Defining, Comparing, and Improving iTRAQ Quantification in Mass Spectrometry Proteomics Data

Abstract

The purpose of this study was to generate a basis for the decision of what protein quantities are reliable and find a way for accurate and precise protein quantification. To investigate this we have used thousands of peptide measurements to estimate variance and bias for quantification by iTRAQ (isobaric tags for relative and absolute quantification) mass spectrometry in complex human samples. A549 cell lysate was mixed in the proportions 2:2:1:1:2:2:1:1, fractionated by high resolution isoelectric focusing and liquid chromatography and analyzed by three mass spectrometry platforms; LTQ Orbitrap Velos, 4800 MALDI-TOF/TOF and 6530 Q-TOF. We have investigated how variance and bias in the iTRAQ reporter ions data are affected by common experimental variables such as sample amount, sample fractionation, fragmentation energy, and instrument platform. Based on this, we have suggested a concept for experimental design and a methodology for protein quantification. By using duplicate samples in each run, each experiment is validated based on its internal experimental variation. The duplicates are used for calculating peptide weights, unique to the experiment, which is used in the protein quantification. By weighting the peptides depending on reporter ion intensity, we can decrease the relative error in quantification at the protein level and assign a total weight to each protein that reflects the protein quantitation confidence. We also demonstrate the usability of this methodology in a cancer cell line experiment as well as in a clinical data set of lung cancer tissue samples. In conclusion, we have in this study developed a methodology for improved protein quantification in shotgun proteomics and introduced a way to assess quantification for proteins with few peptides. The experimental design and developed algorithms decreased the relative protein quantification error in the analysis of complex biological samples. The purpose of this study was to generate a basis for the decision of what protein quantities are reliable and find a way for accurate and precise protein quantification. To investigate this we have used thousands of peptide measurements to estimate variance and bias for quantification by iTRAQ (isobaric tags for relative and absolute quantification) mass spectrometry in complex human samples. A549 cell lysate was mixed in the proportions 2:2:1:1:2:2:1:1, fractionated by high resolution isoelectric focusing and liquid chromatography and analyzed by three mass spectrometry platforms; LTQ Orbitrap Velos, 4800 MALDI-TOF/TOF and 6530 Q-TOF. We have investigated how variance and bias in the iTRAQ reporter ions data are affected by common experimental variables such as sample amount, sample fractionation, fragmentation energy, and instrument platform. Based on this, we have suggested a concept for experimental design and a methodology for protein quantification. By using duplicate samples in each run, each experiment is validated based on its internal experimental variation. The duplicates are used for calculating peptide weights, unique to the experiment, which is used in the protein quantification. By weighting the peptides depending on reporter ion intensity, we can decrease the relative error in quantification at the protein level and assign a total weight to each protein that reflects the protein quantitation confidence. We also demonstrate the usability of this methodology in a cancer cell line experiment as well as in a clinical data set of lung cancer tissue samples. In conclusion, we have in this study developed a methodology for improved protein quantification in shotgun proteomics and introduced a way to assess quantification for proteins with few peptides. The experimental design and developed algorithms decreased the relative protein quantification error in the analysis of complex biological samples. Recent developments in methods and instruments for mass spectrometry enable quantitative proteomics analysis of complex samples with good coverage (1Beck M. Schmidt A. Malmstroem J. Claassen M. Ori A. Szymborska A. Herzog F. Rinner O. Ellenberg J. Aebersold R. The quantitative proteome of a human cell line.Mol. Syst. Biol. 2011; 7: 549Crossref PubMed Scopus (588) Google Scholar, 2Geiger T. Wehner A. Schaab C. Cox J. Mann M. Comparative proteomic analysis of eleven common cell lines reveals ubiquitous but varying expression of most proteins.Mol. Cell. Proteomics. 