INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The amiloride-sensitive epithelial Na⁺ channel (ENaC) is an important pathway for Na⁺ movement across the apical membrane of Na⁺-transporting respiratory epithelia. Functional cloning has identified three homologous genes coding for ENaC subunits (, , and ) in the rat and human (1). Experimental studies using microinjected Xenopus oocytes demonstrated a requirement for coexpression of all three subunits (, , and ) for maximal Na⁺ channel activity; expression of  with  or of  with  led to a reduced whole-cell current (15% to 20% of maximal) and the expression of channels with different biophysical properties (conductance, sensitivity to inhibitors, and ion selectivity) (6). However, a precise stoichiometry for the channel has not yet been determined. The relation between endogenous levels of , , and  human ENaC (hENaC), and the resultant biologic activity in humans in vivo, is unknown.

The active transport of Na⁺, with Cl and water following, plays an important role in at least three relatively common and potentially lethal human lung diseases. Cystic fibrosis (CF) is associated with a marked increase in respiratory epithelial Na⁺ transport (7), and it is believed that this is an important pathogenic mechanism leading to CF lung disease (8, 9). Effective amiloride-sensitive respiratory epithelial Na⁺ transport is essential in the neonatal period to prevent respiratory distress syndrome (RDS). When amiloride is placed in the normally fluid-filled air spaces of newborn animals, they develop respiratory distress and hypoxemia, and fail to clear their lung liquid (10). Knockout mice deficient for the  subunit of ENaC fail to clear their fetal lung fluid and die shortly after birth (11). These findings, along with the knowledge that the immature fetal lung has very low levels of  (12) and  and  (13) ENaC mRNA, and that newborn infants with RDS have an abnormally low amiloride-sensitive electrical potential difference (PD) between their respiratory epithelium and the subcutaneous space (14, 15), suggest that one mechanism leading to neonatal RDS is inadequate respiratory epithelial Na⁺ transport. Additionally, it has been demonstrated that amiloride-sensitive Na⁺ transport plays an important role in the clearance of alveolar pulmonary edema in the adult lung (16, 17), and that the ability to actively concentrate the air space edema fluid correlates with the recovery from high-pressure or high-permeability pulmonary edema (18).

Our knowledge about levels of mRNA for the , , and  subunits of hENaC and their effect on Na⁺ transport within respiratory epithelia in vivo is modest because of limited tissue availability and the lack of adequate assay systems. In the present work, we set out to develop a highly sensitive assay to accurately measure hENaC-subunit mRNA expression in very small and easily obtained respiratory epithelial samples from normal humans. Such an assay would be very valuable for assessing the role of ENaC gene expression in human lung development, as well as in acquired and genetic lung diseases, as described earlier. We selected epithelium of the inferior nasal turbinate as an easily accessible source of tissue for pilot studies, and also because we could readily measure an hENaC- related function at the same site through measurements of the electrical potential difference (PD) between the surface of the nasal epithelium and the subcutaneous space. Nasal PD has been used as an indirect assessment of Na⁺ transport (14, 19), and measurements in this specific area have been shown to reflect ion-transport patterns in more distal regions of the human respiratory tract (20). We also tested the hypothesis that mRNA expression for the subunits of hENaC in nasal epithelia correlates with nasal PD.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cell Culture

Human lung cell lines NCI H661 (large-cell carcinoma) and NCI H441 (papillary adenocarcinoma) were obtained from the American Type Culture Collection (Rockville, MD) and cultured in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS) and 1% penicillin-streptomycin.

