INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CLEARANCE OF SODIUM and water from the mature lung is mediated by a transcellular mechanism, combining diffusion of the luminal Na⁺ through an apical amiloride-sensitive epithelial Na⁺ channel (ENaC) and excretion of intracellular Na⁺ by basolateral Na⁺-K⁺-ATPase (1, 26). Water diffuses either through specific water channels or through the paracellular junctions to equilibrate the osmotic pressure (21). This mechanism contributes to the correct hydration of the apical liquid layer in the proximal and distal airways as well as in alveolar tissue.

A membrane transport complex composed of three homologous subunits called -, -, and -ENaC (4, 5, 18, 19, 35, 37) is responsible for the passive electrodiffusion of the ions through the apical membrane. The subunits are characterized by a large extracellular domain located between two transmembrane regions, the NH₂- and COOH-terminal segments being cytoplasmic (1, 31). The expression of the three subunits is necessary for maximal functional activity (5). They probably associate into functional heterotetramers that contain two -ENaC, one -ENaC, and one -ENaC (11, 16, 33), i.e., a tetrameric organization also found for other members of the same gene superfamily (7).

The three rat ENaC mRNA and protein subunits were detected in many Na⁺-absorptive tissues such as the distal parts of the cortical nephron, the distal colon, and the reabsorptive ducts of the salivary glands and sweat glands (8, 32). In the mature rat, - and -ENaC transcripts were detected in ciliated cells of nasal and bronchial surface epithelium, in bronchiolar Clara cells, and in alveolar type II cells (10, 20). Rat -ENaC transcripts were also detected in nasal and tracheal gland acini. In humans, all three mRNAs were identified in the surface airway epithelium, whereas - and -transcripts were also found in epithelial cells along gland ducts and in gland acini (3, 22). In previous reports, we identified rat airway epithelium as an important site of expression of the three ENaC proteins (32). We also showed that the transcription of all three ENaC subunits was increased around birth, at a moment when the respiratory epithelium switches from chloride secretion to Na⁺ absorption (23, 24, 36, 37). However, these initial studies did not address specifically the distribution and the subcellular localization of the ENaC proteins in the human airways. To understand how the respiratory epithelium is modified during lung development, we have now examined the protein distribution of ENaC in human fetal airway tissues. After having tested several rabbit polyclonal antibodies against rat or human -, -, and -ENaC, two of them raised against human -ENaC and rat -ENaC reacted with the human lung proteins. They were used for immunohistochemical localization using optic and electron microscopy in human fetal and adult lung tissues.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Human fetal and adult tissue material. Ten fetuses ranging from 10 to 36 wk gestational age were obtained from spontaneous abortions or medical inductions. The age distribution is shown in Table 1. All fetuses were well preserved without respiratory abnormality or infection. They were not associated with either polyhydramnios or oligohydramnios. Adult respiratory tissue was obtained during postmortem examination from three patients without hypertension who died from nonpulmonary causes. The Ethics Committee approved these experiments on human tissues. Different tissue specimens were collected from different parts of the fetal and adult airways (tracheae, bronchi, and bronchioles) and immediately fixed. Samples were fixed in 15% Formalin and embedded in paraffin, and 3-µm sections were mounted on gelatin-coated slides and dried overnight at 50°C. Other samples were embedded in optimum cutting temperature compound (Tissue Tek, Miles, IN), frozen in liquid nitrogen, cut at 20°C, and transferred to gelatin-coated slides. For electron microscopy, adult tissues were fixed by immersion for 2 h at room temperature in 4% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS), pH 7.2 (Sigma, St. Louis, MO). Osmium postfixation was omitted. After fixation, the tissue was washed, dehydrated through graded alcohol series, and embedded in Epon.

