Establishment And Application Of A Peptide Mass Spectrometry Response Curve Database

Accurate protein quantification is a basis for accurately describing the regulation process of organisms.As the most promising protein quantitative tool,mass spectrometry has high sensitivity and high-throughput characteristics,and has broad application prospects in the era of large data.Extracted ion chromatogram(XIC)mode for primary mass spectrometry(MS1)and multiple reaction monitoring(MRM)mode based on transitions in the secondary mass spectrometry(MS2)is commonly used mode for mass spectrometry quantification.The peak area of the peptide / transition represent the actual amount of protein.Therefore,both response linearity and intensity are sufficient for a better quantification.Signal intensity of different peptides have significant variation,and their mass spectral response curves are totally different.Therefore,quantitative peptides need to be filtered.There are two strategies to choose good quantitative peptides,one is model prediction,another is selection based on experiments.However,the prediction models rarely consider the linear response curve of ions,limiting both accuracy and the successful application of quantification strategies.In this work,we generated original data and built an experimental database of linear MS response curves for global peptidetransition ions in the mammalian proteome.The database includes dynamic range of over 2,647,773 transitions from 121,318 peptides,covered 11,040 gene products.We choose HeLa cell line as whole cell extraction sample.And due to relatively low abundance,we use TFRE method to enrich the nuclear proteins.The intensity,response curve,and dynamic range of all peptides/transitions were measured in serial dilution experiments,evaluated,scored,and presented in our database.The information is hosted on an database(http://www.firmiana.org/responders/).Researchers can browse and easily select and design best-peptide/transition-responders for MS1 or MS2-based quantification by searching our database through a hierarchy including pathway,gene,protein,peptide,and transition levels.We compared the proteome-scale quantification accuracy between best-responder peptides and iBAQ algorithm.Quantification using best-responder peptides indexed by our method provided better reproducibility and accuracy than the iBAQ algorithm.Besides,we've demonstrated the utility of our method by several experiments.1)We've evaluated the accuracy of quantification based on our method by measuring abundances of Universal Proteome Standard(UPS2),the result shows,the correlation between quantitation amount based on our method and the actual amount is higher than the correlation between MaxQuant iBAQ result and the actual amount.2)We also investigated impact of other parameters including cell lyse buffer,HPLC loading buffer and different C18 columns on the repeatability of our method database.The result showed that the best responders in our method are reliable in quantification and repeatable with different HPLC reagent.3)Furthermore,we performed a series of progression dilution of He La lysate in 5 concentrations combined with E.coli lysate as constant matrix to evaluate the behavior of best responders.4)The application of best responders on different cell line verified the universality of best responders.All these results demonstrated the repeatability and accuracy of proteomic quantification based on the best responders in our database.QconCAT proteins are artificial molecules composed of best-peptide-responders from analyte proteins.They are widely applied in MS-based quantification as internal standards.We chose metabolic pathways as a proof-of-principle experiment and selected best-responder peptides for synthesis of a QconCAT protein covering 32 metabolic proteins.We compared peptide intensity homogeneity of thr QconCAT protein with the natural recombinant protein Zscan21.The AUC of QconCAT peptides expressed 3-fold variation,while peptides of Zscan21 were dispersed over 4 orders of magnitude.Lastly,we used QconCAT to determine the stoichiometry of metabolic pathways,comparing abundances of metabolic proteins in the human heart,liver,lung,and stomach.A stoichiometry map of metabolic pathways in the four organs revealed selective enhancement of major metabolic pathways in different organs.This is a demonstrate of the application of our method.