INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Throughout gestation, the presence of an adequate amount of lung liquid is critical for normal lung growth and development (1). To prepare for an ex utero air breathing existence, the fetal lung must convert from net fluid secretion to net fluid absorption, the latter resulting from the active transport of Na⁺ from the apical to the interstitial side of the respiratory epithelium. The major pathway for Na⁺ absorption across the epithelium's apical membrane is the amiloride sensitive epithelial Na⁺ channel (ENaC). ENaC has three subunits, with the  subunit being essential for channel function (2). The critical role that amiloride-sensitive Na⁺ transport and the  subunit of ENaC play in perinatal lung liquid clearance has been demonstrated in animal studies. The instillation of the Na⁺ channel blockers amiloride (3) or benzamil (4) into the fluid-filled air spaces of full-term guinea pigs resulted in delayed clearance of lung liquid, hypoxemia, and respiratory distress. Mice that are genetically deficient in the  ENaC subunit, are unable to clear liquid from their air spaces and die from respiratory distress (5).

The degree of lung maturity has a profound effect on its epithelium's ability to actively transport Na⁺. For example, the immature, in contrast to the mature intrauterine fetal lamb cannot convert from fluid secretion to amiloride-sensitive fluid absorption in response to -agonists (6). It has also been demonstrated in rodents by Northern analysis of whole fetal lungs that the expression of the , , and  subunits of ENaC mRNA is differentially regulated during development (7) and that there is a surge in ENaC expression in the rat and mouse towards late gestation (7, 8).

The expression of ENaC mRNA in the epithelium of the developing human lung is incompletely understood. Although Northern analysis has shown low levels of ENaC mRNA in early gestation whole human fetal lungs (9) and epithelium cultured from human fetal lungs (10), there are no studies that have evaluated the ontogeny and cellular localization of ENaC using in situ hybridization. There is also apparently contradictory data regarding the biologic importance of lung ENaC expression at the time of birth in humans. Clinical studies in premature human infants (11) have shown that the amiloride-sensitive drop in potential difference (PD) between the nasal epithelium and the subcutaneous space, a surrogate for ENaC activity, is decreased in those infants who develop respiratory distress syndrome (RDS), compared with those without RDS. Active fluid absorption, and hence presumably epithelial Na⁺ transport, is also a major mechanism involved in recovery from pulmonary edema in adult patients (12). In contrast to these observations, adult patients who have pseudohypoaldosteronism arising from a genetic mutation in their ENaC subunit do not have a history of respiratory distress at birth (13).

We therefore performed the following study to determine the ontogeny and cellular expression of the  subunit of ENaC using in situ hybridization during normal and abnormal human fetal lung development. We obtained samples from the earliest stages of normal human lung development, from otherwise normal infants who had been born prematurely, and from term newborns with disorders known to be associated with marked hypoplastic and dysplastic lung development. In this study, which is the first such study in humans, fetal lungs were analyzed for ENaC mRNA expression using a nonradioisotopic high resolution in situ hybridization technique. In addition to the sense and antisense in situ hybridization analyses, structural and epithelial lung characteristics were assessed using hematoxylin-eosin (H&E) staining and cytokeratin (CK) immunohistochemistry. Surprisingly, we detected strong ENaC expression in the epithelium of the embryonic stage lung, within the lungs of preterm infants who had RDS secondary to premature birth, and term infants who had hypoplastic lungs.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Normal Lung Samples

Lung tissue was obtained from 13 autopsies from fetuses 5- to 45-wk postconception. Fetal lung material was obtained from archived autopsy material at The Hospital for Sick Children, Toronto, Canada. Normal fetal lung tissues were obtained from pregnancies terminated by prostaglandin induction or dilatation and evacuation, from spontaneous abortions, or from stillbirths. All normal early gestation tissues: embryonic (n = 2, gestational age [ga] = 5 wk), pseudoglandular (n = 2, ga = 10 wk), and canalicular stage (n = 3, ga = 19 to 20 wk) were obtained from pregnancies terminated for therapeutic reasons performed by dilatation and evacuation. Saccular stage tissues were obtained from the autopsy of two stillborn infants (ga = 28 to 32 wk) after their mothers were involved in motor vehicle accidents. Normal alveolar stage lung tissue (n = 4, ga = 38 to 41 wk) was obtained from four cases of infant death at birth resulting from umbilical cord asphyxia (n = 2) or placental abruption (n = 2). All of these above lung samples were normal for their respective gestational age.

