision Instruments, Sarasota, FL) while the cells were bathed in 37°C Hank's balanced salt solution (GIBCO BRL) supplemented with 1.8 g/l of sodium bicarbonate and equilibrated with a 5% CO₂-balance air gas mixture. The FDLE monolayers were maintained under open-circuit conditions, and their short-circuit current (I_sc) was determined every 10 min with a voltage-current clamp (Physiologic Instruments, San Diego, CA) until stabilized (~20 min). The amiloride-sensitive I_sc was determined by the addition of 0.1 mM amiloride (in DMSO with a final dilution of 1:1,000; Sigma-Aldrich) to the apical side of monolayers. The transepithelial resistance (R) was calculated by dividing the transepithelial potential difference by the I_sc.

Northern analyses. RNA was extracted from FDLE monolayers with 4 ml of TRIzol Reagent (GIBCO BRL) according to the manufacturer's instructions. The final pellet was dissolved in water treated with dimethyl pyrocarbonate (Sigma-Aldrich), and then 20 µg of total RNA were size-fractionated on a 1% agarose-1× MOPS-2% formaldehyde gel. RNA was subsequently transferred to Hybond N+ nylon membranes (Amersham, Oakville, ON). The blots were then ultraviolet cross-linked and hybridized with ³²P random-primed rat -ENaC (-rENaC), -rENaC, and -rENaC cDNA fragments [bp 74-403 for -rENaC, bp 2025-2401 for -rENaC, and bp 2161-2520 for -rENaC (6)] in ExpressHyb solution (Clontech, Palo Alto, CA) following the manufacturer's instructions. After a wash in 0.1× sodium citrate-sodium chloride plus 0.1% sodium dodecyl sulfate at 50°C for 1 h, the blots were exposed to autoradiography film at 80°C. Autoradiographic bands were quantified with an Agfa (Duoscan) scanner and the National Institutes of Health Scion Image version 1.6 quantitation program. The mRNA levels were normalized to 18S rRNA content by hybridizing the blots with a full-length mouse 18S rRNA ³²P random-primed cDNA probe (American Type Culture Collection, Manassas, VA).

Data analysis. We used the INSTAT version 3.0 statistical program (GraphPad, San Diego, CA) to perform ANOVA followed by Tukey's post hoc test to examine significant differences between experimental groups. Probability (P) values of <0.05 were considered to be significant. All data are expressed as means ± SE, and unless otherwise specified, comparisons were made against the data corresponding to the 3% O₂ untreated control groups. Each n represents a single monolayer from one primary culture of FDLE cells. There were a minimum of three primary culture preparations used for each experimental group.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Iron, but not serum iron binding proteins, is required for O₂ induction of FDLE cell Na⁺ transport. To study the iron dependency of O₂-induced Na⁺ transport in FDLE cells, we incubated FDLE monolayers in 21% O₂ with various concentrations of the iron chelator deferoxamine for 48 h. Deferoxamine prevented the induction of amiloride-sensitive I_sc by 21% O₂ in a dose-dependent manner (IC₅₀ ~3 µM; Fig. 1). As Rafii et al. (23) have previously described, amiloride- insensitive I_sc was not affected by switching FDLE cells from a 3 to a 21% O₂ environment. In a separate set of experiments, incubation of FDLE cells with 10 µM deferoxamine for the 48-h O₂ induction period prevented the induction of amiloride-sensitive I_sc by 21% O₂ (Fig. 2). Neutralizing the effect of deferoxamine by coincubation of the deferoxamine-treated FDLE cells with excess iron (20 µM ferrous ammonium sulfate) restored the ability of FDLE cells to be induced by O₂ (Fig. 2). Deferoxamine treatment had no effect on the amiloride-sensitive I_sc of monolayers incubated in 3% O₂ (Fig. 2). Treatment of the cells with deferoxamine did not lead to a lower R in the monolayers, indicating that the effect of deferoxamine was not due to a nonspecific toxic effect on these cells (Table 1). Baseline I_sc in deferoxamine-treated cells, however, was significantly lower than that in the 21% O₂ control group (P < 0.05; Table 1).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Deferoxamine inhibited fetal lung distal epithelial (FDLE) cell Na+ transport in a dose-dependent manner. FDLE monolayers were incubated with indicated concentrations of deferoxamine in 21% O2. After 48 h, amiloride-sensitive short-circuit current (Isc) for these monolayers was measured (IC50 ~3 µM; n = 4-19 monolayers/group).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   O2 induction of Na+ transport in FDLE cells required iron. FDLE monolayers were initially cultured in 3% O2 for 24 h (control group), and then some monolayers were treated with (+) and without () 10 µM deferoxamine (Def) or 10 µM Def plus 20 µM ferrous ammonium sulfate (Fe). Subsequently, monolayers were either left in 3% O2 or switched to 21% O2. After 48 h, amiloride-sensitive Isc for these monolayers was measured (n = 12-28 monolayers/group). * P < 0.05 compared with 3% O2 control group.

