Key words

1 Introduction

The quantitative proteomics field has advanced tremendously during the past decade. This includes the method workflows, automation schemes, analysis regimens, and scope of applications. Improvements have been made in sample throughput and analytical metrics (e.g., precision, sensitivity) in MS-based quantitation [1,2,3]. The reproducibility of the methodologies and techniques has also been enhanced, as has its utility in interlaboratory studies [2, 4, 5]. These merits are collectively needed for the widespread assessment of protein disease biomarkers toward clinical diagnostic implementation [6].

While MS-based proteomic measurements are increasingly performed in a label-free manner [7, 8] (see Chaps. 8, 20–24), higher identification confidence and improved analytics can be obtained through the use of stable isotope-labeled protein or peptide standards (see Chaps. 8, 11). These standards can be inserted at a number of points of the analytical workflow and can be constructed in various formats with different isotope-labeling patterns (commonly 13C/15N, with a minimum mass shift of 6 Da from its unlabeled counterpart) [9]. The classical workflow is “bottom-up,” where the enzymatically cleaved peptides serve as molecular surrogates of the proteins of interest. The types of standard(s) used have included recombinant proteins [10, 11], tryptic peptides [12,13,14], winged (referred to also as extended, flanked, or cleavable) peptides [15, 16], concatenated peptides (abbreviated QconCATs for quantification concatemers) [17, 18], and protein epitope signature tags (PrESTs) [19,20,21]. In all cases, the labeled standards are designed to resemble the chemical structure of its endogenous (natural or NAT) analogue. Their quantitative effectiveness has been evaluated and compared/contrasted in different sample matrices (e.g., plasma, cerebrospinal fluid, urine), using both protein [22] and peptide [23] standards. These studies have demonstrated the utility of labeled standards, with the selection ultimately guided by experimental design and quantitative application.

Protein quantification with standards can be accomplished in a number of ways. While single-point measurements (using area ratios of unlabeled peptides, from a sample, to labeled peptides, from a spiked-in standard) are the easiest to perform, this method is not analytically accurate for large-scale analyses (due, for example, to plate-to-plate variation) and over wide dynamic ranges. Better quantitative methods utilize standard curves prepared in buffer or in a pooled control sample. These can be generated in a forward or reverse manner [24,25,26,27]. Forward curves involve a dilution series of NAT peptide concentrations (derived synthetically or from an endogenous sample) and constant SIS peptide concentrations, while reverse curves are the converse involving a dilution series of SIS peptide concentrations with constant NAT across the calibrant levels. There are advantages and disadvantages of curve-based quantitative strategies, as described and demonstrated previously [28]. One promising and applicable approach uses external standard addition (ExSTA), where forward standard curves are prepared in buffer and constant SIS peptide concentrations are spiked into the experimental samples. The peptide concentrations in the endogenous sample can then be determined by applying the experimental response ratios (i.e., NAT vs. spiked-in SIS) to the regression equation of their peptide-specific standard curves (plot of NAT/SIS peptide response vs. NAT peptide concentration ). This strategy has been demonstrated to be a robust means for precisely quantifying endogenous proteins in human plasma samples. Moreover, it is well suited for routine LC-MS/MS processing, large-scale proteomic applications, and sample-limited analyses.

Detailed here is a procedure for preparing samples (both controls and experimental), via the ExSTA approach, for robust, MS-based quantification of human plasma proteins without the use of upfront depletion, enrichment, or LC fractionation. The controls are generated in a BSA/PBS (bovine serum album/phosphate-buffered saline) surrogate matrix. These surrogate controls encompass standard-curve samples comprising eight concentration -level calibrants (spanning a 1000-fold range ) and three quality-control (QC) samples (at low, medium , and high levels). The mixtures utilized in these samples are NAT and SIS tryptic peptides, with the SIS mixes also implemented in the plasma sample analysis (n = 20). The peptides correspond to a panel of medium-to-high abundance human plasma proteins (125 in total, with one peptide per protein ) and were initially carefully selected according to a series of peptide/protein selection rules [29, 30]. The peptide mixtures and the described methods have been validated according to the Clinical Proteomic Tumor Analysis Consortium (CPTAC) guidelines [31, 32] and are verified to be fit-for-purpose for Tier 2, research-based analysis [33]. The assays utilize reversed-phase liquid chromatography (RPLC) in conjunction with tandem MS (MS/MS), performed in the selected/multiple reaction monitoring (SRM/MRM) or parallel reaction monitoring (PRM) acquisition mode. Performing the procedures described herein should produce precise quantitative results for putative biomarker analysis (at the discovery or verification stage) while opening the door for expanded target panels and sample sizes.

