A Comparative Analysis of the Influence of Human Salivary Enzymes on Odorant Concentration in Three Palm Wines

The influence of human salivary enzymes on palm wines’ odorant concentrations were investigated by the application of aroma extracts dilution analysis (AEDA) and by the calculation of odour activity values (OAVs), respectively. The odorants were quantified by means of stable isotope dilution assays (SIDA), and the degradation profiles of odorants by human saliva were also studied. Results revealed 46 odour-active compounds in the flavour dilution (FD) factor range of 4-256, and all were subsequently identified. Of the 46 odorants, 41 were identified in the Elaeis guineensis wine, 36 in Raphia hookeri wine and 29 in Borassus flabellifer wine. Among the odorants, the highest FD-factors were obtained from acetoin, 2-acetyl-1-pyrroline and 3-isobutyl-2-methoxypyrazine. Among the 13 potent odorants identified, five aroma compounds are reported here as important contributors to palm wine aroma, namely 3-isobutyl-2-methoxy-pyrazine, acetoin, 2-acetyl-1-pyrroline, 3-methylbutylacetate and ethyl hexanoate. Meanwhile, salivary enzymic degradation of odorants was more pronounced among the aldehydes, esters and thiols.


Introduction
Wine aroma lingers for a considerable time after consumption. Precise sensory evaluation of aroma persistence is rare, mainly due to the fact that determination of perception duration and the exact end point poses some difficulties. Wine aroma perception is a complex sensation triggered when volatile compounds are transported to the olfactory epithelium during wine tasting and consumption.

OPEN ACCESS
ability of these enzymes to degrade some selected odorous esters, thiols and aldehydes was reported earlier [10]. The influence of saliva macromolecules, such as proteins on the volatility of several odorants has been documented [16,17]. Saliva affects odorant concentration by means of chemical and biochemical reactions between its components and food volatiles. Evidence has been found for the partial hydrolysis of several odor-active acetates [18], as well as ethylic esters, according to their chemical structures. It has been reported that some compounds such as benzaldehyde, diacetyl, ethyl hexanoate and heptyl acetate are affected by the interaction between mucin and the type of solute present [19]. Mucins are high molecular mass glycoproteins responsible for the typical viscosity and elasticity of saliva. They have binding sites, preferentially occupied by sucrose and these sites are also available to trap volatiles [19]. In fact, mucin can bind specific aroma compounds, principally aldehydes [19,20], to form Schiff bases. While our previous study [5] elucidates the influence of human saliva on wine from E. guineensis, there are no similar studies on other wines obtained from other palm trees. The present study is aimed at correlating and comparing the effect of human salivary enzymes on key odorants of wines from three different palm trees (Elaeis guineensis, Raphia hookeri, and Borassus flabellifer). Table 1 shows the results of odour qualities and the retention indices of solvent-extracted palm wines and palm wines incubated with saliva (pH 7.8-8.0, 10 min). A total of 46 odorants of which 41 were identified in the oil palm wine (Elaeis guineensis, EW) and another 36 in raphia wine (Raphia hookeri, RW). On the other hand, the Borassus flabellifer wine (BW) yielded only 29 odorants. With the exception of very few odorants, the aroma profiles of EW and RW were quite similar. However, the aroma profile of Borassus wine (BW) showed distinct differences to those of EW and RW respectively. The numbers of odorants identified in wines incubated with saliva varied slightly. While a total of 30 odorants were identified in EW, RW and BW yielded 27 and 26 odorants, respectively.

