2.1.1. Evolution of Pigment Composition
The evolution of pigment composition in different model systems was monitored. They were prepared using two different solvents: standard wine (
A, reference model system, and
AC and
AS, added with the tannin
C and
S, respectively) or a standard medium resulting from the fermentative metabolism of glucose (
AF, reference model system, and
ACF and
ASF, added with the tannin
C and
S, respectively). In both types of model systems (in absence or presence of standard medium resulting from the fermentative metabolism of glucose), the disappearance of the anthocyanins over time was observed. The qualitative evolution of the pigments was the same for each type of model system regardless of the presence or absence of enological tannin and independently of the type of enological tannin employed. However, some important differences were observed depending on the solvent employed to prepare the model system. In the case of the model systems prepared with standard wine (
Figure 1a,c for model system
A) the disappearance of the grape anthocyanins (mainly delphinidin 3-
O-glucoside and petunidin 3-
O-glucoside, peaks 1 and 3 in
Figure 1, respectively) was clearly observed, whereas no formation of anthocyanin-derived pigments could be observed. On the contrary, in the model systems prepared using the fermentative medium as solvent (
Figure 1b,d, for model system
AF), the disappearance of the grape anthocyanins occurred along with the formation of anthocyanin-derived pigments, as will be explained later.
Quantitative differences were also observed between the two types of model systems. In the case of the model systems prepared with standard wine, the disappearance of the anthocyanins was slower and the final percentages of anthocyanins in relation to the initial concentration were higher than in the case of the model systems prepared in the fermentative medium (
Figure 2). To be precise, whereas in the standard wine model systems the total anthocyanin content at the end of the study represented more than 10% of the initial content, in those prepared in the fermentative medium the final percentage was lower than 1%. The greater complexity of the fermentation medium in relation to the standard wine makes possible the involvement of anthocyanins in different reactions and, consequently, their transformation is more important than in standard wine model systems. In fact, as indicated above, the formation of anthocyanin-derived pigments was observed in the model systems prepared in the fermentative medium, namely the formation of A-type and B-type vitisins of peonidin 3-
O-glucoside (peak 6 in
Figure 1) and of malvidin 3-
O-glucoside (peak 7 in
Figure 1). The synthesis of these kinds of anthocyanin-derived pigments in the model systems prepared in the fermentative medium was possible since the precursors of these pigments (pyruvic acid for A-type vitisins and acetaldehyde for B-type vitisins) were available as a result of the fermentative metabolism of glucose.
Thus, the higher decrease in the levels of the anthocyanins in the fermentative medium was partly due to their involvement in the formation of anthocyanin-derived pigments, such as A-type and B-type vitisins. It is worth noting that, in both types of model systems, peonidin 3-O-glucoside was the anthocyanin that showed, at the end of the study, the higher percentages in relation to the initial contents. Moreover, A-type and B-type vitisins derived from this anthocyanin were also detected in the model systems prepared in the fermentative medium. Thus, the disappearance of this anthocyanin could be mainly related to the formation of anthocyanin-derived pigments. On the contrary, no derivative pigments from petunidin 3-O-glucoside or delphinidin 3-O-glucoside were detected in any of the studied model solutions although they were the anthocyanins showing the most important disappearance. This could be pointed out by a higher reactivity of these anthocyanins in relation to the others with probably faster degradation rates than the rest of the anthocyanins.
In both types of model systems, a small elevation of the baseline (hump) was observed at the end of the chromatograms of the latest sampling points, which might be related to the formation of oligomeric compounds, among others. However, there is not a tight relationship between the large decrease in the levels of the monoglucosides and the appearance of this hump. This indicated that most of the anthocyanin disappearance was due to the formation of non-colored compounds.
