Quantification of Proteins and Metabolites by MassSpectrometry without Isotopic Labeling or SpikedStandardsWeixun Wang,†Haihong Zhou,†Hua Lin, Sushmita Roy, Thomas A. Shaler, Lander R. Hill,Scott Norton, Praveen Kumar, Markus Anderle, and Christopher H. Becker*SurroMed, Inc., 2375 Garcia Avenue, Mountain View, California 94043A new method is presented for quantifying proteomic andmetabolomic profile data by liquid chromatography-massspectrometry (LC-MS) with electrospray ionization. Thisbiotechnology provides differential expression measure-ments and enables the discovery of biological markers(biomarkers). Work presented here uses human serumbut is applicable to any fluid or tissue. The approach relieson linearity of signal versus molecular concentration andreproducibility of sample processing. There is no use ofisotopic labeling or chemically similar standard materials.Linear standard curves are reported for a variety ofcompounds introduced into human serum. As a measureof analytical reproducibility for proteome and metabolomesampling, median coefficients of variation of 25.7 and23.8%, respectively, were determined for ∼3400 molec-ular ions (not counting their numerous isotopes) from 25independently processed human serum samples, corre-sponding to a total of 85 000 individual molecular ionmeasurements.This contribution reports on a new liquid chromatography-mass spectrometry (LC-MS) approach to quantitative profilingof large numbers of proteins (proteomics) and small molecules(metabolomics) for the purpose of differential expression mea-surements and discovery of biomarkers. In this situation, manyor most monitored proteins and small molecules are unanticipatedat the time of laboratory study, thus eliminating the possibility ofprior investigation of relative sensitivity factors (RSFs). Further-more, methods based on introducing a known amount of achemically analogous extraneous substance as an internal standard(i.e., “spiking” of a standard reference material) are not practical,whether the analogue is chemically identical and isotopicallylabeled1-6or based on chemical similarity.7,8It is worth noting, however, that quantification using spikingwith isotopically labeled compounds (known as the isotope dilutionmethod) has a long history.1-6This approach, including quantita-tive isotopic tracer work, has helped generate many valuablecontributions to science, especially in biochemical pathwayanalysis.Recently, a quantification approach for proteomics has beendevised on the basis of a kind of isotope dilution method in whicha specific amino acid in a sample and reference mixture areisotopically labeled and subsequently separated by solid-phasecapture, wash, and releaseswhat has become known as the ICATand related methods (isotope-coded affinity tags).9-11Drawbacksof the method include sample processing complexity, materiallosses, reagent expense, required presence of a specific aminoacid, nonspecifically captured peptide background, frequent dif-ficulty in obtaining useful tandem mass spectrometry (MS/MS)fragmentation patterns, and inability to address metabolomics.There is definite reason to explore other approaches, as done inthis paper.Although it has been observed that electrospray ionization(ESI) provides signals that can be linear with concentration,12-14historically there has been concern about nonlinearities and ionsuppression effects,15especially in the circumstance of complexbiological matrixes, such as serum. Nevertheless, we haveundertaken a broad study to examine whether analyte ion signalscan in general reflect concentrations in a linear way, especiallyfor the case of complex matrixes, and thereby form the basis ofan analytical method for quantifying proteome and metabolomeprofile data for differential expression.* Corresponding author. Phone: (650) 230-1845. Fax: (650) 230-1960.E-mail: [email protected].†These two authors contributed equally to this work.(1) Hevesy, G.; Hofer, E. Nature 1934, 134, 879.(2) Rittenberg, D.; Schoenheimer, R. J. Biol. Chem. 1939, 130, 703-732.(3) Pinajian, J. J.; Christian, J. E.; Wright, W. E. J. Am. Pharm. Assoc. 1953,42, 301-304.(4) Caprioli, R. M. Biochem. Appl. Mass Spectrom. 1972, 735-776.(5) Hamberg, M. Anal. Biochem. 1973, 55, 368-378.(6) Oda, Y.; Huang, K.; Cross, F. R.; Cowburn, D.; Chait, B. T. Proc. Natl. Acad.Sci. U.S.A. 1999, 96, 6591-6596.(7) Nelson, R. W.; Krone, J. R.; Bieber, A. L.; Williams, P. Anal. Chem. 1995,67, 1153-1158.(8) Bucknall, M.; Fung, K. Y. C.; Duncan, M. W. J. Am. Soc. Mass Spectrom.2002, 13, 1015-1027.(9) Gygi, S. P.; Rist, B.; Gerber, S. A.; Turecek, F.; Gelb, M. H.; Aebersold, R.Nat. Biotechnol. 1999, 17, 994-999.(10) Ji, J.; Chakraborty, A.; Geng, M.; Zhang, X.; Amini, A.; Bina, M.; Regnier, F.J. Chromatogr., B. 2000, 745, 197-210.(11) Cagney, G.; Emili, A. Nat. Biotechnol. 2002, 20, 163-170.(12) Purves, R. W.; Gabryelski, Li. L. Rapid Commun. Mass Spectrom. 1998,12, 695-700.(13) Voyksner R. D.; Lee H. Rapid Commun. Mass Spectrom. 1999, 13, 1427-1437.(14) Chelius D.; Bondarenko P. J. Proteome Res. 2002, 1, 317-323.(15) Muller, C.; Schafer, P.; Stortzel, M.; Vogt, S.; Weinmann, W. J. Chromatogr.,B. 2002, 773,47-52.Anal. Chem.2003,75,4818-48264818Analytical Chemistry, Vol. 75, No. 18, September 15, 200310.1021/ac026468x CCC: $25.00 © 2003 American Chemical SocietyPublished on Web 08/08/2003The quantification method reported here relies on the changesin analyte signals directly reflecting their concentrations in onesample relative to another. Samples are not mixed; neither arethe samples otherwise manipulated beyond that required for theLC-MS analysis itself. The sample preparation and LC-MSconditions need to be carefully controlled, however, for optimalresults.This quantification technology employs overall spectral inten-sity normalization by employing signals of molecules that do notchange concentration from sample to sample. In this way, a simplecorrection can be applied for any drift over time in overall LC-MS response or differences in sample concentrations. We havedeveloped a computer application, called MassView software,which among other functions performs normalization by compar-ing two or more spectra to determine the constant intensity ratiobetween those unchanging analytes; this forms the basis foridentifying the nonchanging concentrations (see below). Analternative would be to spike the samples with some (almost any)exogenous compound(s) and use the intensity of that