Covalent Adduct Formation between ß-Lactoglobulin and Flavor Compounds under Thermal Treatments That Mimic Food Pasteurization or Sterilization.

Thermal processing (e.g., pasteurization and sterilization) is a critical step ensuring the microbial safety of our foods. Previous work from our laboratory has examined the covalent reactions occurring between proteins and a broad selection of flavor compounds under ambient storage temperatures (25-45 °C). However, similar research on reactions of flavor compounds with a protein under thermal processing conditions has not been investigated. In the current study, covalent adduct formation between ß-lactoglobulin (BLG) and 46 flavor compounds encompassing 13 different classes of functional groups was investigated under pasteurization and sterilization conditions by UPLC-ESI-QTOF-MS. BLG was chosen as a representative protein for this study because it is structurally well characterized, its molecular weight is well suited for ESI-MS analysis (18.2 kDa), and it is broadly used in the food industry. Schiff base, aza-Michael addition, and disulfide linkages were the main types of covalent interactions occurring across the reactive samples. Among them, isothiocyanates, aldehydes, and thiol-containing compounds were generally very reactive. Increasing the severity of the thermal treatment [high-temperature-short-time (HTST) pasteurization, in-container pasteurization (IC), and ultra-high-temperature (UHT) sterilization conditions] accelerated the reactions of BLG with flavor compounds, which revealed reactivity of three flavor compounds not previously observed to react at room temperature (eugenol, 4-vinyl phenol, and 3-nonen-2-one). Ketones [other than 2-hydroxy-3-methyl-2-cyclopenten-1-one (cyclotene), diketones, and unsaturated ketones], alcohols, acids, alkenes (terpenes), esters, lactones, 3-acetylpyridine, methyl anthranilate, vanillin, 2-methylthiophene, and dimethyl sulfone did not show measurable reactivity with BLG under the thermal processing conditions examined. An overall view of the data shows that the HTST heat treatment (72 °C for 15 s) had the least effect on the extent of reaction while in-container pasteurization conditions (63 °C for 30 min) produced a similar extent of reaction as the UHT (130 °C 30 s) heat treatment. These varying extents of adductation are in reasonable accord with what one might expect, given that the rates of most classes of chemical reactions occurring near ambient temperature increase by a factor of 2-4 for each increase of 10 K in temperature. Unfortunately, our methodology did not permit us to obtain meaningful data using the most aggressive standard sterilization thermal conditions (110 °C for 30 min) because extensive aggregation/coagulation removed essentially all of the BLG protein from the reaction mixtures prior to MS analysis.

Encapsulation of Orange Oil Using Fluidized Bed Granulation.

The primary objective of this research is to determine how granulation compares to spray drying/agglomeration for producing larger, more dense flavoring particles. Granulation can yield large, dense particles and thereby negate the need for a two-step process (spray drying followed by agglomeration) to achieve improved flow/handling properties of dry flavorings. In this study, a 55% solids slurry (blend of OSAn-modified starch and maltodextrin 15DE) was prepared and then single-fold orange peel oil was added at 20 or 25% of the carrier solids level. The 20% flavoring emulsion was spray dried (SD), and a portion of the resultant powder then agglomerated (Agg) in a bottom spray, fluidized bed. A second emulsion of the same carrier composition but using 25% orange oil based on carrier solids was prepared and subjected to fluidized bed granulation (FBG). Particle size, density, orange oil retention and oxidative stability on storage were determined. Overall, it is observed during this study that FBG produces orange oil encapsulates that possess better properties, such as more resistance to oxidation, a better retention of orange oil and a higher density than SD or SD/Agg powders.

Influence of pH, Temperature, and Water Activity on Covalent Adduct Formation between Selected Flavor Compounds and Model Protein ß-Lactoglobulin.

This study investigates the influence of pH, temperature, and water activity on the occurrence of covalent adduct formation between select flavor compounds and a model food protein (ß-lactoglobulin). These reactions potentially result in the loss of flavor during processing and storage, reducing consumer acceptability. Foods present a diverse reaction environment encompassing a wide range of aw, pH, and storage temperature, which potentially influence protein: flavor reaction rates. Liquid chromatography/mass spectrometry (LC/MS) data showed that covalent adducts were formed more slowly at low pHs (3) than basic pHs (8) (for citral, allyl isothiocyanate, and dimethyl trisulfide). No reactivity was observed for benzaldehyde at pH 3, but substantial reactivity was found at pHs 7 and 8. The amount of adducts formed increased with an increase in storage temperature. Higher temperatures (45 °C) led to the formation of products that were not observed at lower temperatures (4 and 20 °C). An increase in water activity (0.11-0.75) led to an increase in formation of adducts for allyl isothiocyanate. There were no observable differences in adduct formation as a function of aw for benzaldehyde, citral, and dimethyl disulfide. However, this lack of observed effect may be due to the rate of reaction being too slow to be detected in the timeframe of this study.

Covalent Adduct Formation Between Flavor Compounds of Various Functional Group Classes and the Model Protein ß-Lactoglobulin.

The formation of covalent bonds between 47 flavor compounds belonging to 13 different classes of functional groups and ß-lactoglobulin (BLG) has been evaluated using electrospray ionization protein mass spectrometry. Covalent bond formation was determined by the appearance of ions in the mass spectra corresponding to BLG + flavor molecule(s). The observed processes for covalent bond formation were Schiff base, Michael addition, and disulfide linkages. Some reactions resulted in protein cross-linking. Aldehydes, sulfur-containing molecules (especially thiols), and functional group-containing furans were the most reactive flavor components. The thiol-containing compounds cleaved one or both electrophilic disulfide linkages in BLG to form disulfide linkages and the sulfides formed covalent bonds with the free cysteine group. Ketones were generally stable, but α-diketones (e.g., diacetyl) were reactive. Some bases (e.g., pyrazines and pyridines) were interactive, while the nucleophilic allylamine was reactive. Hydrocarbons, alcohols, acids, esters, lactones, and pyrans did not give observable levels of adduct formation within the period studied. The formation of covalent bonding (flavor protein) is potentially responsible for the loss of flavor, limiting the shelf-life of many foods.

Method To Characterize and Monitor Covalent Interactions of Flavor Compounds with ß-Lactoglobulin Using Mass Spectrometry and Proteomics.

This study develops a method to measure the covalent bonds formed between the side chains and terminal amino acids of ß-lactoglobulin (BLG) and selected flavor molecules (benzaldehyde, citral, or allyl isothiocyanate) using electrospray ionization mass spectrometry (ESI/MS) and tandem mass spectrometry (MS/MS). This technique made it possible to measure increases in molecular weight of BLG as the reaction takes place (BLG + flavor compound). The observed mass shifts on the reaction corresponded to either Schiff base or Michael addition reactions between the chosen flavor compounds and BLG. In the case of citral, sodium dodecyl sulfate polyacrylamide gel electrophoresis analysis revealed that these reactions lead to protein cross-linking. A proteomic approach using MS/MS to identify the sites of post-translational modification between benzaldehyde and BLG revealed that the lysine groups were the reaction sites. Interestingly, benzaldehyde was found to react with several different lysine groups but never more than one of them per BLG molecule (BLG contains 15 lysine groups/molecule). Furthermore, adducts with benzaldehyde were not observed at two lysine groups. Allyl isothiocyanate was found to react with several sites on each BLG molecule. The ESI/MS methodology in tandem with proteomics yields a detailed view of flavor/BLG interactions that may offer insights on minimizing these undesirable reactions in the future.