2012; 11 (M111.014050)Abstract Full Text Full Text PDF Scopus (578) Google Scholar, 3Nagaraj N. Wisniewski J.R. Geiger T. Cox J. Kircher M. Kelso J. Pääbo S. Mann M. Deep proteome and transcriptome mapping of a human cancer cell line.Mol. Syst. Biol. 2011; 7: 548Crossref PubMed Scopus (757) Google Scholar, 4Arabi A. Ullah K. Branca R.M. Johansson J. Bandarra D. Haneklaus M. Fu J. Ariës I. Nilsson P. Den Boer M.L. Pokrovskaja K. Grander D. Xiao G. Rocha S. Lehtiö J. Sangfelt O. Proteomic screen reveals Fbw7 as a modulator of the NF-kappaB pathway.Nat. Commun. 2012; 3: 976Crossref PubMed Scopus (69) Google Scholar). Several techniques for quantification by mass spectrometry exist, both using isotopic labeling and label free methods (5Bantscheff M. Schirle M. Sweetman G. Rick J. Kuster B. Quantitative mass spectrometry in proteomics: a critical review.Anal. Bioanal. Chem. 2007; 389: 1017-1031Crossref PubMed Scopus (1257) Google Scholar, 6Ong S.E. Mann M. Mass spectrometry-based proteomics turns quantitative.Nat. Chem. Biol. 2005; 1: 252-262Crossref PubMed Scopus (1317) Google Scholar). Quantification by isotopic labeling can be done on precursor ion level or by quantifying isobaric label fragments in fragment spectra. Isotope-coded affinity tag (7Gygi S.P. Rist B. Gerber S.A. Turecek F. Gelb M.H. Aebersold R. Quantitative analysis of complex protein mixtures using isotope-coded affinity tags.Nat. Biotechnol. 1999; 17: 994-999Crossref PubMed Scopus (4345) Google Scholar), isobaric tags for relative and absolute quantification (iTRAQ) 1The abbreviations used are:iTRAQisobaric tags for relative and absolute quantificationSILACstable isotope labeling by amino acids in cell cultureCVcoefficient of variancePSMpeptide spectrum matchHCDhigher-energy collisional dissociationCIDcollision induced dissociationFDRfalse discovery ratePQPQprotein quantification by peptide quality controlRMSEroot 1The abbreviations used are:iTRAQisobaric tags for relative and absolute quantificationSILACstable isotope labeling by amino acids in cell cultureCVcoefficient of variancePSMpeptide spectrum matchHCDhigher-energy collisional dissociationCIDcollision induced dissociationFDRfalse discovery ratePQPQprotein quantification by peptide quality controlRMSEroot B. K. S. N. S. S. S. S. P. S. M. F. A. protein quantitation in using isobaric Cell. Proteomics. 3: Full Text Full Text PDF PubMed Scopus Google Scholar), and isotope labeling by amino acids in cell S.E. B. I. A. Mann M. isotope labeling by amino acids in cell as a and accurate to expression Cell. Proteomics. 1: Full Text Full Text PDF PubMed Scopus Google are the most used labeling methods based on iTRAQ for relative quantification of to samples a Quantification by mass spectrometry is a and to the in the quantitative in labeling protein precursor ion data and data analysis Aebersold R. R.M. The and of Biotechnol. PubMed Scopus Google Scholar). The quality of quantitation methods can be in of and is affected by that the of is by that and isobaric tags for relative and absolute quantification isotope labeling by amino acids in cell of variance peptide spectrum collisional induced discovery protein quantification by peptide quality error error relative isobaric tags for relative and absolute quantification isotope labeling by amino acids in cell of variance peptide spectrum collisional induced discovery protein quantification by peptide quality error error relative Several have that iTRAQ labeling is with are R.M. P. of and in mass spectrometry data 2011; PubMed Scopus Google Scholar, and in iTRAQ Cell. Proteomics. Full Text Full Text PDF PubMed Scopus Google Scholar, S. of iTRAQ and quantitative proteomics using of for 2012; PubMed Scopus Google Scholar, M. J. C. I. iTRAQ in and complex the and the PubMed Scopus Google Scholar). suggested that this of is by peptides with that are mixed iTRAQ in complex samples M. J. C. I. iTRAQ in and complex the and the PubMed Scopus Google Scholar). iTRAQ data to variance The of variance of the on the intensity, with for R.M. P. of and in mass spectrometry data 2011; PubMed Scopus Google Scholar, and in iTRAQ Cell. Proteomics. Full Text Full Text PDF PubMed Scopus Google Scholar, M. M. D. T. Sweetman G. Kuster B. and iTRAQ quantification on LTQ Orbitrap mass Cell. Proteomics. 