View this table: [in this window] [in a new window]   Table 1.   Immunolocalization of - and -subunit proteins of epithelial Na+ channel along fetal and adult human airways

Preparation of antibodies. Several polyclonal antibodies were raised against rat -ENaC, as described in previous publications (1, 17, 18, 31, 32). The polyclonal antiserum against -ENaC was raised against the last 17 cytoplasmic residues of the human -subunit (1, 36). The polyclonal antiserum against -ENaC was obtained after immunization of a rabbit with an hapten formed with keyhole limpet hemocyanin and the extracellular -ENaC peptide Y¹²⁷GVKI SRKRRI AGS¹⁴³ (32). After immunization, antisera were regularly analyzed by ELISA against pure peptides (Fig. 1, A and B). When a specific immune response was detected, the antisera were characterized by biochemical and histological techniques. The antibodies used in the present study correspond to the positive ELISA antisera, which were able either to immunoprecipitate in vitro translated proteins or to detect the protein with Western blot or immunohistochemical analysis.

View larger version (106K): [in this window] [in a new window]   Fig. 1.   Characterization of rabbit polyclonal antisera against epithelial Na+ channel (ENaC): -ENaC (A, C, and E) and -ENaC (B, D, and F). Immune sera were positive by ELISA against pure antigen, whereas preimmune sera were negative (A and B). Immunolocalization of - (C and E) and -ENaC subunits (D and F) in rat kidney (C and D) and rat lung (E and F). Paraffin sections, immunoperoxidase.

Immunohistochemistry. Control immunohistochemistry of -, -, and -ENaC antisera was done with 6-µm frozen rat lung sections. Other complementary characterization of these antisera has been published elsewhere (1, 17, 18, 32, 36). Some additional characterizations are presented in Fig. 1. Human airway paraffin sections were deparaffinized with xylene and successively rehydrated in graded ethanol baths, distilled water, and 0.1 M PBS, pH 7.2, before treatment with 0.4% pepsin in 0.01 N HCl for 10 min at room temperature. Hydrogen peroxide bath was used for 5 min at room temperature to remove endogenous peroxidase activity. A blocking reagent (6% goat serum) was added for 5 min. The slides were then rinsed twice with PBS and pretreated for 10 min with pepsin (0.04% in 0.01 N HCl). After two rinses, the tissue sections were incubated for 1 h with primary antibodies diluted to PBS as follows: anti--ENaC, 1:100 and anti--ENaC, 1:100. Immunohistochemical staining was carried out using the streptavidin-biotin LSAB2 technique (DAKO, Glostrup, Denmark).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Characterization of the antibodies. The antibodies used in the present study have been characterized already in several previous studies (1, 17, 18, 32, 36). After rabbit immunization, antisera were regularly analyzed by ELISA against pure peptides. Figure 1, A and B, shows the positive signal of immune sera against pure peptides in ELISA reaction. No signal was detected with preimmune serum. Specificity of the antibodies was also checked by histology in rat and human tissues known to express ENaC. Antisera were positive in rat lung and kidney sections (Fig. 1, C-F) (23, 40), i.e., two tissues where an active amiloride-sensitive Na⁺ absorption has been described. A positive signal was also detected in human keratinocytes, in absorptive cells in the surface epithelium of the colon, and at apical membrane of sweat glands (data not shown). Initial experiments were also carried out with an anti--ENaC able to recognize the NH₂-terminal segment of the rat -ENaC in immunohistochemistry. However, it did not cross-react with human lung -ENaC. Therefore, the present study was focused on the expression during fetal development of the - and -ENaC proteins.

Developmental expression of - and -ENaC subunits in fetal airways. At 10 wk of gestation, the human fetal trachea and the bronchi are lined with undifferentiated and polarized columnar epithelial cells and no glands are developed yet. Between 11 and 16 wk of gestation, ciliated and secretory cells progressively differentiate in the surface epithelium along proximal airways. At 13-14 wk, the first glands grow out from the basal aspect of this epithelium into the lamina propria. Until 16 wk of gestation, the branching distal airways exhibit a pseudoglandular pattern and give rise to the future conducting airways. As shown in Table 1, no - or -ENaC immunoreactivity was detected during the early stages of development (16 wk).