Abnormal Lung Samples

Archived autopsy samples were obtained from eight newborns with different perinatal lung diseases who were born at gestational ages ranging from 29 to 41 wk. Three were premature infants who developed typical RDS after birth. Two infants succumbed from their respiratory disease within hours of life, and the third infant died 4 d after birth as a result of intraventricular hemorrhage. Only one infant received exogenous surfactant therapy and antenatal steroids and none had received assisted ventilation. All three lung samples showed typical hyaline membrane disease with coexistent immature pulmonary parenchyma consistent with their gestational age. The remaining five samples were derived from infants with pulmonary hypoplasia. Two infants had oligohydramnios; one with associated cystic renal dysplasia and the other with renal agenesis (Potter's syndrome). Both of these infants died within hours as a result of respiratory insufficiency caused by pulmonary hypoplasia. Despite their full term gestation, the lungs were approximately 45% the normal lung weight for their body size. The radioalveolar count was decreased with a reduced number of alveolar units and extension of airways close to the pleural surface (14, 15). Three infants with congenital diaphragmatic hernia (CDH) had displacement of abdominal contents within the pleural space and severe bilateral pulmonary hypoplasia. Despite their full-term gestation, the lungs were only 10 to 15% of normal weight for their body size. All three infants with CDH died within 3 h of life and postmortem examination revealed alveolar hypoplasia and microatelectasis with large sized airways close to the pleural membrane.

Anatomic Terminology

Reference to "large airways" in this study includes all bronchi (1°, 2°, 3°) that are characterized by their large diameter, which is more than 1 mm in fully developed lungs, tall columnar epithelium, and/or the presence of cartilage, submucosal glands, and goblet cells. "Small airways" in this study includes all generations of bronchioles, including the terminal and respiratory bronchioles. These bronchioles are characterized by a small diameter of less than 1 mm in fully developed lungs, low columnar-to-cuboidal epithelium, with the concomitant absence of cartilage, submucosal glands, and goblet cells. The "distal lung unit" designation refers to all structures distal to the small airways such as the alveolar ducts, alveolar sacs, and alveoli. The characteristic distal lung unit air space is not present until the late canalicular stage; however, the small airway end points in the early canalicular stage have been shown in previous studies to be precursors of the future saccules and alveoli (16). The peripheral structures in the canalicular sections are thus also included in this "distal lung unit" designation.

Kidney Samples

Fetal kidney material was obtained from an archived sample at The Hospital for Sick Children from a fetus 26 wk postconception, following the autopsy of a stillborn infant after the mother had been involved in a motor vehicle accident. Adult kidney samples were obtained from surgical specimens resected for malignancy (research protocol approved by The Toronto Hospital Committee for Research on Human Subjects). Kidney samples were chosen as a positive control since the pattern of ENaC mRNA expression in the kidney has already been characterized by in situ hybridization (17).

Ethics Approval

All human subject material was utilized with approval of the Human Subjects Review Committee of The Hospital for Sick Children, Toronto Canada.

Sample Preparation

Tissues were fixed in 10% formalin and embedded in paraffin using routine methods. Serial sections were used for sense and antisense in situ hybridizations, H&E staining, and CK immunohistochemistry.

ENaC cRNA Probe Preparation

A 319 bp cDNA fragment corresponding to nt 2169 to 2488 of human ENaC (GenBank accession no. X76180) was subcloned into pGEM  3Zf (+/) (Promega, Madison, WI) in both the sense and antisense orientations. Plasmids containing the ENaC probe sequence were linearized with SacI and blunted with T4 DNA polymerase. The cRNA probes were synthesized by in vitro transcription with T7 RNA polymerase and digoxigenin (DIG)-labeled uridine triphosphate (UTP) (Boehringer Mannheim, Dorval, PQ, Canada) according to the manufacturer's protocol.

