Drug Transporter Protein Quantification of Immortalized Human Lung Cell Lines Derived from Tracheobronchial Epithelial Cells (Calu-3 and BEAS2-B), Bronchiolar–Alveolar Cells (NCI-H292 and NCI-H441), and Alveolar Type II-like Cells (A549) by Liquid Chromatography–Tandem Mass Spectrometry

INTRODUCTION

Cell culture-based in vitro systems of the respiratory tract have been employed for the permeability screening of new drug libraries and are expected to shorten the development time for new drugs as an alternative to time-consuming animal testing. The lung anatomically consists of the trachea, bronchi, and alveoli, and the cellular composition of respiratory epithelium varies by region. To date, several pulmonary diseases have been reported that are specific to each region. For instance, tracheomalacia and Mounier–Kuhn syndrome are tracheal diseases, bronchitis, and asthma are bronchial diseases, and emphysema is an alveolar disease. Therefore, it is important to evaluate the region specificity in in vitro studies to predict the disposition of drug candidates. Primary cultured lung cells are considered to be the best experimental model for in vivo studies and are used to study the impact of toxins in the lung or to test new drugs. However, they are more difficult to obtain and have limited growth activity. Therefore, there is an increasing interest in immortalized lung cell lines as an alternative to primary cultured lung cells. Several immortalized lung cell lines have been established for this aim, for example, Calu-3 and BEAS2-B cell lines have been commonly used for the studies of metabolism and the interaction of cells with xenobiotics as models of tracheobronchial epithelial cells.1 A549 cells are the most commonly used in alveolar epithelial models.2, 3 NCI-H292 and NCI-H441 are derived from the human lung carcinoma and have characteristics similar to those of alveolar and bronchiolar epithelial cells.4, 5 It is widely known that A549, BEAS2-B, and NCI-H292 have less polarized characteristics unlike Calu-3 and NCI-H441, nevertheless A549 and BEAS2-B were used for transport studies as alternatives of primary cultured cells.6-8 However, it still remains unclear whether these immortalized commercially available lung cell lines can have similar or greater usefulness than primary cultured lung cells for the drug development.

Drugs administered as aerosols, such as glucocorticoids and β₂-agonists, might interact with drug transporters in the lung.9 The drug transporters in the lung epithelium play an important role in the uptake and efflux of drugs, affecting drug concentrations in the lung and thus altering the efficacy, toxicity, local, and/or systemic pharmacokinetics. Indeed, Al-Jayyoussi et al.10 reported that P-gp-mediated efflux model P-gp substrates, such as rhodamine 123 and loperamide, when instilled into the airways using an isolated perfused intact rat lung. Therefore, the knowledge of the molecular basis of transporter functions will provide valuable insights into the understanding of differential drug distribution in pulmonary organs, as well as systemic circulation of inhaled drug via the lung. The comprehensive quantification of transporters expression is one of the key indicators to characterize immortalized lung cell lines for understanding and/or predicting the impact of transporters on the drug distribution via the lung epithelium. Because immortalized lung cell lines are deliberately immortalized using viral vectors, as well as derived from carcinomas and exhibit tumorigenic characteristics as chromosomal alterations, the expression levels of transporters are expected to vary between immortalized lung cell lines and primary cultured cells.11, 12 Endter et al.11 compared the mRNA expression profile of drug transporters in immortalized lung cell lines with that of primary cultured cells using quantitative reversed-transcription polymerase chain reaction analyses. They found some disparities in the gene expression between immortalized lung cell lines and primary cultured lung cells. However, there are some evidence that the mRNA expression does not always correlate with the protein expression and molecular function.13 Therefore, the comprehensive protein quantification was performed to characterize immortalized lung cell lines in the present study. We employed selected reaction monitoring (SRM)-based analysis using liquid chromatography–tandem mass spectrometry developed by Kamiie et al.14 This method enables simultaneous determination of the protein expression of dozens of functional proteins, such as transporters, receptors, and enzymes. To date, this analytical method has provided quantitative information with high reliability and robustness15 and has revealed the protein profiles of membrane proteins in the mouse liver and brain capillaries,14 human platelets,16, 17 monkey brain microvessels,18 human brain microvessels,19 human liver,13, 20 human hepatocytes,21 human lung, and human pulmonary epithelium.22 Therefore, the use of this quantitative approach would be a rational strategy for evaluating the protein expressions of drug transporters with high specificity and sensitivity.