View larger version (48K): [in this window] [in a new window]   Fig. 2.   Immunolocalization of - and -ENaC subunits in human fetal airways at 18 wk. Surface epithelium of trachea shows numerous ciliated, secretory, and basal cells. - (A) and -subunits (B and C) are localized along apical plasma membrane of ciliated cells. -Subunit (B and C) was also detected in a few basal cells at opening of glandular duct. Cells lining glandular duct are also labeled. Bronchioles show ciliated and Clara cells; the latter were identified using anti-CC10 protein antibody (f). - (D) and -subunits (E and F) were localized at the apical domain of Clara cells and in ciliated cells. Bar, 25 µm; frozen sections, FITC immunostaining (A, B, D, E, and f); paraffin sections, immunoperoxidase (C and F).

Expression of - and -ENaC subunits in adult airways. In the adult surface epithelium of human trachea and bronchi, the distribution of both - and -ENaC subunits was similar to that observed in late- gestational fetuses. In paraffin or frozen sections, the - and -ENaC staining was observed in the ciliated cells, which showed homogeneous apical immunostaining (Fig. 3, A-C). No - and -ENaC staining was identified in basal cells or in mucous goblet cells. In the submucosal glands, the - (Fig. 3D) and -subunits (Fig. 3E) were detected in the serous cells. Serous cells were identified by an anti-lysozyme antiserum, i.e., a specific marker of these cells (Fig. 3F). In the bronchioles, - (Fig. 3G) and -subunits (Fig. 3H) were detected at the apex of the Clara cells. Clara cells were identified by an anti-CC10 antiserum, i.e., a specific marker of these cells (Fig. 3I).

View larger version (79K): [in this window] [in a new window]   Fig. 3.   Immunolocalization of ENaC subunits in human adult airways. - (A) as well as -ENaC subunit analyzed in frozen (B) or in paraffin section (C) are homogeneously localized along and beneath apical plasma membrane of ciliated cells in surface airway epithelium. No - and -ENaC immunolocalization is identified in basal cells or in mucous goblet cells. In submucosal glands - (D) and -subunits (E) are detected in serous submucous gland cells, which can be identified as serous cells by using specific labeling, an anti-lysozyme antiserum (F). Mucous acini (*) are not labeled. In bronchioles where numerous nonciliated Clara cells are identified using specific anti-CC10 antiserum (I), - (G) and -subunits (H) are identified at apex of Clara cells. Bar, 25 µm; frozen sections, FITC immunostaining (A, B, and G-I); paraffin sections, immunoperoxidase (C-F).

View larger version (172K): [in this window] [in a new window]   Fig. 4.   Ultrastructural immunolocalization of ENaC subunits in human adult airways. In differentiated and mature ciliated cells, immunogold labeling for - (B) and -subunits (C) are detected beneath apical plasma membrane and along microvilli. Serous cells contain dark and heterogeneous granules where - (E) and -subunits (F) are mainly localized in secretory granules of serous cells. In Clara cells, - (H) and -subunits (I) are localized in cytoplasm and along plasma membrane of secretory granules. In control sections with preimmune sera, no immunolabeling can be seen in ciliated cells (A), serous cells (D), and Clara cells (G). Bar, 0.4 µm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The results reported in this study demonstrate for the first time an early expression of - and -ENaC proteins in the human fetal airways. Both proteins were detected in the apical membrane of the ciliated cells and also in bronchiolar Clara cells as well as in serous cells of the submucosal glands. All these localizations, including those in the serous cells of the submucosal glands, have been reported at an RNA level by other investigators using in situ hybridization (3, 10). Taken together with these previous studies, the present work identifies the different sites of ENaC expression in the airways. In the human distal lung sections, in which bronchioles were examined, canalicular and alveolar structures could also be observed. However, because of a delay between abortion and examination of the fetus, the distal lumens were not optimally preserved. In contrast, the epithelial cells lining the bronchi and bronchioles were preserved by the rigid cartilage and muscle layers, whereas the distal lung was collapsed. In that zone, flattened epithelial cells were hardly distinguishable from fibroblasts and endothelial cells. For this reason, we limited our analysis to the proximal lung and to the distal bronchioles. We believe that use of alternative animal models, such as rat or mouse, will be necessary to address the issue of the time-dependent expression of ENaC in alveoli. Our data provide, however, original information regarding the expression of ENaC in the human lung during development.

