The cardiac phospho-proteome during pressure overload in mice

Mice and experimental conditions

10-week-old C57BL/6 N mice were used for both 1-day and 1-week TAC/Sham experiments.

The animal experiment protocols were approved under animal license number G-228/18.

TAC/Sham surgical procedure

Mice were anaesthetised with 3% isoflurane in a whole-body chamber, maintained with 2% isoflurane via a face mask, with preoperative analgesia with buprenorphine (0.1 mg/kg body weight) and Rimadyl (10–15 mg/kg body weight), subcutaneously. Mice were placed on a heated surgical table (37 °C), and the ventral thorax shaved and disinfected. An incision was made from the fourth rib to the sternum, and the sternum opened by an upper median sternotomy to the level of the third rib. The thymus was separated and the aortic arch exposed. The aorta was ligated/constricted using a 7-0 silk suture around a 26-gauge needle between the carotid arteries, and the needle removed. Sham surgeries followed the same procedure without aortic ligation. Atropine sulphate (0.1 mg/kg body weight, subcutaneously) was administered, the sternum closed with interrupted sutures, and skin incision closed with additional sutures and sealed with Histoacryl tissue glue. Mice were monitored postoperatively in a 28 °C incubator until fully awake, and postoperative analgesia included buprenorphine (0.1 mg/kg body weight, subcutaneously) and metamizole (200 mg/kg body weight, via drinking water) administered as post-operative analgesia.

Sample preparation for phospho/proteomic analysis

Mice were euthanised by cervical dislocation, the chest disinfected with 70% ethanol, and the heart excised and washed in ice-cold PBS. Hearts were dissected at the mid-ventricular level, with the basal parts of ventricle processed further. Samples were lysed in 5 volumes of lysis buffer (100 mM Tris HCl pH 8.5, 7 M Urea, 1% Triton X-100, 5 mM Tris(2-carboxyethyl)phosphinhydrochloride, 55 mM 2-Chloroacetamide, 10 Units/mL DNase I, 1 mM Magnesium chloride, 1 mM Sodium orthovanadate, 1 x protease inhibitors, 1 x Complete mini EDTA free protease inhibitor). Samples were homogenised and DNA sheered in a Bioruptor for 45 minutes (20 seconds on, 40 seconds off). Residual cell debris was removed by centrifugation and 1% Benzonase was added to the supernatant for 30 minutes at room temperature. A Bradford assay was performed and protein concentration normalised to 1 mg for each sample (in lysis buffer). To the 1 mg of protein sample, four volumes of methanol, one volume of chloroform and 3 volumes of ultrapure water were added sequentially, and samples were centrifuged at 3740 g for 15 minutes. The upper layer was removed, and three volumes of methanol added to the remaining volume, mixed, and centrifuged. The liquid phase was removed from the pellet, and the air-dried protein pellets were submitted for phospho/proteomic analysis.

Sample preparation of purified cardiomyocytes, cardiac endothelial cells and fibroblasts for cell type specific proteomic analysis

To obtain purified cardiac cell populations for proteomic analysis, cardiomyocytes (CM), endothelial cells (EC), and fibroblasts (FB) were isolated from murine hearts using a combination of enzymatic digestion, Langendorff perfusion method (CMs), and magnetic-activated cell sorting (MACS) (ECs & FBs).

Isolation of cardiac endothelial cells and fibroblasts

Mice were euthanised and hearts collected as before. The tissue was weighed and minced into small fragments in an enzyme solution containing 500 U/mL Collagenase I (Worthington, LS004176) and 150 U/mL DNAse I in RPMI medium and digestion at 37 °C for 30 minutes. The resulting cell suspension was filtered through a 70 µm cell strainer and washed with fetal calf serum (FCS) and MACS buffer.

For endothelial cell isolation, the pellet was resuspended in MACS buffer, incubated with CD146 microbeads (Miltenyi Biotec, 130-092-007) (20 µL per sample), and subjected to MACS separation using MS columns (Miltenyi Biotec, MS columns, 130-042-201) on a magnetic stand. Bound endothelial cells were eluted and pelleted by centrifugation at 400 × g for 6 minutes at 4 °C, then processed for proteomic analysis.

