Mouse maintenance and mouse lines

Genetically modified mice were maintained under Specific Pathogen Free (SPF) conditions at the ETH Zurich Phenomics Center (EPIC). They were housed according to Swiss guidelines and received food and water ad libitum. All experiments with mice had been approved by the local veterinary authorities (Cantonal Veterinary Office Zurich). Pkdccfl/fl [9] females were bred with Pkdccfl/+/Col1a2-Cre± [10] males to obtain progeny containing Pkdccfl/fl/Col1a2-Cre−/−, Pkdccfl/+/Col1a2-Cre−/−, Pkdccfl/+/Col1a2-Cre± and Pkdccfl/fl/Col1a2-Cre± genotypes. The knockout mice were compared to mice carrying the Pkdcc floxed alleles, but lacking Cre recombinase. PDGFRa-H2B-eGFP mice were obtained from The Jackson Laboratory, Bar Harbor, ME. All mice were in C57BL/6 genetic background and of mixed sex. The exact sample size used in each experiment is indicated in the figure legends.

Genotyping

Mouse genotyping was performed by polymerase chain reaction (PCR) analysis of genomic DNA, which had been isolated from tail biopsies obtained after sacrifice (embryos and neonates) or ear biopsies (adult mice) using the KAPA2G FAST Genotyping Mix (#KK5621, Roche, Rotkreuz, Switzerland). The following primers were used:

Primer

Sequence forward primer

Sequence reverse primer

mPkdcc

CAC ACG CTC AAT CAT ACC ACA CC

GGT CAT TAG GTC ACA GGG TAG GG

mCol1a2-Cre

TTA GCA CCA CGG CAG CAG GAG GTT

CAG GCC AGA TCT CCT GTG CAG CAT

mPDGFRa-eGFP

CCC TTG TGG TCA TGC CAA AC

GCT TTT GCC TCC ATT ACA CTG G

ACG AAG TTA TTA GGT CCC TCG AC

Protein extraction from lung tissue for proteomics analysis

Lung tissue samples from mice at E18.5 were soaked in 4 M guanidine chloride, 250 mM HEPES pH 7.8, supplemented with 1 × PMSF. Samples were processed using pressure cycling technology (PCT) as previously described [11], sonicated, (3 × 10 cycles; 30 s ON, 30 s OFF) at 4 °C, and centrifuged for 10 min at 10,000 × g and 4 °C. The supernatants containing the extracted proteins were transferred to a new Eppendorf tube, and the buffer was adjusted to 2.5 M guanidinium chloride, 250 mM HEPES pH 7.8. Samples were stored at – 80 °C before proceeding according to the quantitative proteomics workflow.

Quantitative proteomics

For 8plex-TMT quantitative proteomics analysis, we applied protein-level labeling, following a previously described workflow [12]. Samples were analyzed after trypsin digest. The protein samples (50 μg per condition) were first denatured by incubation for 15 min at 65o C. Cysteine residues were reduced by adding 3.5 mM Tris (2-carboxyethyl) phosphine (TCEP) and incubation for 45 min at 65 oC, and then alkylated by adding 5 mM of chloroacetamide (CAA) for 30 min at 65 oC. The proteins in each sample were labeled at a 1:4 protein: TMT (w/w) ratio with TMT reagents (TMT10plex labeling Kit; Thermo Fisher Scientific, Waltham, MA) for 1.5 h at room temperature (RT), after which the labeling reactions were quenched with 100 mM NH4HCO3 for 30 min. The labeled samples were then pooled and precipitated by adding seven sample volumes of ice-cold acetone and one sample volume of ice-cold methanol and incubated for 2 h at – 80 oC. The samples were centrifuged at 4700 × g at 4 oC for 30 min, washed with 5 ml ice-cold methanol, and centrifuged again. The pellet was air-dried, resuspended in 100 mM NaOH, and adjusted with 1 M HEPES, pH 7.8 to 1 mg/ml protein in 100 mM HEPES, pH 7.8. The protein samples were digested with trypsin (Trypsin Gold, V5280, Promega, Madison, WI; 1:100 enzyme: protein ration (w/w)) for 16 h at 37 oC.

