Kockmann - MSQC1

Targeted proteomics coming of age - SRM, PRM and DIA performance evaluated from a core facility perspective

  • Organism: Homo sapiens
  • Instrument: TSQ Vantage, QTRAP5500, Q Exactive, Q Exactive HF and tripleTOF5600
  • SpikeIn: Yes
Abstract
Quantitative mass spectrometry is a rapidly evolving methodology applied in a large number of -omics type research projects. During the past years, new designs of mass spectrometers have been developed and launched as commercial systems while in parallel new data acquisition schemes and data analysis paradigms have been introduced. These advances allow in depth analysis of complex proteomes with high quantitative accuracy. Despite or even due to this fast evolution, life scientists aiming to exploit the full potential of proteomics are facing increasing confusion in identifying the most suitable method for their research question. Core facilities provide access to technologies, data, and interpretation, but also actively support the researchers in finding and applying the best-suited analytical approach. In order to dispose of a solid fundament for this decision making process, facilities need to constantly compare and benchmark the various approaches. In this article we compare the quantitative accuracy and precision of current state of the art targeted proteomics approaches (SRM, PRM and DIA) across multiple liquid chromatography mass spectrometry (LC-MS) platforms, using a readily available commercial standard sample.
Experiment Description
LC separation conditions Desalted peptides were loaded onto in-house made columns (fused silica capillary, 150 mm × 75 μm ID, 375 µm OD) either packed with ReproSil-Pur C18 AQ 1.9 μm (Dr. Maisch HPLC GmbH, Ammerbuch-Entringen) or Magic C18 AQ 3 µm, 200 A, resin, (Michrom BioResources, Auburn, CA) and either coupled to an Eksigent nanoLC-Ultra 1D plus (ABSciex, Zug Switzerland), or to an EASY-nLC 1000 (ThermoScientific, Wohlen, Switzerland). The columns were fitted to a spray tip (360 µm OD, 20 µm ID, and 10 µm ID at tip, PicoTip™ emitter, New Objective, MS Wil GmbH, Wil, Switzerland). Peptides were separated using platform-specific linear gradients mixed from a binary solvent system (solvent A: ddH2O, 0.1% FA; solvent B: ACN, 0.1% FA , Biosolve, Valkenswaard, The Netherlands). A full description of the LC setups can be found in supplementary data file 1. MS settings Qtrap 5500 (ABSciex, Concord, Canada) and TSQ Vantage (ThermoScientific, Bremen, Germany) were run in unscheduled SRM mode using a fixed dwell time of 20 ms. Transitions corresponding to the 14 L:H peptide pairs are listed in the MSQC1 product information (obtainable at www.sigma-aldrich.com). The resulting cycle time of the methods was 1.7 seconds. Q1 and Q3 were operated at unit resolution (0.7 Da). Collision energies (CE) were calculated according to the following equations: CE = 0.034 * (m/z) + 3.314 and CE = 0.044 * (m/z) + 3.314 for doubly and triply charged precursor ions, respectively. The declustering potential (DP) was set to 80. Entrance (EP) and collision cell exit potential (CXP) values were fixed at 10 and 12, respectively. These instrument parameters resemble the ones used by Sigma-Aldrich to determine L:H as reported in the Certificate of Analysis (obtainable at www.sigma-aldrich.com). The Q-Exactive (ThermoScientific, Bremen, Germany) was operated in scheduled PRM mode. The acquisition method combined a single full MS scan and 28 targeted MS2 scans corresponding to the 14 heavy, light peptide pairs. Full MS scan covered the m/z range of 350-1000 at a resolution of 17’500 (AGC target: 3e6, maximum IT: 50 ms). Each targeted MS2 scan was performed at a resolution of 70’000 (AGC target: 1e5, maximum IT: 120 ms) during the 4 min window around the expected RT. The cycle time for this method was 3 seconds. Precursors were isolated using a selectivity of 2 m/z and fragmented using a NCE of 25. The TripleTOF 5600 (ABSciex, Concord, Canada) was operated in DIA mode. Each instrument cycle consisted of a full MS scan (scan range: 350-1250 m/z, accumulation time: 250 ms) and a set of 32 product ion scans covering the mass range of 400-1200 m/z in 25 Da windows (1 Da