Multiplexed targeted mass spectrometry-based assays for the. quantification of N-linked glycosite-containing peptides in serum

Transcription

1 Supplemental Material Multiplexed targeted mass spectrometry-based assays for the quantification of N-linked glycosite-containing peptides in serum Stefani N. Thomas 1*, Robert Harlan 1,2, Jing Chen 1, Paul Aiyetan 1, Yansheng Liu 3, Lori J. Sokoll 1, Ruedi Aebersold 3,4, Daniel W. Chan 1, and Hui Zhang 1 1 Department of Pathology, Clinical Chemistry Division, Johns Hopkins University School of Medicine, Baltimore, MD 2 Current address: Center for Resources in Integrative Biology, Johns Hopkins University Schools of Medicine and Public Health, Baltimore, MD 3 Department of Biology, Institute of Molecular Systems Biology, ETH Zurich, 8093 Zurich, Switzerland 4 Faculty of Science, University of Zurich, 8057 Zurich, Switzerland These authors contributed equally to the study. *Corresponding Author Materials and Methods Page S-2 Supplemental References Page S-6 Supplemental Table Legends Page S-6 Supplemental Table 3 Page S-9 Supplemental Figures 1-2 Page S-18 S-1

2 Materials and methods Materials. Affi-gel hydrazide resin and sodium periodate were from Bio-Rad (Hercules, CA). Iodoacetamide, sodium acetate, sodium chloride, aniline, ammonium bicarbonate, human serum and LC-MS Ultra-grade formic acid were from Sigma-Aldrich (St. Louis, MO). Urea, tris(2- carboxyethyl)phosphine, 7 kda MWCO Zeba spin desalting columns, disposable automation research pipette tips (DART), trifluoroacetic acid, Optima LC/MS-grade water and Optima LC/MS-grade acetonitrile were from Thermo Fisher Scientific (Waltham, MA). Frits for the DART tips (2 mm diameter, 1 mm thick, µm pores) were from POREX (Fairburn, GA). Polysulfoethyl A TopTips µL for SCX chromatography were from Glygen (Columbia, MD). Sequencing-grade trypsin was from Promega (Madison, WI), and PNGase F was from New England Biolabs (Ipswich, MA). Assay characterization: Response curves and repeatability experiments. Commercially available human serum (Sigma-Aldrich) was used to prepare the background matrix consisting of tryptic N-glycopeptides. The background matrix was used for the reverse response curves (background matrix spiked with SIS at variable concentrations) and for the preparation of the samples for the mini-validation of repeatability experiments. Given the limitation of a maximum yield of 20 µg of glycopeptides per preparation of 40 µl of serum, N-glycopeptides prepared using an automated format of the SPEG method 1 with multiple DART pipette tips packed with hydrazide resin were combined to achieve the required amount of N-glycopeptides for the response curves. Reverse response curves were generated for each peptide to determine the linear range of its corresponding assay. The biological matrix (serum) was the source of background analytes and the light (endogenous) peptides. An advantage of reverse vs. normal curve S-2

3 quantification is that there is minimal contribution to the heavy peptide signal from endogenous analytes. Reverse response curves have been demonstrated to be suitable for the complete assay characterization required for biomarker candidates in complex matrices. 2-4 A 7-point concentration curve was established (0.0576, 0.288, 1.44, 7.2, 36, 180 and 900 pmol on column) using SIS. Blanks not containing SIS were also prepared. The final preparation of each sample contained background matrix, SIS and irt peptides. The matrix was diluted to 0.25 µg/µl with 2% ACN/0.2% formic acid, and a dilution mix was prepared containing µl matrix (0.25 µg/µl), 110 µl irt standard diluted 1:10 and µl 0.1 % formic acid (final concentration of matrix in dilution mix was µg/µl and final dilution of irt was 1:100). One µg of background matrix peptides was used per injection (6 µl injection volume). The dilution mix was also used as the blank (no SIS added). The sample for the highest point on the curve was prepared by taking 84 µl of SIS (660 pmol/µl) and drying in a vacuum centrifuge followed by resuspension in µl of dilution mix for a final SIS stock concentration of 150 pmol/µl. Beginning with the highest point on the curve (150 pmol/µl), a 1:5 dilution was prepared by adding 14 µl of the 150 pmol/µl solution to 56 µl of the dilution matrix. Six 1:5 serial dilutions were then prepared by adding 14 µl of the previously diluted sample to 56 µl of the dilution matrix. The response curves were run in triplicate in the following order: three blanks, one iteration of the curve from the lowest to the highest concentration point, two blanks, and two washes to minimize carryover. A mini-validation of repeatability was conducted to provide data on assay reproducibility. The SIS peptides were prepared at three levels (Low: pmol/µl, Medium: 7.5 pmol/µl and High 12 pmol/µl) and spiked into the background matrix. Analysis was conducted in triplicate across five days. To avoid artificially minimizing variability, the run order was randomized. To S-3

