Acta Scientiarum Polonorum Technologia Alimentaria

Background. Peptides are important components of foods mainly due to their biological activity. The basic method of their identification is reversed phase high-performance liquid chromatography coupled with electrospray-ionization mass spectrometry (RP-HPLC-ESI-MS). Retention time (t_R) prediction in silico is very helpful in analysis of multicomponent peptide mixtures. One of problems associated with RP-HPLC-ESI-MS is deterioration of mass spectra quality by trifluoroacetic acid (TFA). This problem can be avoided through the use of chromatographic columns designed for work with low TFA concentrations in mobile phase. The objective of this study was to determine the correlations between peptide retention times predicted with the use of a program available on-line and experimental retention times obtained using the column working with low TFA concentrations.

Material and methods. The set of synthetic peptides and bovine α-lactalbumin fragments (18 peptides) was used in the experiment. Theoretical retention times were calculated using Sequence Specific Retention Calculator (SSRC) program. The experimental retention times were measured via RP-HPLC-ESI-MS method using column working with low TFA content. The dependence between theoretical and experimental t_R was expressed via empirical equations.

Results. The best correlation between theoretical and experimental retention times of peptides containing at least four amino acid residues has been obtained when third order polynomial (R² = 0.9536). Prediction quality for di- and tripeptides was significantly lower. The method described can be applied for cysteine-containing peptides although our sample preparation procedure did not include modification of this amino acid, taken into attention by SSRC program.

Conclusions. The results of this study validate the usefulness of a third degree polynomial as a simple function describing the correlation between peptide retention times predicted by an on-line application and experimental retention times. The above function can effectively predict retention times in situations when experimental conditions differ from the computational environment (various columns, mobile phase composition, use or resignation from chemical modifications during sample preparation, various HPLC equipments). On-line available t_R predicting application with correction based on user’s own data may be a useful tool in food peptidomics.