Desalting of unfractionated peptides

Before peptide fractionation, peptide samples were desalted with Sep-Pak C18 columns (Waters Corporation, Milford, MA). Columns were activated with 0.9 ml of 100% methanol, cleaned with 0.9 ml of 80% acetonitrile (ACN), 0.1% formic acid (FA), and equilibrated with 3 × 0.9 ml of 3% ACN, 1% trifluoroacetic acid (TFA). Next, the samples were acidified with 1% TFA and loaded on the column. The columns were washed with 3 × 0.9 ml 0.1% FA, after which the samples were eluted with 3 × 200 μl 80% ACN, 0.1% FA. The eluted peptides were dried under vacuum and stored at – 20 oC.

Peptide fractionation and LC–MS/MS

Peptide mixtures were fractionated using a Dionex UltiMate 3000 UHPLC (Thermo Fisher Scientific) coupled to an Acclaim™ PA2 nano HPLC column (3 μm, 150 × 0.3 mm, Thermo Fisher Scientific). Samples were resuspended in 5 mM NH4HCO3, pH 10, and fractionated with the following gradient: 2 min 5% B; 50 min 35% B; 58 min 70% B; 65 min 70% B; 70 min 5% B with eluent A (5 mM NH4HCO3) and eluent B (100% ACN) at a flow rate of 5 μl/min. Forty-five fractions were collected using a Dionex AFC-3000 fraction collector in a 96 deep-well plate and subsequently pooled into 22 samples. The peptide fractions were analyzed on a Q Exactive HF-X mass spectrometer coupled to an LC Evosep One system. They were loaded onto Evotips (Evosep, Odense, Denmark), according to the manufacturer’s instructions. Briefly, the Evotips were washed with Solvent B (80% ACN, 0.1% FA) and centrifuged for 1 min at 700 × g. Next, the tips were soaked for ~ 1 min in 1-propanol, equilibrated with Solvent A (0.1% FA), and centrifuged for 1 min at 700 × g. The samples were loaded and centrifuged for 1 min at 700 × g. Subsequently, they were washed with Solvent A and centrifuged. Finally, 100 μl Solvent A were added to the Evotips to prevent them from drying before injecting into the mass spectrometer. After the samples were loaded, they were analyzed with a pre-programmed 44 min gradient per injection using an Acclaim™ PepMap™ RSLC C18 column (2 μm, 75 μm × 150 mm, Thermo Fisher Scientific) at RT. Data was recorded in data-dependent acquisition (DDA) mode. A precursor MS1 scan (m/z 350–2000) was acquired at a resolution of 120,000 with an AGC target 3e6 and a maximum fill time of 50 ms. The 20 most abundant precursor ions were selected from each MS scan for a subsequent higher-energy collision-induced dissociation (HCD) fragmentation with a normalized collision energy (NCE) of 30%. Fragmentation was performed at resolution 45,000 with an AGC target of 1e5 and an injection time of 96 ms, using a precursor isolation window of 0.7 m/z and a dynamic exclusion of 20 s after single isolation and fragmentation of a given precursor.

Data analysis and normalization

Raw files were searched by Sequest HT from Proteome Discoverer 3.0 (Thermo Fisher Scientific) against the mouse UniProt database (sp_canonical TaxID = 10090, v2022-01-30; 17067 sequences). The following parameters were used for database searches: semi-ArgC for enzyme specificity, allowing one missed cleavage; carbamidomethyl (C) and TMT6plex (K) as fixed modifications, and acetyl (N-term), TMT6plex (N-term), pyroQ (N-term), deamidation (NQ), oxidation (MP), and phosphorylation (Y) were set as variable modifications: precursor mass error tolerance of 10 ppm and fragment mass error at 0.02 Da. Percolator was used for decoy control and FDR estimation (0.01 high confidence peptides, 0.05 medium confidence). TMT6plex-modified N-terminal modified and tyrosine phosphorylated peptides were excluded from protein quantification, data were normalized to ‘Total Peptide Amount’, scaling performed with mode ‘On All Average’, and differential abundance determined with ‘Protein Abundance Based’ ‘Protein Ratio Calculation’ and applying Proteome Discoverer’s ‘ANOVA (Individual Proteins)’ setting.