4 minimize carryover, one wash was inserted after the Low and Medium samples, and two washes were inserted after the High samples. Assay characterization data and standard operating procedures for assays with acceptable performance metrics are publicly accessible on the CPTAC Assay Portal ( and can be located by searching for the protein or peptide of interest based on the protein name, protein UniProt Accession ID, or the peptide sequence. MRMPlus. MRMPlus was developed in our laboratory to compute QC metrics for targeted mass spectrometry-based assays. MRMPlus was implemented in the platform-independent Java programing language (SDK 6). MRMPlus takes as input Skyline-derived transition output files in a tab-delimited format. Other inputs include user-defined program parameters inputted via MRMPlus graphical user interface (GUI) and experiment-associated metadata information (Run Order, Replicate Name, Analyte Type Blank or Serial Dilution, Replicate Number, and Calibration Point) in tab-delimited files. MRMPlus computes QC measures for response curve and repeatability experiments that are conducted in accordance with the guidelines established by the Clinical Proteomics Tumor Analysis Consortium (CPTAC) Assay Development Working Group (ADWG) ( 5,6 Measures computed include LOD, LLOQ, Linearity, Carry-Over, Partial Validation of Specificity, ULOQ, Intra-day Assay CV, Inter-day Assay CV, and Total Assay CV. The response curves were generated by plotting the mean heavy-to-light peak area ratios (for the three replicates) against the theoretical concentration of the heavy peptide. The concentrations (in fmol units) were then back-calculated based on the regression values of the fits to the S-4

5 response curves. The limit of detection (LOD) was calculated based on the average signal from the 3 blank injections + 3x the standard deviation (SD) of the peak area ratios observed for the lowest concentration. The lower limit of quantification (LLOQ) was the lowest concentration at which the percent coefficient of variation (%CV) was <20%. Assay linearity was assessed by fitting a simple linear regression using three calibration points other than the middle calibration point from the mid-range of the curve. The observation was considered to be linear if the average of the middle calibration point concentration was within 5% of that predicted from the best fit line passing through the other points. Carryover was evaluated by dividing the peak area of the analyte peptide in the blank after the highest concentration by the peak area in the high concentration sample, expressed as a percent. Partial validation of specificity was conducted by determining whether, for samples above the LLOQ, any transition deviates >30% from the mean peak area ratio of the other transition ratios. The upper limit of quantification (ULOQ) was defined as the highest concentration that lies on the linear part of the response curve. From the assay repeatability data, intra-assay variability was calculated at each concentration (Low, Medium, and High) as the CV of the three replicates on each of the five days. Inter-assay variability was calculated at each concentration by determining the CV of the first, second, and third injection of each concentration across the five days. The total CV was calculated as the square root of the sum of the squared average intra-assay CV and the squared average inter-assay CV. MRMPlus is freely available its source codes and compiled binaries can be freely downloaded from and respectively. S-5

6 Supplemental References (1) Chen, J.; Shah, P.; Zhang, H. Anal Chem 2013, 85, (2) Campbell, J.; Rezai, T.; Prakash, A.; Krastins, B.; Dayon, L.; Ward, M.; Robinson, S.; Lopez, M. Proteomics 2011, 11, (3) Li, W.; Cohen, L. H. Anal Chem 2003, 75, (4) Whiteaker, J. R.; Zhao, L.; Lin, C.; Yan, P.; Wang, P.; Paulovich, A. G. Mol Cell Proteomics 2012, 11, M (5) Whiteaker, J. R.; Halusa, G. N.; Hoofnagle, A. N.; Sharma, V.; MacLean, B.; Yan, P.; Wrobel, J. A.; Kennedy, J.; Mani, D. R.; Zimmerman, L. J.; Meyer, M. R.; Mesri, M.; Rodriguez, H.; Paulovich, A. G. Nat Methods 2014, 11, 703. (6) Schoenherr, R. M.; Saul, R. G.; Whiteaker, J. R.; Yan, P.; Whiteley, G. R.; Paulovich, A. G. Mol Cell Proteomics 2015, 14, 382. Supplemental Table Legends (Supplemental Tables 1, 2, 4 and 5 are provided as separate Excel files) Supplemental Table 1. Stable isotope-labeled formerly N-linked glycosite-containing peptide targets for scheduled PRM assays. K^ = 13 C 15 6 N 2 -Lysine. R* = 13 C 15 6 N 4 -Arginine. N= deamidated Asn residue (site of N-linked glycosylation). ox = oxidation. The transition nomenclature in the "Transitions" column is decimal-separated precursor ion charge, product ion type, and product ion charge. The transitions are listed in order of decreasing rank. S-6

