PRIDE Assigned Tags:Biological Dataset
TCellXTalk enables the detection of co-modified peptides for the study of protein post-translational modification cross-talk in T cells
Protein function is regulated by post-translational modifications (PTMs) that may act individually or interact with others in a phenomenon termed PTM cross-talk. Multiple databases have been dedicated to PTMs, including recent initiatives oriented to the in silico prediction of PTM interactions. The study of PTM cross-talk ultimately requires experimental evidence about whether certain PTMs co-exist in a single protein molecule. However, available resources do not assist researchers in the experimental detection of co-modified peptides. Here we present TCellXTalk, a comprehensive database of phosphorylation, ubiquitination and acetylation sites in human T cells that supports the experimental detection of co-modified peptides using targeted or directed mass spectrometry. We demonstrate the efficacy of TCellXTalk and the strategy presented here in a proof of concept experiment that enabled the identification and quantification of 15 co-modified (phosphorylated and ubiquitinated) peptides in CD3 proteins of the T-cell receptor complex. To our knowledge, these are the first co-modified peptide sequences described in this widely studied cell type. Furthermore, quantitative data showed distinct dynamics of co-modified peptides upon T cell activation, demonstrating differential regulation of co-occurring PTMs in this biological context. Overall, TCellXTalk enables the experimental detection of co-modified peptides in human T cells and puts forward a novel and generic strategy for the study of PTM cross-talk.
Sample Processing Protocol
Jurkat T lymphocytes (clone E6-1) were obtained from the American Type Culture Collection (Manassas, VA, USA) and cultured in Roswell Park Memorial Institute (RPMI) 1640 medium (Sigma, St Louis, MO, USA), supplemented with 10 % (v/v) complement-inactivated fetal bovine serum (FBS) (Gibco, Grand Island, NY, USA). The acetylome dataset generated for the database, was obtained using cells treated with 1% (v/v) deacetylation inhibition cocktail (Santa Cruz Biotechnology, Santa Cruz, CA, USA) for 24 h at a density of 1.5 x 106 cells/mL. In the T cell activation experiment, Jurkat T cells were activated using the ImmunoCult Human CD3/CD28 kit (Stemcell, Vancouver, Canada) and harvested 15 or 120 min after activation, or left untreated (control group). The experiment was conducted in three biological replicates. In all cases, cells were washed in PBS, frozen and stored at -80 ºC until processed. Cells were lysed and proteins were digested by LFASP as described (Casanovas 2017). For acetylome analysis, 12 mg of starting protein material were digested and fractionated by basic pH reversed-phase liquid chromatography as previously described (Casanovas 2017). Acetylated peptides were enriched by immunoaffinity purification (IAP) using anti-K-Ac antibody beads (PTMScan pilot acetyl-lysine motif (Ac-K) kit, Cell Signaling Technology Danvers, MA, USA) for 2 h at 4 °C with rotation, with an amount of antibody equivalent to 1/10 of the commercial vial. K-Ac enriched peptides were washed and eluted as previously described for K-ε-GG peptides (i.e., peptides containing the Gly-Gly remnant on the modified lysine residue that results from tryptic digestion of ubiquitinated proteins) (Casanovas 2017) and desalted using C18 pipette tips (PolyLC, Columbia, MD, USA) prior to LC-MS/MS analysis. For the analysis of co-modified peptides (i.e., peptides bearing at least one phosphorylation and one ubiquitination site) in activated Jurkat cells, we digested 1.2 mg of total protein per sample and K-ε-GG peptides were enriched by IAP using cross-linked anti-K-ε-GG antibody beads from the PTMScan ubiquitin remnant motif (K-ε-GG) kit (Cell Signaling Technology, Danvers, MA, USA) as previously described (Casanovas 2017).Peptides were analyzed by LC-MS/MS using an HPLC system composed of an Agilent 1200 capillary nano pump, a binary pump, a thermostated microinjector and a microswitch valve coupled to an LTQ Orbitrap XL mass spectrometer (Thermo Scientific) equipped with a nanoelectrospray source (Proxeon, Odense, Denmark, now Thermo Scientific). For acetylome analysis, 30 % of the IAP eluate was analyzed using the conditions and settings