2 Materials

2.1 Sample and Solution Preparation Supplies

  1. 1.

    Analytical balance.

  2. 2.

    Adjustable pipettes and pipette tips.

  3. 3.

    Polypropylene Falcon tubes (15 and 50 mL) and screw caps (all from Fisher Scientific; Suwanee, GA, USA).

  4. 4.

    LoBind (i.e., low-binding) microcentrifuge tubes (1.5 mL; Eppendorf; Hauppauge, NY, USA).

  5. 5.

    Glass beaker (250 mL).

  6. 6.

    Glass graduated cylinder (100 mL).

  7. 7.

    Ultralow temperature freezer (capable of −80 °C).

  8. 8.

    Vortex mixer.

  9. 9.

    Minicentrifuge (capable of ≥1000 × g).

  10. 10.

    Benchtop centrifuge (compatible with 96-well plates).

  11. 11.

    Incubator.

  12. 12.

    Aluminum foil.

  13. 13.

    96-well deep well plates (1.1 mL) and accessories (e.g., silicone sealing mat, Axygen; Union City, CA, USA; AluminaSeal II™ sealing film; EXCEL Scientific; Victorville, CA, USA).

  14. 14.

    96-well Oasis HLB (i.e., hydrophilic-lipophilic balance) μElution plates (2 mg sorbent, 30 μm particles; part number 186001828BA; Waters; Milford, MA, USA).

  15. 15.

    Skirted 96-well (150 μL) PCR collection plates (Eppendorf).

  16. 16.

    Positive pressure vacuum manifold (compatible with 96-well plates, e.g., part number 186006961; Waters).

  17. 17.

    Lyophilizer (or SpeedVac) and accessories (e.g., wide-mouth borosilicate glass flasks).

  18. 18.

    Autosampler vials and screw caps.

2.2 Control and Experimental Sample Preparations

  1. 1.

    10 mg/mL BSA in PBS: In a 1.5 mL LoBind microcentrifuge tube, dissolve 10 mg BSA in 1 mL of PBS solution (see Note 1). The pH of this solution should be approximately 7.5.

  2. 2.

    Reagent A: 1 M Tris (pH 8.0). In a 250 mL glass beaker, dissolve 12.12 g of Tris in 90 mL of LC-MS grade water and adjust the pH to 8.0 with dropwise addition of 12 M HCl. Transfer the solution to a 100 mL glass graduated cylinder and bring the volume to 100 mL with LC-MS grade water (see Note 1). Vortex briefly.

  3. 3.

    Reagent B: 9 M urea in 300 mM Tris (pH 8.0). In a 15 mL screw cap polypropylene tube, dissolve 5.4 g of urea in 3 mL of Reagent A and 3 mL of LC-MS grade water (see Note 1). Vortex until fully solubilized (may take up to 5 min; see Note 2).

  4. 4.

    Reagent C: 9 M urea and 20 mM DL-dithiothreitol (DTT) in 300 mM Tris (pH 8.0). In a 1.5 mL LoBind microcentrifuge tube, dissolve 15.4 mg of DTT in 200 μL of Reagent B, to prepare a 500 mM DTT solution. Using the 500 mM DTT solution, dilute this 25× by transferring 40 μL to 960 μL of Reagent B (see Note 1). Vortex briefly.

  5. 5.

    Reagent D: 100 mM iodoacetamide (IAA). In a 1.5 mL LoBind microcentrifuge tube, dissolve 18.5 mg of IAA in 1 mL of LC-MS grade water. Vortex until completely solubilized, and then wrap the tube in aluminum foil to prevent deactivation by light (see Note 1).

  6. 6.

    Reagent E: 100 mM Tris (pH 8.0). In a 15 mL screw cap polypropylene tube, add 1 mL of Reagent A to 9 mL of LC-MS grade water (see Note 1). Vortex briefly.

  7. 7.

    1 mg/mL trypsin. In a 1.5 mL LoBind microcentrifuge tube, dissolve 1 mg of trypsin in 1 mL of Reagent E (see Note 1).

  8. 8.

    2% aqueous formic acid (FA) (v/v). In a 15-mL screw cap polypropylene tube, add 200 μL FA to 9.8 mL of LC-MS grade water.

  9. 9.

    0.1% aqueous FA (v/v). In a 50-mL screw cap polypropylene tube, add 50 μL of FA to 49.95 mL of LC-MS grade water. Vortex briefly.