Hydroxyproline analysis

Collagen content of the lung was analyzed as described previously [12, 13]. Harvested lungs were weighed, snap frozen and then homogenized in phosphate buffered saline (PBS). 1 ml of the homogenate was treated with 10% trichloric acid (TCA), then hydrolyzed with 6 M hydrochloric acid for 18 h at 110 oC, and the pH was adjusted to 7. The oxidation process was started by 20 min incubation with 1 ml of chloramine T reagent at RT and stopped by addition of 1 ml of 3.15 M perchloric acid. Samples were then incubated in Ehrlich reagent (p-dimethylaminobenzaldehyde added to methyle cellusolve) for 20 min at 55–65 oC. Finally, the absorbance of each sample was measured at 557 nm, and a standard curve was calculated using known concentrations of reagent grade hydroxyproline (Sigma) as described before.

Collagen cross-link and protein analysis

Analysis of collagen and of collagen cross-links was performed as reported previously [15]. Briefly, both lungs from two animals were pooled and treated with sodium borohydride (Sigma, 25 mg NaBH4/ml in 0.05 M NaH2PO4/0.15 M NaCl pH 7.4, 1 h on ice, 1.5 h at RT) to stabilize reducible acid-labile cross-links, digested for 12 h at 37 oC with high purity bacterial collagenase (C0773; Sigma, 50 U/ml) and hydrolyzed in 6 N HCl at 110 °C for 24 h. The hydrolysates were precleared by solid phase extraction and analyzed on an amino acid analyzer (Biochrome30, Biochrome, Cambridge, UK). Quantification was based on ninhydrin-generated leucine equivalence factors (DHLNL, HLNL: 1.8). The nomenclature used in the manuscript refers to the reduced variants of cross-links (DHLNL, HLNL). For protein analysis, specimens were digested with bacterial collagenase. After centrifugation, the soluble fraction containing collagen was subjected to hydrolysis and amino acid analysis. Collagen content was calculated based on a content of 14 mg hydroxyproline in 100 mg collagen. The residual fraction was extracted with hot alkali (0.1 N NaOH, 95 °C, 45 min). After centrifugation, the supernatant containing non-collagen/non-elastin proteins and the insoluble residue containing elastin were subjected to hydrolysis and amino acid analysis.

Tissue and cell processing

Tissue samples were fixed overnight at 4 °C in acetic ethanol (25% acetic acid glacial, 75% ethanol) or 4% paraformaldehyde (PFA) (#P6148, Sigma), embedded in paraffin, and sectioned (3.5 µm thickness). Alternatively, fresh tissue was immediately frozen in tissue freezing medium® (#14020108926, Leica Biosystems, Wetzlar, Germany) and sectioned (5 µm thickness).

Immunofluorescence staining

Sections from lung tissue, which had been fixed with PFA or acetic ethanol, were dewaxed using xylene and rehydrated using an ethanol gradient. PFA sections were subjected to an antigen retrieval step performed by incubation in citrate buffer (10 mmol/l citric acid pH 6) at 95 °C for 1 h, followed by three washes with PBS containing Tween (PBS-T; 137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 2 mM KH2HPO4; 0.1% Tween). All samples were blocked with 10% bovine serum albumin (BSA) (#P06-1391100, PAN Biotech, Aidenbach, Germany) in PBS for 30 min at RT, followed by incubation with the primary antibody overnight at 4 °C and for 2 h with the secondary antibody at RT. Hoechst 33342 (1:1000 diluted) was used to counterstain nuclei. The samples were mounted with Mowiol-DABCO.

