Culture support was prepared according to Bals and collaborators  except that 50 μl of 0.01% collagen solution (Sigma–Aldrich, Saint-Quentin, France) were used to coat a 6.54 mm ThinCert™ - TC Inserts (Greiner bio-one, Courtaboeuf, France).
Newborn pig trachea cell culture
The NPTr cells  (between 30 and 50 passages) were cultured in Dulbecco’s modified Eagle medium (DMEM) (Invitrogen, Cergy Pontoise, France) supplemented with 10% fetal calf serum (FCS) (Sigma-Aldrich), 20 IU/ml of penicillin and 20 mg/ml of streptomycin (Invitrogen). Cells were plated onto 24-well plastic plates (Greiner bio-one, Courtaboeuf, France) and incubated at 37°C in 5% CO2 in a humidified atmosphere. Sub-passages were made when cells reached 100% confluence. After trypsinization, collected cells were seeded onto coated ThinCert™ - TC Inserts (Greiner bio-one). A total of 0.8 ml of fresh medium was added to the lower reservoir and 0.25 ml of a 105 cells/ml suspension was added to the upper reservoir. As a control, cells were also plated onto conventional 24-well plastic plates for twenty-two days.
Culture after seeding cells on the insert
After seven days of culture at 37°C in 5% CO2 in a humidified atmosphere, when cells were completely confluent, medium was removed from the upper reservoir. The cells were gently washed with Ca/Mg-free phosphate buffered saline (PBS) every two days at the apical side. Half of the basolateral culture medium was replaced every other day. The culture was kept in ALI conditions for twenty-two days.
Parallel experiments were carried out in order to evaluate the cells’ capacity to fully differentiate using other culture media. NPTr cells were cultured with different types of media (Table 1). Again, after seven days of culture, when cells were completely confluent, the apical medium was removed. The culture medium in the lower reservoir was replaced by serum-free 50% Dulbecco's Modified Eagle's Medium (DMEM)–50% DMEM/Ham's F-12 (HAMF12) medium (Sigma-Aldrich) supplemented with 10−7 M dexamethasone (Sigma-Aldrich), 20 IU/ml of penicillin and 20 mg/ml of streptomycin (Invitrogen) or DMEM/HAMF12 supplemented with insulin (5 mg/ml), transferrin (5 mg/ml), selenium (5 ng/ml), epidermal growth factor (5 mg/ml) (all supplied by Sigma-Aldrich), and 20 IU/ml of penicillin and 20 mg/ml of streptomycin (Table 1). Every two days, the basal medium was changed and the apical surface washed with Ca/Mg-free PBS. The cultures were kept for twenty-two days in ALI conditions to induce cell differentiation. Finally, another experiment was performed to evaluate the impact of retinoic acid on NPTr cell differentiation. The procedure was identical to the one described above except that the culture medium in the lower reservoir was replaced either by serum free airway epithelial cell medium (AECM) (Promocell, Heidelberg, Germany) supplemented as recommended by the supplier with bovine pituitary extract (0.004 ml/ml), epidermal growth factor (10 ng/ml), insulin (5 μg/ml), hydrocortisone (0.5 μg/ml), epinephrine (0.5 μg/ml), triiodo-L-thyronine (6.7 ng/ml), transferrin holo (human) (10 μg/ml), and retinoic acid (0.1 ng/ml) (all supplied by Promocell) or DMEM/HAMF12 supplemented with 5% FCS, 20 IU/ml of penicillin and 20 mg/ml of streptomycin, dexamethasone 10−7 M, and retinoic acid (0.1 ng/ml) (Table 1).
Transepithelial electrical resistance measurements
Transepithelial electrical resistance (TEER) measurement provides an indirect measure of the formation of tight junctions . Among the cell-cell junctions (tight junctions, adherens junctions, gap junctions, and desmosomes), tight junctions are the most important for maintaining epithelial integrity. TEER is also used as a marker of disruption of epithelial cells. TEER was measured using a MILLICELL® ERS volt-ohm meter (Millipore, Molsheim, France). On day 0 and every fourth day up to day 22 in ALI conditions, 150 μl of medium was added apically into the insert and the measurement taken. Apical medium was then aspirated to restore ALI conditions. Prior to testing the culture’s TEER an empty culture insert was used as a blank and subtracted from each subsequent sample reading. Data are presented as resistance values (Ω cm2).
