Cells, Cell Culture, and Reagents
SUM159 cells were propagated in Nutrient Mixture F-12 supplemented with 5% fetal calf serum, 0.5 μg/ml hydrocortisone, and 10 μg/ml insulin (all from Sigma), 100 IU/ml penicillin, 100 μg/ml streptomycin, and 100 μg/ml Normocin (InvivoGen). MCF10A-HER2/HER3  were propagated in DMEM/F12 medium (Invitrogen) supplemented with 5% horse serum (Hyclone), 20 ng/ml human EGF (Peprotech), 0.5 μg/ml hydrocortisone, 100 ng/ml cholera toxin, and 10 μg/ml insulin (all from Sigma), 100 IU/ml penicillin, 100 μg/ml streptomycin and 100 μg/ml Normocin (InvivoGen). These cells were profiled for cell line-specific highly-polymorphic short tandem repeat loci (STRs) (Microsynth). Previously published tools were used for the inducible RNAi studies . For siRNA experiments, 175,000 cells were seeded in 6-well plates. The following day, cells were transfected using DharmaFECT and the human ETS1 siRNA- SMARTpool (L-003887-00-0005) or the non-targeting pool as control (D-001810-10-05). Experiments were performed according to the manufacturer’s protocol and using siRNA at final concentrations of 25 nM and 12.5 nM for SUM159 and MCF10A-HER2/HER3, respectively. Cells were harvested at the times indicated in the figure legends.
SHP099 and MEK162 were obtained from Novartis (Basel, Switzerland and Cambridge, USA). Compounds were prepared as 10 mM stock solutions in DMSO and stored protected from light at –20 °C.
Cytokine Array and IL-8 ELISA
Cells were cultured overnight in 6-well plates at 250,000 cells/well and the culture medium containing the inhibitor(s) described above was then added. Cell supernatants were collected 48 h later and cells lysed with RIPA buffer (50 mM Tris-HCl pH 8, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) supplemented with 1 × protease inhibitor cocktail (Complete Mini, Roche), 0.2 mM sodium orthovanadate, 20 mM sodium fluoride, and 1 mM phenylmethylsulfonyl fluoride (Fig. 1A–C). For washout experiments, cells were seeded in 96-well plate at 10’000 cells/well for MCF10A HER2/HER3 and 5’000 cells/well for SUM159. The day after, culture medium containing DMSO or 10 μM SHP099 was added. After 24 h, cell supernatants were collected, and fresh media was added (washout), and cell supernatant was harvest again after 24 h . For protein quantification, cells were fixed and stained for SRB (Fig. 1D). For SHP2 knockdown cell lines, a similar time plan was used, following 5 days of doxycycline treatment (Fig. 1E, F). Cytokine arrays were performed using a Proteome Profiler Human Cytokine Array Kit (R&D Systems) according to the manufacturer’s protocol. IL-8 ELISA was performed using the Legend Max Human IL-8 ELISA Kit with Pre-coated Plates (BioLegend) according to the manufacturer’s protocol.
Treated cells were lysed in RIPA buffer (50 mM Tris-HCl pH 8, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) containing 1 × protease inhibitor cocktail (Complete Mini, Roche), 0.2 mM sodium orthovanadate, 20 mM sodium fluoride, and 1 mM phenylmethylsulfonyl fluoride. Samples were supplemented with Laemmli buffer and boiled for 5 min at 95 °C on a heating block. Proteins (30 µg) were loaded onto a 10% polyacrylamide gel and subsequently transferred to a PVDF membrane (Immobilon-P, Millipore). The membrane was blocked for 1 h at room temperature with 5% BSA in TBS supplemented with 0.05% Tween 20 (TBS-T). Primary antibodies (anti-phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204) Cell Signaling, #9101; anti-ERK2 Santa Cruz Biotechnology, sc-1647) were diluted in TBS-T and incubated with the membrane overnight at 4 °C. Secondary antibodies (IRDye 680RD or 800RD) were incubated with the membrane for 1 h at room temperature. Blots were imagined using a LI-COR Odyssey CLx imager.
SUM159 cells were cultured in 500-cm2 Square TC-treated Culture Dishes (Corning) to 65% confluence. For each sample, three plates have been treated with DMSO for the control sample and with 10 µM of SHP099 at 1 min, 5 min, 15 min and 30 min time points. After rinsing the plates with ice-cold PBS twice, cells were collected into 2% sodium deoxycholate in 50 mM HEPES buffer (pH 8.5) using a cell scraper. The cell lysate was sonicated using a Branson digital tip sonicator for 1 min at 70% amplitude on ice, followed by a 5-min incubation at 95 °C. Proteins were reduced and alkylated for 30 min in the dark in 5 mM TCEP or 5 mM chloroacetamide, respectively. The samples were fourfold diluted with 50 mM HEPES buffer (pH 8.5) and digested overnight with LysC (Wako Chemicals) at a 1:100 ratio. The next morning, samples were supplemented with trypsin (ThermoFisher) at a 1:100 ratio and incubated for 24 h at 37 °C. The digestion was quenched by adding 10% TFA to a final concentration of 1%. The peptides were cleared by centrifugation for 5 min at 7000 × g, purified using a SEP-PAK (Waters), and eluted in 50% acetonitrile in water.