The following antibodies were used for immunostaining: rabbit anti-VLK 404 (Whitman laboratory, Harvard University, Boston, MA), goat anti-cytokeratin 19 (Hybridoma Product TROMA-III; Developmental Studies Hybridoma Bank (DSHB), Iowa City, IA), rabbit anti-SPC (#sc-13979, Santa Cruz, Santa Cruz, CA), rat anti-Ki-67-FITC (#11-5698-82, Thermo Fisher Scientific), rabbit anti-fibromodulin (#ab81443, Abcam, Cambridge, UK), mouse-anti podoplanin (#8.1.1.; DSHB), rabbit-anti-SOX9 (#AB5535 Millipore, Darmstadt, Germany), rat anti-CD31-phycoerythrin (PE) (#553370, BD Pharmingen, San Diego, CA), rabbit anti-matrilin-4 (#ab106379, Abcam), rabbit anti-cleaved caspase 3 (#9661, Cell Signaling, Danvers, MA), guinea pig anti-pan keratin (GP14, Progen Biotechnik GmbH, Heidelberg, Germany), goat anti-PDGFR alpha (AF1062, R&D Systems, Minneapolis, MN), donkey anti-rabbit-Cy3 (#711-165-152, Jackson ImmunoResearch, West Grove, PA), donkey anti-rat-Cy3 (#712-165-150, Jackson ImmunoResearch), bovine anti-goat-Cy3 (#805-165-180, Jackson ImmunoResearch), goat anti-guinea pig-Cy2 (#106-225-003, Jackson ImmunoResearch) and donkey anti-goat-Cy2 (#705-225-147, Jackson ImmunoResearch).

Haematoxylin and eosin (H&E), Sirius red and Herovici staining

Acetic ethanol- or PFA-fixed paraffin sections were deparaffinized, rehydrated using a xylene/ethanol gradient, stained with haematoxylin (#3870, JT Baker®, Phillipsburg, NJ) and eosin-Y alcoholic (#102439, Merck, Darmstadt, Germany), Sirius Red (Direct Red 80, #365548, Sigma), or using the Herovici procedure [16], and mounted with Eukitt® (#03989, Sigma).

Quantification of Sirius red und Herovici staining

ECM patterns based on Sirius Red staining were analyzed in Fiji v1.53t using TWOMBLI (version April 2022) [17]. Three representative images were used as a test set to determine optimal parameters: contrast saturation (0.35), line width (5), curvature window (40), minimum branch length (15), and maximum display HDM (200). Gap analysis was excluded.

Herovici-stained areas were quantified using QuPath Version 0.40 [18]. A pixel classifier based on an artificial neural network (ANN_MLP) was trained on a representative subset of the analyzed images to identify areas of young collagen. Young collagen was calculated in relation to total stained tissue area.

Isolation and culture of primary mouse embryonic fibroblasts (MEFs) or lung fibroblasts

MEFs or primary lung fibroblasts were isolated from E18.5 embryos. Embryos were decapitated. For MEF isolation, the head, liver and heart were removed, and the remaining tissue was minced with a razor blade. For lung fibroblasts, the lung was isolated and also minced with a razor blade. Minced tissue was incubated with 2 × trypsin (#59418C, Sigma) for 15 min in a water bath at 37 °C. The suspension was then centrifuged, and afterwards the supernatant was discarded. Cells were then cultured in 6-well plates in DMEM (#6429, Sigma) containing 10% fetal bovine serum (FBS, #A4766801 Thermo Fisher Scientific) and 1% penicillin–streptomycin (Sigma) at 37 °C, 5% CO2.

RNA isolation and RT-qPCR analysis

RNA was isolated using TRIzol (#15596026, Thermo Fisher Scientific) using the manufacturer’s protocol. cDNA was synthesized using iScript (#1708891, BioRad, Hercules, CA). RT-qPCR was performed using the LightCycler®480 SYBR Green I Master reaction mix (Roche), and data (Ct-values) were collected using the LightCycler®480 software. All samples were measured in duplicates and gene expression was determined using the 2−ΔΔCt method. Data were normalized to the expression levels of the gene encoding ribosomal protein 29 (Rps29).

