Protein Aroma Interactions

In terms of chemical interactions, reversible weak hydrophobic interactions, as well as stronger ionic effects and irreversible covalent bonds, may be formed (e.g., the reaction of aldehydes with the NH2 and SH groups of proteins) between aroma compounds and proteins. The interactions one would expect are influenced by the type and amount of protein (amino acid composition), types of flavoring components. 

Aroma interactions with lipids, proteins and carbohydrates affect the retention of volatiles within the food and, thereby, the levels in the gaseous phase. Consequently, the interactions affect the intensity and quality of food aroma. Since such interactions cannot be clearly followed in a real food system, their study has been transferred to model systems which can, in essence, reliably imitate the real systems. Consider the example of emulsions with fat contents of 1%, 5% and 20%, which have been aromatized with an aroma cocktail for mayonnaise consisting of diacetyl, (Z)-3-hexenol, (E,Z)-2,6-nonadienol, allyl isothiocyanate and allyl thiocyanate. The sample with 20% of fat has the typical and balanced odor of mayonnaise (Fig. 5.32 a). If the fat content decreases, the aroma changes drastically. The emulsion with 5% of fat has an untypical creamy and pungent odor since there is a decrease in the intensities of the buttery and fatty notes in the aroma profile (Fig. 5.32 b). In the case of 1% of fat, pungent, mustard-like aroma notes dominate (Fig. 5.32 c). 

Flavor perception results from interactions between a consumer and stimulants in a food. For the aroma part of flavor, the stimulants are volatiles that bind to receptor proteins found on the olfactory epithelium. These stimulants can reach the receptors by two routes, orthonasal or retronasal. The retronasal route is used when odorants are drawn from the mouth during eating through the nasal pharynx to produce aroma. 

During AEDA, interactions between the odorants are not taken into consideration, since every odorant is evaluated individually. Therefore, it may be possible that odorants are recognized which are possibly masked in the food flavor by more potent odorants. Furthermore, the odor activity values only partially reflect the situation in the food, since OAVs are mostly calculated on the basis of odor thresholds of single odorants in pure solvents. However, in the food system, the threshold values may be influenced by nonvolatile components such as lipids, sugars or proteins. The following examples will indicate that systematic sensory model studies are important further steps in evaluating the contribution of single odorants to the overall food aroma. 

In this study the physical parameters involved in interaction of a major class of meat flavorants, methyl pyrazines, with soy proteins were determined at meat roasting temperatures. Beef diffusate, the water soluble, low molecular weight fraction that constitutes about IX of beef, was shown to contain the necessary precursors to obtain a desirable, thermally generated meat aroma (8). Diffusate was heated under controlled conditions and generated volatiles were transferred to a gas chromatograph for separation and quantitation. Methyl pyrazines, either from heated diffusate or from standard solutions, were measured in the presence of purified soy proteins and the thermodynamics of binding were determined. 

Partitioning of volatile substances between the liquid and gas phases is mainly governed by aroma compound volatility and solubility. These physicochemical properties are expected to be influenced by wine constituents present in the medium, for instance polysaccharides, polyphenols, proteins among others. Consideration of the physicochemical interactions that occur between aroma compounds and wine constituents is necessary to understand the perception of wine aroma during consumption. The binding that occurs at a molecular level reflects changes at a macroscopic level of the thermodynamic equilibrium, such as volatility and solubility, or changes in kinetic phenomena. Thus, thermodynamic and dynamic approaches can be used to study the behaviour of aroma compounds in simple (model) or complex (foods) media. 

Many of the wine macro-components (e.g. carbohydrates, proteins, polyphenols), come from the skins and the pulp of grapes and from the cell walls of the yeast. Although this varies, the molecular weight of the majority of macromolecules is over 10,000 D and their final concentration ranges from 0.3 to 1 g/L (Voilley et al. 1991). Most macromolecules will be eliminated by clarification and stabilization treatments of the wine. Because of their interactions with wine aroma... 