Immunofluorescence staining was performed directly on cells cultured onto the ThinCert™ - TC Inserts (Greiner bio-one), on ThinCert™ - TC Insert frozen sections and on lung tissue frozen sections as described below.
Cell cultures were washed three times with PBS prior to fixation for 15 min with 3% paraformaldehyde (Sigma-Aldrich). After one wash with PBS containing 0.1 M glycin (Fisher Scientific, Illkirch, France) cells were treated for permeabilization with 0.2% Triton X-100 (Sigma-Aldrich) over 15 min. Finally, inserts were washed three times with Ca/Mg-free PBS before staining.
Insert frozen sections
Insert membranes were removed from the ThinCert™ membrane supports, then immersed in Tissue-Tek® O.C.T. Compound (Sakura Finetek, Flemingweg, The Netherlands), snap-frozen, and stored at −80°C. Serial transverse sections (7 μm thick) of the membrane were cut at −20°C using a LEICA CM3050 microtome (Leica, Nanterre, France),collected onto treated glass slides (SuperFrost Plus, Menzel-Glaser, Braunschweig, Germany), air-dried, fixed in acetone (Sigma-Aldrich) for 10 min at 4°C, and then stored at −80°C until use. Insert frozen sections were washed three times with Ca/Mg-free PBS before staining.
Lung tissue frozen sections
Small pieces of lung tissue (6 mm × 6 mm) were collected from a two-month-old healthy pig provided by INRA experimental unit (Nouzilly, France). The pig was cared for in accordance with the guidelines of the Institutional Animal Care and Use committee at INRA. The pieces were then immersed in Tissue-Tek® O.C.T. Compound (Sakura Finetek), snap-frozen, and stored at −80°C. Serial transverse sections (7 μm thick) of the membrane were cut at −20°C using a LEICA CM3050 microtome and treated as described above for the insert frozen sections.
In the case of filter cultures, the reagents were added to the apical filter chamber. Each incubation period with the selected antibodies was performed at room temperature for 20 min in the dark. The goblet cells were stained indirectly by using monoclonal anti-human gastric mucin 5 AC clone 45 M1 antibodies (dilution 1/200) (Sigma-Aldrich) followed by AF488-labeled secondary antibodies (dilution 1/600) (Invitrogen) (see Table 2). Tight junctions were stained with purified monoclonal mouse anti-human ZO-1 antibodies (dilution 1/100) (BD Biosciences, Rungis, France). For cilium staining, cells were treated with Cy3-labeled monoclonal antibodies recognizing β-tubulin (dilution 1/500) (clone TUB 2.1, Sigma-Aldrich). β-tubulin is often expressed as a cytoskeletal protein, however, its apical expression is a marker of ciliated cells . 4’, 6’-diamidino-2-phenylindole (DAPI) (Life Technologies Inc., Carlsbad, CA, USA) at 0.5 μg/ml was used as counterstaining before the cells were washed three times with Ca/Mg-free PBS. Controls were incubated with primary isotype control antibodies followed by secondary antibodies (Table 2). All samples were observed with a Nikon Eclipse 80i microscope connected to Nikon intensilight C-HGF and the imaging software NIS Elements D (Nikon Instruments Europe BV, Amsterdam, The Netherlands).