7 Supplemental Table 2. Response curve assay performance metrics calculated by MRMPlus for 43 assays. LOD.Mean (mean limit of detection), LOD.StDev (standard deviation of limit of detection), LOD.CalibPointUsed (calibration point used for limit of detection calculation), LLOQ.CalibPoint (calibration point used for lower limit of quantification calculation), LLOQ.CV (coefficient of variation of lower limit of quantification), LLOQ.Mean (mean lower limit of quantification), LLOQ.StDev (standard deviation of lower limit of quantification), Curve.Slope (slope of response curve), Curve.Intercept (intercept of response curve), Curve.R 2 (R 2 value of equation fit to response curve), Curve.Diff (Predicted-Observed; %) (percent difference between predicted and observed value based on response curve), Curve.SlopeStdErr (%) (standard error of response curve slope), Carryover(%) (percent carryover), PVSpec.hasValues>LLOQ (partial validation of specificity result has values >LLOQ true or false), PVSpec.pointOfMaxDev (calibration point with maximum deviation based on partial validation of specificity test), PVSpec.maxDeviation (%) (percent of maximum deviation based on partial validation of specificity test), ULOQ.CalibPoint (calibration point for upper limit of quantification), ULOQ.CV (coefficient of variation of upper limit of quantification); ULOQ.Mean (mean of upper limit of quantification), ULOQ.StDev (standard deviation of upper limit of quantification), NA (parameter could not be calculated), and NaN (0 or missing value in calculation). For the rows highlighted in red, 2 of the 3 transitions returned NA or NaN values for some of the calculations; thus, abnormally high values are reported for the LLOQ-related parameters. Values in columns C, D, H, I, T, and U are in fmol units. Supplemental Table 3. Summary of assay repeatability performance (Intra-Assay CV, Inter- Assay CV and Total Assay CV) of 43 assays. The bold Asn residue in each peptide sequence S-7

8 indicates the site of N-linked glycosylation. CVs > 20% are in red font. Peptides in red font (LNSSTIK and FNETTEK) were excluded from the panel of 41 peptides used for biomarker assessment in the patient serum samples. The transitions used in the PRM assays are listed in parentheses below the peptide sequence. Supplemental Table 4. Example MRMPlus input files for the calculation of assay characterization parameters based on response curves: Worksheet 1) Dilution file Number of calibration points included in the response curve, and the analyte concentration corresponding to each point on the curve; Worksheet 2) Metadata file Run order of all files, Replicate/file name, Analyte type (Blank or Serial dilution), Replicate number, and Calibration point; and Worksheet 3) Preprocessed file Data exported from Skyline: peptide sequence, replicate/file name, precursor charge, product ion charge, fragment ion, light precursor m/z, light product ion m/z, light retention time, light area, heavy precursor ion m/z, heavy product ion m/z, heavy retention time, and heavy area. Supplemental Table 5. Mean peak area ratios and fold-changes of the 41 formerly N-linked glycosite-containing peptides selected for relative abundance assessment in the serum of prostate cancer patients (AG, NAG) and non-cancer patients (Neg). Peak area ratios are reported as the peak area ratios of the heavy peptide vs. the light peptide. NAG, AG and Neg columns indicate mean peak area ratios. The 4 peptides with significantly different relative abundance between the NAG and AG serum samples are highlighted in gray. S-8

9 Supplemental Table 3. Summary of assay repeatability performance (Intra-Assay CV, Inter- Assay CV and Total Assay CV of 43 assays). O75882 Attractin IDSTGNVTNELR (2.y10.1, 2.y8.1, 2.y6.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P00450 Ceruloplasmin EHEGAIYPDNTTDFQR (3.y8.1, 3.y7.1, 3.b6.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P00738 Haptoglobin VVLHPNYSQVDIGLIK (3.y7.1, 3.y6.1, 3.b4.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P01009 Alpha-1-antitrypsin YLGNATAIFFLPDEGK (2.y8.1, 2.y5.1, 2.b9.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) S-9

10 P01033 Metalloproteinase inhibitor 1 FVGTPEVNQTTLYQR (2.y11.1, 2.y9.1, 2.y8.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P01042 Kininogen-1 LNAENNATFYFK (2.y10.1, 2.y8.1, 2.y6.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P01591 Immunoglobulin J chain ENISDPTSPLR (2.y8.1, 2.y7.1, 2.y6.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P01833 Polymeric-immunoglobulin receptor VPGNVTAVLGETLK (2.y12.1, 2.y9.1, 2.y8.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P01857 Immunoglobulin gamma-1 chain C region EEQYNSTYR (2.y7.1, 2.y6.1, 2.y5.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) S-10