  10. 10.

    0.1% FA in 30% LC-MS grade acetonitrile (ACN)/water (v/v/v). In a 1.5 mL LoBind microcentrifuge tube, add 3 μL of FA to 899 μL of LC-MS grade ACN and 2.098 mL of LC-MS grade water. Vortex briefly.

  11. 11.

    0.1% FA in 50% LC-MS grade ACN/water (v/v/v). In a 15 mL screw cap polypropylene tube, add 5 μL of FA to 2.497 mL of LC-MS grade ACN and 2.497 mL of LC-MS grade water. Vortex briefly.

  12. 12.

    NAT peptide mixture. Rehydrate a lyophilized NAT mix aliquot (see Note 3) in 60 μL of 0.1% FA/30% ACN (v/v). Vortex and centrifuge briefly, and then store on ice. This stock solution represents level H of the 8-point standard curve and is to be later diluted serially to produce the calibrant levels and three curve QC samples.

  13. 13.

    SIS peptide mixture. Rehydrate a lyophilized SIS mix aliquot (see Note 3) in 450 μL of 0.1% FA/30% ACN (v/v). Vortex and centrifuge briefly, and then store on ice. This creates a concentration-balanced solution (i.e., MS response signals have been tested to be within an order of magnitude of their unlabeled peptide analogues) for direct addition to experimental and control samples. For alternative SIS peptide configurations, refer to Note 4.

2.3 LC-MS Equipment

  1. 1.

    LC system: 1290 Infinity UHPLC (Agilent Technologies; Santa Clara, CA, USA).

  2. 2.

    Analytical column: Zorbax Eclipse Plus RRHD C18 RP-UHPLC (150 × 2.1 mm i.d., 1.8 μm particles; Agilent Technologies).

  3. 3.

    Eluent A: 0.1% aqueous FA in LC-MS grade water.

  4. 4.

    Eluent B: 0.1% FA in LC-MS grade ACN.

  5. 5.

    6495 QqQ (Agilent Technologies) with standard-flow ESI source.

  6. 6.

    Q Exactive™ Plus (Thermo Fisher Scientific) with standard-flow ESI source.

  7. 7.

    Mass spectrometer calibrant: ESI-tuning mix.

2.4 Data Collection and Analysis Software

  1. 1.

    MassHunter Quantitative Analysis (Agilent Technologies).

  2. 2.

    Xcalibur™ (Thermo Fisher Scientific).

  3. 3.

    Skyline-daily.

  4. 4.

    Microsoft Excel.

3 Methods

3.1 Digest Preparations (for Control and Experimental Samples)

  1. 1.

    Using a 96-well deep well plate, add 20 μL of Reagent C to the highlighted wells in Fig. 1a (red wells, A1-H1, A2-H2, and A3-D3 for the experimental samples; green wells, A11 and A12 for the control samples).

  2. 2.

    Add 10 μL of undiluted human plasma to the red wells (i.e., A1-H1, A2-H2, and A3-D3).

  3. 3.

    Add 10 μL of the BSA/PBS solution (at 10 mg/mL) to the green wells (i.e., A11 and A12).

  4. 4.

    Cover the plate with a silicone sealing mat, and then mix, centrifuge briefly, and incubate for 30 min at 37 °C. The concentrations of urea and DTT during the denaturation/reduction are 4.5 mM and 10 mM, respectively.

  5. 5.

    Following this incubation, alkylate the free sulfhydryl groups by adding 20 μL of Reagent D to each of the aforementioned wells (22 in total). The reaction concentration of IAA is 33 mM.

  6. 6.

    Cover the plate with a silicone sealing mat, and then mix, centrifuge briefly, and incubate for 30 min at ambient temperature (i.e., 20–25 °C) in the dark. During this incubation, prepare the trypsin solution as described in item 7 of Subheading 2.2.

  7. 7.

    After alkylation, prepare the protein solutions for digestion by adding 272 μL of Reagent E and 35 μL of trypsin (at 1 mg/mL for a 20:1 w/w substrate/enzyme ratio; see Note 5) to the 22 aforementioned wells. The reagent concentrations at digestion are 0.5 M urea (>1 M inhibits trypsin activity), 1 mM DTT, and 6 mM IAA in 100 mM Tris (pH 8.0).

  8. 8.

    Cover the plate with a silicone sealing mat, and then mix, centrifuge briefly, spin-down, and incubate for 18 h at 37 °C. Prior to the completion of this incubation period (e.g., 2 h beforehand), we recommend to prepare the NAT/SIS peptide standard solutions outlined in Subheading 2.2.