Primers

Sequence forward primer

Sequence reverse primer

Pkdcc

CAA GCT GCT CAA AGA GAT GGT

TGG TAG CAA TAG CCA TAG AGC TG

Fmod

CAG GGC AAC AGG ATC AAT G

CTG CAG CTT GGA GAA GTT CAT

Matn4

GGC GAT CCA GTA CGC TAT GAA

GGC CAA ACT CCT GGA TGA GA

Timp1

GCC CCC TTT GCA TCT CTG GCA T

TGC GGC ATT TCC CAC AGC CT

Col1a1

TGT TCA GCT TTG ACC TCCC GGCC T

TCT CCCC TTG GGT CCC TCG ACT

Col3a1

TCC CCT GGA ATC TGT GAA TC

TGA GTC GAA TTG GGG AGA AT

Hrg

CAC CAA CTG TGA TGC TTC TGA

AGT AGT AGA CTG TGG CCG TTC C

C4b

CCT GGG TGT TCA GCT TCT GT

CAG GAA CCA CCC TTT GGG TT

Cfh

TTA CCG TGA ATG TGG TGC AGA

GCT CCA AAG GCC ATT TTC TGA

Fgg

GGT CAC CCA GAC ACC ATG AG

GGT TGG GCA GAA ACT ACC GA

Rbp4

ACA AGG CTC GTT TCT CTG GG

TGT GAA AGT GCC CAC CAT GT

Fth1

ACC TGG AGT TGT ATG CCT CCT

AGG AAG ATT CGG CCA CCT

Rps29

GGT CAC CAG CTC TAC TG

GTC CAA CTT AAT GAA GCC TAT GTC C

Ccnd1

ACT GCC GAG AAG TTG TGC AT

AAG CAG TTC CAT TTG CAG CAG

Smo

GCA AGC TCG TGC TCT GGT

TCC ACT CGG TCA TTC TCA CA

Ptch1

TGG AGC AGA TTT CCA AGG GGA

GCC CCA AAT ATG AGG AGA CCC

Gli1

GTA TGA GAC AGA CTG CCG CT

GCT CAC TGT TGA TGT GGT GC

Azgp1

TCA CCC CAG ACA TCA ACT CCT A

GGT CTA AGG GGA TCC AAG CTG

Pdpn

GGA GGG CTT AAT GAA TCT ACT G

GTT GTA CTC TCG TGT TCT CTG

Scgb1a1

AAG ATC GCC ATC ACA ATC AC

CTT CAG GGA TGC CAC ATA AC

Fgb

AAG CTG CCG ATG ATG ACT ACG

CGA TAG CCC CCT CCA CTG ATA

Plod1

CCA CAA AAG AGA CTG AGG GC

CAT CCA CAC TCC AGT CCT CC

Plod2

TGA TGG ATT CCA CAG ATT TAT GA

CCC CTC CGA TAC TGT TCA TT

Plod3

ATT GCT GGT GAT CAC TGT GG

TCC TTC TTG AGC CAC CTG AC

Image acquisition and quantification

Fluorescence stainings were imaged using an Axioskop 2 fluorescence microscope (Carl Zeiss, Inc., Oberkochen, Germany), and the corresponding software (Carl Zeiss, Inc.). Image acquisition was performed with an Axiocam HRc camera (Carl Zeiss, Inc.) connected to the microscope. Histochemical stainings were imaged with a Pannoramic 250 slide scanner (3DHISTECH, Budapest, Hungary). Quantifications were performed using the Fiji software [19].

Computer tomograph (CT) scan

The heads from E18.5 CTRL and Pkdcc−/− embryos were fixed in 4% PFA for 24 h and then washed and stored in PBS. They were scanned on a vivaCT80 (Scanco Medical AG, Brüttisellen, Switzerland). 1000 projection images were obtained at an energy of 45 kVp, a current of 177 µA and an integration time of 350 ms. Density calibration of the scanner is checked weekly. Images were reconstructed at an isotropic voxel size of 10.4 µm. After filtration (Gauss filter, sigma 1.2, support 1), a threshold of 150 mgHA/cm3 was applied to segment bone from soft tissue. Because of the low threshold, noise in the soft tissue remained, which was removed partially by a component labelling that excluded components containing less than 20 voxels. Images were visualized in 3D using the software of the scanner (µCT Ray, Scanco Medical AG).

Statistical analysis

Statistical analysis was performed using Prism 8 software (GraphPad Inc., San Diego, CA). Differences between groups were analyzed using Mann–Whitney test (n ≥ 4 or unpaired t test (n = 3)).