Nasal Epithelial PD

Eight healthy men who were nonsmokers, did not have a history of nasal allergy or respiratory infection within the previous 2 wk, and were not taking any medications were recruited for the study. Nasal PD was measured in both nares, underneath the inferior turbinate, at 0.5, 1.0, 1.5, 2.0, and 3.0 cm posterior to the anterior tip of the turbinate, according to techniques developed by Knowles and colleagues (19) and previously used by us (L.E., P.D.) in patients with CF (21). The response of the nasal PD to a Na⁺-channel blocker was measured in one nare at the site of maximal PD (maxPD) by infusing onto the site a solution containing 100 µM amiloride (Merck Frosst, Kirkland, PQ, Canada). The study was approved by our institution's human research ethics board, and consent was obtained from all subjects.

Tissue Sampling from Nasal Epithelium

In preliminary experiments, we sampled nasal epithelium at the site of amiloride infusion. However, for unknown reasons, we could not obtain sufficient RNA. Therefore, nasal epithelium was obtained from the nare contralateral to that used for the amiloride infusion, at the site of this contralateral nare's maxPD. The sample was obtained without the use of local anesthetics, through a technique identical to that utilized in the clinical evaluation of patients for primary ciliary dyskinesia. Under direct visual inspection, a single tissue scraping was taken, using a Rhinoprobe (Arlington Scientific, Arlington, TX) disposable plastic scoop (1.5-mm diameter, 0.5-mm depth). Scrapes were immediately dispersed in 600 µl of RLT Lysis Buffer (RNeasy RNA preparation kit; Qiagen, Santa Clarita, CA) by repeated passage through a 29-gauge needle. To determine the cell populations obtained by this method, comparable scrapes were taken and subjected to histologic examination with the Pap smear technique.

Preparation and Quantitation of Total RNA

Total RNA from cultured cells and nasal scrapes was isolated with the QIAshredder and RNeasy kit (Qiagen), following the manufacturer's instructions. RNA from cultured cells was quantitated through standard UV absorbance measurements. RNA from nasal scrapes was quantitated by slot blot analysis (22). Briefly, aliquots of nasal epithelial-cell RNA were applied in triplicate to Hybond-N⁺ membrane (Amersham Life Science, Amersham, UK) alongside known quantities of NCI H661-cell RNA. Blots were hybridized overnight at 65° C to a ³²P-labeled complementary DNA (cDNA) probe to human 18S rRNA in Expresshyb buffer (Clontech, Palo Alto, CA). Filters were washed twice for 10 min in 2× standard saline citrate (SSC)/0.1% sodium dodecyl sulfate (SDS), and then twice for 30 min each in 0.1× SSC/0.1% SDS at 65° C. Bands on the resulting autoradiograms were quantitated densitometrically (LKB Ultroscan XL), and RNA concentrations from nasal epithelial-cell RNA were calculated using a standard curve derived from the NCI H661-cell RNA data.

Internal Control RNA

Cytokeratin 18 (CK18) cDNA and , , and  hENaC cDNAs were amplified from phage  DNA extracted from a human kidney cDNA library (Clontech) by PCR. PCR reactions (100 µl) contained 1.5 µg cDNA template, 1× thermopolymerase buffer (20 mM Tris-HCl, pH 8.8; 10 mM KCl; 10 mM (NH₄)₂SO₄; 2 mM MgSO₄; and 0.1% Triton X-100), 100 µM deoxynucleotide triphosphates (dNTPs), 0.1 µM sense and antisense oligonucleotide primers, and 2 units of Vent polymerase (New England Biolabs, Mississauga, ON, Canada). Amplification was done in a GeneAmp 2400 thermocycler (Perkin Elmer, Mississauga, ON) under the following conditions: for  and  hENaC, 95° C for 1 min (pre-PCR); 35 cycles of 94° C for 1 min, 55° C for 1 min, and 72° C for 30 s; and 10 min at 72° C. For  hENaC and CK18, 95° C for 1 min (pre-PCR); 40 cycles of 94° C for 30 s, 58° C for 30 s, and 72° C for 30 s; and by 10 min at 72° C. Primers (Table 1) were extended at the 5'-end to include restriction sites to facilitate subsequent cloning into plasmids, and were designed to amplify 300- to 400-bp fragments containing suitable restriction sites for deletion construction. The cDNA for  hENaC was subcloned to EcoRI/BamHI-cut pGEM-3Zf(+) (Promega, Madison, WI); all other cDNAs were subcloned to SalI/ XbaI-cut pBluescript II KS (Stratagene, La Jolla, CA). Deletion constructs were created by removal of the following restriction fragments from the cDNAs:  hENaC, ScaI to KpnI (bp 424 to 519);  hENaC, AatII to PpuMI (bp 1027 to 1104);  hENaC, BalI to BalI (bp 1435 to 1620); CK18, StuI to BstXI (bp 714 to 790). The resulting deletion constructs were linearized by restriction-enzyme digestion 3' to the cDNA, and were used as templates for in vitro transcription by T7 ( hENaC) or T3 (,  hENaC and CK18) RNA polymerase (Promega), following the supplier's instructions. The resulting cRNAs were treated with ribonuclease (RNase)-free deoxyribonuclease (DNase) RQ1 (Promega), purified on an RNeasy column, and quantitated spectrophotometrically at 260 nm. In vitro-transcribed cRNAs were diluted and stored in aliquots at 70° C. 