  9. 9.

    After the 18-h incubation, quench the proteolytic digestion by adding 343 μL of 2% FA. This brings the total volume of each well to 700 μL.

  10. 10.

    Cover the plate with a silicone sealing mat and then mix and centrifuge briefly.

  11. 11.

    Combine the two BSA/PBS digests into a 1.5 mL LoBind microcentrifuge tube. This solution will be used in the control sample preparations (for the calibration curve and the QC samples; Subheading 3.3).

  12. 12.

    Store the plate (containing the plasma samples) and microcentrifuge tube (containing the surrogate matrix) on ice until standard spiking (Subheading 3.3).

Fig. 1
figure 1

Map of 96-well plate illustrating the suggested placement of the control and experimental samples. Part (a) is for the digest preparations and (b) for the standard spiking to the digests. In both cases, the green/blue wells are for the control samples and red wells for the experimental samples (n = 20 human plasma). Regarding the controls in (b), the green wells are for the calibration curve sample digests (A, or low NAT peptide concentration, to H, or high), and the blue wells are for the curve QC digests (low, QC-A; medium , QC-B; and high, QC-C)

Full size image

3.2 Standard Solution Preparations (for the Control Samples)

  1. 1.

    Label seven 1.5 mL LoBind microcentrifuge tubes from Standard G to A. These form a 7-concentration dilution series for the 8-point standard curve , with the highest concentration level being Standard H. This standard reflects the undiluted, concentrated NAT stock that was prepared earlier (in item 12 of Subheading 2.2).

  2. 2.

    To the Standard G microcentrifuge tube, add 14 μL of Standard H to 21 μL of 0.1% FA/30% ACN. Vortex and centrifuge briefly, and then store on ice. Subsequent concentration levels (i.e., Standards F to A) are to be prepared as outlined in Table 1.

  3. 3.

    For the curve-based QC samples, label five LoBind microcentrifuge tubes as follows: QC-C, ID-1, QC-B, ID-2, and QC-A. The IDs reflect intermediate dilutions and are used only in these QC-sample preparations.

  4. 4.

    Prepare QC-C by adding 18 μL of Standard H to 42 μL of 0.1% FA/30% ACN. Vortex and centrifuge briefly, and then store on ice.

  5. 5.

    The subsequent ID (ID-1 and ID-2) and QC (QC-B and QC-A) samples are to be prepared as outlined in Table 2. The ID samples can be discarded following the QC-B and QC-A preparations.

Table 1 Serial dilution of the NAT peptide mix toward an 8-point standard curve (1000-fold range)
Full size table
Table 2 Preparation of three curve QC samples (high, QC-C; medium , QC-B; low, QC-A)
Full size table

3.3 Standard Spiking to Control/Experimental Digests

The steps outlined in this section are to be conducted in a fresh 96-well deep well plate.

  1. 1.

    To prepare the curve samples (comprises eight concentration levels, 1:2.5:2.5:4:2:4:2:2.5 dilutions from standard H to A), add the following solutions in succession to wells A12–H12 (see green wells in Fig. 1b):

    1. (a)

      40 μL of the BSA/PBS digest (from step 12 of Subheading 3.1)

    2. (b)

      10 μL of the SIS peptide mixture (rehydrated earlier in item 13 of Subheading 2.2)

    3. (c)

      10 μL of a NAT standard mix (i.e., add standard A to well A12, down to standard H to well H12; prepared in step 2 of Subheading 3.2)

    4. (d)

      540 μL of 0.1% aqueous FA.

  2. 2.

    To prepare the curve QC samples, add the following solutions in succession to wells H10 and A11 to H11 (see blue wells in Fig. 1b):

    1. (a)

      40 μL of the BSA/PBS digest (from step 12 of Subheading 3.1)

    2. (b)

      10 μL of the rehydrated SIS peptide mixture

    3. (c)

      10 μL of a NAT QC mix (QC-A to wells H10, A11, B11; QC-B to C11, D11, E11; QC-C to F11, G11, H11; prepared in step 4 of Subheading 3.2)

    4. (d)

      540 μL of 0.1% aqueous FA.

  3. 3.

    To prepare the experimental plasma samples (n = 20), add the following solutions in succession to wells A1–H1, A2–H2, and A3–D3 (see red wells in Fig. 1b and Note 6):

    1. (a)

      40 μL of a human plasma digest (from step 12 of Subheading 3.1)

    2. (b)

      10 μL of the rehydrated SIS peptide mixture

    3. (c)

      10 μL of 0.1% FA/30% aqueous ACN

    4. (d)

      540 μL of 0.1% aqueous FA.