7: Full Text Full Text PDF PubMed Scopus Google Scholar, S. J. A. iTRAQ quantitative proteomic analysis on a ion mass 2007; PubMed Scopus Google Scholar). of iTRAQ for quantification are in the of the and to a relative protein are for the iTRAQ peptide data to a reliable protein to the protein quantification by the variance have based on peptide data J. J. O. S. T. K. M. C. S. proteomic analysis of a of using isobaric PubMed Scopus Google Scholar, a for protein PubMed Scopus Google Scholar), weighting the peptide data to a for protein PubMed Scopus Google Scholar, and biological in isobaric tags for relative and absolute quantitation 2007; PubMed Scopus Google Scholar, G. J. B. for and accurate protein quantification of isobaric peptide data LTQ PubMed Scopus Google Scholar, R.M. K. C. of and for Quantitative on LTQ Orbitrap 2012; PubMed Scopus Google or the variance and in iTRAQ Cell. Proteomics. Full Text Full Text PDF PubMed Scopus Google Scholar). Quantitative of complex human samples are to to biological and of the human proteome and a of in in peptides and a of peptides that can and in the mass spectrometry for discovery the is to quantitative or in protein or clinical is to as accurate and precise quantitative the data as as well as to estimate the of the quantification. for quantitative proteomics analysis is for to in protein and proteins to the biological and clinical Aebersold R. R.M. The and of Biotechnol. PubMed Scopus Google Scholar, B. Aebersold R. and a quantitative proteomics Biotechnol. PubMed Scopus Google Scholar). a protein as and to in clinical using techniques is and The of 2011; PubMed Scopus Google Scholar). of iTRAQ labeling in and to is to assess the and of the common to study variance and bias in mass spectrometry based protein quantification is to a set of proteins a sample and the and bias of the of peptides. of proteins the of at a set of peptides and how in the used in of the that iTRAQ quantification R.M. P. of and in mass spectrometry data 2011; PubMed Scopus Google Scholar, and in iTRAQ Cell. Proteomics. Full Text Full Text PDF PubMed Scopus Google Scholar, S. of iTRAQ and quantitative proteomics using of for 2012; PubMed Scopus Google Scholar, M. J. C. I. iTRAQ in and complex the and the PubMed Scopus Google Scholar). the of data be to the of a biological which thousands of proteins A. B. J. S.A. D. K. S. J. Aebersold R. A. B. bias in experiment design for mass spectrometry-based quantitative Cell. Proteomics. 2007; Full Text Full Text PDF PubMed Scopus Google Scholar). In the peptides in a complex human cell line sample are used to estimate of the quantitative and experimental is to a discovery study with high complex human proteome samples. The quality of the protein is mass in this also the of peptide and the of methods for sample are such as variance and bias of peptide quantification by iTRAQ are in high complex samples. methods for the protein quantification are by on the peptide level to quality and by weighting the peptide to for the of at G. J. B. for and accurate protein quantification of isobaric peptide data LTQ PubMed Scopus Google Scholar). We have the to bias and variance in protein quantification by iTRAQ for how to estimate the of protein which be in both discovery and of biological Based on the we experimental design each labeling set duplicate and we how duplicates are used for calculating peptide that can be used in the of protein is to protein quantification by iTRAQ in data of cell line samples with and a clinical data set of lung cancer tissue samples. Several mass spectrometry as in peptide as well as sample Mass spectrometry data was using three instruments Orbitrap 4800 MALDI-TOF/TOF and for for and fragmentation as well as the of investigated for the Orbitrap on experimental is in of lung cancer cell line A549 by and by by of peptides at with iTRAQ tags to the iTRAQ peptides by a liquid chromatography or by by isoelectric focusing on a by as J. J. K. B. B. Lehtiö J. Quantitative proteomics peptide isoelectric focusing for of cell lung cancer PubMed Scopus Google Scholar). of peptides or peptide the analyzed on three platforms; LTQ Orbitrap Velos, 4800 and 6530 analysis by LTQ Orbitrap peptides using by a The LTQ Orbitrap was in a data for fragmentation by induced and collisional and analyzed by the and with was used for peptide and protein analysis on 4800 peptides on by to a the MALDI-TOF/TOF data was using the A. S.P. The a that and to peptides mass Proteomics. 