Transmission electron microscopy
The filter membranes with NPTr cells were fixed by incubation for 24 h in 4% paraformaldehyde and 1% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) (Sigma-Aldrich) and post-fixed by incubation for 1 h with 2% osmium tetroxide (Electron Microscopy Science, Hatfield, PA, USA). They were then dehydrated in a graded series of ethanol solutions, cleared in propylene oxide, and embedded in Epon resin (Sigma-Aldrich) which was allowed to polymerize for 48 h at 60°C. Ultra-thin sections were cut and placed on 300 mesh copper grids and then stained with 5% uranyl acetate and 5% lead citrate (Sigma-Aldrich). The grids were then observed with Jeol 1230 TEM (Tokyo, Japan) connected to a Gatan slow scan digital camera and digital micrograph software (Gatan, Pleasanton, CA, US) for image acquisition.
Scanning electron microscopy
The filter membranes with NPTr cells were washed in PBS, fixed in 4% paraformaldehyde (Sigma-Aldrich) and 1% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) (Sigma-Aldrich) and post-fixed by incubation for 1 h with 2% osmium tetroxide. Then, specimens were dehydrated in a graded series of acetone and dried in hexa-methyl-disilazan solution (HMDS) (Sigma-Aldrich). Dried specimens were coated with a thin layer of platinum with ion beam coater PECS (Gatan France, Evry, France) and observed with Zeiss Ultra + Field Emission Gun Scanning electron microscope (FEGSEM) (Carl Zeiss S.A.S, Le Pecq, France).
Real time polymerase chain reaction assays and validation of reference genes
NPTr cells were lysed and total RNA was isolated using RNeasy Mini kit (Quiagen, Courtaboeuf, France). Quantitative real-time PCR (qPCR) was performed using cDNA synthesized as previously described . Primers were designed using Clone Manager 9 (Scientific & Educational Software, Cary, NC, USA) and were purchased from Eurogentec (Liège, Belgium) (Table 3). Diluted cDNA (10X) was combined with primer/probe sets and IQ SYBR Green Supermix (Bio-Rad, Hercules, CA, USA) according to the manufacturer’s recommendations. The qPCR conditions were 98°C for 30 seconds, followed by 37 cycles with denaturation at 95°C for 15 seconds and annealing/elongation for 30 seconds (annealing temperature, Table 3). Real time assays were run on a Bio-Rad Chromo 4 (Bio-Rad, Hercules, CA, USA). The specificity of the qPCR reactions was assessed by analyzing the melting curves of the products and size verification of the amplicons. To minimize sample variations, we used an identical amount of cells and high quality RNA. The quality of RNA was assessed by capillary electrophoresis (Agilent 2100 Bioanalyzer, Agilent Technologies, Massy, France) and RNA integrity numbers (RIN) were calculated. RIN were always ≥8.7 demonstrating the high quality of the RNA. Samples were normalized internally using simultaneously the average cycle quantification (Cq) of the three most suitable reference genes in each sample to avoid any artifact of variation in the target gene. These three most suitable reference genes were selected among eight commonly used reference genes which were investigated in each tissue using qPCR with SYBR green. The genes included beta-actin (ACTB), beta-2-microglobulin (B2MI), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), hydroxymethylbilane synthase (HMBS), hypoxanthine phosphoribosyltransferase-1 (HPRT-1), ribosomal protein L-19 (RPL-19), succinate dehydrogenase complex subunit A (SDHA) and TATA box binding protein 1 (TPB-1). The stability of these reference genes in all the selected tissues was investigated using the geNorm application . The threshold for eliminating a gene was M ≥0.5 as recommended . The correlation coefficients of the standard curves were >0.995 and the concentration of the test samples was calculated from the standard curves, according to the formula y = −M*Cq + B, where M is the slope of the curve, Cq the first positive second derivative maximum of amplification curve calculated using PCR Miner  and B the y-axis intercept. All qPCRs displayed efficiency between 90% and 110%. Expression data are expressed as relative values after Genex macro analysis (Bio-Rad, Hercules, CA, USA) .
Data for the comparison of differences in relative mRNA expression between NPTr cells (W – 0 and W – 3) were expressed as relative values. Because data were independent and non-normally distributed, the Mann–Whitney test was selected for statistical analysis (GraphPad Prism software version 3.00, GraphPad Software Inc., San Diego, CA, USA). Differences between groups were considered significant when p < 0.05.