  4. 4.

    Cover the plate with a sealing film (AlumaSeal II™) and then mix and centrifuge briefly.

  5. 5.

    Store on ice until performing sample cleanup by solid-phase extraction .

3.4 Extraction and Reconstitution for LC-MS/MS

  1. 1.

    Desalt and concentrate the peptide solutions by solid-phase extraction using a positive pressure vacuum manifold on a 96-well Oasis HLB μElution plate. The extractions are to take place in 37 wells (see Fig. 1b for example locations) and involve the following sequential steps (see Note 7):

    1. (a)

      Wash with 600 μL of LC-MS grade methanol.

    2. (b)

      Condition with 600 μL of 0.1% aqueous FA.

    3. (c)

      Load with 510 μL of each sample (i.e., standard curve samples A–H, QC samples A–C, plasma samples 1–20).

    4. (d)

      Wash with 600 μL of LC-MS grade water or 0.1% FA.

    5. (e)

      Elute with 75 μL of 0.1% FA/50% ACN into a skirted 96-well microplate.

  2. 2.

    Cover the plate of eluate with a sealing film (AluminaSeal II™) and centrifuge briefly.

  3. 3.

    Puncture the sealing film with a hole (using an 18-gauge needle) at the designated well positions (37 in total), and then cover with parafilm.

  4. 4.

    Freeze the plate and lyophilizer container at −80 °C.

  5. 5.

    Once frozen, remove the parafilm, and then dry down (with lyophilizer) to completion overnight.

  6. 6.

    Rehydrate the 37 samples in 34 μL of 0.1% FA (this will give a final concentration of 1 μg/μL based on an initial plasma protein concentration of 70 mg/mL).

  7. 7.

    Cover the plate with a sealing film, mix, and centrifuge briefly.

  8. 8.

    Remove the sealing film and replace with a silicone sealing mat.

  9. 9.

    Place the sample plate and an autosampler vial containing ~200 μL eluent A (for blank injections) into the LC autosampler (see Note 8).

  10. 10.

    Prepare for the LC-MS/MS analysis (Subheading 3.5).

Fig. 2
figure 2

Peptide concentrations to be entered in Skyline’s “Document Grid: Replicates.” These fmol/μL values reflect the dilution series of this methods protocol, which comprises an 8-point standard curve (1000-fold range) and 3-point QCs (labeled -A to -C). No concentrations are to be entered for the experimental samples

Full size image

3.5 LC System Setup

  1. 1.

    Thermostat the column and autosampler compartments to 50 and 4 °C, respectively.

  2. 2.

    Set up the autosampler for acquisition at a 20 μL/min draw, 40 μL/min eject, and vial/well bottom sensing activated.

  3. 3.

    To facilitate the RPLC peptide separations, use mobile phase compositions of 0.1% aqueous FA in water for A and 0.1% aqueous FA in ACN for B.

  4. 4.

    Use the following ACN gradient (time in min, eluent B composition in %): 0, 2; 2, 7; 50, 30; 53, 45; 53.5, 80; 55.5, 80; 56, 2.

  5. 5.

    To help preserve instrument cleanliness, divert the solvent stream (2% eluent A) to waste at the conclusion of the ACN gradient (at 56 min), and then hold for 4 min to allow column equilibration.

  6. 6.

    Use a 0.4 mL/min flow rate.

  7. 7.

    Use 10 μL injection volumes for all sample types.

3.6 MRM/MS System Setup

  1. 1.

    Operate the ESI source in the positive ion mode with capillary voltage at 3.5 kV.

  2. 2.

    Use the following general parameters: 300 V nozzle voltage, 11 L/min sheath gas at 250 °C, 15 L/min drying gas at 150 °C, 30 psi nebulizer, and iFunnel RF pressures of 200 V (high) and 110 V (low). Ultrahigh-purity nitrogen serves as the carrier gas in all settings.

  3. 3.

    Using the MRM mode (for the 6495 QqQ, see Note 9), enter the specific acquisition parameters. This encompasses the precursor/product ion m/z values, retention times (RTs), and collision energies (see Table 3).

  4. 4.

    Use the following general MRM parameters for all transitions in the final method: unit mass resolution (for first and third quadrupole), 1.5 min RT windows, 380 V fragmentor voltage, 5 V cell accelerator potential, and 10 ms dwell times. The total cycle time for this acquisition method run on this LC-MS platform is 700 ms.