Northern Blot Analysis of Probes

Northern blot analysis was carried out to test the specificity of antisense and sense cRNA probes, using RNA isolated from adult human lung, kidney, and liver. Tissues were collected from surgical specimens resected for malignancy (research protocol was approved by The Toronto Hospital Committee for Research on Human Subjects). Twenty micrograms of total RNA were separated on a 1% agarose-formaldehyde gel and transferred to nylon membrane. Hybridization to DIG-labeled cRNA probes was carried out in DIG EASY HYB solution (Boehringer Mannheim). Hybridization and washes followed the manufacturer's instructions. Chemiluminescent detection was performed using the DIG Nucleic Acid Detection System (Boehringer Mannheim) followed by a 20-min exposure to Kodak X-OMAT film.

In Situ Hybridization

To enhance signal and facilitate probe penetration, sections were transferred into a pressure cooker containing 1 L of 0.1 M TRIS at pH 8.0 and heated at maximum power for 18 min in a microwave. After the heating step, the slides were left in the pressure cooker with the lid on for 15 min, and then an additional 30 min with the lid off. Prehybridization and hybridization were carried out as described previously (18, 19), with the exception that incubations were carried out at 37° C, and the RNase digestion step was omitted. Slides were washed sequentially in 2× SSC at 37° C, 1× SSC at 37° C, 0.5× SSC at 47° C, and 0.1× SSC at 47° C for 15 min each wash (1× SSC = 15 mM tri-sodium citrate/150 mM NaCl at pH 7). Detection was achieved with the DIG Nucleic Acid Detection kit (Boehringer Mannheim) following the manufacturer's instructions. Slides were mounted in routine fashion.

CK Immunohistochemistry

Sections of lung and kidney were evaluated for the presence of the epithelial-specific cell marker CK using mouse monoclonal antibody CAM5.2 against low molecular weight CK (Becton Dickinson Immunocytometry Systems, San Jose, CA) diluted 1:30, followed by biotinylated horse antimouse IgG (Vector Laboratories, Burlingame, CA) at 5 µl/ml and Elite Avidin Biotin Complex (Vector Laboratories). The reaction was visualized using a diaminobenzidine substrate.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Characterization of ENaC cRNA Probe

The DIG-labeled ENaC antisense cRNA probe detected a single band of 3.7 kb on a Northern blot of RNA isolated from adult human lung and kidney, but not from liver (data not shown), in agreement with previous studies (9, 20). No signal was detected using the sense cRNA probe. Using in situ hybridization, ENaC mRNA was detected in fetal kidney distal nephron (Figure 1E) and the adult kidney distal nephron (data not shown), but it was absent in the glomerulus (data not shown). This is consistent with previous results (17). There was minimal background using the ENaC sense cRNA probe (Figure 1F). We further demonstrated that the DIG-labeled probe signal could be competed out by adding increasing amounts of unlabeled ENaC antisense cRNA to the hybridization step, and that RNase treatment of sections prior to hybridization also eliminated all signal (data not shown). These experiments confirm the specificity of the probe used for ENaC mRNA.

View larger version (103K): [in this window] [in a new window]   Figure 1.   Expression of ENaC mRNA in embryonic human lung bud at 5 wk gestation by nonisotopic in situ hybridization. The ENaC antisense probe labeled the epithelial cells lining the human fetal lung bud (A). No signal was detected using the ENaC sense probe (B). In (C ), epithelial cells lining the bud were identified by immunohistochemistry for low molecular weight CK (hematoxylin counterstained). Lung bud morphology was further visualized by H&E staining in (D). In situ hybridization detection of ENaC mRNA in fetal human kidney was performed as a positive control (E, F ). The ENaC antisense cRNA probe labeled the epithelial cells lining the distal nephron (arrows) (E ), but no signal was detected in the glomerulus (data not shown). Minimal background signal was detected using the ENaC sense cRNA probe (F ).

ENaC mRNA Expression in Human Lung Development

Embryonic stage. ENaC mRNA was detected in sections from the embryonic stage of human lung development by in situ hybridization. The signal was uniformly distributed throughout all the epithelial cells lining the lung bud with a trend for localization in the basal and apical aspects of the cell (Figure 1A). The H&E stains demonstrated that the embryonic lung consisted of a solitary bud lined by undifferentiated pseudostratified epithelium (Figure 1D), which was positive for CK expression as assessed by immunohistochemistry (Figure 1C).

