Effect of pan-PTP4A inhibition on patient-matched primary and recurrent Glioblastoma models using phosphoproteomics

Summary
Characterization of pan-PTP4A inhibition via JMS-053 on patient-matched primary GBM (BT594) and post-treatment recurrent GBM (BT972) using phospho-proteomics.

Sample preparation
2.5 million tumor cells (BT594 and BT972) were plated at a density of 0.5 million cells/mL of NeuroCult NS-A proliferation kit (Human). Cells were either treated with 1 µM JMS-053 (active pan-PTP4A inhibitor) or 1 µM JMS-038 (inactive control compound) for five minutes or 30 minutes at 37°C and 5% CO2. Following incubation, samples were immediately washed with cold PBS containing 1X Halt Protease and Phosphatase Inhibitor (Thermo Fisher Scientific no. 78442). Following three cycles of centrifugation and washing with HALT-containing PBS, samples were pelleted. Except for BT594 tumor cells treated with JMS-053 for 30 minutes, all samples were cultured, treated and processed in experimental duplicates. Cell pellets were suspended in 200 μL of lysis buffer comprising 8M urea (Sigma-Aldrich) and 100 mM ammonium bicarbonate (Sigma-Aldrich) each. Cells were then vortexed at 2800 rpm using Mini S-2 Vortex Mixer (Fisher Scientific) for ten seconds, followed by ten seconds of incubation on ice. This procedure was repeated six times. The lysate was then centrifuged at 21,000 ×g for five minutes at 4°C. Protein reduction was conducted using 5 mM of tris (2-carboxyethyl) phosphine (Sigma-Aldrich) for 45 minutes at 37°C. Subsequently, 10 mM of iodoacetamide (Sigma-Aldrich) was added for protein alkylation for 45 minutes at room temperature (dark). Following alkylation, cell lysate was diluted five-fold with 100 mM of ammonium bicarbonate to lower urea concentration. Based on protein amount, Sequencing Grade Modified Trypsin (Promega) was then added in (trypsin: protein[w/w] at 1:50) for overnight digestion at 37°C. Trifluoroacetic acid (Thermo Fisher Scientific) was added to reduce pH, and desalting was conducted with SOLA Solid Phase Extraction 10 mg 96-well plates (Thermo Fisher Scientific). Peptides were eluted using 400 mL 80% Acetonitrile - 0.1% trifluoroacetic acid. Eluted peptides were speed-vacuum dried using Labconco CentriVap Benchtop Vacuum Concentrator (Kansas City, MO).
Labeling, phospho-peptide enrichment and LC/MS/MS
Tandem Mass Tag (TMT)11-plex Isobaric Label Reagent Set (Thermo Fisher Scientific) were used to label peptides from each condition following the manufacturer’s protocol. Briefly, dried peptide samples were resuspended in 100 mM of triethylammonium bicarbonate (TEAB) (Sigma-Aldrich) to 1 µg/µL. Then, 0.2 mg TMT reagents were mixed with 50 µg of peptide samples at 4:1 (w/w) ratio and incubated at room temperature for one hour. Following incubation, each TMT reaction was quenched with 2 µL of 5% hydroxylamine (Sigma-Aldrich) for 15 minutes at room temperature. Labeled samples were pooled together at equal ratio and then speed-vacuum dried. Desalting was then performed to purify salt-free TMT-label peptide samples. Phosphopeptides-enrichment were conducted using Ni-NTA magnetic agarose beads (Qiagen). Briefly, 138 µl of Ni-NTA magnetic agarose beads were washed three times with 400 µL of ultrapure nuclease-free water, and then incubate with 400 µL of 100 nM EDTA for 30 minutes. After three washes with 400 µL of ultrapure nuclease-free water, the beads were incubated with 100 nM FeCl3 for 30 minutes. Then beads were washed three times with 400 µL of ultrapure nuclease-free water and one time with 400µL of 80% acetonitrile - 0.1% trifluoroacetic acid. 550 µg TMT-labeled peptides were resuspended in 550 µL of 80% acetonitrile - 0.1% trifluoroacetic acid and added to the Fe-activated beads, rotating for 30 minutes. After capture, the supernatant was saved for analysis of unphosphorylated peptides and phosphopeptides were eluted using 50 µL of 1.4% of ammonia hydroxide solution followed by another 50 µL of ultrapure nuclease-free water.Enriched phosphopeptides were separated into nine fractions by high pH reverse-phase liquid chromatography (RPLC) using a home-made C18 column (200 µm x 30 cm bed volume, Waters BEH 130 5 µm resin) from 11%-32% acenotritile-20 mM ammonium formate (pH10) at flow rate of 5 µL/minute. Each fraction was then loaded onto a home-made trap column (200 µm x 5 cm bed volume) packed with POROS 10R2 10 µm resin (Applied Biosystems), followed by a home-made analytical column (75 µm x 25 cm bed volume) packed with Reprosil-Pur 120 C18-AQ 1.9 µm particles (Dr. Maisch) with integrated Picofrit nanospray emitter (New Objective). LC-MS experiments were performed on a Thermo Fisher Ultimate 3000 RSLCNano UPLC system that ran a 3 hr gradient (11%-38% acetonitrile-0.1% formic acid) at 70 nL/min, coupled to a Thermo QExactive HF quadrupole-Orbitrap mass spectrometer. A parent ion scan was performed using a resolving power of 120,000 and then up to the 30 most intense peaks were selected for MS/MS (minimum ion counts of 1,000 for activation), using higher energy collision induced dissociation (HCD) fragmentation. Dynamic exclusion was activated such that MS/MS of the same m/z (within a range of 10ppm; exclusion list size=500) detected twice within 5s were excluded from analysis for 30s.
Data analysis
LC-MS data were searched against a UniProt human protein database (Ver 2017-06, 42,173 entries) for protein identification and quantification by Protein Discover software (Thermo Fisher Scientific). From 300970 MS/MS spectra acquired in all nine fractions, 84496 peptide-to-spectrum matches (PSMs), 20443 unique peptide groups (with Peptide FDR < 0.01), and 4,732 unique proteins (Protein FDR < 0.01) were identified. Among the identified peptide groups, 17,127 hits were phosphopeptides. The density of these phosphopeptides was normalized by total TMT reporter intensity values from a separate LC/MS/MS run of the phosphoenrichment flow-through. Phosphopeptide fold-change between JMS-053 and JMS-038 for both cell lines was calculated, and the mean value of TMT reporter intensity for each phosphopeptide was calculated as well. These values were then uploaded to Perseus software to generate significance B values.
Contact.
Contact Dr. Chirayu Chokshi regarding any questions about the data or analysis ([e-mail]).