Fibroblasts were collected from the MACS flow-through fraction. The cell suspension was incubated with fibroblast feeder removal beads (Miltenyi Biotech, 130-095-531) (20 µL per sample) for 30 minutes at room temperature, followed by an additional MACS purification step. The fibroblast fraction was subsequently pelleted, resuspended in MACS buffer, and processed for proteomic analysis.

Langendorff isolation of adult murine cardiomyocytes

Cardiomyocytes were isolated via the Langendorff perfusion method. Excised hearts were immediately placed in perfusion buffer (10 mM BDM, 55 mM Glucose in H2O). The aorta was cannulated and secured with a silk suture. Retrograde perfusion was initiated at a rate of 3 mL/min to clear residual blood. An enzyme solution containing Liberase DH (5 mg/mL), trypsin (1%), and CaCl₂ (100 mM) was perfused for 5–6 minutes at 37 °C to digest extracellular matrix proteins. The heart was released from the cannula, minced and homogenised. The resulting cell suspension was passed through a 100 µm strainer to remove tissue debris. Enzymatic digestion was stopped using 10% FCS, 12.5 µM CaCl2 in perfusion buffer. CM were pelleted by low-speed centrifugation (17 × g for 2 minutes) and gently resuspended in 5% FCS, 20 µM CaCl₂ in perfusion buffer. The suspension was allowed to sediment for 10 minutes, after which the cell pellet was collected for downstream analysis.

Protein extraction and sample preparation for mass spectrometry

Isolated CMs, ECs, and FBs were lysed in a buffer containing 10 mM Tris, 150 mM NaCl, 4% glycerol, 0.5 mM sodium disulfite, 1% Triton X-100, 0.1% sodium deoxycholate, 0.05% SDS, 1 mM DTT, and protease inhibitors. Cell lysates were subjected to shock-freezing in liquid nitrogen and later thawed on ice. Following centrifugation at 12,000 × g for 10 minutes at 4 °C, the soluble protein fraction was collected.

Mass spectrometrySample preparation

Pellets for whole proteome and phospho-proteomic analysis were resuspended in digestion buffer (100 mM Tris-HCl pH 8.5, 1% sodium deoxycholate, 5 mM Tris(2-carboxyethyl)phosphin-hydrochlorid and 30 mM chloroacetamide), trypsin was added (1:50 (w/w)) and proteins digested overnight at room temperature (RT). Digestion was stopped with 1% TFA, sodium deoxycholate was precipitated and samples were centrifuged for 10 min at 17000 g. Desalting was performed by using 30 uM Oasis HLB 96-well plates, buffer A was 0.1% formic acid and buffer B 80% acetonitrile with 0.1% formic acid. Pooled fractions were dried under vacuum centrifugation and reconstituted in 10 μL 1% formic acid, 4% acetonitrile and stored at −80 °C. Phosphopeptide enrichment was done as in Leutert et al.18. Peptides were taken up in IMAC loading solvent (80% acetonitrile, 0.4% TFA). An aliquot was used for full proteome analysis, remaining peptides were subjected to phosphopeptide enrichment using the KingFisher Apex TM platform (ThermoFisher) and magnetic Fe-NTA beads (Cube Biotech). Phosphopeptides were eluted with 0.2% dimethylamine in 80% acetonitrile. Peptides were labelled with TMT16plex and fractionated by high-pH reversed-phase fractionation19.

For cell specific proteomic analysis, reduction of disulfide bonds was performed with dithiothreitol (56 °C, 30 min, 10 mM in 50 mM HEPES, pH 8.5), followed by alkylation with 20 mM 2-chloroacetamide (RT, 30 min). Samples were subjected to the SP3 protocol and peptides were eluted by tryptic digestion overnight at 37 °C20. Peptides were recovered twice in HEPES buffer on a magnet and labelled with TMT6plex or TMT16plexaccording the manufacturer’s instructions19,21. Samples were combined and desalted using an OASIS® HLB µElution Plate (Waters). Pooled samples were subjected to high-pH reverse phase fractionation on an Agilent 1200 Infinity high-performance liquid chromatography system equipped with a Gemini C18 column (3 μm, 110 Å, 100 × 1.0 mm, Phenomenex) with a Gemini C18, 4 × 2.0 mm SecurityGuard cartridge as a guard column. 48 fractions were collected and pooled into 12 fractions and dried down.