11 P01859 Immunoglobulin gamma-2 heavy chain EEQFNSTFR (2.y7.1, 2.y6.1, 2.y5.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P01861 Immunoglobulin gamma-4 chain C region EEQFNSTYR (2.y7.1, 2.y6.1, 2.y5.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P01877 Immunoglobulin alpha-2 heavy chain TPLTANITK (2.y7.1, 2.y6.1, 2.y8.2) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P02749 Beta-2-glycoprotein I VYKPSAGNNSLYR (2.y10.1, 2.y9.1, 2.y8.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P02750 Leucine-rich alpha-2-glycoprotein LPPGLLANFTLLR (2.y11.1, 2.y8.1, 2.y12.2) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) S-11

12 P03952 Plasma kallikrein/klkb1 IYPGVDFGGEELNVTFVK (2.y13.1, 2.y11.1, 2.y16.2) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) IYSGILNLSDITK (2.y11.1, 2.y8.1, 2.y7.1) P04004 Vitronectin NGSLFAFR (2.y6.1, 2.y5.1, 2.y4.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P04114 Apolipoprotein B-100 FNSSYLQGTNQITGR (2.y10.1, 2.y9.1, 2.y8.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) YDFNSSMLYSTAK (2.y10.1, 2.y9.1, 2.y8.1) S-12

13 P05155 Plasma protease C1 inhibitor DTFVNASR (2.y6.1, 2.y5.1, 2.y4.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P07339 Cathepsin D GSLSYLNVTR (2.y7.1, 2.y6.1, 2.y5.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P07357 Complement component C8 alpha chain GGSSGWSGGLAQNR (2.y10.1, 2.y8.1, 2.y7.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P0C0L4 Complement C4-A FSDGLESNSSTQFEVK (2.y11.1, 2.y10.1, 2.y9.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) GLNVTLSSTGR (2.y9.1, 2.y7.1, 2.y6.1) S-13

14 P10909 Clusterin EDALNETR (2.y6.1, 2.y5.1, 2.y4.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) LANLTQGEDQYYLR (2.y10.1, 2.y9.1, 2.y8.1) P13473 Lysosome-associated membrane glycoprotein 2 LNSSTIK (2.y6.1, 2.y5.1, 2.y3.1) Intra-Assay CV (%) NaN Inter-Assay CV (%) NaN Total Assay CV (%) NaN Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) VQPFNVTQGK (2.y8.1, 2.y7.1, 2.y6.1) P13598 Intercellular adhesion molecule-2 AAPAPQEATATFNSTADR (2.y12.1, 2.y10.1, 2.y8.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) S-14

15 P16070 CD44 antigen AFNSTLPTMAQMEK (2.y9.1, 2.y8.1, 2.y8.2) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P19823 Inter-alpha-trypsin inhibitor heavy chain H2 GAFISNFSMTVDGK (2.y10.1, 2.y8.1, 2.y7.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P29622 Kallistatin FLNDTMAVYEAK (2.y10.1, 2.y8.1, 2.y7.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P41222 Prostaglandin-H2 D-isomerase SVVAPATDGGLNLTSTFLR (2.y13.1, 2.y12.1, 2.y11.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P43251 Biotinidase FNDTEVLQR (2.y8.1, 2.y7.1, 2.y6.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) S-15

16 P43652 Afamin DIENFNSTQK (2.y8.1, 2.y7.1, 2.y6.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) FNETTEK (2.y6.1, 2.y5.1, 2.y4.1) P55290 Cadherin-13 INNTHALVSLLQNLNK (3.y6.1, 3.b6.1, 3.b7.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) P80188 Neutrophil gelatinase-associated lipocalin SYNVTSVLFR (2.y8.1, 2.y7.1, 2.y6.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) Q08380 Galectin-3-binding protein ALGFENATQALGR (2.y11.1, 2.y9.1, 2.y8.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) S-16

17 Q7Z685 Complement C2/Peptidase S1 QSVPAHFVALNGSK (3.y8.1, 3.y7.1, 3.y6.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) Q96IY4 Carboxypeptidase B-like protein QVHFFVNASDVDNVK (3.y7.1, 3.y6.1, 3.y5.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) Q9HDC9 Adipocyte plasma membrane-associated protein AGPNGTLFVADAYK (2.y10.1, 2.y8.1, 2.y7.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) Q9Y5Y7 Lymphatic vessel endothelial hyaluronic acid receptor 1 ANQQLNFTEAK (2.y8.1, 2.y7.1, 2.y6.1) Intra-Assay CV (%) Inter-Assay CV (%) Total Assay CV (%) S-17

18 Supplemental Figure 1. Formerly N-linked glycosite-containing peptide PRM assay development and characterization workflow schematic. S-18

19 Supplemental Figure 2. Representative extracted ion chromatograms of endogenous formerly N-linked glycosite-containing peptides quantified in serum from patients with AG and NAG prostate cancer and the corresponding heavy peptides. S-19

20 S-20