It must nevertheless be stressed that given the very low concentrations of these compounds, the natural variability of these concentrations and their reactivity (especially the free aglycones), and so on, monitoring during curing is very difficult to implement and requires special attention regarding the sample size in order to avoid the sometimes erratic biases and fluctuations observed. This implies specific studies that are often lacking.
Approaches that consist in studying the kinetic parameters of the different aroma precursor glucosides on purified β-glucosidase(s) are undoubtedly more demonstrative for establishing once and for all whether these glucosides are substrates of the enzyme, but clearly imply obtaining glucosides and purified enzyme, which represents a good deal of work in itself.
It is also possible to work on raw enzyme extracts rather than on purified enzymes in order to simplify the task and to be sure to have all the β-glucosidase activities present in the green bean (if several exist). It should nevertheless be remembered that the enzyme(s) may be put into contact with products that affect the kinetic parameters (activators/inhibitors), which would otherwise not come in contact in the green bean or during curing.
From experience, we observed during the development of our enzymatic test that extracting the enzyme with low buffer/beans ratios (i.e., 4:1 v/w) led to rapid and considerable losses in β-glucosidase activity (even using a protective agent, such as antiprotease, antioxidants, etc.), while very high buffer/beans ratios (2000:1 v/w) made it possible to achieve high stability for over 4 h of storage at ambient temperature without any protective agent in the extraction buffer (unpublished results). These observations suggest the presence of compounds that inhibit β-glucosidase activity in moderately diluted green vanilla extracts.
OTHER COMPOUNDS
In most cases, the reaction mechanisms that lead to the appearance and/or disappearance of the other aroma compounds during curing remain to be determined.
For example, aldehydes are not found in green vanilla beans (Perez-Silva, 2006), but appear in advanced stages of curing when their content peaks, and then disappear with only traces found in cured vanilla. Some of these aldehydes could be products of the autooxidation of oleic acid and linoleic acid (found in large quantities in green vanilla).
Aliphatic acids on the contrary, of which linoleic acid is the major representative, are found in the green beans and their concentration tends to decrease during curing. Only acetic acid is not found in the green beans, but appears at a concentration of around 700 ppm after the first heat treatment, then decreases and stabilizes at around 200 ppm. The formation of acetic acid could be of microbial origin (Perez-Silva, 2006).
Aliphatic alcohols also appear after heat treatment, whereas the esters and hydrocarbons are already found in the green beans and their concentrations gradually decrease during curing.
As previously mentioned, among the compounds from the phenylpropanoid pathway, a certain number of aglycones are present despite the (apparent) lack of hydrolysis of the corresponding glucoside (as with p-cresol), or are present in far higher proportions than their concentration in glucosylated form suggested (as with vanillic acid), or are even present when the glucosylated form is absent (as with guaiacol). In fact, although the glucosylated forms are known to be relatively chemically unreactive, the aglycone released after hydrolysis may on the contrary become the substrate for chemical and/or enzymatic reactions, and thus be oxidized, reduced, decarboxylated, methylated, and so forth, and interconversions of one phenolic into another are therefore possible and even common.
For example, concentrations of vanillic acid higher than those expected can be explained by the oxidation of vanillin, and vanillic acid may also undergo decarboxy-lation into guaiacol (Perez-Silva, 2006), which could also explain the lack of glucoside of guaiacol. A solution of vanillin kept at 80°C at pH 5 in the presence of oxygen for several days also reveals the presence, in addition to the aforementioned molecules, of 2-methoxyhydroquinone, which can form quinones and semiquinones, which may lead to dimers (Perez-Silva, 2006). Different molecules of this kind, including divan-illin, have been identified in cured vanilla (Figure 12.4) and are considered to contribute positively to the flavor of the product (Gatfield et al., 2006; Schwarz and Hofmann, 2009). Gatfield et al. (2006) suggest that the formation of divanillin is the result of the action of a peroxidase, a form of which has recently been purified (Marquez et al., 2008). This kind of reaction can also lead to the formation of large polymers that are at least partly responsible for the brown color of the vanilla.
These different reactions, which all start from vanillin, partly explain the considerable losses of this compound during curing (Odoux, 2000; Gatfield et al., 2007) that were already mentioned in Chapter 11.
Other reactions with no direct relationship to aroma formation may also explain these losses of vanillin, such as sublimation, coevaporation with water (Frenkel and Havkin-Frenkel, 2006), or the formation of a covalent bond with the lignin of the bean (Gatfield et al., 2007).
References
Adedeji, J., Hartman, T.G., and Ho, C.-T. 1993. Flavor characterization of different varieties of vanilla beans. Perfumer and Flavorist 18:25–33.
Arana, F.E. 1943. Action of a β-glucosidase in the curing of vanilla. Food Research 8:343–351. Dignum, M.J.W., van der Heijden, R., Kerler, J., Winkel, C., and Verpoorte, R. 2004. Identification of glucosides in green beans of Vanilla planifolia Andrews and kinetics of vanilla β-glucosidase. Food Chemistry 85:199–205.
Dignum, M.J.W., Kerler, J., and Verpoorte, R. 2001. Vanilla production: Technological, chemical and biosynthetic aspects. Food Reviews International 17:199–219.
Dignum, M.J.W., Kerler, J., and Verpoorte, R. 2002. Vanilla curing under laboratory conditions. Food Chemistry 79:165–171.
Ehlers, D. and Pfister, M. 1997. Compounds of vanillon (Vanilla pompona Schiede). Journal of Essential Oil Research 9:427–431.
Frenkel, C. and Havkin-Frenkel, D. 2006. The physics and chemistry of vanillin. Perfumer & Flavorist 31:28–36.
Galetto, W.G. and Hoffman, P.G. 1978. Some benzyl ethers present in the extract of vanilla (Vanilla planifolia). Journal of Agricultural and Food Chemistry 26:195–197.
Gassenmeier, K., Riesen, B., and Magyar, B. 2008. Commercial quality and analytical parameters of cured vanilla beans (Vanilla planifolia) from different origins from the 2006–2007 crop. Flavour and Fragrance Journal 23:194–201.
Gatfield, I., Hilmer, J.M., Weber, B., Hammerschmidt, F., Reiβ, I., Poutot, G., and Bertram, H.-J. 2007. Chemical and biochemical changes occurring during the traditional Madagascan vanilla curing process. Perfumer & Flavorist 32:20–28.
Gatfield, I., Reiβ, I., Krammer, G., Schmidt, C.O., Kindel, G., and Bertram, H.-J. 2006. Novel taste-active component of fermented vanilla beans. Perfumer & Flavorist 31:18–20.
Goris, A. 1924. Sur la composition chimique des fruits verts de vanille et le mode de formation du parfum de la vanille. Comptes Rendus de l’Académie des Sciences 179:70–72.
Goris, A. 1947. Formation du parfum de la vanille. Industrie de la Parfumerie 2:4–12. Hanum, T. 1997. Changes in vanillin and activity of β-glucosidase and oxidases during post harvest processing of vanilla bean (Vanilla planifolia). Bulletin Teknologik dan Industri Pengan 8:46–52.