These results are consistent with the data published by Venkatesh and Katzberg (34), showing the expression of all three ENaC RNAs as early as 21 wk of gestation. At 24 wk, they reported mRNA contents equal to 13, 26, and 32% of the adult values for -, -, and -ENaC, respectively. Although the immunohistochemistry is not quantitative, the level of expression of the two subunits that have been tested was definitely lower at the end of the second trimester than during any later periods of gestation (Table 1).

Unfortunately, we were unable to carry out the same study with -ENaC because our antibodies raised against rat -ENaC did not react with the human lung subunit. Previous experiments performed in the rat lung (32) have shown that -ENaC is indeed colocalized with -ENaC and -ENaC in the airways. Moreover, the expression of the -subunit parallels the expression of the -subunit (10). From this perspective, the expression patterns of -ENaC and -ENaC are expected to be similar. Because the stoichiometry of each subunit within a functional complex is a fixed value (7, 11), it is likely that the less abundant transcripts will be the limiting factor for expression of the complex. Recent data from Otulakowski et al. (25) suggest that in human airways the limiting factor corresponds to -ENaC. Our experiments therefore identify probably the main sites of expression of the highly Na⁺-selective and highly amiloride-sensitive channel in airways. It remains possible that other channels, characterized by different biophysical or pharmacological properties and formed by distinct proteins, could also participate in lung Na⁺ homeostasis.

The role conferred to -ENaC is usually more important than the role conferred to - or -ENaC, which is consistent with the fact that there are two copies of -ENaC per copy of - or -ENaC (11). This is also suggested by knockout of the -ENaC gene in the mouse, which is associated with an early death caused in part by defective neonatal lung liquid clearance (13). The situation might be different in humans where inactivation of -ENaC observed during pseudohypoaldosteronism type I is not associated with major lung defects.

The detection of the ENaC proteins in the lung during early stages of development does not of course prove a strict parallel to function. Several disparities between molecular and functional data have been reported in the literature (1, 26), and the concept of silent Na⁺ channels has also been recognized for years (27). Therefore, it is possible that the channels are present in the epithelium but are silent. We suspect that the activation of such a silent pool and hence the activation of lung liquid clearance might be helpful to clear excess lung liquid during some pathological situations. The effect of known Na⁺ channels stimulators, such as -adrenergic agonists (1, 2, 21), glucocorticoids (1, 6, 36), oxygen (28, 29, 38), or growth factors, is certainly worth investigating in pathogen conditions such as premature delivery or respiratory distress syndrome.

In the adult surface epithelium, both - and -ENaC proteins were detected by optical microscopy at the apical membrane of ciliated cells, whereas they were not identified in the mucous goblet cells. Electron microscopy data also suggest an apical or subapical localization of the - and -ENaC subunits. No signal was found near the basolateral membrane. Such localization is similar to that of the cystic fibrosis transmembrane conductance regulator (CFTR) protein, which is also present in apical vesicles under the apical plasma membrane of the ciliated cells (30). In the submucosal glands, - and -ENaC were specifically expressed in the serous gland cells and were identified at the level of the secretory granules, which membrane fuses to the apical membrane during exocytosis. This localization at the level of secretory cells is consistent with the detection of the corresponding RNA by in situ hybridization (3, 10). Moreover, a similar staining was observed with the two different antibodies that recognize two distinct regions of two different ENaC subunits. - and -ENaC were also detected by electron microscopy in the lumen of secretory granules (Fig. 4F) and were absent with control preimmune antisera (Fig. 4D). Although a vesicular localization of these membrane proteins might appear surprising, it is not unique and has been reported previously for CFTR (14).

Interestingly, our results suggest that in human airways, ENaC and CFTR proteins could be colocalized, even if their temporal expression throughout development differs (12). Further work will be necessary to quantify the relative levels of expression of the ENaC proteins along the bronchial tree and distal air spaces of the lung and to test whether distal airways can contribute significantly to the distal lung liquid clearance observed at birth or during pathological conditions with alveolar edema.

In conclusion, our work shows that - and -ENaC proteins are expressed not only in the apical membrane of the ciliated cells but also in cells lining glandular ducts, in the serous gland cells, and in Clara cells. Early expression of ENaC proteins in human fetal airways suggests that Na⁺ absorption might begin significantly before birth, even if secretion is still dominant.