described in Casanovas et al. (2017). For experimental detection of co-modified peptides of CD3 proteins we used a directed MS strategy. In short, 15 % of the K-ε-GG IAP eluate was injected in 0.5 % (v/v) TFA, 5 % (v/v) methanol and the peptides were separated with a C18 pre-concentration cartridge (Agilent Technologies, Santa Clara, CA, USA) connected to a C18 100 μm x 150 mm column (Nikkyo Technos Co, Tokyo, Japan) at 400 nL/min using a 120 min linear gradient from 0 to 35 % solvent B (Solvent A: 0.1 % (v/v) formic acid; solvent B: ACN 0.1 % (v/v) formic acid). The LTQ Orbitrap XL was operated in data-dependent mode at a target mass resolution of 60,000 (at m/z 400). The scan range of each survey scan was m/z 400-1650, and up to 5 peaks with charge ≥2 and an intensity above the 500 threshold were selected and fragmented by collision induced dissociation with normalized collision energy of 35. Precursor ions were selected for MS/MS using an inclusion list of potential co-modified peptides (i.e., at least one phosphorylated and one ubiquitinated residue) generated with TCellXTalk using the option “trypsin with exceptions” and 0 missed cleavages for the in silico digestion of the proteins CD3ε, CD3γ and CD3ζ (Table S1). The inclusion list comprised ions with charges +2, +3 and +4, and an m/z value between 400 and 1650 (Table S1). The maximum injection times for the survey scan and the MS/MS scan were 1 s and 500 ms, respectively, and the respective ion target values were set to 1E6 and 1E4. To reduce the redundant selection of precursor ions, the fragmented ions were dynamically excluded for 6 s. To induce a higher degree of peptide backbone fragmentation and sequence information (Olsen 2004), the method included an additional data-dependent fragmentation step (MS3) when a fragment corresponding to the neutral loss of a phosphoric acid was among the 3 most abundant fragments in the MS/MS spectrum
Data Processing Protocol
The raw data of the acetylome dataset were processed using Proteome Discoverer (version 188.8.131.528, Thermo Scientific). The fragmentation spectra were searched with the Sequest HT engine against the UniProtKB/Swiss-Prot human database (canonical and isoform sequences, release 2016_05) using the following parameters: trypsin, maximum 3 missed cleavages, 20 ppm precursor mass tolerance, 0.6 Da fragment mass tolerance, cysteine carbamidomethylation as a fixed modification, and oxidation of methionine and acetylation of lysine as dynamic modifications. Peptide spectral matches were filtered for “search engine rank 1” and 1 % false discovery rate using Percolator (Kall 2007, Spivak 2009). The localization probability for each K-Ac site was calculated using ptmRS (version 1.4) (Taus 2011) and only K-Ac peptides with sites localized with a probability ≥75 % were annotated as modified peptides in the database. Because acetylation blocks tryptic cleavage (Zee 2012), K-Ac sites mapped to a peptide C-terminal lysine residue were excluded from the database except when the site coincided with the protein C-terminal residue. For quantitative analysis of co-modified peptides in CD3 proteins, peptide abundance was determined by a label-free approach using the Progenesis QI for proteomics software (v3.0, Nonlinear Dynamics). Peptides were identified with Proteome Discoverer as described above considering oxidation of methionine, GG addition to lysine, and phosphorylation of serine, tyrosine or threonine as dynamic modifications. Peptide-spectrum matches were filtered for “search engine rank 1” and 5 % false discovery rate using Percolator. The peptide identifications were imported and matched to the corresponding features across the different runs. PTM site probabilities derived from ptmRS were integrated in the Progenesis output file and data was further processed using custom scripts. Peptides with a K-ε-GG site mapped to a peptide C-terminal lysine were removed and only peptides with all sites localized with a probability ≥75 % were kept. MS3 spectra were interpreted manually for identification of co-modified peptides.
Corresponding dataset(s) in other omics resources
Casanovas A, Gallardo Ó, Carrascal M, Abian J. TCellXTalk facilitates the detection of co-modified peptides for the study of protein post-translational modification cross-talk in T cells. Bioinformatics. 2018 PubMed: 30219844