Table 3 Specific acquisition parameters for the bottom-up LC-MRM/MS analysis
Full size table

3.7 PRM/MS System Setup

  1. 1.

    Operate the ESI source in the positive ion mode.

  2. 2.

    Use the following general parameters: 350 °C capillary temperature, 50 L/min sheath gas, 20 L/min auxiliary gas at 350 °C, 0 sweep gas, 3 kV spray voltage, and S-lens RF level at 50. Ultrahigh-purity nitrogen serves as the carrier gas in all settings.

  3. 3.

    Additional PRM parameters include 17.5 k resolution, 2e5 automatic gain control (AGC) target, 60 ms maximum injection time (IT), and 30 loop count.

  4. 4.

    Using the PRM mode, enter the specific acquisition parameters. This comprises an inclusion list consisting of peptide precursor m/z values and RTs (inserted as “Start” and “End” times over 1.5 min window; see Table 3).

3.8 LC-MS/MS Platform Performance Test

  1. 1.

    Purge the pumps with 50% mobile phase B for 5 min at 10 mL/min.

  2. 2.

    Once the UHPLC system is re-equilibrated to the starting conditions (i.e., 2% eluent B), run a solvent blank (at 0.1% FA) to flush the 20 μL loop and connecting tubing.

  3. 3.

    Using the ESI tuning mix, run the automated tuning program to tune or realign a set of m/z values for positive ESI-MS/MS analysis.

  4. 4.

    Following a successful instrument tune, confirm the peptide RTs by injecting 10 μL of a calibrant standard (see Note 10, for scheduling, and Note 11, for data analysis ).

  5. 5.

    Input the updated RTs into the acquisition method and use 1.5 min in all target acquisition windows.

3.9 Sample Injection

  1. 1.

    After the performance tests (for LC RTs and mass analyzer tuning, see Subheading 3.7) have been successful, a sample batch can be queued for analysis. Worklists vary, but in our practice, the first sample to be injected (all at 10 μL) is the first replicate of the curve QCs (order: A in well H10, B in C11, then C in F11).

  2. 2.

    Inject two solvent blanks after the three QC samples (replicate 1), and then inject a single replicate of the standard curve samples from low concentration (Standard A in well A12) to high concentration (Standard H in H12).

  3. 3.

    Inject two solvent blanks after the calibration-curve samples; then inject the second replicate of the curve QC samples (in the order: A in A11, B in D11, then C in G11).

  4. 4.

    Inject two solvent blanks after the three QC samples (replicate 2), and then inject the experimental samples (n = 20), individually, from well A1 to D3.

  5. 5.

    Inject two solvent blanks after the experimental samples, and then inject the final replicate of the curve QCs (order: A in B11, B in E11, C in H11).

  6. 6.

    After the completion of the sample batch, set the LC to an isocratic flow of 0.02 mL/min with 50% eluent B. Divert the solvent stream to waste to help preserve the MS ion source until the next worklist is initiated.

3.10 Data Analysis

Although the analysis of the MRM or PRM data can be performed with several vendors’ software programs, for the purposes here, a guide for executing this in Skyline-daily is outlined (see [34] for login and installation instructions).

  1. 1.

    Starting with an empty Skyline document set in the “Proteomics Interface,” build the MRM or PRM target list. This can be inserted or copied directly from Excel, provided that the column names and order are the same as in the “Skyline Transition List” window. After pasting, click on “Check for Errors,” and then save the populated Skyline Window.

  2. 2.

    In the main screen, import the raw data by clicking on “File,” “Import,” and then “Results.” In the “Import Results” window, select “Add single-injection replicates in files” and enable “Show chromatograms during import.” After selecting “OK,” locate and highlight the appropriate data files (i.e., .d for Agilent, .raw for Thermo) to be imported.

  3. 3.

    In the document grid, select “Views” and “Replicates.” In that tab, input the “Analyte Concentration” values for the curve calibrants and QC samples, as illustrated in Fig. 2.

  4. 4.

    In the main screen, select “View,” “Document Grid,” and then “Peptide Quantification.” Enter the peptide-specific concentration multipliers, as outlined in Table 4 (see Note 12).

  5. 5.

    In the “Quantification” tab, adjust the peptide settings to allow the quantification of the experimental samples by linear regression using the NAT/SIS peptide ratios with a regression weighting of 1/x2 (x = concentration).

  6. 6.