For phosphotyrosine enrichment, 6.5 mg aliquots of peptides from each sample were subjected to immunoprecipitation using the PTMScan Phospho-Tyrosine Motif Kit from Cell Signaling (P-Tyr-1000) and the phosphorylated peptides were enriched and eluted according to the manufacturer’s instructions. Eluted peptides were labeled with TMT10plex isobaric labeling reagents (Thermo Fisher) as described in the manufacturer’s instructions, followed by off-line high pH fractionation.
For the global proteome and the phosphoserine/threonine enrichment, a peptide fraction of 500 µg from each sample was labeled with TMT reagents and pooled. A 100 µg aliquot of the peptide mixture was subjected to off-line high pH fractionation for global proteome measurements, while the rest was used for TiO2-based phosphorylated peptide enrichment as described in Borisova et al. , followed by off-line high pH fractionation of phosphorylated peptides. The high pH off-line fractionation was carried out on a YMC Triart C18 0.5 × 250 mm column (YMC Europe GmbH) using the Agilent 1100 system (Agilent Technologies). A total of 96 fractions was collected for each experiment and concatenated into 48 fractions as previously described . For each LC-MS analysis, approximately 1 µg aliquots of peptides were loaded onto PepMap 100 C18 2 cm trap (Thermo Fisher) using the Proxeon NanoLC-1000 system (Thermo Fisher). On-line peptide separation was performed on a 15 cm EASY-Spray™ C18 column (ES801, Thermo Fisher) by applying a linear gradient of increasing ACN concentration at a flowrate of 150 nL/min. Orbitrap Fusion Lumos Tribrid (Thermo Fisher) mass spectrometer was operated in the data-dependent mode. The ions for the survey scan were collected for a maximum of 100 ms to reach the AGC target value of 20’0000 and the scan recorded using an Orbitrap detector at a resolution of 120’000. The top 10 most intense precursor ions from the Orbitrap survey scan were selected for higher-energy C-trap dissociation (HCD) at 38% normalized collision energy scan. To reach an AGC value of 50’000 ions, the maximum ion accumulation time for the MS2 scan was set to 180 ms for the proteome measurements and 150 ms for phosphorylated peptide measurements. The TMT reporter ions were quantified using an MS2 scan recorded using the Orbitrap analyzer at a resolution of 50’000. Thermo RAW files were processed using Proteome Discoverer 2.1 software (Thermo Fisher) as described in the manufacturer’s instructions. Briefly, the Sequest search engine was used to search the MS2 spectra against the Homo sapiens UniProt database (downloaded on 04/04/2017) supplemented with common contaminating proteins. For total proteome analysis, cysteine carbamidomethylation and TMT tags on lysine and peptide N-termini were set as static modifications, whereas oxidation of methionine residues and acetylation protein N-termini were set as variable modifications. For phosphorylated peptide-enriched sample analysis, serine, threonine, and tyrosine phosphorylation were set as variable modifications while other modifications were set as for the proteome analysis. The assignments of the MS2 scans were filtered to allow 1% FDR. For reporter quantification, the S/N values were corrected for isotopic impurities of the TMT reagent using the values provided by the manufacturer. The sums across all TMT reporter channels were normalized assuming equal total protein content in each sample for proteome analysis whereas, for phosphorylated peptide analysis, normalization was based on the total amount of phosphorylated peptides.
Total RNA was extracted from SUM159 samples with TRIzol reagent (Invitrogen), processed and hybridized to GeneChip Human Gene 1.0 ST arrays (Affymetrix, Santa Clara, CA), and scanned according to the manufacturer's instructions. CEL files for MCF10a samples were downloaded from Gene Expression Omnibus repository GSE34525 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE34525).
All gene arrays were processed in R (http://www.r-project.org/) using Bioconductor and the package oligo . Robust multi-array mean was performed using the following command: expr <- rma(read.celfiles(filenames)). Probes with the largest interquartile range were selected as representative of corresponding genes (using array annotation from Bioconductor package hugene10sttranscriptcluster.db). Microarray data for SUM159 are accessible from the Gene Expression Omnibus repository (GSE182033). Differential gene expression between cells engineered with Crtl shRNA, SHP2 sh1, and SHP2 sh2 was calculated using the package limma .
Motif Activity Response Analysis (MARA)
We used the MARA  to model genome-wide gene expression patterns in terms of computationally predicted transcription factor binding sites. We compared the activity means and standard deviations of several regulatory motifs under control and SHP2-knockdown conditions.
Quantitative Real-Time PCR
Total RNA was extracted using the RNeasy Plus Mini Kit (Qiagen) and reverse transcribed using the iScript cDNA Synthesis Kit (BioRad). The resulting cDNA was used for TaqMan-based quantitative real-time PCR using the PrimeTime Gene Expression Master Mix (IDT) for quantification of CXCL18, PTPN11, and ETS1, with HPRT1 as a control gene. The following PrimeTime qPCR Probe Assays (IDT) were used: Hs.PT.58.39926886.g, Hs.PT.56a.20552233, Hs.PT.58.39917763, Hs.PT.58v.45621572.
In each of the studies presented, the results shown represent at least three independent experiments. Values are reported as means ± standard deviation. Data were tested for normal distribution and ANOVAs tests were applied using GraphPad Prism 7.04. The P values < 0.05 were considered statistically significant.
Proteomic data are available via ProteomeXchange with identifier PXD017219. Microarray data were described elsewhere  and can be downloaded from the Gene Expression Omnibus repository under GSE34525 and GSE182033 for the MCF10A and SUM159, respectively.