Key points are not available for this paper at this time.
The proposed model is based on the measurement of the retention times of 346 tryptic peptides in the 560- to 4,000-Da mass range, derived from a mixture of 17 protein digests. These peptides were measured in HPLC-MALDI MS runs, with peptide identities confirmed by MS/MS. The model relies on summation of the retention coefficients of the individual amino acids, as in previous approaches, but additional terms are introduced that depend on the retention coefficients for amino acids at the N-terminal of the peptide. In the 17-protein mixture, optimization of two sets of coefficients, along with additional compensation for peptide length and hydrophobicity, yielded a linear dependence of retention time on hydrophobicity, with an R2 value about 0.94. The predictive capability of the model was used to distinguish peptides with close m/z values and for detailed peptide mapping of selected proteins. Its applicability was tested on columns of different sizes, from nano- to narrow-bore, and for direct sample injection, or injection via a pre-column. It can be used for accurate prediction of retention times for tryptic peptides on reversed-phase (300-Å pore size) columns of different sizes with a linear water-ACN gradient and with TFA as the ion-pairing modifier. The proposed model is based on the measurement of the retention times of 346 tryptic peptides in the 560- to 4,000-Da mass range, derived from a mixture of 17 protein digests. These peptides were measured in HPLC-MALDI MS runs, with peptide identities confirmed by MS/MS. The model relies on summation of the retention coefficients of the individual amino acids, as in previous approaches, but additional terms are introduced that depend on the retention coefficients for amino acids at the N-terminal of the peptide. In the 17-protein mixture, optimization of two sets of coefficients, along with additional compensation for peptide length and hydrophobicity, yielded a linear dependence of retention time on hydrophobicity, with an R2 value about 0.94. The predictive capability of the model was used to distinguish peptides with close m/z values and for detailed peptide mapping of selected proteins. Its applicability was tested on columns of different sizes, from nano- to narrow-bore, and for direct sample injection, or injection via a pre-column. It can be used for accurate prediction of retention times for tryptic peptides on reversed-phase (300-Å pore size) columns of different sizes with a linear water-ACN gradient and with TFA as the ion-pairing modifier. The application of MS to biomolecular analysis has revolutionized protein research within the past decade (1Mann M. Hendrickson R.C. Pandey A. Analysis of proteins and proteomes by mass spectrometry..Annu. Rev. Biochem. 2001; 70: 437-473Google Scholar). This can be mostly attributed to the development of ionization techniques that are compatible with biomolecules, i.e. MALDI (2Karas M. Hillenkamp F. Laser desorption ionization of proteins with molecular masses exceeding 10,000 daltons..Anal. Chem. 1988; 60: 2299-2301Google Scholar, 3Hillenkamp F. Karas M. Beavis R.C. Chait B.T. Matrix-assisted laser desorption/ionization mass spectrometry of biopolymers..Anal. Chem. 1991; 63: 1193A-1203AGoogle Scholar) and ESI (4Fenn J.B. Mann M. Meng C.K. Wong S.F. Whitehouse C.M. Electrospray ionization for mass spectrometry of large biomolecules..Science. 1989; 246: 64-71Google Scholar), as well as improved instrumentation. However, although modern mass spectrometers provide high mass accuracy and sensitivity, the protein complexity and concentration range usually found in biological samples still present a challenge. The problem has been traditionally attacked by separation of complex protein mixtures by two-dimensional gel electrophoresis, with subsequent protein in-gel digestion, followed by ESI or MALDI MS. This remains one of the most popular sample preparation procedures, especially suitable for protein identification and quantitation. However, the method is best suited for higher abundance proteins with masses greater than 12–14 kDa, and some categories of molecules, such as membrane proteins (1Mann M. Hendrickson R.C. Pandey A. Analysis of proteins and proteomes by mass spectrometry..Annu. Rev. Biochem. 2001; 70: 437-473Google Scholar) or species with extremes in isoelectric points, are handled poorly. There are also difficulties in adapting the method to high-throughput applications. Alternative analytical approaches are based on pre-fractionation of protein mixtures or cell lysates before the final MS steps of analysis (5Blonder J. Goshe M.B. Moore R.J. Pasa-Tolic L. Masselon C.D. Lipton M.S. Smith R.D. Enrichment of integral membrane proteins for proteomic analysis using liquid chromatography-tandem mass spectrometry..J. Proteome Res. 2002; 1: 351-360Google Scholar, 6Verma R. Chen S. Feldman R. Schieltz D. Yates J. Dohmen J. Deshaies R.J. Proteasomal proteomics: Identification of nucleotide-sensitive proteasome-interacting proteins by mass spectrometric analysis of affinity-purified proteasomes..Mol. Biol. Cell. 2000; 11: 3425-3439Google Scholar, 7Optiteck G.J. Ramirez S.M. Jorgenson J.W. Moseley M.A.I. Comprehensive two-dimensional high-performance liquid chromatography for the isolation of overexpressed proteins and proteome mapping..Anal. Biochem. 