Abstract
A robust peptide quantification method was developed where overlapping peptide isotopic distributions were fit with predicted peptide isotopic envelope mixture models (IEMMs). Application to two difficult quantitative problems was demonstrated. The first was the quantification of deamidation, where masses of isotopic peaks differ by 1 Da, and the second was (18)O labeling, where the isotopic peaks are shifted 2 and 4 Da. In both cases, peptide quantification cannot be performed by simple integration of extracted ion chromatograms, because the isotopic envelopes of mass-shifted peptides are normally not resolved. To test the methodology for quantification of deamidation, several synthetic peptides and their corresponding deamidated forms were mixed at various ratios (1:0, 1:2, 2:1, 4:1, 10:1, and 20:1) and analyzed using the IEMM method, resulting in a high correlation (R(2) = 0.96) between measured and known percentages of deamidation. The IEMM method was then incorporated into a workflow for deamidation quantification in a large-scale proteomics experiment. A series of normal (3 day, 2 year, 35 year, and 70 year) and cataractous (93 year) human lenses were analyzed using two-dimensional liquid chromatography tandem mass spectrometry, and deamidation quantities of several gammaS-crystallin peptides ([N14-Q16], N53, [Q63-Q70], and N143) were determined. Two peptides (N53 and [Q63-Q70]) had more extensive deamidation in the water-insoluble portions of normal lens samples, and deamidation at N143 was more extensive in the 93 year water-insoluble cataractous sample. The utility of the technique for analysis of (18)O-labeled peptides was examined using mixtures of labeled BSA peptides in known (16)O/(18)O ratios (10:1, 4:1, 1:1, 1:4, and 1:10). The methodology allowed for accurate measurements of ratios of (16)O/(18)O peptides over the wide range of relative abundances.