Competitive QRT-PCR

A 1:2 serial dilution of target total RNA (from 512 ng to 8 ng for cultured-cell RNA, 24 ng to 1.5 ng for nasal-scrape RNA) was prepared in diethylpyrocarbonate (DPC)-treated water in 0.2 ml MicroAmp reaction tubes (Perkin Elmer) in a final volume of 10 µl. The dilutions were heated to 65° C for 5 min and then chilled on wet ice. Aliquoted RT mixture, including the primer and internal-control cRNA, was added to each tube. Each reaction contained 1× first-strand synthesis buffer (50 mM Tris-HCl, pH 8.3; 75 mM KCl; 3 mM MgCl₂), 10 mM dithiothreitol (DTT), 0.5 mM dNTPs, 1 unit/µl RNAguard (Pharmacia Biotech, Baie d'Urfé, PQ), and 50 units Superscript II reverse transcriptase (Life Technologies, Burlington, ON). For quantitation of  hENaC, each reaction contained 250 ng of random hexamer primer; for  or  hENaC or CK18, each reaction contained 5 pmol of -8, -4, or CK-4 primer (Table 1), respectively. Internal-control cRNA was added to the RT mixture such that each tube of the target RNA dilution series received a fixed amount of control cRNA. The amounts of internal-control cRNAs used per reaction in assays of cultured-cell RNA were as follows:  hENaC, 100 fg (1.36 attomol) for NCI H441;  hENaC, 2.5 fg (2.16 × 10² attomol) for NCI H441, 10 fg (8.64 × 10² attomol) for NCI H661;  hENaC, 1 fg (7.46 × 10³ attomol) for NCI H441, 20 fg for NCI H661; and CK18, 10 pg (105 attomol) for NCI H441, 1 pg (10.5 attomol) for NCI H661. In assays of nasal-scrape RNA, these amounts were:  hENaC, 10 fg (0.136 attomol);  hENaC, 1.25 fg (1.08 × 10² attomol);  hENaC, 0.5 fg (3.73 × 10³ attomol); and CK18, 150 fg (1.58 attomol). Preparations for RT reactions were mixed and incubated for 45 min at 42° C, heated at 95° C for 5 min, and chilled to 4° C in a GeneAmp 9600 thermocycler (Perkin Elmer).