Pseudoglandular stage. In the pseudoglandular stage samples (Figure 2), airway development was evident and the epithelium remained undifferentiated. The ENaC signal was present in a uniform pattern in both large and small airway epithelia.

View larger version (157K): [in this window] [in a new window]   Figure 2.   Expression of ENaC mRNA in pseudoglandular stage human lung at 11 wk gestation by nonisotopic in situ hybridization. (A) Antisense cRNA probe detected a uniform pattern of ENaC mRNA expression in epithelial cells lining the developing airway (a) epithelium. (B) No hybridization signal was detected using the sense probe.

Canalicular stage. In the canalicular stage samples (Figure 3), the large airway epithelium morphologically resembled the postnatal lung, with the appearance of ciliated and secretory cells (Figure 3G). In the small airways, the transition from columnar to cuboidal cells demarcated the future respiratory portion of the lung. The ENaC mRNA signal was present throughout the airway epithelium and notably absent from the vascular tissues (Figure 3A). The large (proximal), more differentiated airway epithelium exhibited a more patchy signal compared with the signal in the distal (small) airway epithelium (Figure 3D).

View larger version (144K): [in this window] [in a new window]   Figure 3.   Expression of ENaC mRNA in canalicular stage human lung at 18-wk gestation by nonisotopic in situ hybridization. Sections representative of the distal (A-C ), medium (D-F ), and large (G-I ) airway regions are shown. In distal lung (A), a uniform pattern of ENaC mRNA expression was detected in the epithelial cells lining the distal airway, but no signal was detected in vascular structures (v). ENaC mRNA in the epithelial cells lining the medium (D) and large (G) airways (arrows) was expressed in a more patchy fashion. No signal was detected using the ENaC sense cRNA probe (B, E, H ). Immunohistochemistry for low molecular weight CK labeled epithelial cells lining all airways (C, F, I ).

Saccular stage. In the saccular stage samples (Figure 4), the interstitial tissue was greatly decreased relative to earlier gestational ages, and saccules were clearly present. The large airway epithelium was well differentiated. The cuboidal epithelium lining the distal lung unit had begun to flatten forming an early air-blood interface. Cuboidal, alveolar type II (ATII) pneumocytes were detected. Flattening of cuboidal epithelial cells may also represent the differentiation of ATII into ATI cells. Submucosal glands were present. The ENaC signal was present in most of the superficial airway epithelia and in the serous cells of the submucosal gland acini. The ENaC signal in the distal lung unit was localized to the corner cells in an ATII cell pattern.

View larger version (155K): [in this window] [in a new window]   Figure 4.   Expression of ENaC mRNA in saccular stage human lung at 32-wk gestation by nonisotopic in situ hybridization. Sections representative of alveolar (A-C ) and large airway (D-F ) regions are shown. ENaC mRNA was detected in discrete cells (thin arrow) of the terminal sac regions consistent with future expression in ATII cells (A), and in serous cells of the submucosal glands (thick arrow) and epithelial cells linking the large airways (D). No signal was detected using the ENaC sense cRNA probe (B, E). Immunohistochemistry for low molecular weight CK (C, F ) labeled epithelial cells of large and small airways and submucosal glands.

Alveolar stage. In the alveolar stage samples (Figure 5), the alveolus was thin-walled and the alveolar epithelial differentiation was advanced. Submucosal glands were prevalent in the large airways (Figure 5D). The ENaC mRNA was strongly expressed in large airway epithelium and in the serous cells of the submucosal gland acini (Figure 5D). Alveolar expression of ENaC mRNA was consistent with an ATII cell pattern (Figure 5A). Expression in small airways (Figure 5G) appeared to be somewhat weaker than in large airways or alveolar regions.

