Collagenase type I (288 U/mg), trypsin (8500 U/mg), DNAse I (2080 KU/mg) and calcium chloride were purchased from Sigma Aldrich (St Louis, MO). Monoclonal rat anti-ki67 primary antibody was purchased from DAKO North America, Inc. (Carpenteria, CA). Superscipt II reverse transcriptase, dNTPs, Quant IT™ Oligreen R ssDNA assay kit, Rat IgG2a isotype control, popo 3 iodide dimeric cyanine nucleic acid stain (C4H58I4N6O2), and Alexa Fluor 647-goat anti-rat and -goat anti-rabbit secondary IgGs were purchased from Invitrogen (Carlsbad, CA). Phycoerythrin (PE) - anti-human monoclonal α-smooth muscle actin antibody, Mouse IgG2a PE isotype control, recombinant human PDGF-AA (221-AA) and porcine TGFβ1 (101-B1) were purchased from R & D Systems (Minneapolis, MN). BD Compbeads compensation particle sets for optimising the flow cytometry fluorescence signals were purchased from BD Biosciences. Rabbit polyclonal anti-SMAD 2/3 primary antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). RNAse H was purchased from New England Laboratories (Ipswich, MA). NucAway Spin columns and Taqman Gene Expression Assays for α-smooth muscle actin and β-2-microglobulin were purchased from Applied Biosystems.
Mice bearing the PDGF-Rα-GFP allele were originally obtained from Dr. Philippe Soriano (Fred Hutchinson Cancer Center, Seattle, WA), and details regarding their characterization and phenotype have been reported elsewhere [30, 31]. Briefly, the human histone H2B gene was subcloned into a pEGFPN1 cloning vector to generate the H2B-GFP construct encoding a nuclear form of GFP . The resultant H2B-GFP construct was inserted into targeting (knock in) vectors. By homologous recombination in the mouse embryos, these vectors replaced a fragment in the wildtype pdgf-rα allele corresponding to the signal peptide and the first two immunoglobulin domains. Thus nuclear GFP expression in the transgenic mice is under the control of the endogenous pdgf-rα promoter. Soriano's group observed that GFP expression in the PDGF-Rα-GFP mice recapitulated endogenous spatial and temporal PDGF-Rα expression . We used mice that were heterozygous for the pdgf-rα-gfp allele; thus the mice still carry one functional pdgf-rα allele. The heterozygous mice phenotypically resemble wildtype (GFP-) mice . Mice were housed in Thoren cages with access to food and water ad libitum in a thermally regulated environment and a 12-hour light/dark cycle. Protocols for animal use were approved by the Iowa City Veterans Affairs Medical Center animal use committee.
Isolation of mouse lung fibroblasts
Prior to the fibroblast isolation, mice were screened for the PDGF-Rα-GFP allele using either PCR or fluorescence microscopy. For PCR, 1 μl of tail DNA was mixed with dNTPs, MgCl2, 10× PCR buffer, and the following GFP primers: EGFP forward - 5' - GCC ATG CCC GAA GGC TAC GTC - 3' EGFP reverse - 5' - GGG TGC TCA GGT AGT GGT TGT C - 3' The amplification reaction was performed in a thermocycler unit (Techne Limited, Cambridge UK). To screen the animals using fluorescence microscopy, a small piece of the lung was compressed on a glass slide using a coverslip, and the presence or absence of GFP fluorescence was observed using an epifluorescence microscope.
Primary mouse lung fibroblasts were isolated from 4 and 12-day old mice using a previously reported method, with minor modifications . After the animals had been anesthetized with 25 μl of a mixture of 90 mg/ml ketamine and 10 mg/ml xylazine, their lungs were removed, pooled together and digested for a total period of 1 h, with removal of the dispersed cells every 10 minutes. The enzyme solution consisted of Hanks buffered salt solution (HBSS, without Ca or Mg) with 300 μg/ml collagenase type I, 250 U trypsin (8500 U/mg solid), 75 μg/ml DNAse I and 5 mM calcium chloride. After each digestion interval, the solution containing dispersed cells was passed through 100 μm mesh into Ham's F-12 k media containing 10% FBS, and the undigested lung tissue placed in fresh enzyme solution. Once digestion was complete (after 3-4 digestions), erythrocytes were separated from the fibroblasts by sedimentation through a 40% Percoll solution. To select for fibroblasts, cells were adhered to 100 mm tissue culture dishes for 1 h at 37°C. Adherent cells were harvested by trypsinization followed by gentle scraping and centrifugation.
