Short peptides are of extreme interest in food research, due to their wide range of biological activity they are well-known to exert. However, they still represent an often unresolved analytical issue. The main aim of this project is the development of an analytical platform for a considerable advancement in short peptides identification. For the first time, short sequences presenting both natural and post-translationally modified amino acids are going to be comprehensively studied, with a particular attention to extraction, purification, data acquisition, and compound identification. Data acquisition and compound identification are the most critical aspects, as the standard technologies borrowed from proteomics are not efficient for short peptide sequences. For this purpose, specific short peptide databases can be generated. Short peptide databases will have a dual purpose. First, they will be employed as inclusion list for a suspect screening mass-spectrometric analysis, overcoming the limits of data-dependent acquisition mode and allowing the fragmentation of such low-abundance substances. Moreover, the databases will be implemented in Compound Discoverer 3.1, a software dedicated to the analysis of short molecules, for the creation of a data processing workflow specifically dedicated to short peptide identification. For this purpose, a detailed study of short peptide fragmentation pathways will be carried out for the first time, with the purpose of elucidating the differences between natural and modified short peptide sequences and the highly studied medium-sized tryptic peptides.
The main goal of the project is the development of an analytical workflow for the comprehensive characterization of natural and modified short bioactive peptide sequences. In our previous work on the development of a purification and enrichment strategy, a graphitized carbon black sorbent was found to be very effective in retaining short peptides, whose polarity causes C18-SPE to be highly ineffective, by hydrophobic and electrostatic interactions of short peptide sequences with the delocalized aromatic systems of the GCB. In comparison to the state of art, in which extensive manual search of short peptide sequences in the large datasets deriving from high-resolution mass-spectrometry is required, the issue of data analysis and untargeted identification of short peptides will be met.
First, a detailed study of the characteristic fragmentation pathways of short peptides will be achieved using analytical standard peptides which will be chosen with the purpose of covering all physicochemical properties (molecular weight, polarity, and acid-base properties). A comprehensive database of natural and modified short amino acid sequences will be generated by MATLAB. Other than the 20 natural amino acids, 13 modified residues and lactic acid will be employed, according to the most common modified residues. The thirteen selected residues are hydroxyproline, hydroxylysine, methyllysine, dimethyllysine, trimethyllysine, acetyllysine, succinyllysine, methylarginine, citrulline, pyroglutamic acid, phosphoserine, tyrosine sulfate, and methionine sulfoxide. As regards data collection, LC-HRMS analysis will be performed. Preliminary experiments will be carried out for determining the optimum separation strategy. Depending on the nature of the endogenous sequences (and, in particular, the medium GRAVY index of the peptides), C18-RP or HILIC separation could furnish better results. HILIC columns, and in particular those with zwitterionic phases, have been proven to guarantee good results in separating small polar metabolites. Whereas RP columns usually provide higher performance in terms of peak shape and reproducibility, HILIC separation could still provide information on extremely hydrophilic peptides, which uniquely present residues with charged side chains (i.e., glutamic acid, aspartic acid, lysine, histidine, and arginine).
As regards MS data acquisition, the analysis of short endogenous has still several issues to be addressed. When employing orbitrap-based instrumentations, in fact, untargeted approaches are commonly performed with data-dependent acquisition (DDA) methods, since more valuable data-independent acquisition (DIA) approaches are highly time-consuming and show weak performances in such slow instruments. However, DIA approaches, like all ion fragmentation (AIF), would grant the MS/MS fragmentation of all eluting precursor ions in a predefined isolation window, including low-abundance species like short peptides. On the other hand, DDA methods, in which top N ranked precursor ions are sequentially isolated and fragmented, would repeatedly cause high-abundance species to suppress less concentrated coeluting compounds, which would not be fragmented and, eventually, identified. When precursor ion databases are available, suspect screening MS approaches constitute a valuable alternative to DIA, granting the selective fragmentation of precursor ions present in the inclusion list and overcoming the limits of DDA mode. Thus, many low-abundance species, that would have normally been neglected, could be fragmented and manually validated.
Finally, a customized data analysis workflow on Compound Discoverer software will be set up to greatly ease the manual spectral validation of short peptides. The customized data analysis will grant easier, faster, and deeper identification of natural and short peptides, based on the match of m/z to those present in the aforementioned database and match of MS/MS spectra to the typical product ions deriving from the 34 considered amino acid residues. The proposed approach would be the first customized data processing workflow specifically dedicated to this class of compounds with applicability that is independent of the matrix under exam.