MATERIALS AND METHODS

RESULTS

DISCUSSION

The present study is the first to comprehensively characterize the protein expression levels of drug transporters in five commonly used immortalized cell lines, such as Calu-3 and BEAS2-B derived from tracheobronchial epithelial cells, NCI-H292 and NCI-H441 derived from bronchiolar–alveolar cells, and A549 derived from alveolar type II-like cells in human. This study was exclusively focused on the evaluation of characteristics in these five commercially available cell lines because they are easily obtainable without taking patent issues into account, although there are other well-established lung cell models such as 16HBE14023 and VA10.24 In addition, to evaluate these cell lines, we listed the protein expression levels determined for primary cultured lung cells, such as the trachea, bronchi, and alveoli reported previously,22 in Table 1. Figure 1 shows a comparison of the protein expression levels of selected transporters between immortalized cell lines and primary cultured cells. There were some inconsistencies found in protein expressions compared with PCR assay as shown in Table 2. For instance, among organic cation transporters (OCTs)/OCTNs, Endter et al.11 and Courcot et al.12 showed OCTN1 and OCT3, respectively, were most abundantly expressed in BEAS2-B, but OCTN2 showed the highest protein expression in our protein assay. Drug transporters that have undergone post-transcriptional processes are partly localized in the intracellular membranes as well as in the plasma membrane because of their intracellular trafficking.25 In this study, we isolated the plasma membrane for protein expression analysis, whereas mRNA expression is evaluated using whole cell compartments, thus inconsistent data between mRNA levels and our protein expression may be explained on the basis of sample preparation processes prior to expression analysis.

Organic cation transporters/OCTNs are considered to be involved in the uptake of a wide variety of inhaled drugs, such as beclomethasone, budesonide, fluticasone, albuterol, and formoterol.26, 27 Our previous study showed that OCNT1 is most abundantly expressed in lung tissues and primary cultured pulmonary cells,22 and it is reported that immunofluorescence analysis showed that OCTN1 is primarily expressed at the apical portion of the human epithelial cells.27 Therefore, OCTN1 could be a key transporter to determine the absorption of inhaled drugs in the lung. Therefore, the provision of appropriate cell lines is important for understanding and/or predicting the in vivo behavior of OCTN1 substrates. As illustrated in Figures 1 g–1l, OCTN1 protein expressions in NCI-H292 and NCI-H441 were within twofold of those in primary cultured tracheal, bronchial, and alveolar epithelium. So far, it is reported that NCI-H441 has the ability to form polarized monolayer in culture,28 but NCI-H292 lacks polarized characteristics.29 Therefore, NCI-H441 is more likely to have similar characteristics of primary cultured cells compared with NCI-H292 for the evaluation of OCTN1 substrate distribution in the lung, regardless of regions.

OCT3 and OCTN2 were expressed at a protein level in all immortalized lung cell lines (Calu-3, BEAS2-B, NCI-H292, NCI-H441, and A549) tested in the present study. Salomon et al.30, 31 reported that OCT3 and OCTN2 were expressed in A549 at a protein level, which shows the consistency of our data with Salomon's data for A549. Also, Mukherjee et al.32 observed OCT3 and OCTN2 expressions in Calu-3, whereas OCT2 could not be detected, which was similar with our data. On the contrary, MDR1 protein expressions were not observed in tested five immortalized lung cell lines except A549 in this study, whereas MDR1 expression in the lung has been previously reported by RT-PCR,33-35 and MDR1 has been found to be functionally active in Calu-3 cells36-39 and in ex vivo rodent lungs.10 Furthermore, PEPT2 was reported to be responsible for the uptake and transport Gly-Sar in Calu-3 monolayers,40 but this study showed no detection in PEPT2 protein expression. This finding suggested a detection limit in mass-based assays compared with immunoassays and PCR. This implies that there would be several functionally active transporters but their protein expression levels were below the detection limit; therefore, further approaches for improving sensitivity are required for less abundant proteins in mass spectrometry-based assays.