Aldehydes and Esters
In contrast to the pyrazines and alcohols, there were significant degradations in aldehydes and esters incubated with saliva. Figure 1 gives an insight into the influence of saliva on aldehydes, esters and alcohols. Methional was reduced by approximately 19% after 10 min of incubation with saliva. Moreover, the decrease of methional was related to the formation of methionol (Figure 2), indicating that reduction after incubation is the obvious reaction occurring with saliva. 3-Methylbutanal was found to follow the same reaction, being reduced to 3-methylbutanol. The experiment on 3-methylbutanal was repeated three times with three different samples of saliva from one panellist on three different days. It was found that the enzymic degradation of 3-methylbutanal could vary by 10%-18% from one day to the other, indicating that reductive salivary activity for one panellist is not fully consistent. This effect has already been previously observed for the degradation of model homologous aliphatic aldehydes in the presence of saliva [12]. After thermal treatment of the saliva (100 °C, 10 min), no degradation of the aldehydes was observable (data not shown). At 100 °C, other effects could have occurred such as protein denaturation, protein aggregation and precipitation or change of the physicochemical properties of the saliva, in particular viscosity. These events could also explain some of the obtained results. The factors inducing salivary reduction of the investigated aldehydes cannot be confirmed at present. In a previous study on white and red wines, Friel and Taylor [19] reported significant interaction between aldehydes and saliva mucin. They showed that aldehydes can bind to mucin to form Schiff bases. Generally, the two major metabolic pathways for aldehydes in human beings are; oxidation to carboxylic acids and reduction to the corresponding alcohols, with the first being catalysed by NAD-linked alcohol dehydrogenases and by NADP-linked aldehyde reductases [23]. The esters (ethyl hexanoate and 3-methylbutyacetate) were more degraded than the aldehydes. The most probable factor responsible for the degradation of esters is hydrolysis, as many esterolytic enzymes can be found in human saliva [24]. The decrease of esters in wine with human saliva has been attributed to carboxylesterases [14,18]. Nevertheless, the action of mucin cannot be excluded. It's possible that saliva mucin can also establish hydrophobic bonds with aroma compounds, causing a decrease in concentration as previously demonstrated for ethyl hexanoate and heptyl acetate by Friel and Taylor [19]. The behaviour observed for these two compounds is enhanced by the interaction between mucin and solute salivary components. According to Friel and Taylor [19], salivary salts may modify the number of available binding sites of mucin and may also result in the formation of hydrophobic inclusion sites that can trap volatiles within the solution structure. This could also explain the decrease in the level of esters. Although, the presence or absence of bacteria in the saliva medium was not investigated in this study, previous reports have shown that bacteria in the saliva are capable of hydrolysing/or oxidising different aroma compounds [25].  (Figure 1) and 3-mercapto-2-methylpentanone were greatly degraded. The ability of thiols to function as peroxidase substrates has been described [26,27]. Interestingly, peroxidase activity assays have also been performed by the use of guaiacol an important aroma compound in foods as the substrate [28]. Also, the influences of pH, hydrogen peroxide and thiocyanate on thiols have been investigated [26], the two latter compounds being general constituents of human saliva [27].

Odour Activity Values (OAVs)
To estimate the respective contribution of the odorants to the wines' aroma profile, the OAVs of the odorants were calculated on their nasal odour thresholds in water ( Table 2). The OAVs showed that 3-isobutyl-2-methylpyrazine, acetoin and to lesser extent, 2-acetylpyrroline, 3-methylbutylacetate and ethylhexanoate, contributed intensely to the fruity-moody aroma of the wines. Interestingly, odorant compounds with high concentration in the wines such as 3-methylbutanol, 2-phenylethanol and phenylacetic acid produced relatively low orthonasal OAVs. These compounds contribution to the overall orthonasal aroma quality of the wines would be low. Results also revealed some compounds (ethyl hexanoate, 2-acetylpyrroline and 3-methylbutylacetate) with OAVs higher than their corresponding orthonasal odour threshold in water. While these compounds might not play much important role during sniffing of the wines, they might have significant impact during the consumption of the wines. A series of interaction phenomena, such as additive, synergistic or suppressive effects, are well-documented, so that the presented OAVs do not allow a direct prediction of the odorants' contribution to the wine aroma sensation.

Raw Materials
Three bottles (4.5 L) of each wine obtained from three different palm trees (E. guineensis, R. hookeri and B. flabellifer) were freshly purchased directly from the production farm in a sterilized containers encrusted in ice. The samples were bottled, pasteurized and later dispensed into 45 mL glass-tubes and stored at −20 °C prior to analysis. The alcohol contents of the palm wine samples were 3.7% (E. guineensis), 4.0% (R. hookeri) and 3.2% (B. flabellifer), respectively.