View larger version (96K): [in this window] [in a new window]   Figure 5.   Expression of ENaC mRNA in alveolar stage human lung at 41-wk gestation. Sections representative of alveolar (A-C), large airway (D-F ), and small airway (G-I) regions are shown. ENaC mRNA was detected in the alveoli (A) in a pattern consistent with an alveolar type II cell pattern (thin arrow). In large airways, ENaC mRNA was detected in a patchy fashion in epithelial cells lining the airway and in the serous cells of the submucosal glands (thick arrow). ENaC mRNA expression in small airways (G) was uniform over the superficial epithelium, but it appeared to be weaker. No signal was detected using the ENaC sense cRNA probe (B, E, H ). Immunohistochemistry for low molecular weight CK (C, F, I ) labeled epithelial cells lining the alveolus, airway, and submucosal glands.

ENaC mRNA Expression in Newborn Lung Disease

Infant prematurity. In samples derived from infants exhibiting respiratory distress caused by pulmonary immaturity secondary to infant prematurity, the interstitial thickness and alveolar complexity observed in lung sections were comparable with that of the normal fetal lung at the early saccular stage of development (Figure 6C). ENaC mRNA signal was patchy in the large airway epithelium (Figure 6D). Although alveolar development is not complete, small discrete areas of labeling were present within immature alveolar units; this signal may correspond to future ATII cells (Figure 6A). Expression in sections derived from the infant who had received antenatal steroids (not shown) was not detectably different from infants who had not received steroids.

View larger version (118K): [in this window] [in a new window]   Figure 6.   Expression of ENaC mRNA in perinatal lung disease. In a case of pulmonary immaturity from an infant born at 28-wk gestation (A-F ), ENaC mRNA was detected in the alveolus (A), in discrete cells consistent with the expected location of future alveolar type II cells. ENaC mRNA was also detected in epithelial cells lining large airways (D). The increased interstitial thickness and decreased alveolar complexity demonstrated by CK immunolabeling (D, F ) was consistent with that seen in normal fetal lung in the early saccular stage of development (not shown). A similar pattern of ENaC mRNA was detected in a case of pulmonary hypoplasia associated with Potter's syndrome in a term infant (G- L), with ENaC mRNA again detected in a subset of cells in the alveolus (G) and epithelial cells lining large airways ( J ). There was an increased amount of mesenchymal tissue detected in sections immunolabeled for CK (I, L) compared with that in normal term infants (Figure 3C), reflecting the retarded alveolar development in newborns with Potter's syndrome. No signal was detected using the ENaC sense cRNA probe (B, E, H, K ).

Oligohydramnios. In lung samples derived from cases of pulmonary hypoplasia associated with oligohydramnios, there was histologic evidence of a delayed alveolar development (Figure 6I). The interstitium was much thicker and the mesenchymal cells more abundant than what is seen in normal lung at term gestation. Despite the morphologic immaturity, the ENaC mRNA signal was observed in small discrete areas of distal lung unit epithelium (Figure 6G) and the superficial airway epithelium (Figure 6J).

Congenital diaphragmatic hernia (CDH). In lung samples derived from cases of pulmonary hypoplasia associated with CDH, the alveoli were hypoplastic (data not shown). Mesenchymal cells were more abundant in these lungs compared with normal lungs at the same term gestation. The ENaC mRNA signal was distributed in a patchy fashion throughout the large airway epithelium, and in small discrete areas within the immature alveolar parenchyma.

ENaC in situ hybridization results for both normal and abnormal human fetal lung studies are summarized in Table 1. Only background signal was present in all in situ hybridization experiments performed using the sense cRNA probe.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our initial studies of the ontogeny of ENaC expression using Northern analysis in rats (7, 21) led us to speculate that ENaC mRNA expression was linked to the maturation of the respiratory epithelium. In contrast, our study of the ontogeny and cellular localization of ENaC mRNA during normal and abnormal lung development showed that ENaC subunit mRNA is expressed in the epithelium from the earliest stage of human lung development. As the human lung develops, strong but patchy expression was noted in superficial epithelium of larger airways, whereas expression in bronchioles appeared to be weaker but more uniform. As the lung matures, strong expression is also noted in serous cells of submucosal glands and the distal lung unit in a pattern consistent with the known distribution of ATII epithelium. The expression of ENaC mRNA was readily detected in newborns who had lung disease secondary to premature birth or from lung hypoplasia. Expression of alveolar epithelial cell proteins may be either linked (for example, surfactant protein(SP)-A) (22, 23) or unlinked (SP-C) (24, 25) to the maturation of respiratory epithelium. Our observations suggest that in humans, ENaC mRNA is constitutively expressed throughout lung development and is separate (unlinked) from the process of pulmonary epithelial maturation and differentiation.