1998; 258: 349-361Google Scholar, 8McCormack A.L. Schieltz D.M. Goode B. Yang S. Barnes G. Drubin D. Yates III, J.R. Direct analysis and identification of proteins in mixtures by LC/MS/MS and database searching at the low-femtomole level..Anal. Chem. 1997; 69: 767-776Google Scholar, 9Link A.J. Eng J. Schieltz D.M. Carmack E. Mize G.J. Morris D.R. Garvik B.M. Yates III, J.R. Direct analysis of protein complexes using mass spectrometry..Nat. Biotechnol. 1999; 17: 676-682Google Scholar). This often involves proteolytic digestion, followed by one- or multi-dimensional chromatographic separation of the resulting peptides, with subsequent detection by MS/MS. Such a method may yield considerable simplification of the problem, because the fractions from on- or off-line HPLC separations have reduced complexity compared with the original sample. Indeed, the combination of HPLC-ESI (MS or MS/MS) has proved to be a “work horse” for large-scale high-throughput proteomics (9Link A.J. Eng J. Schieltz D.M. Carmack E. Mize G.J. Morris D.R. Garvik B.M. Yates III, J.R. Direct analysis of protein complexes using mass spectrometry..Nat. Biotechnol. 1999; 17: 676-682Google Scholar, 10Aebersold R. Goodlett D.R. Mass spectrometry in proteomics..Chem. Rev. 2001; 101: 269-295Google Scholar), because of its ability to deal with complex samples and to be fully automated (11Gygi S.P. Rist B. Gerber S.A. Turecek F. Gelb M.H. Aebersold R. Quantitative analysis of complex protein mixtures using Biotechnol. 1999; 17: Scholar, A. Ramirez J. for proteomics: with two of Chem. 2001; Scholar, R. mass spectrometric protein and proteome Mass Scholar). However, the for of the HPLC and the mass are usually of the HPLC to the mass as used for may for approaches, as well as time In the of HPLC to a MALDI is has the of the two optimization of and time on the mass spectrometric The capability to J. R. E. M. M. J. based on the of sample Chem. 2001; Scholar) samples in with MALDI MS M. mass with a laser desorption/ionization and Mass 2000; Scholar, A. A. A. MALDI mass for proteomic Chem. 2000; Scholar, A.L. The of peptide using a high-performance mass Chem. 2000; Scholar, M. Chait B.T. identification of proteins with a mass Chem. 2001; Scholar, high laser desorption/ionization mass spectrometry for of Mass 2001; Scholar) the HPLC-MALDI MS combination for detailed of protein that and injection for the analysis of Identification of two in tryptic peptides from the sample protein and analysis of on on Mass and Scholar, A. R. S. J. F. Mass spectrometric of proteins from the Cell. Scholar). separations with of a the of J. laser desorption/ionization mass spectrometry using a Chem. 2000; Scholar, J. of to and Proteome Res. 2002; 1: Scholar), and a has been used for off-line of nano- and to the mass and B. MS analysis of complex peptide mixtures using on Mass and Scholar, and B. for off-line of to MALDI MS and on Mass and Scholar). However, such a is usually in the because chromatographic often have an of at such off-line of separations to MALDI MS can be based on and of and laser desorption/ionization mass Chem. Scholar, S. R.C. J. Hillenkamp F. Jorgenson J.W. and identification of peptides in by liquid laser desorption/ionization mass spectrometry and Chem. 1998; 70: Scholar, S.P. Rist B. Aebersold R. A. A. Quantitative proteomic analysis using a MALDI mass Chem. 2001; Scholar). the HPLC can be Direct sample using in laser desorption/ionization mass spectrometry for Mass 2000; Scholar), S. J. G. liquid chromatography to laser desorption/ionization mass spectrometry using an Mass 2000; Scholar, J. D. E. for mass on Mass and Scholar), or J. and off-line for HPLC and MALDI on Mass and Scholar) on a MALDI in using a a suitable MALDI in a In HPLC has been used as a separation the additional that be derived from the chromatographic retention the of HPLC and modern MS are such can in be used to peptides, such as that are by mass measurement Pasa-Tolic L. Lipton M.S. R. Smith R.D. of for the accurate prediction of peptide liquid chromatography times in proteome Chem. Scholar). the of of can the of protein identification by peptide mass There have been predictive approaches of peptide retention times in reversed-phase used peptide of These are based on the that the chromatographic of a peptide is on its amino HPLC of HPLC of Scholar). In sets of retention coefficients for individual amino acids were from of the retention times of peptides of of peptide retention times in liquid chromatography on the of amino S. A. Scholar, retention and of peptides in high-performance liquid Scholar, S. The isolation of peptides by high-performance liquid chromatography using Biochem. Scholar, of peptide retention times in high-performance liquid chromatography linear gradient Scholar, of peptide retention 1988; Scholar). method for the of individual amino was based on measurement of the retention times of model peptides, for using was by of the amino acids found in proteins D. of peptide retention times in reversed-phase high-performance liquid chromatography of retention coefficients of amino of model Scholar, D. of peptide retention times in reversed-phase high-performance liquid chromatography of and peptide retention times and the retention times of Scholar). the introduced a for accurate prediction of retention time of peptides than amino acids of peptide length on