Metabolite profiling using gas chromatography coupled to mass spectrometry (GC–MS) is one of the most frequently applied and standardized methods in research projects using metabolomics to analyze complex samples. However, more than 20 years after the introduction of non-targeted approaches using GC–MS, there are still unsolved challenges to accurate quantification in such investigations. One particularly difficult aspect in this respect is the occurrence of sample-dependent matrix effects. In this project, we used model compound mixtures of different compositions to simplify the study of the complex interactions between common constituents of biological samples in more detail and subjected those to a frequently applied derivatization protocol for GC–MS analysis, namely trimethylsilylation. We found matrix effects as signal suppression and enhancement of carbohydrates and organic acids not to exceed a factor of ~2, while amino acids can be more affected. Our results suggest that the main reason for our observations may be an incomplete transfer of carbohydrate and organic acid derivatives during the injection process and compound interaction at the start of the separation process. The observed effects were reduced at higher target compound concentrations and by using a more suitable injection-liner geometry.

Protein carbonylation, a marker of excessive oxidative stress, has been studied in the context of multiple human diseases related to oxidative stress. The variety of post-translational carbonyl modifications (carbonyl PTMs) and their low concentrations in plasma challenge their reproducible identification and quantitation. However, carbonyl-specific biotinylated derivatization tags (e.g., aldehyde reactive probe, ARP) allow for targeting carbonyl PTMs by enriching proteins and peptides carrying these modifications. In this study, an oxidized human serum albumin protein model (OxHSA) and plasma from a healthy donor were derivatized with ARP, digested with trypsin, and enriched using biotin-avidin affinity chromatography prior to nano reversed-phase chromatography coupled online to electrospray ionization tandem mass spectrometry with travelling wave ion mobility spectrometry (nRPC-ESI-MS/MS-TWIMS). The presented workflow addresses several analytical challenges by using ARP-specific fragment ions to reliably identify ARP peptides. Furthermore, the reproducible recovery and relative quantitation of ARP peptides were validated. Human serum albumin (HSA) in plasma was heavily modified by a variety of direct amino acid oxidation products and adducts from reactive carbonyl species (RCS), with most RCS modifications being detected in six hotspots, i.e., Lys10, Lys190, Lys199, Lys281, Lys432, and Lys525 of mature HSA.

Bovine milk contains a variety of endogenous peptides, partially formed by milk proteases that may exert diverse bioactive functions. Milk storage allows further protease activities altering the milk peptidome, while processing, e.g., heat treatment can trigger diverse chemical reactions, such as Maillard reactions and oxidations, leading to different posttranslational modifications (PTMs). The influence of processing on the native and modified peptidome was studied by analyzing peptides extracted from raw milk (RM), ultra-high temperature (UHT) milk, and powdered infant formula (IF) by nano reversed-phase liquid chromatography coupled online to electrospray ionization (ESI) tandem mass spectrometry. Only unmodified peptides proposed by two independent software tools were considered as identified. Thus, 801 identified peptides mainly originated from α_S- and β-caseins, but also from milk fat globular membrane proteins, such as glycosylation-dependent cell adhesion molecule 1. RM and UHT milk showed comparable unmodified peptide profiles, whereas IF differed mainly due to a higher number of β-casein peptides. When 26 non-enzymatic posttranslational modifications (PTMs) were targeted in the milk peptidomes, 175 modified peptides were identified, i.e., mostly lactosylated and a few hexosylated or oxidized peptides. Most modified peptides originated from α_S-caseins. The numbers of lactosylated peptides increased with harsher processing.

Sample preparation is the most critical step in proteomics as it directly affects the subset of proteins and peptides that can be reliably identified and quantified. Although a variety of efficient and reproducible sample preparation strategies have been developed, their applicability and efficacy depends much on the biological sample. Here, three approaches were evaluated for the human milk and plasma proteomes. Protein extracts were digested either in an ultrafiltration unit (filter-aided sample preparation, FASP) or in-solution (ISD). ISD samples were desalted by solid-phase extraction prior to nRPC-ESI-MS/MS. Additionally, milk and plasma samples were directly digested by FASP without prior protein precipitation. Each strategy provided inherent advantages and disadvantages for milk and plasma. FASP appeared to be the most time efficient procedure with a low miscleavage rate when used for a biological sample aliquot, but quantitation was less reproducible. A prior protein precipitation step improved the quantitation by FASP due to significantly higher peak areas for plasma and a much better reproducibility for milk. Moreover, the miscleavage rate for milk, the identification rate for plasma, and the carbamidomethylation efficiency were improved. In contrast, ISD of both milk and plasma resulted in higher miscleavage rates and is therefore less suitable for targeted proteomics.

Thermal treatments of dairy products favor oxidations, Maillard reactions, and the formation of sugar or lipid oxidation products. Additives including flavorings might enhance these reactions or even induce further reactions. Here we aimed to characterize protein modifications in four flavored milk drinks using samples along the production chain—raw milk, pasteurization, mixing with flavorings, heat treatment, and the commercial product. Therefore, milk samples were analyzed using a bottom up proteomics approach and a combination of data-independent (MS^E) and data-dependent acquisition methods (DDA). Twenty-one small carbonylated lipids were identified by shotgun lipidomics triggering 13 protein modifications. Additionally, two Amadori products, 12 advanced glycation end products (AGEs), and 12 oxidation-related modifications were targeted at the protein level. The most common modifications were lactosylation, formylation, and carboxymethylation. The numbers and distribution of modification sites present in raw milk remained stable after pasteurization and mixing with flavorings, while the final heat treatment significantly increased lactosylation and hexosylation in qualitative and quantitative terms. The processing steps did not significantly affect the numbers of AGE-modified, oxidized/carbonylated, and lipid-carbonylated sites in proteins.