Preparation of freshly isolated mouse lung fibroblasts for Flow Cytometry analysis of αSMA and Ki67
Freshly isolated mouse lung fibroblasts were proportioned into aliquots of 5 × 105 cells and washed several times by mixing with 1× phosphate-buffered saline (PBS - 0.145 M NaCl, 0.0015 M KH2PO4, 0.0027 M KCl, 0.0086 M Na2HPO4, pH 7.4) followed by centrifugation. Following rinsing, the fibroblasts were fixed in 2% paraformaldehyde/PBS (v/v) for 20 min with shaking at 4°C. After additional rinses in PBS, the fibroblasts were permeabilized in 0.1% (w/v) saponin in PBS at 4°C for 10 min. After the permeabilization step, non-specific binding was reduced by incubating the fibroblasts in 5% non-fat dry milk/0.1% saponin/PBS for 2 h at 4°C. Following the blocking step, monoclonal rat anti-ki67 primary antibody (1:300 dilution, clone TEC-3) or rat IgG2a was added to the fibroblasts, and the cells incubated overnight at 4°C. The next day, the cells were rinsed in 0.1% saponin/PBS. In a sequential manner, Alexa Fluor 647-goat-anti rat secondary IgG (1 μg/ml) and phycoerythrin (PE) - anti-α-smooth muscle actin primary antibody (0.0625 μg/ml) or PE-mouse IgG2a were added and the cells incubated for 1 h at 4°C. Finally, the cells were rinsed in 0.1% saponin/PBS, and resuspended in PBS for flow cytometry.
Collection and analysis of Flow Cytometry data
Flow cytometry was performed at the University of Iowa Flow Cytometry facility using the LSR II instrument (BD Bioscience). An unstained sample was used to gate the forward and side scatter pattern of the cell population of interest. Optimization of signal as well as compensation for overlap between the PE and GFP or the Alexa Fluor 647 signals was achieved using BD Comp beads - polystyrene microparticles coated with mouse IgGκ or FBS (negative control, BD Bioscience) - stained in the same manner as the fibroblasts. The resultant data were analyzed using Cell Quest Pro (BD Biosciences), Microsoft Excel and Sigma Stat software (Jandal Scientific, San Rafael, CA). Each "n" represents a cohort of 2 to 5 mice, of the same age, whose lungs had been pooled and digested together. A minimum of 9000 gated events were analyzed for each "n." Background compensation was determined from the IgG controls prior to calculating the proportions of the different fibroblast populations.
Evaluating the purity of the fibroblast isolate
The fibroblast isolate was evaluated for epithelial, endothelial and macrophage contamination using flow cytometry. Epithelial cells were detected after fixation using monoclonal mouse anti-cytokeratin 18 (Sigma), followed by Alexa Fluor 647-conjugated goat anti-mouse secondary IgG. Anti-cytokeratin independent staining was assessed using only the secondary antibody. Endothelial cells were stained prior to fixation using PE-conjugated monoclonal rat anti-mouse CD31 or -mouse IgG1 isotype control (BD Biosciences). Macrophages were detected after fixation and permeabilization using Alexa Fluor 647-conjugated-rat anti-mouse mannose receptor c type 1 (CD206) or -IgG2a (Serotec). Proportions of cells staining positive for either cytokeratin 18, CD31 or CD206 were determined after correcting for primary antibody-independent fluorescence.
Preparation of Lung tissue for Confocal Microscopy studies
Confocal microscopy studies utilized either 100 or 7 μm lung tissue sections obtained from mice at postnatal days 4 and 12. Preparation of the 100 μm sections has been described elsewhere . For the preparation of the 7 μm sections, mice were anesthetized and their chests opened to expose their lungs. The lungs were perfused transcardially with 1× PBS and then inflated to total lung capacity with a uniform quantity (for a particular age) of 75% Optimal Cutting Temperature (OCT) Medium, dissected out of the thoracic cavity, and frozen in a bath of liquid nitrogen and isomethylbutane. The frozen lung tissue was cut into 7 μm sections using a cryostat, mounted on glass slides and stored at -20°C prior to fixation.