To investigate whether iron-containing or iron binding proteins within serum are necessary for the O₂ induction of Na⁺ transport in FDLE cells, we conducted our O₂ induction experiments in the presence and absence of serum. As in monolayers incubated in the presence of serum, FDLE cells exposed to 21% O₂ in serum-free medium had a higher amiloride-sensitive I_sc than those kept in a 3% O₂ concentration (Fig. 3). The monolayers incubated under serum-free conditions, however, had a significantly lower R than those of the serum-treated 21% O₂ control group (P < 0.05), although the baseline I_sc in these monolayers was comparable to that of the serum-treated control cells (Table 1).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   O2 induction of Na+ transport in FDLE cells was independent of serum proteins. FDLE monolayers were initially cultured in DMEM plus 10% fetal bovine serum (FBS) in 3% O2 for 24 h, and then medium was changed to either DMEM or DMEM plus 10% FBS and cells were maintained at either 3 or 21% O2. After 48 h, amiloride-sensitive Isc for these monolayers was measured (n = 9-12 monolayers/group). * P < 0.05 compared with 3% O2 control group.

Effect of CoCl₂ and other transition metals on O₂ induction of FDLE cell Na⁺ transport. Transition metals with a low binding affinity for O₂, such as Co²⁺, are known to displace iron in the protoporphyrin ring of heme proteins and thereby chemically mimic the hypoxic conditions in cells. To study the effect of Co²⁺ on amiloride-sensitive Na⁺ transport, we incubated FDLE monolayers with various concentrations of CoCl₂ for 48 h in 21% O₂. Co²⁺ inhibited amiloride-sensitive I_sc in a dose-dependent manner (IC₅₀ ~100 µM; Fig. 4). In a separate set of experiments, we incubated FDLE monolayers with 100 µM CoCl₂, NiCl₂, or ZnCl₂ for the 48-h duration of the O₂ induction period. Co²⁺ prevented 21% O₂ induction of amiloride-sensitive I_sc (Fig. 5). FDLE cells incubated with Ni²⁺, another divalent transition metal with similar hypoxia-inducing properties as Co²⁺ (11, 26), similarly failed to respond to induction by 21% O₂ (P < 0.05 vs. 3% O₂ control group; n = 12 monolayers/group; data not shown). However, monolayers incubated with Zn²⁺, a transition metal shown not to displace iron in heme proteins (27), did not block 21% O₂-induced amiloride-sensitive I_sc (Fig. 5). This latter finding suggests that the effect of Co²⁺ is not due to a nonspecific effect of transition metals. Although treatment of the monolayers with Co²⁺ led to lower baseline I_sc, R was not affected by this treatment (Table 1).

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Cobalt inhibited FDLE cell Na+ transport in a dose-dependent manner. FDLE monolayers were incubated with indicated concentrations of CoCl2 in 21% O2. After 48 h, amiloride-sensitive Isc for these monolayers was measured (IC50 ~100 µM; n = 4-16 monolayers/group).

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Cobalt but not zinc inhibited O2-induced Na+ transport in FDLE cells. FDLE monolayers were initially cultured in 3% O2 for 24 h, and then some monolayers were incubated with 100 µM of either CoCl2 or ZnCl2. Subsequently, monolayers were either left in 3% O2 or switched to 21% O2. After 48 h, amiloride-sensitive Isc for these monolayers was measured (n = 8-24 monolayers/group). * P < 0.05 compared with 3% O2 control group.

Effect of iron chelation and Co²⁺ on ENaC mRNA expression. To test whether the effects of iron chelation and Co²⁺on Na⁺-channel activity correlated with changes in the ENaC mRNA levels, we performed Northern analyses on several FDLE monolayers exposed to either deferoxamine or Co²⁺. The - and -ENaC mRNA levels changed in parallel with the changes in FDLE amiloride-sensitive I_sc (Fig. 6). Deferoxamine blocked O₂-induced - and -ENaC mRNA expression. The effect of deferoxamine on -ENaC mRNA levels was reversed by coincubation of deferoxamine-treated cells with ferrous ammonium sulfate. As with deferoxamine, Co²⁺ was able to abrogate O₂-induced - and -ENaC mRNA expression. The changes in -ENaC mRNA levels (data not shown) in response to deferoxamine, iron, and Co²⁺ showed a statistically insignificant pattern similar to that for - and -ENaC (Fig. 6).

