A Calibration Method That Simplifies and Improves Accurate Determination of Peptide Molecular Masses by MALDI-TOF MS

详细信息    查看全文

• 作者：Johan Gobom ; Martin Mueller ; Volker Egelhofer ; Dorothea Theiss ; Hans Lehrach ; and Eckhard Nordhoff
• 刊名：Analytical Chemistry
• 出版年：2002
• 出版时间：August 1, 2002
• 年：2002
• 卷：74
• 期：15
• 页码：3915 - 3923
• 全文大小：464K
• 年卷期：v.74,no.15(August 1, 2002)
• ISSN：1520-6882

文摘

The use of delayed ion extraction in MALDI time-of-flightmass spectrometry distorts the linear relationship between m/z and the square of the ion flight time (t²) withthe consequence that, if a mass accuracy of 10 ppm orbetter is to be obtained, the calibrant signals have to fallclose to the analyte signals. If this is not possible,systematic errors arise. To eliminate these, a higher-ordercalibration function and thus several calibrant signals arerequired. For internal calibration, however, this approachis limited by signal suppression effects and the increasingchance of the calibrant signals overlapping with analytesignals. If instead the calibrants are prepared separately,this problem is replaced by an other; i.e., the ion flighttimes are dependent on the sample plate position. For thisreason, even if the calibrants are placed close to thesample, the mass accuracy is not improved when a higher-order calibration function is applied. We have studied thisphenomenon and found that the relative errors, whichresult when moving from one sample to the next, aredirectly proportional to m/z. Based on this observation,we developed a two-step calibration method, that overcomes said limitations. The first step is an externalcalibration with a high-order polynomial function used forthe determination of the relation between m/z and t², andthe second step is a first-order internal correction forsample position-dependent errors. Applying this method,for instance, to a mass spectrum of a mixture of 18peptides from a tryptic digest of a recombinant proteinresulted in an average mass error of 1.0 ppm with astandard deviation of 3.5 ppm. When instead using aconventional two-point internal calibration, the averagerelative error was 2.2 ppm with a standard deviation of15 ppm. The new method is described and its performance is demonstrated with examples relevant to proteome research.