2007; Full Text Full Text PDF PubMed Scopus Google in the analysis on 6530 was using to 6530 with a by the the data using the peptide using with for to with peptides. the to human protein discovery was by the data a of both and and set to at the protein level using Claassen M. S.P. M. Schmidt A. Aebersold R. discovery for proteomics data by mass Cell. Proteomics. Full Text Full Text PDF PubMed Scopus Google Scholar). to a protein used in the of iTRAQ reporter ions for isotope by iTRAQ reporter ion is to as peptide the data analysis in the R. R. for data analysis and Scholar). of peptide to assess data and iTRAQ The of the iTRAQ also and have The peptides to with reporter ions in at of samples at both and The peptide used to assess the quantitative and The error of peptide quantification was by error reflects a of the the by mass spectrometry and the To on intensity, the to be on the The was done by the with the of the line of The was by a to and The can be as quantification error in both variance and To be to study the variance the peptide data was that the peptide was samples. The the variance in quantification. was as the of peptide measurements samples by the peptide The bias was investigated by the iTRAQ peptide the peptide was for the on the peptide iTRAQ we to a iTRAQ to the in to the which is based on each the protein level on the the and we used the relative error for of the protein quantification. The relative error was as the of the protein the protein by the To the by the of and the are for the and iTRAQ the of the cell line data duplicate for the clinical data set the of internal was and with the of The relative protein quantification was by the peptide to peptide or peptide used and to by methods also with calculating a or of the relative peptide that to the The is based on the developed by G. J. B. for and accurate protein quantification of isobaric peptide data LTQ PubMed Scopus Google Scholar), with for the variance of peptides by a weight in the protein quantification is the of peptides a To the peptide weights, the peptide to and each the of the peptide to was The for the as by the error of the peptide that In the by the peptide are based on the of reporter ion and the weight is as by the error each The peptide data used to the was the iTRAQ and is to as the To that the data bias the protein based on of the samples in the iTRAQ also The of the relative protein quantities was by the relative The was on data of cell line samples with The data of and the and each and are The was used as internal data to the in the data The protein quantities based on the relative error of duplicate The was on a clinical data set of lung cancer tissue samples The samples analyzed in three iTRAQ internal in of the internal in The internal used to and the protein in The for calculating based on duplicates and to proteins as well as the for and is in The of quantification and peptide quality J. Johansson M. Branca R.M. A. Lehtiö J. by Quantification and Cell. Proteomics. 2011; Full Text Full Text PDF PubMed Scopus Google the to protein quantities to the The data the be using the The experimental for the data set can be in and the the analysis of the samples and are in and the of peptide and protein for the three instruments by using the The by using the are in and protein data for can be in and the of peptides the of peptides and proteins in for the on of the with a a data that sample and enable a set of peptides to be The of peptides with the mass spectrometry the Orbitrap as as the and The in the of and was the peptide amount, samples on the In the the be for that by a to the of The the of unique peptide and of proteins with the instruments using at protein isoelectric peptide spectrum The is based on using peptide amount, with a in a The quality of the peptide quantification was by error The both bias and variance and the of the error peptide iTRAQ The the reporter ion with the of the iTRAQ The in the of to intensity, in are the at the of is the of of the each is based on the of in S. of the was used to the The at the of the The reflects the of for and is for the of quantification. By using the most are the most by the in the error in quantitation is and as the peptide The measurements of error by both variance and bias the the To be to study the variance in the peptide the peptide to sample of the samples to in a of the iTRAQ in this the the peptide the iTRAQ was was the The was that we have a bias in the and the with To assess bias of iTRAQ iTRAQ of high level to was a bias in on The bias to be at the of with and as and the is with as and the is The for peptides was and to peptide bias can be on the is at which is the of also can be the and is that the of is in this the is at the was to peptide amount, and The can be in as well as in In instrument data by the peptide quantities the Orbitrap and have peptide quantities have reveals that the are improved for Orbitrap and by using peptide and by In is on peptide and of for are iTRAQ reporter for the instruments used in this In the is to the peptide in the is to the data of the data of the at