peptide retention in reversed-phase 1988; Scholar). Such have high for the peptides used to but is be for In the application of a prediction model to a mixture of tryptic peptides from proteins a predictive M. M. J. of chromatographic retention and protein identification in liquid Chem. 2002; Scholar). for large of peptides and retention times to sets of peptides derived from protein digests. was used for optimization of a model of retention using an Pasa-Tolic L. Lipton M.S. R. Smith R.D. of for the accurate prediction of peptide liquid chromatography times in proteome Chem. Scholar). The large of by a accurate of retention coefficients for individual amino that a predictive model can be improved by the of of the of amino acids along the peptide as well as the amino This a of for accurate prediction of retention times for peptides This model was based on analysis of a MS on a 17-protein tryptic mixture 346 peptides in the 560- to 4,000-Da mass The of the was to a model that peptide from linear gradient was used for the peptide as is a linear peptide retention times and peptide HPLC of HPLC of Scholar). The were selected to be as as the of off-line separation and subsequent MALDI MS with TFA the of subsequent in the ionization The of the of peptides used to a model is for the of retention peptides amino a range of sizes and be well The in the present was to an mixture of the tryptic of 17 proteins The of of proteins were peptide identification and a of However, is that of the proteins for the and are and as such one that the of peptides is in MS analysis of the HPLC fractions of the 17-protein of peptides by by in a the was used in were reduced in the at molecular mass and with The samples were with and with peptide of was the and steps MS analysis on different a protein mixture was from the and of The mixture was by of the proteins in a of affinity-purified was also A. of of Biol. Chem. Scholar). of protein were by MALDI MS to protein mixture of 17-protein of was by in TFA of the mixture of was the a tryptic of was with TFA and of was sample of protein was in and in the and injection for the analysis of Identification of two in tryptic peptides from the sample protein and analysis of on on Mass and Scholar). an sample of A. R. S. J. F. Mass spectrometric of proteins from the Cell. Scholar) was in-gel with to the by A. B. M. Mann M. peptide by a combination of and mass Mass 1997; 11: Scholar). The peptides were from the and in of TFA to MS a tryptic was from the proteins of a was at 10,000 for to cell and the was for using molecular mass from was from of as for the protein The was and in of of the resulting peptide mixture was to MS and were used for the preparation of was at separations were using different HPLC in was used for analysis of the of samples were a and with a linear gradient of in or at was at using a and were used for and liquid The was with MALDI and by a J. and off-line for HPLC and MALDI on Mass and Scholar) a at was used for in the fractions were as the of tryptic peptides were in the chromatographic were and to MALDI MS In to applicability of the model to different was used for separation of the sample mixture separation to In one of a was with a of the protein mixture a of of protein at of in The sample was using a gradient in was used for separation of the 17-protein mixture on a of was a and with a linear gradient of in at was at using a In fractions were and were in on a and by MALDI of the was used for analysis of the sample and injection for the analysis of Identification of two in tryptic peptides from the sample protein and analysis of on on Mass and Scholar). of the sample of tryptic of protein was a and on a with a linear gradient of in were used for liquid The was with and a at used as a of peptide retention the of a was in a the was a retention time to the However, that was two the retention time was the of the than of peptides in the gradient were found in one or two but were a of In the two fractions that the most for a peptide were The of of the proteins was confirmed by peptide mass protein was with in on a MALDI and to MALDI MS The from individual as well as the chromatographic fractions were by MS with m/z range and by in the mass M. mass with a laser desorption/ionization and Mass 2000; Scholar). In are by of the sample with from a laser with injection of from the the a mass of and accuracy within a in the in MS and and for were used for peptide mass and protein peptide of was used for by protein with a of m/z values The for of the of peptide was using a on an using for was also using the of but the be to has for to its and of the 17-protein mixture is in of peptides were in fractions by masses as by MALDI MS There were and of peptides in the m/z and The identities of of were confirmed by were used for the development of of the peptide from as an of to D. of peptide retention times in reversed-phase high-performance liquid chromatography of retention coefficients of amino of model Scholar), the retention time for a gradient of the of the retention coefficients for the amino and the time for of and the time for a peptide The value of was from the retention time of a is a gradient time that to the time for the of the in to the The to the of the HPLC the of the and is the of the from the gradient to sample on the of the the for gradient or the for the the and