    Manually inspect the extracted ion chromatograms (XICs) and calibration curves for the control samples to ensure correct peak selection and accurate integration (see Fig. 3 and Note 13 for projected observations).

  7. 7.

    Review the accuracy of the control-based samples, as calculated automatically by Skyline, against the batch acceptance criteria (see Note 14).

  8. 8.

    Using the control XICs as reference, manually inspect the experimental sample XICs (e.g., for peak symmetry, retention time, interference) and its relation to the curve (see Fig. 4).

  9. 9.

    Export the collective Skyline results for further review and analysis . Data extraction pertains to the peaks (e.g., SIS and NAT responses, retention times) and quantified peptides (see Note 15).

Table 4 Target details for the quantification calculations
Full size table
Fig. 3
figure 3

Theoretical plot of control and experimental sample analysis by bottom-up LC-MS/MS. (a) XICs of expected NAT/SIS peptide ratios (NAT in red, SIS in blue) in a buffer-generated, standard curve . (b) Placement of 8-point calibrant levels (in green diamonds) and 3-point QCs (in blue diamonds) relative to a plasma sample measurement. Quantitation of this sample is then accomplished through linear regression analysis of the peptide relative response (i.e., NAT/SIS peak area ratio)

Full size image
Fig. 4
figure 4

Bottom-up LC-MRM/MS data for two representative, moderate abundance targets from the 125 protein panel. Illustrated are two Skyline screenshots for peptides ATVVYQGER (from β-2-glycoprotein 1) in (a) and VGSALFLSHNLK (from kallistatin) in (b). The tabs refer to the monitored panel, two example XICs (for curve calibrant E and an experimental sample), the standard curve (with calibrant levels marked with gray squares, QC samples with green diamonds, experimental sample with red arrow), and the results grid. The concentrations in the human plasma sample (Bioreclamation, lot BRH1447352) shown were calculated to be 72.7 μg/mL for β-2-glycoprotein 1 and 6.9 μg/mL for kallistatin

Full size image

4 Notes

  1. 1.

    The stability of the prepared reagents (i.e., Reagents A to E, BSA/PBS, and trypsin) is predicated based on their time of preparation and storage conditions. Reagent A (stock Tris solution) can be stored at 4 °C for 1 month. The remaining reagents noted above are to be prepared fresh, on day of use. Reagent C (comprising urea and DTT in Tris), Reagent D (IAA), and trypsin are to be prepared immediately prior to use and stored on ice until dispensed.

  2. 2.

    Do not heat the urea solution in an effort to hasten its solubilization. Elevated temperature (i.e., >37 °C) accelerates urea decomposition, with its product (isocyanic acid) resulting in carbamylation at ɛ-amine lysine residues and protein N-termini [35]. To reduce this artifact, dissolve the urea at room temperature with the aid of vortexing only.

  3. 3.

    The NAT and SIS peptides were synthesized (at the University of Victoria-Genome BC Proteomics Centre; Victoria, BC, Canada) on a robotic peptide synthesizer using Fmoc chemistry. The labeled protected amino acids (13C6/15N2 l-lysine, CNLM-4754-H; 13C6/15N4 l-arginine, CNLM-8474-H), used in the C-terminal labeling of tryptic peptide residues, were obtained from Cambridge Isotope Laboratories, Inc. (Andover, MA, USA). After synthesis, the standard peptides were purified (by HPLC) and characterized (for composition and concentration) and then formulated into mixtures, in a manner similar to that described previously [36]. Lyophilized aliquots of the mixtures are stored at −80 °C, with established stabilities of at least 6 months. Overall, the peptides selected are proteotypic and abide by a series of sequence-specific selection rules (e.g., unique within plasma proteome, devoid of cysteine and methionine, 6–20 residues in length), which makes them well suited to serve as external or internal standards.

  4. 4.

    If an alternative SIS peptide labeling scheme is desired, as would be the case with double isotopologue peptide standards (i.e., isotope labeling at a C-terminal amino acid as well as an internal residue), the protocol described herein would need to be amended. The methods article by Eshghi et al. serves as a useful reference guide for the quantitative application of the double SIS peptide approach to different sample types [37]. Regardless of the type of peptide standard , these should be prepared, stored, and handled according to established guidelines for best practices in MS-based proteomics [38].

  5. 5.

    The ideal quantity of trypsin added was previously determined in empirical measurements. The 20:1 (w/w) substrate/enzyme ratio implemented enables efficient tryptic digestion in pooled human plasma samples. Appropriate substrate/enzyme ratios for alternative sample types will need to be determined before application to real experimental samples.