Dufour and Bayonove (1999a) reported two criteria for polysaccharide discrimination acidity and protein content. Neutral peptic substances (type II arabinogalac-tans and arabinogalactans-proteins) represent 40% of the polysaccharides in wine and acidic pectic polysaccharides, (e.g. homogalacturonans and rhamnogalacturo-nans) account for 20% of them. Because of the difficulty in purifying wine polysaccharides, most of the studies on interactions between wine polysaccharides and aroma compounds have been carried out with exocellular and cell wall mannoproteins (thus mainly glycoproteins) of Saccharomyces (see effect of yeast and derivatives in the next section). 

Other than studies on the role of proteins released by yeast during autolysis (mannoproteins) on wine aroma, little work has been reported on interactions of other proteins with aroma compounds. One study investigating such interactions was published by Druaux et al. (1995). They used synthetic wines and bovine serum albumin (BSA) as a model protein. This protein was found to bind 5-decalactone and there was greater binding when in water than in a model wine environment (pH 3.5 and 10% ethanol). To our knowledge this is the only study focused on elucidating the effect of proteins (others than mannoproteins) on the aroma release in wine or model wine. 

Food flavor is a very important parameter influencing perceived quality. The volatile compounds contributing to the aroma of foods possess different chemical characteristics, such as boiling points and solubilities and the sensory properties of food cannot be understood only from the knowledge of aroma composition. This can be explained by interactions between flavor compounds and major constituents in food such as fat, proteins and carbohydrates (1). A number of different interactions has been proposed to explain the association of flavor compound with other food components. This includes reversible Van der Waals interactions and hydrogen bonds, hydrophobic interactions. The understanding of interactions of flavor with food is becoming important for the formulation of new foods or to... 

Fats act as precursors to flavor development by interacting with proteins and other ingredients when heated. Off-flavors are not normally perceived in full-fat systems because most of them are fat-soluble. However, in the absence of fat, the vapor pressure of the aroma chanical in water is increased, resulting in a very intense perception of the off-flavor chemical. Flavor release is a critical factor governing smell and taste. The majority of flavor components are dissolved to some extent in the lipid phase of food— releasing the flavor slowly in the mouth and resulting in a pleasant aftertaste [75]. 

The structure of the food matrix is also known to affect the release of volatile compounds having an impact on flavors and aroma. Changes in flavor result from the interactions of lipid-derived carbonyl compounds by aldolization with the amino groups of proteins. Undesirable flavors are produced when beef or chicken are fried in oxidized fats by the interaction of secondary lipid oxidation... 

Recent studies of photooxidized butter and butter oil identified by aroma extract dilution analysis, 3-methylnonane-2,4-dione, a potent volatile compound derived from furanoid fatty acids (see Section C.4) (Figure 11.7). Six different furanoid fatty acids were established as dione precursors, and were found in various samples of butter made from either sweet cream (116 76 mg/ kg), or from sour cream (153-173 mg/kg), or from butter oil (395 mg/kg). Similar precursors of the dione were identified in stored boiled beef and vegetable oils. This flavor defect arising by photooxidation of butter or butter oil is apparently different from the light-activated flavor in milk,that involves the interaction of sulfur-containing proteins and riboflavin. However, more sensory comparisons are needed to distinguish between these two flavor defects due to light oxidation. 

The volatiles produced from an extruded corn-based model system has been studied by Ho et al (1989). In this model system, zein (a corn protein), corn amylopectin, and corn oil were mixed and extruded using a single-screw extruder at 120 or 165 C. The extruded samples were ground and extracted with ethyl ether, the volatiles were analyzed by GC and GC/MS. For comparison, the mixture was also baked in an oven at 120 or 180°C for 30 min. The results so obtained are presented in Table 2. In all cases examined, the higher temperature resulted in the formation of higher amounts of all volatiles. It is also interesting to note how different ingredients interacted to produce different aroma volatiles. 

Mannoproteins are complex hydrocolloids released from yeast cell walls during autolysis (Goncalves et al., 2002 Charpentier et al., 2004). According to Feuillat (2003), mannoproteins are important to wine quality as these contribute to protein and tartrate stability, interact with aroma compounds, decrease the astringency and bitterness of tannins, and increase the body of wine. For instance, Dupin et al. (2000) reported that mannoproteins prevent protein haze formation. Using a model wine. Lubbers et al. (1994) noted that yeast cell walls bound volatile aroma compounds, especially those more hydrophobic, and could potentially change the sensory characteristics of wines through losses of these aromas. 

---