the of the from to sample the gradient time for a separation can be or greater R. Pasa-Tolic L. L. R. Smith R.D. proteomics using liquid chromatography with high-performance ESI mass Chem. 2001; Scholar). The value can be in a gradient by a gradient with a of concentration at the time as sample found for the used in peptides than can be two at and in at in the and However, because the of peptides was retention times be by the peptides with retention times than for of were from the The model proposed by D. of peptide retention times in reversed-phase high-performance liquid chromatography of retention coefficients of amino of model Scholar) was as a for This model is based on the measurement of retention times of a of peptides, in a than based on These also applicability of the model to different with pore the as the used in the of peptide as the of the retention coefficients for the individual amino using the values by D. of peptide retention times in reversed-phase high-performance liquid chromatography of retention coefficients of amino of model Scholar). individual retention coefficients for amino as the optimization the R2 value for the of retention time for on the 346 peptides on the The values of the retention coefficients in were found to be to by D. of peptide retention times in reversed-phase high-performance liquid chromatography of retention coefficients of amino of model Scholar), as However, R2 value was greater than because the peptides in N-terminal amino and model was from a of N-terminal and of retention coefficients for individual amino from different and applicability for the of peptides used in present that the retention for in the present to peptides were used for of the R2 values to different that the retention for in the present to peptides were used for of the R2 values to different that the retention for in the present to peptides were used for of the R2 values to different in a that retention coefficients of individual amino acids depend on the of also a with as to a of In peptides were because used with to reduced proposed for were compared to the of the ion-pairing the that of than in the measured retention times from the values by the This to the the peptide than were of peptides amino at the were mostly found for peptides at the were with peptides or high and amino the retention time were most accurate for proteolytic of amino acids, and large peptides, of from the than for the separation of in the followed by retention of on the or a in the is on the of the followed by in an used the model to some of of TFA with of peptides, proteolytic peptides N-terminal amino the separation used in amino acids at the N-terminal of the peptide be from the with the of the This a in retention time for peptides with a and for with a for such an a of retention coefficients was introduced along with coefficients the of from the values of for optimization were by the retention coefficients of individual amino acids from the value of of a coefficients were as and for the and amino from the and the of the peptides were The individual N-terminal retention coefficients and the coefficients were to provide a R2 values for coefficients were found to be close to the optimization and The values of the N-terminal retention coefficients for amino are in values were found for amino acids and for high values were found for and of This is a of the of amino acids that the of the N-terminal amino This and the chromatographic of the in the separation and is to to the and to its for an in retention time for peptides N-terminal by of peptide retention times in liquid chromatography on the of amino S. A. Scholar, S. The isolation of peptides by high-performance liquid chromatography using Biochem. Scholar, D. of peptide retention times in reversed-phase high-performance liquid chromatography of retention coefficients of amino of model Scholar). of amino at the N-terminal and amino acids may retention by the ion-pairing ability of a peptide. of the of retention coefficients the R2 value from to that the of 346 peptides for of the retention coefficients as the and were in the N-terminal were in and It be to the of in to a accurate of for and the were to the of the amino in and the and the retention coefficients The is the of the and the of TFA for the and especially the In the retention time of a peptide is on its of peptide length on peptide retention in reversed-phase 1988; Scholar) found that the accuracy of peptide retention time prediction about D. of peptide retention times in reversed-phase high-performance liquid chromatography of retention coefficients of amino of model Scholar) a linear retention time the of peptide and of the of retention times of peptides to in length can be of the peptides, a the of peptide length and based on the of the was and an R2 value of was for the of 346 peptides for optimization model were to the of about the of N-terminal on peptide two amino acids at the were as an found that of peptides with a from the retention times the N-terminal from the final Such a The model was tested for chromatographic columns of sizes using different and of the linear for with R2 are in The 17-protein mixture was on the but with a gradient and fractions were of the peptides in HPLC were used to the retention time and The of the was the found in the separation was on a of using a of and a