Pharmacoproteomic characterisation of human colon and rectal cancer - CRC65 Kinobeads
Most molecular cancer therapies act on protein targets but data on the proteome status of patients and cellular models are only beginning to emerge. Here, we profiled the proteomes of 65 colorectal cancer (CRC) cell lines to a depth of >10,000 proteins using mass spectrometry. Integration with proteomes of 90 CRC patients, as well as transcriptomes of 145 cell lines and 89 patients defined integrated CRC subtypes, highlighting cell lines representative of each tumour subtype. Modelling the responses of 52 CRC cell lines to 577 drugs as a function of proteome profiles enabled predicting drug sensitivity for cell lines and patients. Among many novel associations, MERTK was identified as a predictive marker for resistance towards MEK1/2 inhibitors and immunohistochemistry of 1,000 CRC tumours confirmed MERTK as a prognostic survival marker. We provide the proteomic and pharmacological data to the community to e.g. facilitate the design of innovative prospective clinical trials.
Sample Processing Protocol
Kinobeads (KB) pulldowns (biological triplicates) were performed as described elsewhere (Medard et al., 2015), with minor modifications. Total Cell Lysates (TCLs) were quickly thawed at 37°C in a water bath and subsequently kept on ice. For each pulldown, 3 mL of 2.2 mg/mL TCL was transferred to pre-cooled ultracentrifuge tubes (Beckman & Coulter) and cleared by ultracentrifugation at 167,000 xg and 4°C for 20 min. Meanwhile, 35 µL of settled KBγ per TCL were distributed to separate wells of deep 96-well filter plates (Porvair Sciences), washed twice with 1 mL CP buffer (50 mM Tris/HCl pH 7.5, 5% Glycerol, 1.5 mM MgCl2, 150 mM NaCl, 1 mM Na3VO4) each and subsequently equilibrated using 1 mL CP buffer each, supplemented with 0.4% Igepal CA-630. Solvents were drained from the beads via gravitational flow through the filter unless stated otherwise. Afterwards, the plates were placed on regular 96-well plates for waste collection, centrifuged for 2 min at 240 xg and 4°C and sealed with the corresponding bottom sealing mat. Next, approximately 1.8 mL (equivalent of 4 mg of protein) of each TCL was transferred to its corresponding well, the plates were sealed with their top sealing mats and incubated for 60 min at 4°C on a head-over-end shaker. After lysate incubation, the plates were centrifuged for 3 min at 240 xg and 4°C, followed by drainage of the TCLs from the beads and three washing steps with CP buffer containing 0.4% Igepal CA-630 and two washing steps with CP buffer containing 0.2% Igepal CA-630. Subsequently the plates were centrifuged for 2 min at 240 xg on a 96-well plate for waste collection and sealed again with their bottom sealing mat. For elution of proteins bound to KBγ, 60 µL of 2x NuPAGE LDS sample buffer (Thermo Fisher) per well containing 50 mM DTT were distributed to the filter plates, the plates were sealed with their top sealing mats and incubated for 30 min at 50°C and 700 rpm in a thermomixer. Afterwards, the top and bottom sealing mats were removed, the filter plates were placed onto 96-well plates for eluate collection and centrifuged for 2 min at 670 xg and 4°C. Free sulfhydryl groups were then alkylated for 30 min at room temperature protected from light by addition of chloroacetamide to a final concentration of 55 mM. Finally, detergents and salts were removed from eluates by running a short electrophoresis (approximately 0.5 cm) using 4−12% NuPAGE gels (Thermo Fisher), followed by tryptic in-gel digestion according to standard procedures.
Data Processing Protocol
MaxQuant v.184.108.40.206 was used to search our Kinobeads raw data against UniProtKB (v25.11.2015; 92,011 sequences), concatenated with a list of common contaminants supplied by MaxQuant (245 sequences) in two separate runs with identical settings. We set the digestion mode to fully tryptic, allowing for cleavage before proline (Trypsin/P) and a maximum of two missed cleavages. Carbamidomethylation of cysteines was set as a fixed modification and oxidation of methionines, as well as acetylation of protein N-termini were set as variable modifications, allowing for a maximum number of 5 modifications per peptide. Candidate peptides were required to have a length of at least 7 amino acids, with a maximum peptide mass of 4,600 Da. The fragment ion tolerance was set to 120 ppm for FTMS. A first search with a precursor ion tolerance of 20 ppm was used to recalibrate raw data based on all peptide-spectrum-matches (PSMs) without filtering using hard score cut-offs. After recalibration, the data were searched with a precursor ion tolerance of 4.5 ppm, while chimeric spectra were searched a second time using MaxQuant’s “Second peptides” option to identify co-fragmented peptide precursors. We used “Match between runs” with an alignment time window of 30 min and a match time window of 1.1 min to transfer identifications between raw files of the same and neighbouring fractions (± 1 fraction). Using the classical target-decoy approach with a concatenated database of reversed peptide sequences, data were filtered using a PSM and protein false discovery rate (FDR) of 1%. Protein groups were required to have at least one unique or razor peptide, with each razor peptide being used only once during the calculation of the protein FDR. No score cut-offs were applied in addition to the target-decoy FDR. We used unique and razor peptides for quantification, discarding the unmodified counterparts of peptides harbouring oxidated methionines and acetylated protein N-termini.
Frejno M, Zenezini Chiozzi R, Wilhelm M, Koch H, Zheng R, Klaeger S, Ruprecht B, Meng C, Kramer K, Jarzab A, Heinzlmeir S, Johnstone E, Domingo E, Kerr D, Jesinghaus M, Slotta-Huspenina J, Weichert W, Knapp S, Feller SM, Kuster B. Pharmacoproteomic characterisation of human colon and rectal cancer. Mol Syst Biol. 2017 13(11):951 PubMed: 29101300