  6. 6.

    The current protocol is amenable to be scaled up to 30 samples. For larger-scale analysis, the procedure will need to be adapted. This affects the quantity of surrogate digest initially prepared (in Subheading 3.1) and the quantity of SIS peptide mixture to be reconstituted (in Subheading 2.3).

  7. 7.

    For all extraction steps, add solvent or sample before applying vacuum. The flow through and washes (extraction steps in Subheading 3.4) are to be collected in an empty basin and can be discarded. In the sample loading step, start at the lowest vacuum setting and then gradually increase the vacuum to load the full sample onto the sorbent bed. Flow rates for sample loading and elution should not exceed 1 mL/min. Also, for improved throughput and precision, all SPE steps should be performed with a multichannel pipette or with a liquid handling robot (e.g., Tecan Freedom EVO 150).

  8. 8.

    Ensure that the sample container of the LC autosampler is configured to read a 96-well plate in one tray (for the digests) and vials in a second tray (with one designated for the solvent blank, e.g., at position A1). This latter tray will need to be loaded first with a vial plate to accommodate the autosampler vial(s).

  9. 9.

    The specific acquisition parameters (i.e., transitions, collision energies, and retention times) will need to be optimized if alternative mass spectrometers are used in place of the 6495 or Q Exactive instrument. It is recommended that these be selected based on empirical measurements using the SIS peptide mixtures as MS infusion solutions.

  10. 10.

    Although we have observed minimal RT shifts in method transfer between laboratories when identical LC-MS conditions/parameters are employed, it is recommended that a broader detection window of 3 min be initially used for peptide RT verification. This test should require one run only, with standard E being an example standard for injection. After the peptide rescheduling, confirm that the cycle/dwell times enable 10–15 data points across the chromatographic peaks for optimum and accurate peak shapes. It must be noted that although the RT verification is conducted in buffer, in our experience, minimal shifts (maximally 0.5 min) are typically observed when the assay is transferred to human plasma.

  11. 11.

    Analysis of the peptide scheduling data can be accomplished with vendor software, Skyline [39], or other (e.g., Qualis-SIS [40]). If Skyline is used, as advocated here, refer to Skyline’s tutorial section for guidance on method refinement for efficient acquisitions. After the retention times have been confirmed for all targets, export the data file to Excel and then update the acquisition method accordingly.

  12. 12.

    The analyte concentration refers to the initial peptide concentration in the standard mixture and is used as a basis for the quantitation calculations. The concentration multipliers are used to automatically convert the calculated values to protein concentrations (expressed in fmol/μL of human plasma).

  13. 13.

    The NAT (synthetic and endogenous) and SIS peptides have identical physicochemical properties and therefore exhibit the same behavior during extraction , separation , and mass analysis (e.g., ionization, fragmentation). The resulting XICs for each peptide pair should therefore co-elute in all sample analyses, with the responses of the NAT standards differing from SIS in the curves and QC samples only (see Fig. 3a). Regarding the ratios, the calibrant levels should approximate a 1:2.5:2.5:4:2:4:2:2.5 dilution from standard H to A (e.g., standard E is 25-fold diluted from standard H), with the QC samples being x-fold from standard H (i.e., 3.33-fold, in QC-C; 33.3-fold, in QC-B; 250-fold, in QC-A; see Fig. 3b). The relative responses and ratios should be confirmed for all sample types.

  14. 14.

    The accuracies of the standard curve and curve QC samples are reported in Skyline as a percent theoretical value. For each peptide , each measured NAT standard in the curve should be within 20% of the theoretical concentration, and at least 6 of the 9 measured NAT concentrations in the QC samples should be within 20% of the theoretical concentrations. In addition, for each peptide, at least 1 of 3 of the QC samples (QC-A, QC-B, or QC-C) should be within 20% of its theoretical concentration. Finally, for each peptide quantified, at least 6 of the 8 measured in the standard curve should be within 20% of the expected concentration.

  15. 15.

    To convert the peptide concentrations (in fmol/μL) to protein concentrations (in ng/mL), multiply the calculated concentrations from Skyline by the molecular weight of a given protein (see Table 4) and then divide by 1000. For example, the concentration of ATVVYQGER in a de-identified human plasma sample (from Bioreclamation, lot BRH1447352) was determined to be 1897.5 fmol/μL, which equates to 72.6 μg/mL for β-2-glycoprotein 1 (also referred to as apolipoprotein H or Apo-H).