overlap). An accumulation time of 100 ms was used to record product ion scans, resulting in a cycle time of 3.4 seconds. The collision energy for each window was determined based on the appropriate collision energy for a 2+ ion centered in the respective window with a spread of +/- 15. The Q Exactive HF (ThermoScientific, Bremen, Germany) was operated in DIA mode. Data independent scans covering a mass range from 400-1200 m/z in 32x 25 Da windows were acquired at a resolution of 30’000 (AGC target: 3e6, max. IT: auto, fixed first mass: 120 m/z, fixed NCE: 28). Each instrument cycle (3 seconds) was completed by a full MS scan monitoring 380-1260 m/z at a resolution of 30’000 (AGC target: 3e6, max. IT: 45 ms). Data dependent analysis (DDA) of pure MSQC1 was conducted using the following instrument settings: On the TripleTOF 5600 precursor scans covering 385-1250 m/z were acquired at 200 ms accumulation time. The most abundant (top35) precursors (charge state +2-+5 above 150 cps) were selected for product ion spectra recording from 200-1800 m/z at 70 ms accumulation time (TOF in high resolution mode) leading to a cycle time of 2.7 seconds. CE was calculated as stated above for the DIA measurements. Precursors were excluded for 20 s after the third occurrence. On the Q Exactive HF the following DDA method was applied: Full MS scans covering 300-1700 m/z were conducted at a resolution of 60000 (AGC target: 3e6, max IT: 15 ms). The (top12) most intense ions above an intensity threshold of 4.2e4 were selected for ddMS2 scans covering 200-2000 m/z (AGC target: 1e5, max IT: 119 ms). The isolation window was set to 1.2 m/z and ions were fragmentation at a NCE of 28. Dynamic exclusion of precursors was set to 10 s. The cycle time of this method was 2.2 seconds. Spectral libraries were generated by matching resulting MS2 spectra against the canonical human reference proteome (derived from Uniprot) using the Mascot search engine [22] with the following parameters: tryptic enzyme specificity allowing no missed cleavages, fixed carbamidomethyl on cysteine, 20 ppm precursor tolerance, 0.5 Da fragment ion tolerance, target+decoy search. Peptides passing the 2% peptide FDR based on target+decoy statistics were incorporated into a BiblioSpec library [23]. For signal response measurements in DIA/SWATH-MS mode, we selected all library peptides between 6-25 aa matched to the uniprot entries listed in Supp. Table 1. The top5 fragment ions, from the y- or b-ion series from the 3rd to the last-1 ion of charge state +1 and +2 were used for the initial targeted data extraction. Based on the results, only the top3 fragment ions were kept for final peptide quantification .
Sample Description
To obtain the MSQC1 stock solution, MS Qual/Quant QC Mix (Catalog #: MSQC1, Lot # 081M6281) was purchased from Sigma-Aldrich (Buchs, Switzerland) and reconstituted in 20 µl of 20% ACN, 0.1% formic acid (FA) and stored as 10 µl aliquots at -20°C. MSQC 1 working solution was prepared by diluting 10 µl stock solution with 40µl of 0.1% FA. To derive a complex sample matrix, 10 µl of MSQC1 working solution was mixed with 10µl (≈10 µg) tryptic yeast digest (Saccharomyces cerevisiae), 1µl iRT peptide stock solution (Biognosys, Schlieren, Switzerland), and 19 µl 0.1% FA. This sample is denoted as standard sample throughout the manuscript. Additionally, we prepared a six point dilution series containing relative MSQC1 amounts of 0.025, 0.05, 0.2, 1, 2, and 5 with respect to the standard sample. The amount of yeast digest and iRT peptides was kept constant across all dilution steps. The reference L:H ratio vs. the on column amount of SIL peptide can be found in Supp. Figure 1. For all injections with complex background the total amount of peptides on column was ~1µg.
Created on 3/7/16 11:59 AM

Raw instrument data is accessible for registered users through http://fgcz-bfabric.uzh.ch/bfabric as project 1959.

The R code and a condensed data set used for further statistical analysis is available through mscq1.

Tobias Kockmann, Ph.D.
Functional Genomics Center Zurich
Winterthurerstr. 190
CH-8057 Zurich
Switzerland

http://www.fgcz.ch/the-center/people/kockmann.html

Clustergrammer Heatmap