Unless otherwise specified, PCR reactions contained 1× Thermopol reaction buffer, 0.2 mM dNTPs, 0.4 µM sense and antisense primers (Table 1), 5 µl cDNA, and 0.5 units Vent(exo-) polymerase (New England Biolabs) in a final volume of 25 µl. For PCR of hENaC, the MgSO₄ concentration was increased to 4 mM and the primer concentration was reduced to 0.2 µM. Amplification of cRNA for  hENaC required two rounds of PCR with nested primers: -3 and -4 in the first round and 1340 and 1679 in the second round. Five microliters of a 1:100 dilution of first-round product was used as template in the second round. Amplification was done in a Perkin Elmer 9600 thermocycler. For  and  hENaC and CK18, a total of 34 cycles of amplification included two cycles of denaturation at 94° C for 1 min, annealing at 55° C for 1 min, and elongation at 72° C for 2 min, followed by 32 cycles in which the denaturation, annealing, and elongation times were reduced to 30 s, 30 s, and 1 min, respectively. For  hENaC, 26 cycles of 94° C for 30 s, 55° C for 30 s, and 72° C for 1 min were performed in each round.

Quantitative Analysis of PCR Products

Fifteen microliters of PCR product were loaded onto an 8% nondenaturing polyacrylamide gel and subjected to electrophoresis. After staining with ethidium bromide (0.5 µg/mL), the gel was photographed with Polaroid Type 665 positive/negative film (VWR Canlab, Mississauga, ON). The amplified DNA bands were quantitated from the negative by densitometry. The ratio of the integrated optical density (IOD) obtained from the target and internal control DNA bands was plotted on a log-log scale against the amount of target RNA present in the RT reaction mixture. The amount of hENaC or CK18 mRNA in the target RNA was calculated from the plot.

Statistical Analysis

Linear regression analysis was performed on plots of the ratio of amplification products against input RNA. Experiments in which the r² value was < 0.97 were discarded and the assay was repeated. QRT-PCR was done in duplicate or triplicate for each target in each RNA sample. Results are expressed as mean ± SE unless otherwise indicated. The correlation of hENaC mRNA levels with nasal PD measurements was determined with Pearson's correlation test and linear regression analysis with Instat v2.05a (GraphPad Software, San Diego, CA).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Quantitation of hENaC and CK18 mRNA in Cultured Human Respiratory Epithelia

The sensitivities of the RT-PCR conditions for each target were optimized by reverse transcribing and amplifying serial dilutions of the internal control cRNAs alone. We could routinely detect amplified bands of the expected size from less than 2.5 fg of CK18 cRNA. The detection limit for  hENaC was about 2 fg, and for  and  hENaC was below 1 fg (data not shown). None of the primer sets used amplified a band from human genomic DNA under the conditions described, presumably because the primers are located in different exons in the corresponding genes. This assumption has subsequently been shown to be correct for  hENaC (23) and for  hENaC (G.O. and H.O., unpublished data). This eliminated the need to stringently remove genomic DNA from RNA preparations.

Figure 1 illustrates a typical determination of  hENaC mRNA in NCI H441 total RNA. By visual inspection of the ethidium bromide-stained gel (Figure 1A), the intensities of the target and control bands are indistinguishable in Lane 3, which contained 128 ng of NCI H441 RNA. This suggests that 128 ng RNA contained approximately 1.36 attomol  hENaC mRNA (i.e., 10.6 attomol/µg). Exact quantitation was achieved by densitometric quantitation of the bands, followed by analysis using linear regression (Figure 1B). In competitive PCR, amplification efficiency is determined by the primer sequences, and many previous studies have shown that target and internal-control amplify with identical efficiencies when a single set of primers is used to amplify both targets (24). As expected when target and control RNAs amplify with equal efficiencies, the plotted relationship was linear over a 100-fold range. The representative analysis yielded a calculated concentration of 10.8 attomol hENaC mRNA per µg NCI H441 total RNA.