MRP1 confers resistance to several chemotherapeutic agents, including vincristine, daunorubicin, and methotrexate.41, 42 It has been reported that MRP1 expression was diminished in the bronchial epithelium in COPD patients compared with healthy subjects.43 Therefore, MRP1 could play an important role in the pathophysiology of COPD as well as the efflux transport of its substrates. These suggestions can provide the justification of selecting immortalized lung cell lines in the development and evaluation of MRP1 substrate distribution in the lung. In addition, as reported previously, among the ABC transporters, MRP1 was most abundantly expressed in lung tissues and primary cultured lung cells at the protein level. Considering that MRP1 is localized at the basolateral side of the human bronchial and bronchiolar epithelial layer,44 it is suggested that MRP1 plays a role in pumping its substrates out of the epithelial cells toward the circulating blood. There was a considerable difference observed in protein expressions from region to region in the lung (Trachea; 0.488, Bronchi; 4.07, and Alveoli; 2.95 fmol/μg protein).22 This finding implies that appropriate immortalized lung cell lines should be selected for each region. In the present study, we found that MRP1 protein expression in BEAS2-B was the closest to that in primary cultured tracheal cells (Fig. 1d). Because BEAS2-B was derived from the trachea–bronchial epithelium, this cell line would be appropriate as an alternative for primary cultured tracheal cells. Moreover, Calu-3 and A549 showed protein expressions most similar to those in primary cultured bronchial (Figs. 1b and 1n) and alveolar cells (Figs. 1c and 1o), respectively. Considering the regional derivation of both cell lines, A549 would be used for alveoli and Calu-3 for bronchi as appropriate cell lines from their derivation. Taking polarized characteristics into account, Hamilton et al.36 showed the basolateral membrane localization of MRP1 in polarized Calu-3 cells, whereas localization of MRP1 is not expected in BEAS2-B and A549 because of their less polarized characteristics.29 Therefore, the localization of transporters, as well as their protein expressions, would be necessary for the selection of appropriate immortalized lung cell lines by the region.

Organic anion transporting polypeptides (OATPs) are known to transport endogenous and exogenous organic anions and amphipathic drugs, including β blockers, angiotensin receptor blockers, steroids, and others.45 Because OATP2B1 was detected in primary cultured tracheal, bronchial, and alveolar cells (0.191, 0.419, and 0.152 fmol/μg protein, respectively),22 it is expected that it could be one of the useful transporter proteins for drug targeting to the lung or to achieve pulmonary drug absorption after inhalation. However, OATP2B1 levels were below the limit of detection in all immortalized lung cell lines tested. These findings suggest that an in vitro system of immortalized lung cell lines causes underestimation in the absorption in the pulmonary tract for OATP2B1 substrates. Therefore, the evaluation of transporter protein levels in advance of experiments is necessary to prevent the misinterpretation of their functional activities.

As the previous study reported, our gradient centrifugation method showed that about fourfold enrichments in Na⁺–K⁺ ATPase was observed from the whole cell fraction to the plasma membrane in a blood–brain barrier cell model.46 Na⁺–K⁺ ATPase was also used in this study as a plasma membrane marker, and enrichments of membrane fractions can be evaluated by relative comparison of such markers protein expression. The differences of Na⁺–K⁺ ATPase were within 1.5-fold among immortalized lung cell lines; thus, it is conceivable that purity of the isolated plasma membrane from these cell lines were similar, whereas it cannot be denied that our gradient centrifugation method included unintended subcellular fractions except plasma membrane.

Quantification results in this study showed that protein expression profile differed in immortalized lung cell lines compared with their primary counterparts. For instance, A549 showed upregulation of MRP2, BCRP, OCT1, OCT3, OAT4, and PEPT1 and downregulation of MRP5, MRP6, OCT2, and OATP2B1, compared with primary alveolar cells. This disparity could be caused by aforementioned chromosomal alteration, but the culture conditions would be important to the characteristics of immortalized lung cell lines. We cultured five commercially available immortalized cells for 7 days in order to compare protein expressions of transporters in cell lines with that in primary cultures cells previously reported under the same culturing periods.22 Haghi et al.38 reported that transepithelical electrical resistance and permeability values of Calu-3 were shown to plateau from 5 days onwards; therefore, Calu-3 used in this study was expected to have polarized characteristics, but the possibility cannot be completely excluded that the determined protein levels of drug transporters could alter in the response to culturing periods. Endter et al.11 reported the mRNA expression level of the drug transporter in Calu-3 on days 8 and 15, after the seeding. As a result, mRNA expressions of MDR1, MRP5, MRP6, PEPT1, and OAT4 were higher, and those of MRP4 and OCT3 were lower at day 15 compared with those at day 8. Furthermore, the culturing of immortalized lung cell lines were carried out by liquid-covered culture (LCC) in this study under the same conditions for primary cultured cells in our previous study,22 but Grainger et al.47 observed that Calu-3 cultured using air-interfaced culture (AIC) was more morphologically similar with airway epithelium than LCC. Therefore, further investigations are necessary to elucidate the similarity and disparity of transporters protein expression among different conditions such as AIC and LCC. To date, there has been little information about culturing conditions that affect transporter protein expressions in other four immortalized lung cell lines, and the unexpected upregulation/downregulation of the expression levels of drug transporters could have occurred in the present study. This possibility emphasizes the importance of evaluating transporter protein levels in immortalized lung cell lines in the response to culturing periods used in each laboratory.