View larger version (841514K): [in this window] [in a new window]   Fig. 6.   Def and cobalt abrogated O2-induced - and -RNA expression in FDLE cells. FDLE monolayers were initially cultured in 3% O2 for 24 h, and then some monolayers were incubated with either 10 µM Def, 10 µM Def plus 20 µM Fe, or 100 µM CoCl2. Subsequently, monolayers were either left in 3% O2 or switched to 21% O2. After 48 h, total RNA from these monolayers was extracted for Northern analyses. A: representative blots of -, -, and -epithelial Na+ channels (ENaCs) and 18S rRNA (18S). B and C: densitometric analyses of multiple blots of - and -ENaCs, respectively, normalized first to 18S rRNA and then to corresponding 21% O2 group (n = 2-4 monolayers/group; each monolayer was prepared from a separate harvest of FDLE cells). * P < 0.05 compared with 3% O2 control group.

CO induces amiloride-sensitive Na⁺ transport and ENaC mRNA expression. Our hypotheses of heme involvement in O₂ induction of ENaC would predict that CO, a molecule that mimics binding of O₂ to heme proteins, might stimulate ENaC mRNA expression and Na⁺ transport in FDLE cells. We incubated FDLE monolayers in a 3% O₂-10% CO gas mixture and measured amiloride-sensitive I_sc and steady-state mRNA levels for -, -,and -ENaC after 48 h. As predicted, CO induced amiloride-sensitive I_sc with respect to the 3% O₂ control group (Fig. 7) even though the monolayers were incubated in 3% O₂. Both baseline I_sc and R for the CO-treated group were comparable to those for the 21% O₂ untreated control group (Table 1). Similarly, CO increased the level of - and -mRNA expression compared with that in the 3% O₂ control group (P < 0.05; Fig. 8, A-C). Levels of -ENaC mRNA were unresponsive to induction by CO (Fig. 8, A and D).

View larger version (13K): [in this window] [in a new window]   Fig. 7.   Carbon monoxide (CO) induced Na+ transport in FDLE cells. FDLE monolayers were initially cultured in 3% O2 for 24 h, and then some monolayers were incubated in 10% CO plus 3% O2 while others were incubated in either only 3% or only 21% O2. After 48 h, amiloride-sensitive Isc was measured (n = 12 monolayers/group). * P < 0.05 compared with 3% O2 control group.

View larger version (110131313K): [in this window] [in a new window]   Fig. 8.   CO induced - and - but not -ENaC mRNA expression. FDLE monolayers were initially cultured in 3% O2 for 24 h, and then some monolayers were incubated in 10% CO plus 3% O2 while others were incubated in either only 3% or only 21% O2. After 48 h, total RNA was extracted for Northern analyses. A: representative blot from 3 different experiments. B-D: densitometric analyses of multiple blots of -, -, and -ENaC, respectively, normalized first to 18S rRNA and then to corresponding 21% O2 group (n = 7 monolayers/group; each monolayer was prepared from a separate harvest of FDLE cells). * P < 0.05 compared with 3% O2 control group.

Inhibition of heme synthesis attenuates O₂ induction of Na⁺ transport in FDLE cells. To provide further evidence supporting the role of a heme protein in the O₂ induction of Na⁺ transport in FDLE cells, we incubated FDLE monolayers in the presence of the heme synthesis inhibitor DHA (2 mM) throughout the 48-h period in which they were exposed to 21% O₂. DHA attenuated the ability of 21% O₂ to induce Na⁺ transport (Fig. 9). The treatment of cells with DHA led to a lower baseline I_sc but not to a lower R (Table1), indicating that its effect was not due to a nonspecific toxic property of this compound.

View larger version (13K): [in this window] [in a new window]   Fig. 9.   Heme synthesis inhibitor dioxoheptanoic acid (DHA) attenuated O2-induced Na+ transport in FDLE cells. FDLE monolayers were initially cultured in 3% O2 for 24 h, and then some monolayers were incubated with 2 mM DHA. Subsequently, monolayers were either left in 3% O2 or switched to 21% O2. After 48 h, amiloride-sensitive Isc in these monolayers were measured (n = 23-24 monolayers/group). * P < 0.05 compared with 3% O2 control group. # P < 0.05 compared with 21% O2 control group.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study, we have shown that O₂-induced ENaC mRNA levels and amiloride-sensitive Na⁺ transport is dependent on the presence of adequate amounts of iron and that it is inhibited by micromolar concentrations of Co²⁺. Both - and -ENaC mRNA expression and Na⁺ transport in FDLE cells can be induced by CO, whereas O₂-induced Na⁺ transport can be attenuated by an inhibitor of heme synthesis. Together these results point to involvement of a heme-containing molecule in the O₂ induction of ENaC mRNA expression and Na⁺ transport in FDLE cells.