The of were the for columns gradient were However, different were because of in the gradient times The of a for sample injection additional to the a in and the was and found for the separation using injection via a of the model for different chromatographic and samples were a and with a linear gradient of water-ACN at that were different for samples some in a values may be of of a and was by separations of peptide mixtures with of of protein The were for the of peptides from in different HPLC were for a and as and injection using sample injection using tryptic of protein of of peptides from different tryptic of samples were a and with a linear gradient of water-ACN at that were different for samples some in a values may be of of a and was by separations of peptide mixtures with of of protein The were for the of peptides from in different HPLC were for a and as and in a the was to the of The sample a tryptic of protein was in using a HPLC and injection for the analysis of Identification of two in tryptic peptides from the sample protein and analysis of on on Mass and Scholar). was in that that of of and are for the in the and analytical columns also to the R2 value to tryptic of was in the of its This sample R2 for tryptic from and In the analysis of a tryptic of as a to the of different of peptides on retention a of the the ability to the retention times of peptides derived from a sample of the A. R. S. J. F. Mass spectrometric of proteins from the Cell. Scholar). the the of that be by an R2 value of was for the peptides These additional for the of the It was of to model for peptide retention time prediction to different as the applicability of some to be S. The isolation of peptides by high-performance liquid chromatography using Biochem. Scholar, D. of peptide retention times in reversed-phase high-performance liquid chromatography of retention coefficients of amino of model Scholar, M. M. J. of chromatographic retention and protein identification in liquid Chem. 2002; Scholar). The 17-protein mixture was on chromatographic and are based on with higher and R2 value of were for of separations have been using an combination with an ESI It was to the model for the off-line HPLC-MALDI MS with the ion-pairing used for separation of the 17-protein was the chromatographic linear gradient of water-ACN at with in The higher of the dependence the of the used in However, the R2 value the applicability of the model for separation with different of protein and often such as and peptides in a protein or peptides with or masses from different proteins. the to of However, that about peptide retention times derived from an additional for a of of and of predictive model in peptide mapping of one of the 17-protein of peptides m/z two peptides from were found in the tryptic of the mixture It be to of by in the at present F. Hendrickson mass of peptide for molecular mass Chem. 2001; Scholar), and of the peptides be by mass HPLC a method of the as well as based on the chromatographic the peptides in the to m/z range in peptide mass mapping using the peptides of the analysis that peptides were but in the and retention times by than such be tryptic in the m/z range by the retention times based on the peptide were of were by of the in the two of were of protein was found for protein in the analysis of the Such a capability the of the off-line HPLC-MALDI MS combination for proteomic of retention time prediction for peptide mapping of measured of peptides with close to MS by to in a The retention time prediction can also be used for the of protein The has been to have a on the chromatographic retention of peptides S. The isolation of peptides by high-performance liquid chromatography using Biochem. Scholar), with most peptides be found in the fractions as using gradient and This identification of the on in the sample and injection for the analysis of Identification of two in tryptic peptides from the sample protein and analysis of on on Mass and Scholar). In found a the and the of the to the identification of peptides is a to peptides of usually than gradient was by S. The isolation of peptides by high-performance liquid chromatography using Biochem. Scholar). The peptide that the peptide mapping was to be found in fractions and with its in fractions and of from fractions the of of by the mass of a This is still for identification of peptides, because than were in the 17-protein tryptic However, in to the the peptide has a that can be by its chromatographic N-terminal and a and of N-terminal of 17 from peptides and in peptide Proteome Res. 2002; 1: Scholar, of peptides N-terminal by high-performance liquid laser desorption/ionization Mass 17: Scholar). of a mass and from The of a was and for and of peptides N-terminal by high-performance liquid laser desorption/ionization Mass 17: Scholar). in of the than on the of in fractions and the of can be at the chromatographic that the peptides in most have a at the The mass was by the mass of the tryptic from the mass of the peptide in and the of the was to be of the peptides in was confirmed by have an improved model for prediction of retention times of tryptic peptides in The model was from a of 346 peptides in a HPLC-MALDI MS R2 values of were for separations different chromatographic different and the applicability of the to peptide
Krokhin et al. (Tue,) studied this question.