View larger version (19K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Figure 1.   Quantitation of  hENaC mRNA from NCI H441 cells. (A) Ethidium bromide staining of PCR products separated in 8% polyacrylamide gel. Serial dilutions of RNA were subjected to RT- PCR in the presence of 100 fg (1.36 attomol) of internal control  hENaC cRNA. The amounts of NCI H441 RNA in Lanes 1 to 7 are 512, 256, 128, 64, 32, 16, and 8 ng, respectively. Lanes 8 and 9 are negative controls for RT and PCR reactions. Lane 10 is a DNA marker (1-kb DNA ladder; Life Technologies). Two DNA bands, consistent with the predicted sizes of target and internal control products, were observed after polyacrylamide-gel electrophoresis. Since the same primers were used to amplify both target and control RNAs, the efficiency of the amplification reaction should not differ for the two, and the molar ratio of target to control RNA in the initial RT mixture should equal the molar ratio of the amplified bands. (B) Linear-regression analysis of the ratio of amplified bands versus amount of input RNA. Integrated optical density (IOD) was obtained by densitometric scanning of the Polaroid negative of the gel illustrated in A. Since the intensity of ethidium-bromide fluorescence of a DNA band depends on the mass of DNA present, whereas PCR amplification would maintain the molar ratio of the targets, the IOD of the control band was multiplied by the ratio of the band sizes (i.e., 306 bp/211 bp) to normalize the fluorescence intensity of the control band. The ratio of target- to internal-control-band densities after QRT-PCR was calculated for each dilution. This ratio was then plotted against the amount of target total RNA in each reaction.

A complete analysis of CK18 and hENaC mRNA concentrations in two human lung-carcinoma-cell lines was done in order to validate the method before using it in our human study. NCI H441 is derived from a papillary adenocarcinoma, and NCI H661 from a large-cell carcinoma. Results of this analysis are presented in Table 2. It should be noted that the absolute levels of each mRNA are highly reproducible within each cell line, but differ considerably between the two lines.  hENaC mRNA was undetectable in NCI H661 RNA even when up to 2 µg of total RNA was used, and in the absence of competitor cRNA, despite numerous attempts to derive suitable amplification conditions. On the basis of the assay sensitivity determined by amplifying decreasing amounts of control  hENaC cRNA alone, we predict that  hENaC mRNA is either absent in this cell line or that its concentration is much lower than 0.02 attomol/µg RNA.  hENaC mRNA levels for the two cell lines are somewhat similar, but the mRNA for  hENaC is approximately 20-fold greater whereas the mRNA for CK18 is 24-fold lower in NCI H661 than in NCI H441 cells.

Nasal PD in Healthy Men

The men enrolled in the study were 31 ± 7.1 (mean ± SD) (range: 23 to 45) yr old. In each case, visual appearance of the respiratory epithelium was normal. The individual nasal maxPDs at the site of the Rhinoprobe scrape and at the site of the amiloride infusion are listed in Table 3. The mean basal PDs (averaged PD values from five sites on the inferior turbinate, as described in METHODS, data not shown) and maxPDs (Table 3) in this group agree well with data published by Knowles and colleagues (19) and derived from 149 normal (i.e., non-CF) controls, and with a larger data set (39 healthy volunteers) obtained by members of our group (21), although the range is smaller. In addition, the absolute decrease in PD (PD), and the percent inhibition from baseline response to amiloride, are very similar to data in the foregoing studies. The baseline maxPDs are also similar to those reported for 40 healthy full-term newborns (22.7 ± 0.3 mV) obtained by Barker and coworkers (14). The comparable PD values suggest that our small sample is representative of a larger normal population. Since it was not possible to directly measure amiloride-sensitive PD at the same site as the scrape, we used the percent inhibition of PD by amiloride, measured in the contralateral nare, to calculate the amiloride-inhibitable PD at the site of the scrape. This calculation assumes that the percent inhibition is the same in both nares of an individual.

Histologic Analysis of Rhinoprobe Scrapings

The inferior turbinate of the nasal cavity is covered by a ciliated, pseudocolumnar epithelium, whereas squamous epithelium overlies its anterior tip (28). Since we were performing nasal scrapes at the point of maxPD, we should have obtained mainly nonsquamous epithelium. As expected, relatively few squamous cells were seen, with the majority of cells having a ciliated morphology (Figure 2).