Measurement

Samples for cell type specific proteomics were measured on a Q Exactive™ Mass Spectrometer (Thermo) and whole proteome and phospho-proteomic analysis in an Orbitrap Fusion™ Lumos™ Tribrid™ Mass Spectrometer (Thermo) coupled to an UltiMate 3000 RSLC nano LC system (Dionex). Sample was concentrated on a C18 µ-Precolumn (Acclaim PepMap 100, 5 µm, 300 µm i.d. × 5 mm, 100 Å) and resolved on a nanoEase™ M/Z HSS T3 column (75 µm × 250 mm C18, 1.8 µm, 100 Å). Trapping was carried out with 30 µL/min 0.5% trifluoroacetic acid for 4 minutes. Peptides were eluted via the analytical column (solvent A: 0.1% formic acid in water, 3% DMSO) with a constant flow of 0.3 µL/min, with increasing percentage of solvent B (0.1% formic acid in acetonitrile, 3% DMSO) from 2% to 8% in 6 min, 8% to 28% in 66 min, from 28% to 40% in 10 min, followed by an increase of B from 40–80% for 3 min and a re-equilibration back to 2% B for 5 min.

For the phospho-proteomic samples the percentage of solvent B was increased as follows: from 2% to 4% in 4 min, to 8% in 2 min, to 25% in 64 min, to 40% in 12 min, to 80% in 4 min, followed by re-equilibration back to 2% B in 4 min, for the full proteome analysis, the steps were from 2% to 8% in 4 min, to 28% in 104 min, to 40% in 4 min, to 80% in 4 min, followed by re-equilibration back to 2% B in 4 min. Peptides were introduced into the mass spectrometer via a Pico-Tip Emitter 360 µm OD × 20 µm ID; 10 µm tip (New Objective) and an applied spray voltage of 2.4 kV. The capillary temperature was at 275 °C. Settings for the Q Exactive were: Full mass scan (MS1) with mass range 375–1200 m/z, profile mode, in the orbitrap, resolution of 70000, fill time 10 ms. AGC target 3E6. Data dependent acquisition (DDA) was performed with the resolution set to 17500, fill time 50 ms, AGC target of 9E1 ions with a normalised collision energy of 32, HCD, fixed first mass 110 m/z.

Settings for the Fusion Lumos were: full mass scan with mass range 375 to1400 m/z for the phosphoproteome (375 to 1500 m/z for the full proteome), in profile mode in the orbitrap with resolution of 120000. The filling time was set at maximum of 50 ms for the full proteome with a limitation of 4 × 105 ions. Data dependent acquisition (DDA) was performed with the resolution set to 30000, with a fill time of 110 ms for the phosphoproteome (94 ms for the full proteome) and a limitation of 1 × 105 ions. A normalised collision energy of 34 was applied. MS2 data was acquired in profile mode. Fixed first mass was set 110 m/z.

IsobarQuant with Mascot (v2.2.07) (for cell type samples) and Fragpipe v21.1 with MS Fragger v4.0 (for whole proteome and phosphoproteome) were used to process the acquired data, searched against a Mus musculus proteome database (IsobarQuant: UP000000589, May 2016, 59550 entries; MS Fragger: UP000000589, October 2022, 21968 entries) including common contaminants and reversed sequences22,23. The following modifications were included into the search parameters: Carbamidomethyl on Cysteine and TMT6/16 on lysine as fixed modifications, protein N-term acetylation, oxidation on methionine, phosphorylation on STY (phosphoproteome only) and TMT6/16 on N-termini as variable modifications. A mass error tolerance of 10 ppm (IsobarQuant) or 20 ppm for precursor ions and 0.02 Da (IsobarQuant) or 20 ppm for fragment ions was set (MSFragger). Trypsin was set as protease with a maximum of two missed cleavages, and minimum peptide length set at seven amino acids. The minimum peptide length was set to seven amino acids.