The cellular localization of ENaC mRNA expression has previously been studied in the fetal mouse lung (26), adult rat and mouse lung (26), and human airway epithelium (20) by in situ hybridization using radioactive probes. During fetal mouse development, ENaC mRNA was not detected until 16-day gestation and it was seen only in the central bronchi. Expression spread to more distal airways and developing acinar structures on Days 17 to 19. This is in sharp contrast to our results in developing human lung where ENaC mRNA was detected in all airway epithelia from the earliest stages of lung development. A possible explanation for our findings is that species differences exist and that ENaC expression occurs at a much earlier time point in humans than in rodents. It is well known that there are important interspecies variations in lung development. For example, lung development in humans is more advanced by end gestation compared with that in the rat in which alveolar development predominately occurs after birth (30). In the adult rodent, ENaC had a diffuse expression pattern in the epithelia of the trachea, bronchi, and bronchioles and in nasal and tracheal submucosal glands; expression in distal lung was consistent with the distribution of ATII cells. This is similar to our observations in full-term human neonatal lung, with the exception that we observed patchy expression in large airways, and relatively weaker, uniform expression in bronchioles. These differences may be related to the increased resolution available in our study using non- radioactive probes; for example, lesser resolution in large airways would have given the appearance of a weaker, more uniform signal, comparable to what we have seen in small airways.

The only previous study using in situ hybridization analysis of the human respiratory tract was limited to the nasal cavity and larger bronchi of adults (20) where ENaC expression was detected in the superficial epithelium of the upper airways and in submucosal gland ductular and acinar epithelium. It was suggested, in contrast with the serous cell-specific CFTR expression in gland acini, that ENaC mRNA was expressed in both serous and mucous cells in the gland acini. In our present study, ENaC expression in the full-term human fetal lung was largely consistent with that reported for adult human lung, with the exception that ENaC expression within the submucosal gland acini was limited to the serous cells, whereas the mucous cells were negative.

Our study has shown that ENaC subunit mRNA is detectable in preterm human infants with RDS. These results appear to contrast with previous observations that amiloride-sensitive PD and presumably ENaC activity is diminished in the nasal epithelium of preterm infants with RDS (11, 31). There are several possible explanations for these findings. First, since it is well known that ENaC is composed of three subunits, , , and , and that all three are required for maximal activity (2), it is possible that in cases of incomplete lung growth, expression of one of the other two subunits is deficient and causes impaired Na⁺ transport. The limited amounts of human fetal lung samples available to us prevented us from examining all three ENaC subunits within the scope of this study. Second, it is recognized that a direct relationship between mRNA and functional protein cannot be predicted, i.e., it is possible that ENaC function is regulated by post-transcriptional mechanisms. Alternate potential sites of regulation include: , , and  subunit translation, ENaC subunit assembly and stability, ENaC transport to the membrane and/or control of ENaC function at the membrane surface. It has not been possible to determine the ontogeny of ENaC protein expression. Polyclonal antibodies raised against the biochemically purified renal amiloride-sensitive ENaC (32, 33) or cloned ENaC fusion proteins (17, 34) do recognize epitopes in fetal and adult lung epithelium; however, in each of these studies Western analyses showed that the antibodies recognize multiple bands in native epithelia. Thus, the validity of immunohistologic studies using these antibodies is uncertain. Third, it is possible that alternate Na⁺ transport pathways, other than ENaC, play an important role in Na⁺ and fluid transport in the human fetal lung. Such a hypothesis is supported by biochemical (32) and electrophysiologic (9, 32, 35) studies, which suggest the presence of more than one amiloride binding Na⁺ permeant ion channel in fetal distal lung epithelium. This speculation, combined with our present results, is compatible with the observations that adults with genetic mutations in ENaC (pseudohypoaldosteronism) do not have an obvious history of RDS at birth (13). Regardless of the correct explanation, our present study clearly indicates that much further work is required before we understand how the human lung regulates the biologically important Na⁺ transport across its epithelium.