Stable-Isotope-Labelled Standards
The following labelled internal standards were synthesized according to the cited literature:  [30]; [ 13 C 2 ] acetic acid and phenylacetic acid were obtained from (Aldrich). The concentrations of the labelled internal standards and the response factors (FID) were determined gas chromatographically, using methyl octanoate as the internal standard as described by Buettner and Schieberle [33]. The calibration factors for the labelled compounds were calculated as reported by Sen et al. [31] (Table 3).

Collection of Saliva and Enzyme Assay
Mixed whole resting saliva (10 mL) was collected separately from four panellists 2 h after breakfast and after thorough cleaning of the teeth and was used immediately for analysis. Panellists (four males and four females) were volunteers (non-smokers) from the Technical University of Munich, exhibiting no known illnesses at the time of examination and with normal olfactory and gustatory functions. Subjective aroma perception was normal in the past and at the time of examination, before sampling, each panellist rinsed his/her mouth several times with tap water to avoid contamination.

Interaction of Wine with Saliva
Saliva (10 mL) obtained from panellists was immediately used for analysis. Wine (10 mL) was kept in a flask sealed with a lid, and thermostatted at 37 °C after application of 1 mL whole human saliva [25], the solution was stirred at 37 °C for 10 min [11]. The pH of the wine solution containing saliva was always between 7.5 and 8.0. Then, 10 mL of a saturated CaCl 2 solution was added to inhibit enzymatic processes, and the mixture was immediately subjected to quantitation [33]. Each experiment was performed in triplicate. A reference analysis was performed in parallel by using a sample (10 mL of wine) in exactly the same way but without adding saliva. Furthermore, a blank (10 mL of saliva solution) was run in exactly the same way as was done with the wine samples. Therefore, contamination of samples with odorants originating from saliva was excluded.

Inhibition of Enzymatic Activity
The same experiments were performed after thermal treatment of saliva samples in a closed vessel (100 °C, 10 min). The saliva was cooled to 37 °C and immediately applied for the enzyme assays, as described above.

Quantitation of the Odorants by Stable Isotope Dilution Assays
After the enzyme assay, the solution was immediately spiked with known amounts of the labelled internal standards listed in Table 3, stirred for equilibration (20 min), and extracted with dichloromethane (three times, total volume = 300 mL). The combined organic extracts were dried over anhydrous Na 2 SO 4 , and then concentrated to a total volume of 200 µL [34] and subsequently analysed by multidimensional GC-MS.

Chromatography-Mass Spectrometry
The odorants were quantified by two-dimensional gas chromatography (TD-HRGC), using a Mega 2 gas chromatography (Fisons Instruments, Mainz-Kastel, Germany) as the precolumn system in tandem with a Fisons GC 5160 as the main column system. MS analyses were performed with an ITD-800 (Fisons Instruments) running in the C I mode with methanol as the reagent gas. . The gas chromatographic conditions were as described by Buettner and Schieberle [33]. The concentrated wine extracts were applied by the 'cool'-on column injection technique at 40 °C. After 2 min, the temperature of the oven was raised at 4 °C/min to 50 °C and held for 2 min isothermally at the same temperature. The oven temperature was later raised at 6 °C/min to 180 °C and finally raised to 230 °C at 15 °C/min. The flow rate of the carrier gas (helium) was 2.5 mL/min.

Aroma Extracts Dilution Analysis (AEDA)
The FD factors of the odour-active compounds were determined by AEDA [35] using the following dilution series; the original wine extracts (400 µL) from 600 mL of fresh palm wine was specially diluted with diethyl ether (1:1) until no odorant of wine was detectable by sniffing of the highest dilution. HRGC/O was performed with aliquots (0.5 µL), using capillary FFAP. In total, three experienced sniffers were used to perform the AEDA experiments. Only the odours detected by all the three panellists were considered valid. Their response (sensitivity) to individual compounds did not differ by > 2 FD factors.

Identification of Volatile Compounds
Compounds were identified by comparison with the reference substances on the basis of the following criteria: retention index (RI) on two stationary phases of different polarities, mass spectra obtained by MS (EI) and MS (CI), and odour quality, as well as odour intensity perceived at the sniffing port. Odour intensity was checked by GC/O and by comparing the FID signal caused by a defined amount of each reference aroma compound.

Calculation of Odour Activity Values (OAVs)
The OAVs were calculated by dividing the concentrations of the odorants by their ortho-nasal odour threshold in water.