View larger version (124K): [in this window] [in a new window]   Figure 2.   Epithelial cells obtained with Rhinoprobe nasal scrapings. Cells obtained by scraping under the inferior nasal turbinate were stained with the Pap-smear technique. Scale bar = 20 µm.

Quantitation of hENaC mRNA in Nasal Epithelium

The yield of purified RNA from Rhinoprobe scrapings ranged from 600 ng to 3 µg (determined by slot-blot analysis with 18S rRNA probe, data not shown). Although we initially attempted to obtain RNA from the exact site of amiloride application immediately after infusion, the total RNA yields in five attempts in different individuals ranged from 150 ng to less than 30 ng, which were insufficient for a complete quantitative analysis. In order to maximize the number of assays that could be performed on each sample, the starting amounts of total and internal control RNAs used in the QRT-PCR assays were reduced, and the number of dilutions tested were decreased relative to the cultured-cell assays (see METHODS). Duplicate quantitations of CK18 mRNA were performed on RNA from each nasal scrape. As shown in Figure 3, duplicate determinations within each sample gave consistent results, yet considerable variation in CK18 mRNA levels existed between scrapes taken from different subjects, ranging from 220 to 485 attomol/µg RNA, with a mean of 381 ± 82 attomol/µg. Using a CK18 mRNA length of 1,400 bp (29) and a predicted recovery of 1 µg RNA per 10,000 epithelial cells with the RNeasy protocol (manufacturer's handbook), this would translate to approximately 2,300 CK18 transcripts per cell. This level of expression is comparable to that of another housekeeping gene (actin) in lung and colon epithelial cells (30).

View larger version (60K): [in this window] [in a new window]   Figure 3.   CK18 mRNA levels in human nasal epithelial RNA. A 1:2 serial dilution of nasal epithelial RNA (24, 12, 6, 3, and 1.5 ng) from each of eight subjects was coamplified with CK18-specific primers by QRT-PCR with 150 fg of CK18 cRNA control. Following electrophoretic separation and analysis of the PCR products, CK18 concentrations were calculated for each sample. Bars represent mean ± SE of duplicate quantitations.

mRNA levels for , , and  hENaC were similarly determined in duplicate or triplicate for each nasal-epithelial RNA sample. The mean hENaC mRNA expression (initially derived in attomol/µg RNA) of individual samples was normalized by dividing it by the mean CK18 mRNA concentration of each sample. Expression of the three hENaC subunits relative to the marker CK18 is presented in the form of box plots in Figure 4. The mean level of  hENaC expression for the group of eight subjects was 39 ± 4.0 attomol/fmol CK18, or approximately 90 transcripts per cell. Mean  and  hENaC mRNA levels in the group were 5-fold and 21-fold lower than that of  hENaC, respectively (7.5 ± 0.92 and 1.8 ± 0.25 attomol/fmol CK18 mRNA).

View larger version (10K): [in this window] [in a new window]   Figure 4.   hENaC mRNA levels in human nasal epithelial RNA. QRT- PCR was performed for , , and  hENaC mRNAs, as described in the text. Each RNA sample was assayed in duplicate or triplicate, the results averaged, and the resulting hENaC-subunit mRNA level (in attomol/µg total RNA) divided by the CK18 mRNA concentration of the corresponding sample. Box plots represent the data from eight subjects. The extent of the box represents the 25th and 75th percentiles of the data. The center line in each box is the 50th percentile. Error bars indicate the 10th and 90th percentiles, and open circles indicate data points outside these limits.

Correlation of hENaC mRNA Levels with Nasal PD

We further analyzed hENaC expression data by plotting the normalized expression of each subunit against the calculated amiloride-sensitive PD at the site of the scrape (PD_scrape) (Figure 5). No significant correlation was found between PD and  or  hENaC mRNA levels. The mRNA levels for  hENaC had a negative correlation with PD. A similar correlation was found between  hENaC mRNA levels and the maxPD measured at the site of the scrape (not shown).