When lung cell lines are used, it is necessary to take into consideration the possibility that many chemical compounds are not delivered via a single specific transporter but with the help of multiple transporters. For instance, cisplatin was pumped out via BCRP and MRP1 as well as by MDR1.48 Considering that only MRP1 protein expression could cause the misinterpretation of the drug distribution because the contribution of BCRP in the efflux of such substrates was expected to be higher in A549, Calu-3, NCI-H292, and NCI-H441, which was caused by the higher BCRP protein expression compared with primary cultured cells. Therefore, the knowledge of transporters involved in the delivery of target compounds can be important in the selection of immortalized lung cell lines for the better understanding and/or prediction in transport studies.

The comprehensive profile of the protein expression is of great importance for prediction of the biological function of proteins. Moreover, these functions are known to be associated with not only protein expressions but also the intracellular localization and post-translational modification, such as phosphorylation. Moreover, single-nucleotide polymorphisms (SNPs) are of etiological importance. Furthermore, it has been reported that MRP1 SNPs were linked with the lung function and inflammatory markers, suggesting that the severity of COPD could be predicted by the protein expression of MRP1 polymorphism.49 Therefore, further investigations are necessary for the precise understanding and/or prediction of transporter functions.

Protein expressions of several transporters were reported in immortalized lung cells at protein level using their antibodies. For example, Western blot showed that MRP1, MDR1, and PEPT1 were expressed in Calu-336, 37, 40; moreover OCT1, OCT2, OCT3, OCTN1, and OCTN2 were expressed in A549.30 However, antibody-based quantifications, such as immunoblotting and enzyme-linked immunosorbent assay, have limitations for assessing drug transporters with high-sequence homology because the preparations of specific antibodies are often difficult. Thus, quantitative detection and analysis by proteolytic peptides as surrogates of target proteins would be necessary to complement the results of these immunoassays. Mass spectrometry-based quantification was utilized to measure peptide amounts specific to drug transporters in five commercially available immortalized lung cells in this study. As a result, expressions of MRP1 in Calu-3, and OCT1, OCT3, and OCTN1 in A549 were testified, and their absolute protein expressions were also determined. Therefore, we verified the significance of immortalized lung cells as alternative in vitro tools of transport study in the lung.

This study was aimed at absolute protein quantification of transporters, but further investigations would be necessary. First, we found disparities of transporters protein expressions between immortalized lung cells and primary cultured cells but no conclusive decision could be made concerning the difference of transporter activities among cell lines based on their transporter protein expressions. Therefore, it is important to evaluate to what extent such differences of protein expression would affect the transporter activities in each immortalized lung cell lines. Second, the cellular localization of drug transporters is important to evaluate their transport activities. The present study showed protein expressions including both basolateral and apical sides of plasma membrane because of no fractionation process. Therefore, the establishment of fractionating method for cellular domain would be highly required. Moreover, there have been several markers reported for cellular localization, for example, ACE2 is to be localized on the apical plasma membrane,50 and Collagen-IV and laminin are expressed on the basolateral surface.51 In addition, HTII-280 and CD44v6 are, respectively, expressed on apical and basolateral plasma membrane of alveolar Type II cells.52, 53 These markers would be useful for the compensation of the purity during sample preparation process.

CONCLUSIONS

The protein expression profiles of drug transporters in five immortalized lung cell lines were determined. The data showed that there were similarities in the protein expressions of drug transporters between immortalized lung cell lines and primary cultured cells. In addition, we observed several gaps in protein amounts. Because these disparities can affect functional activities of drug transporters, protein expressions should be taken into consideration in the selection of appropriate immortalized lung cell lines. These findings may contribute to the better understanding of drug transport in immortalized lung cell lines.