Immunofluorescence and Confocal Microscopy Imaging
The 100 μm sections were stained with Cy3-α-smooth muscle actin monoclonal antibody (R & D systems, Minneapolis, MN) and a lipid counterstain, 1,1'-dioctadecyl-3,3,3/,3/-tetramethylindodicarbo-cyanine perchlorate (DiD, Invitrogen, Carlsbad, CA). The staining and imaging processes have already been described in detail .
The 7 μm lung tissue sections were equilibrated at 4°C in 1× PBS for 5 min, and then fixed in 2% paraformaldehyde for 1 h at 4°C. Fixed sections were then rinsed extensively with 1× PBS. All subsequent antibody incubations were performed in a humidified chamber, and rinses in Coplin jars. Following the rinses, the lung tissue was permeabilized in 0.1% Triton X-100/3 mg/ml bovine serum albumin (BSA) (permeabilization buffer) for 30 min at RT. To minimize non-specific antibody binding, the tissue was then placed in blocking solution - permeabilization buffer/5% normal goat serum - overnight at 4°C. After the blocking step, rabbit polyclonal anti-SMAD 2/3 primary antibody (0.5 μg/ml) was added to the blocking solution, and the incubation performed overnight at 4°C. The lung tissue was subsequently rinsed with permeabilization buffer, and then incubated with Alexa Fluor 647-goat anti-rabbit secondary IgG (4 μg/ml) for 1.5 h at RT. After more rinsing, the tissue was incubated for 40 min at 37°C with 1 × 10-7 M popo 3 iodide.
Images of the stained sections were acquired from randomly selected fields using Zeiss LSM510 Laser Scanning Confocal Microscopes (LSCM). The following excitation/emission filters were used: a 488/505-530 nm band pass filter to detect GFP fluorescence, a 533/560 nm long pass filter to detect PoPo 3 iodide fluorescence, a 543/560-615 nm band pass filter to detect Cy3-α-smooth muscle actin antibody, and a 650 nm long pass filter for detecting either DiD or Alexa Fluor 647-goat anti-rabbit secondary IgG. Z-stack images were obtained at 1024 × 1024 pixel density.
Analysis of PDGF-Rα and αSMA in images acquired by laser scanning confocal microscopy
For analysis of images acquired from the 100 μm sections, alveoli were identified and annotated in the 3 μm interval z-stacks using StereoInvestigator (MBF Bioscience, Williston, VT). Each alveolus selected for analysis was traced through the z-stack, and was defined as having an entry ring and a base . Inspection of the tissues revealed that there were different levels of intensity of GFP-fluorescence that were not related to the distance from the laser beam or the detector. Therefore we devised uniform segmentation criteria using IP lab software (BD Bioscience, Rockville, MD), which discriminated between populations of nuclei with high and low intensities of GFP fluorescence. Because the intensity of the GFP was intrinsic to the tissue and not dependent on antibody staining, it is representative of the level of PDGF-Rα gene expression. To examine PDGF-Rα-GFP distribution at the alveolar entry ring and base, "low" and "high" PDGF-Rα-GFP cells were manually counted at each level.
The intensity of α-SMA staining was antibody-dependent, thus we always imaged tissues obtained from P4 and P12 animals that were stained concurrently. To avoid variability related to distances from the laser beam, alveolar entry rings located 1/4 to 1/2 of the way into the z-stack were used for analysis. To quantify the pixel density of αSMA associated with PDGF-Rα expression at the alveolar entry ring, a circle measuring 9 μm in diameter was drawn around each PDGF-Rα-GFP nucleus. The pixel density of αSMA was ascertained relative to this uniform region of interest. The entry ring was defined by staining of αSMA, which occupied at least 75% of the circumferential transition from an alveolar duct into an alveolus. The areal pixel densities of αSMA were expressed in binned segments.