that of for three instruments are by To study the of iTRAQ quantification on peptide and fragmentation was in the Orbitrap and of is a for peptide quantities and of In for of to a good the fragmentation to in a decrease in the of peptide spectrum for and for quantification. fragmentation was fragmentation at and analysis in the Based on the and was to for and quantification the are in and decrease but also decrease the of by with of using In this for the iTRAQ peptide data to a reliable protein a weight based on peptide and a peptides of protein The based on internal data and to protein in the data and To bias depending on which iTRAQ that for the weight iTRAQ used as set to The protein quantities on which that was used and was the the weight based on using as internal The of the of the peptide used for weight was also The that the to peptide measurements in each the quality of the protein quantities The in peptides in each was to as well as to with peptides have of peptides in each the the was set at peptide to peptide measurements with and The was with peptides calculating a as well as to using peptides for the of a the was with the in The protein with the and the relative error was for protein quantification The protein quantities to and generate accurate protein quantities the and can be in that proteins are with a relative error using the as with the In the relative error of protein is to protein weight as the of peptide for proteins with of peptides. in the the relative error of the protein is on the of peptides used for quantification of the proteins with few the of the peptides by protein the relative error for proteins with of peptides the of the peptides on at protein weight the relative error is for proteins with peptides for quantification. The be used to set a on protein weight for accurate protein quantification. The of was for and level as The that have a to the at the and and is for the relative error and protein weight in this data the in this study was with the by G. J. B. for and accurate protein quantification of isobaric peptide data LTQ PubMed Scopus Google Scholar), which the methods to the in this study but a to the clinical lung cancer The of using internal set to was with the on was that the is for the in the experiment, the was used to protein for the experimental The relative error for protein and The the the on the peptide the Orbitrap by and and to peptide and the Orbitrap with the peptide and by the is for the the to with the peptide and a the of proteins is for the three proteins by the Orbitrap proteins by and by of relative error for peptide and The of protein relative error for the experimental in this and is by the The of calculating based on internal set was also to data of cell line samples and lung cancer tissue samples. The used to protein of duplicates in the experiment, as well as to protein of internal in the lung cancer The relative error of the protein and with using a the peptides for calculating the protein and in the the the for the data the the data To the of protein to and the protein a the protein protein weights, of and the relative error was the data set in The relative error was based on the in The relative error can be used as a to assess protein quantification and protein weight can be to quantitative data is in discovery and to proteome The purpose of this study was to generate a basis for the decision of what protein quantities are reliable and find a way for accurate and precise protein quantification by isobaric To investigate this we have used thousands of peptide measurements to estimate variance and bias for quantification by iTRAQ mass spectrometry in complex human cell line samples. Based on the we have suggested a concept for experimental design and a methodology to assess protein quantification. In based proteomics is to have as good protein coverage as for peptides with accurate quantitative measurements be methods peptides a variance but the in the quantitative and in iTRAQ Cell. Proteomics. Full Text Full Text PDF PubMed Scopus Google Scholar). a variance proteins can be in a complex human can be used to peptides with as in of quantification J. J. O. S. T. K. M. C. S. proteomic analysis of a of using isobaric PubMed Scopus Google Scholar, a for protein PubMed Scopus Google Scholar). peptides decrease the of proteins by in the data the is for the that the quantitative on the peptide level is to protein level We have in this study methods to protein by peptides to protein or by using peptides but weight to by absolute to protein The which for introduced by was G. J. B. for