Data analysis

For phospho-proteomics data, raw output files of FragPipe (psm.tsv for phospho data and protein.tsv files for input data) were processed using the R language (ISBN 3-900051-07-0)24. Peptide spectral matches (PSMs) with a phosphorylation probability greater 0.5 and proteins with at least two razor peptides were considered. Phosphorylated amino acids were marked with a * in the amino acid sequences behind the phosphorylated amino acid, labelled with a 1, 2 or 3 for the number of phosphorylation sites in the peptide and concatenated with the protein ID to create a unique ID for each phosphopeptide. Raw TMT reporter ion intensities for all matches with the same phosphopeptide ID were summed. For the input data, the reporter ion intensities were used as given in the protein.tsv output files. Phospho signals were also normalised by input abundance according to the following formula:

Transformed summed TMT reporter ion intensities were cleaned for batch effects using the ‘removeBatchEffects’ function of the limma package and normalised using the vsn package25,26. Missing values were imputed with ‘knn’ method using the Msnbase package27. Differential expression was tested using the limma package. Phospho, normalised phospho and input data were tested separately. The replicate information was added as a factor in the design matrix given as an argument to the ‘lmFit’ function of limma, imputed values were given a weight of 0.05. Hits were defined as a false discovery rate (fdr) smaller 5% and a fold-change of at least 100% and candidates with an fdr below 20% and a fold-change of at least 50%. Clustering with all hit phospho-peptides based on the median protein abundances normalised by median of control condition was conducted to identify groups of proteins with similar patterns across conditions. The ‘kmeans’ method was employed, using Euclidean distance as the distance metric and ‘ward.D2’ linkage for hierarchical clustering. The optimal number of clusters (3) was determined using the Elbow method, which identifies the point where the within-group sum of squares stabilizes. Gene ontology (GO) enrichment analysis (Molecular Function (MF), and Biological Process (BP)) was performed using the ‘clusterProfiler’ package using ‘org.Mm.eg.db’ as the reference database28. The odds ratio (‘odds_ratio’) for each GO term was calculated by comparing the proportion of genes associated with that term in the dataset (‘GeneRatio’) to the proportion in the background set (‘BgRatio’). An odds ratio greater than one indicates an enriched GO term.

For cell type specific proteomic data analysis, initial data processing included filtering out contaminants and reverse proteins. Only proteins quantified with at least two unique peptides (with qupm >= 2) were considered for further analysis. Additionally, only proteins identified and quantified in at least two out of three mass spec runs were retained to ensure robustness. 4435 proteins passed the quality control filters. Batch effects were removed using the ‘removeBatchEffect’ function of the limma package on the log2 transformed raw TMT reporter ion intensities (‘signal_sum’ columns)25. Normalisation was performed using the ‘normalizeVSN’ function of the limma package26. Missing values were imputed with the ‘knn’ method using the ‘impute’ function from the Msnbase package ensuring that incomplete data did not distort the analysis27. Differential expression analysis was performed using a moderated t-test accounting for replicate information by including it as a factor in the design matrix passed to the ‘lmFit’ function25. Imputed values were assigned a weight of 0.01, while quantified values were given a weight of one, ensuring analysis reflected the uncertainty in imputed data. Proteins were annotated as hits if they had a false discovery rate (FDR) below 0.05 and an absolute fold change greater than two. Proteins were considered candidates if they had an FDR below 0.2 and an absolute fold change greater than 1.5. Clustering with all hit and candidate proteins based on the median protein abundances normalised by median of control condition was conducted to identify groups of proteins with similar patterns across conditions. The ‘kmeans’ method was employed, using Euclidean distance as the distance metric and ‘ward.D2’ linkage for hierarchical clustering. The optimal number of clusters (20) was determined using the Elbow method. Whilst sham and TAC operated samples were processed and analysed (and included in the deposited data set), only sham data was reported in the manuscript.