View larger version (14K): [in this window] [in a new window]   Figure 5.   Relation between hENaC mRNA expression and nasal PD. hENaC mRNA levels (attomol/fmol CK18) in nasal epithelial scrapings from eight healthy men were plotted against the calculated amiloride-sensitive PD at the site of the scrape. Although there was a trend for  hENaC to have a weak positive correlation (r2 = 0.34), it was not statistically significant (p > 0.1). In contrast, a significant (p < 0.01) correlation was found for  hENaC mRNA, with linear regression analysis yielding an r2 value of 0.72.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have developed a sensitive, quantitative assay to determine the mRNA levels for the , , and  subunits of hENaC in the respiratory epithelium of healthy, unanesthetized men. This assay has allowed us to make the first direct correlation of ,  and  hENaC mRNA levels with a relevant biologic activity in humans: specifically, the amiloride-sensitive respiratory epithelial PD, which reflects active Na⁺ transport. We found a negative correlation between the amount of  hENaC mRNA and the amiloride-sensitive PD of the nasal epithelium. Although the present study utilized the epithelium lining the inferior nasal turbinate, there is no obvious reason why the epithelium in the lower respiratory tract of humans cannot be utilized in a similar fashion.

The use of RT-PCR for quantitation of nucleic acids can be problematic, owing to the exponential nature of the reaction, whereby small variations in amplification efficiency can have profound effects on the yield of product (31). In addition, PCR amplification is subject to a plateau effect after many cycles, owing to depletion of reaction components, diminished enzymatic activity, and accumulation of products. The use of internal controls has been adopted to correct for tube-to-tube variation in amplification. Housekeeping genes, such as that for -actin, are the most convenient type of control and have been widely used, despite controversy about their validity (32). With such controls, the amplification of two unrelated DNA fragments can differ greatly because of differences in primer annealing and transcription efficiencies. This may also lead to plateauing of the two amplification targets after different numbers of cycles.

In the assays done in our study we used a control specifically created for each target mRNA, which consisted of in vitro-transcribed cRNA generated from a modified cDNA containing a deletion in the corresponding hENaC or CK18 sequence. The small deletion in the control allows separation of the PCR products on a polyacrylamide gel for quantitation of each band. Since the control is an RNA molecule, the efficiency of both the RT and PCR steps is controlled. Fixed amounts of the cRNA internal control are mixed with varying amounts of nasal epithelial RNA prior to RT-PCR, to generate a competitive reaction (27, 33). Since both sequences utilize the same primers, competitive amplification between the target and control takes place, such that both plateau at the same time. This allows us to use a high number of cycles in the PCR reaction and to generate sufficient product to visualize with ethidium bromide.

The sensitivity of the QRT-PCR method presented here allowed us to quantitate a specific mRNA from as little as 50 ng of total RNA, as compared with 1 µg for RNase protection or 10 to 20 µg for Northern hybridization analysis. In addition, this method allows the calculation of an absolute concentration for each mRNA, as opposed to Northern analysis, which allows only a relative quantitation. It is particularly difficult to estimate the relative abundance of different mRNAs using RNase protection or Northern hybridization, since the results would be affected by differences in the specific activities and hybridization efficiencies of the individual probes.

Nasal scrapes may contain nonepithelial cells, such as leukocytes or subepithelial cells, which would not express hENaC, but would contribute to the total RNA concentration measured by hybridization of slot-blots to 18S rRNA. To normalize for possible variations in the amount of epithelial-cell RNA in different nasal scrapes, we also quantitated the epithelial-cell-specific mRNA for CK18. Nasal scrapes in our healthy males were found to consist primarily of ciliated, pseudocolumnar epithelial cells, but did contain some squamous cells. We have no specific information on CK18 expression in squamous epithelium of the turbinate; however, in situ hybridization studies in rats suggest very low levels of hENaC expression in nonciliated squamous epithelium (34).