Because tissue culture serum contains iron-containing proteins, we evaluated the relevance of these proteins by carrying out O₂ induction of FDLE cells in the presence and absence of serum. Na⁺ transport in monolayers cultured in the absence of serum was induced by O₂ similar to that in cells cultured in the presence of serum. However, serum-free culture of FDLE monolayers significantly decreased their R values, suggesting that factors within serum may be important for the integrity of tight junctions. These experiments provide evidence that the phenomenon of O₂ induction does not require iron-containing proteins or other factors within serum.

Other investigators have used divalent transition metals to study the mechanisms whereby changes in O₂ concentration alter cellular physiological functions. It is well known that Co²⁺and Ni²⁺ can displace iron in the protoporphyrin ring of heme molecules (26, 30). When these compounds replace iron within the heme group, there is a reduction in the affinity of the heme molecules for O₂ (26, 30). Therefore, Co²⁺- and Ni²⁺-containing heme molecules are locked in their deoxygenated conformation. Using this information, others (11) have demonstrated that Co²⁺ can mimic the hypoxic induction of the glycoprotein Epo by liver cells and that CO, a molecule that binds to heme proteins and locks it in the oxygenated conformation, can, in turn, abrogate this induction. The current working hypothesis on hypoxic induction for Epo is that when liver cells are exposed to hypoxia or Co²⁺, there is a concomitant activation of the transcription factor hypoxia-inducible factor-1 (HIF-1) (25). HIF-1 binds to an enhancer in the gene encoding Epo, resulting in increased transcription. In contrast to the effect of hypoxia on Epo, hypoxia downregulates FDLE cell Na⁺ transport and ENaC mRNA levels. Thus if HIF-1 is involved in our system, one must speculate that hypoxic induction of HIF-1 increases the expression of a protein that is capable of suppressing ENaC expression. To gain more insight into this issue, we also incubated some monolayers with deferoxamine in hypoxic O₂ concentrations. Under 3% O₂, deferoxamine had no effect on Na⁺ transport. Because deferoxamine inhibits Na⁺ transport in 21% but not in 3% O₂, we suggest that the phenomenon of O₂ regulation of ENaC likely involves activation under a 21% O₂ rather than suppression under a 3% O₂ environment.

CO binds to heme proteins in a manner analogous to O₂ binding to hemoglobin and locks the molecule in its oxygenated conformation (11). Our present observation that - and -ENaC mRNA expression and Na⁺ transport were induced in monolayers incubated with CO under hypoxic conditions supports our hypothesis that a heme-containing protein is involved in O₂ signaling in FDLE cells. In contrast to - and -ENaC, -ENaC mRNA expression was neither suppressed by Co²⁺ nor induced by CO. We, therefore, speculate that the mechanism and pathways of induction of -ENaC by O₂ may be different from those of either - or -ENaC. We also performed additional experiments to further investigate the potential role of heme proteins in O₂ induction. In those studies, O₂ induction was done in the presence of DHA, an inhibitor of aminolevulinate dehydratase (29), an enzyme in the heme synthetic pathway. DHA significantly reduced O₂ induction of FDLE cells, further supporting a role for heme proteins in O₂ induction of Na⁺ transport in FDLE cells. Inability of DHA to completely inhibit O₂-induced Na⁺ transport could be attributed to its inability to completely inhibit intracellular heme synthesis (24).

To date, heme-mediated activation of gene transcription in response to O₂ has been conclusively demonstrated only in the FixL/FixR system of the nitrogen-fixing bacterium Rhizobium meliloti (10). Some previously identified heme proteins, however, have been proposed as potential candidates for this O₂ sensor in eukaryotic systems. These include cytochrome P-450 (8) and NADPH oxidase (1, 13, 31), which under normoxic conditions bind and convert O₂ to superoxide, and NO-inducible guanylate cyclase (28).

Our results do not rule out the possibility that there is also a nonheme O₂ sensor similar to that described in neurons (15) or an iron-sulfur cluster-type molecule such as aconitase that produces reactive oxygen species in response to changes in environmental O₂ concentration. Rather, our results suggest that heme proteins play an essential role in the pathway(s) leading to increased Na⁺ transport by FDLE cells when they are exposed to postnatal O₂ concentrations.

In conclusion, our experiments have demonstrated that induction of Na⁺ transport and ENaC mRNA expression in FDLE cells by O₂ requires iron and heme-containing protein. Further studies are required to determine whether the heme protein-dependent increase in ENaC mRNA levels occurs through increasing the stability or inducing the synthesis of the mRNA message under higher postnatal oxygen concentration. Because FDLE (19) and human distal lung epithelial cells (2) have similar bioelectric properties, these findings may be relevant to the normal transition of the lung from fetal to postnatal life and in the recovery of patients with pulmonary edema.