and accurate protein quantification of isobaric peptide data LTQ PubMed Scopus Google with In the the weight of a is as by the error for peptides that In the error of peptides in the is used to the the is to measurements the In the the weight is to the peptide intensity, of the of reporter ion this the most in the The in the improved the protein quantities to clinical data set of lung cancer tissue samples. In the the weight is based on internal set for each internal set for the is to to experimental for the analysis the data quality as is in this The and experimental and on and based on study be to the of on experimental we duplicate in each iTRAQ can be for data and be to the biological iTRAQ samples. The of the of and a for protein quantification that the protein quantities have relative error protein quantification by calculating the The is for the clinical data set of lung cancer tissue samples. we this is to proteins in the clinical data set with accurate quantification relative which can be for protein quantification by of the peptides with the relative error at the protein level is peptides have high the and to a protein The also that the relative error of protein is on the of peptides used for protein proteins with few peptides for quantification is a on the peptide level by the protein for proteins with peptides the of the peptides on the at protein weight the relative error is for proteins with peptides for quantification. peptides with can be for a protein this is for a peptide the peptide of the Orbitrap data a of this to error at the protein level for of protein is in line with the of To assess the of the we have used and relative error used by R.M. P. of and in mass spectrometry data 2011; PubMed Scopus Google Scholar, and in iTRAQ Cell. Proteomics. Full Text Full Text PDF PubMed Scopus Google Scholar). and relative error both bias and variance and reflects the in the In the variance to be the to the error and bias be in this the of R.M. P. of and in mass spectrometry data 2011; PubMed Scopus Google Scholar, and in iTRAQ Cell. Proteomics. Full Text Full Text PDF PubMed Scopus Google Scholar, S. of iTRAQ and quantitative proteomics using of for 2012; PubMed Scopus Google Scholar, M. J. C. I. iTRAQ in and complex the and the PubMed Scopus Google Scholar). In a biological study we to the sample and labeling by to or of peptide is based on the that the samples are in of protein also the analysis to we can that most to the bias are in the data analysis and the variance the error also in a biological instruments for Orbitrap and for the peptide quantification. reflects the used by the have a high for fragmentation as the Orbitrap the a is also in the of peptides and proteins Orbitrap proteins the instruments the of peptides as well as the sample by in the for the both to error at the peptide and protein level as well as of peptides and to the in this the suggested for the Orbitrap be a of a fragmentation of and as the of instruments but can as a for In the data set the peptide are the the iTRAQ peptides to a protein the protein In a biological data set this is of the each iTRAQ a biological this we also the on cell line and clinical data protein quantification was improved by using the internal duplicate to and relative error The the study is a to assess the quality of protein of the experimental we calculating the peptide and the in each study based on a duplicate the The protein are based on the peptide to generate accurate protein quantities relative We that a the in and are for each data set based on the duplicate in the The with the be used to set a on protein to reliable protein be for proteins with or a few peptides for quantification. proteins with peptides as well as proteins have the relative and the to reliable protein quantification. By this the level of relative error can be set based on the experimental and biological By a on the protein at the peptide intensity, we the of peptides for accurate protein quantification as well as the of to the to data the also be to both labeling methods such as as well as to label free mass spectrometry label free the of on good duplicate study in of peptide and a of peptide quantities for in We have in this study developed a methodology for improved protein quantification in shotgun The suggested experimental design and developed algorithms decrease the relative protein quantification error in the analysis of complex biological samples. this methodology quality of protein data and of quantification for proteins with few peptides. is in biological as in we for quantitative samples. with