Qualitative analyses by in situ hybridization of the mRNA expressed for different ENaC subunits along the human and rat respiratory tract yield results consistent with an important role for ENaC in the absorption of Na⁺ and fluid across the pulmonary epithelium (34). Quantitative studies of ENaC expression have been limited by the low level of protein and RNA expression and the limited amounts of tissue obtainable, particularly in human studies. A previous measurement of hENaC mRNA expression in nasal epithelia was made with ribonuclease protection (36), but this method required the investigators to pool nasal scrapings from five to 10 individuals in order to obtain enough RNA for a single assay. Both techniques suggested a relative abundance of  >  > . Our quantitative results, derived from individual assays of each scrape, are in agreement with this prediction. In contrast, in situ hybridization analysis of nasal epithelium from rats suggested greater expression of the  and  subunits over the  subunit of ENaC (34). Whether this represents a true species difference or methodologic differences is unknown.

The physiologic significance of the correlation of  hENaC mRNA expression with nasal PD is unknown. It is recognized that the mRNA level does not necessarily reflect the protein level within the cell, owing to factors such as translational regulation and protein stability. In addition, the processing and activation pathways for a membrane channel such as hENaC may further modify the channel function as reflected by the amount of amiloride-sensitive PD. Nevertheless, it is intriguing that it was the -subunit mRNA, which is expressed in limiting amounts relative to the mRNAs for the  and  subunits, that correlated with channel function. It is also important to note that the correlation, contrary to what may be expected, was a negative one. As mentioned in the introduction, the stoichiometry of native ENaC has not yet been published, and it is possible that different stoichiometries may exist in different tissues. McNicholas and Canessa (6) have recently reported studies suggesting that the stoichiometry and order of the subunits around the channel pore control the functional properties of the channel. Our results suggest that higher expression of the  subunit may result in the assembly of channels with altered transport properties, leading to a decreased nasal PD. Alternatively, a channel containing more  subunits might be processed differently, altering activation or degradation of the channel. Recent studies on the stability of ENaC in fact suggest that  ENaC is the predominant subunit regulating ubiquitination-mediated degradation of the ENaC channel (37).

Although statistical significance was not reached, the data in Figure 5 suggest that a positive correlation may exist between  hENaC mRNA levels and amiloride-sensitive nasal PD. We could detect no hint of a correlation between  hENaC and PD. Since we examined only healthy adult males in this study, the range of values detected may be too narrow to adequately demonstrate a relationship between  and  hENaC mRNA expression and biologic activity. Future studies, described below, will reexamine these relationships as they relate to lung development, injury, and disease.

The sensitivity of the QRT-PCR assay will make it valuable in studying clinically relevant patient populations. Considerable evidence shows that removal of excess liquid from the air spaces of the lung is primarily driven by active Na⁺ transport across the tight alveolar epithelial barrier (38). There is clinical evidence that patients with pulmonary edema who can reabsorb some alveolar edema fluid within 12 h after intubation and acute lung injury exhibit more rapid recovery from respiratory failure, and a lower mortality (18). This was demonstrated by quantitation of the edema-fluid protein concentration in sequential samples of lung fluid from mechanically ventilated adult patients suffering from either high-pressure or high-permeability pulmonary edema. In addition, a recent study of preterm infants showed that the amiloride-sensitive nasal PD was significantly lower in infants in whom RDS subsequently developed, than in those who did not go on to develop RDS (14). Application of QRT-PCR in similar patient groups would allow us to elucidate the role of transcriptional regulation of hENaC in the resolution of pulmonary edema. Furthermore, the ability to measure hENaC mRNA levels may lead to an improved understanding of the genotype-phenotype correlation in CF lung disease.