Analysis of nuclear SMAD 2/3 in the alveolar region
Localization of SMAD 2/3 to nuclei in cells that either expressed PDGF-Rα (nuclei with GFP) or did not express PDGF-Rα (nuclei without GFP) was ascertained as follows: The fluorescent DNA dye, PoPo3, was assigned a blue pseudocolor. In the blue-only image, we optimized the threshold intensity to accurately select pixels representing all blue nuclei (nuclei stained with PoPo3), and quantified the pixel area covered by all the nuclei (all nuclei comprised both those with and without GFP). Next, in the green-only image, we selected only for the nuclei, which also contained GFP above a fixed threshold intensity, and quantified the total pixel area covered by these nuclei. In order to ascertain the nuclear area occupied by SMAD 2/3 (assigned the color red) but not GFP, uniform segmentation criteria were applied to the red-green-blue merged image, enabling selection of pixel segments which contained blue and red, but not green (magenta). In order to ascertain the nuclear area occupied by SMAD 2/3 and GFP, we applied uniform segmentation criteria to the RGB merged image, enabling selection of pixels which contained green and red (yellow). Finally, we ascertained the proportional pixel areas of SMAD 2/3 in nuclei that did or did not contain GFP using the following formulas: (a) Fraction of SMAD 2/3 within GFP- nuclei = area of magenta pixels ÷ (area of all nuclei minus area of GFP+ nuclei) (b) Fraction of SMAD 2/3 within GFP+ nuclei = area of yellow pixels ÷ area of GFP+ nuclei.
We applied this procedure to 3 different mice at P4 and at P12, and examined all of the alveoli in at least 10 different randomly selected fields per animal. The results were expressed as mean ± SD, and ANOVA on ranks followed by Dunn's posthoc test was used to determine statistical significance.
The neonatal mouse lung fibroblast cell line, MLg 2908 (CCL-206) was obtained from the American Type Culture Collection (Manassas, VA). The cells were maintained at 5% CO2 and 37°C in T-75 flasks with MEM media containing 10% FBS (v/v) and sub-cultured every 4 days. For experiments, the cells were seeded into 100 mm tissue culture dishes. Cells were used between passages 3 and 10.
Assessing the effect of exogenous PDGF-AA and/or TGFβ treatment on αSMA mRNA expression in Mlg Cells
On day 1, the Mlg 2908 cells were plated in 10% FBS/MEM media onto 100 mm tissue culture treated dishes at a density of either 2 × 105(sub-confluent) or 2 × 106 (confluent) cells. On day 2, the cell monolayer was rinsed with Opti MEM I media, and 0.2% BSA/Opti MEM 1 media was added. The concentrations of PDGF-AA and/or TGFβ indicated in the figure legends were added 24 or 48 h after serum starvation. Total RNA was isolated by the guanidinium chloride method  24 or 48 h post treatment.
Reverse Transcription of total RNA
Oligo (dT) primers and dNTPs were added to the total RNA, and the RNA was denatured at 70°C for 15 min, and then cooled on ice for 5 min. Subsequently, dithiothreitol (DTT) and 5× first strand buffer were added to the samples, and they were annealed at 42°C for 5 min. To initiate the reverse transcriptase reaction, Superscript II (Invitrogen) was then added, and the samples heated to 42°C for 50 min. The reaction was subsequently quenched by heating the samples to 70°C for 15 min. To remove any residual RNA, the samples were treated with RNAse H at 37°C for 30 min. The cDNA product was diluted with TE buffer, and placed on NucAway Spin Columns to remove salts and unincorporated nucleotides. Concentrations of cDNA were determined using the Quant IT™ Oligreen R ssDNA Assay Kit.
Real-time quantitative PCR (RT-qPCR)
Quantification of αSMA mRNA was achieved by mixing 5 ng of cDNA with 1.5 μl of Taqman Gene Expression Assay for αSMA (probe number Mm01546133_m1) or β-2-microglobulin (probe number Mm00437762_m1) (housekeeping gene), and 15 μl of Taqman Universal PCR MasterMix (catalogue number 4324018), to a final volume of 20 μl. The probes are designed to amplify exon-intron junctions to minimize amplification of any remaining genomic DNA. The Real-time PCR was performed in triplicate (3 amplification reactions for one reverse-transcribed RNA) using an Applied Biosystems Model 7500 Sequence Detection System. At least four independent experiments were performed. Values for the αSMA expression were normalized to β-2-microglobulin, and calculated in Excel using the 2(-ΔΔCT) relative quantification method .
Data were expressed as means ± SD. Unless otherwise indicated, statistical significance was determined by performing a 2-way Analysis of Variance followed by a Student-Newman-Keuls post-hoc test. Values of p < 0.05 were considered significant.