Regarding the influence of the presence of the enological tannin, in the case of the model systems prepared with standard wine, it was observed that model system
A (control) showed the lowest disappearance of anthocyanins in relation to model systems
AC and
AS (see
Tables S1 and S2 in the Supplementary Materials). However, in the case of the model systems prepared in the fermentative medium, the lowest loss of anthocyanins was observed for those model systems added with the enological tannins, and no differences were observed between the two different enological tannins. The higher complexity of the fermentative medium in relation to standard wine might have caused important differences on the redox status for the different type of model systems. Thus, the addition of the enological tannins can exert different effects depending on the solvent used.
The formation of A-type and B-type vitisins, as explained above, was only observed in the model systems prepared in the fermentative medium. The formation of these anthocyanin-derived pigments in these model systems took place both in the presence and in the absence of enological tannin, although the presence affected their formation. From these results, it can be deduced that the formation of these pigments is mainly conditioned by the presence of pyruvic acid or acetaldehyde in the media. However, based on the absence of these anthocyanin-derived pigments in the models systems prepared in model-wine solution, it seems that the presence of the possible precursors (such as ethanol in the case of acetaldehyde) is not enough to allow their synthesis. With the chromatographic method employed in this study, the A-type and B-type vitisins of each anthocyanin co-eluted in the same peak in the chromatogram and for this reason they were quantified together.
Figure 3 shows the evolution of the content of A-type + B-type vitisins of peonidin and malvidin 3-
O-glucosides in the model systems prepared in the fermentative medium. The vitisins derived from malvidin 3-
O-glucoside were detected in the first sampling point (2 h), whereas those derived from peonidin 3-
O-glucoside were not detected until the second sampling point (five days). Furthermore, the levels of the former ones were higher than those of the latter ones during all the period studied which can be explained by the higher initial levels of malvidin 3-
O-glucoside in relation to those of peonidin 3-
O-glucoside. The model systems added with the enological tannins (
ACF and
ASF) showed the highest levels of vitisins, which might be related to the presence of ellagitannins in the enological tannins. On the one hand, they have been reported to favor the transformation of ethanol into acetaldehyde [
18], one of the substrates for the formation of B-type vitisins. On the other hand, the ellagitannins present in the enological tannins can favor the formation of the A-type vitisins acting as oxidants in the last step of the synthesis, which is an essential step to complete it [
32]. However, the absence of B-type vitisins in the model systems prepared in model-wine solutions (even in the presence of ellagitannins), along with the preliminary results of current studies carried out in our laboratory [
33] where the formation of A- and B-type vitisins is studied in model-wine solution in the presence of individual ellagitannins and in the presence or absence of pyruvic acid, are indicating that the only presence of ethanol and ellagitannins does not lead to a significant formation of these anthocyanin-derived pigments. Thus, it seems that the main way in which ellagitannins could favor the formation of vitisins is related to their possible role as oxidants in the last oxidation step of the synthesis of these anthocyanin-derived pigments.
Moreover, flavanol-anthocyanin acetaldehyde-mediated condensation products were also detected (but could not be quantified due to their low levels) in the model system ACF. The higher levels of proanthocyanidins in the enological tannin C in relation to the enological tannin S could explain why these anthocyanin-derived pigments were only detected in those model systems added with enological tannin C.
2.1.2. Color Evolution
The evolution of the CIELAB color parameters calculated from the visible spectra of the model systems at each sampling point was studied. Regarding lightness (L*,
Table 1), it increased for all the samples over time, which was in accordance with the decrease observed in the total pigment content [
34]. Nevertheless, the magnitude of the increase was different for the different model systems. Thus, at the end of the study and in relation to the initial value, it was observed an increase of ca. three units for the standard wine-model systems and of ca. 12 units for the model systems prepared in the fermentative medium. The difference on the evolution of the lightness could be attributed to the difference on the evolution of the pigment composition [
34] since, in the model systems
ACF,
ASF, and
AF, the loss of anthocyanins was much higher than in
AC,
AS, and
A.
For a same type of model system (model wine or fermentative medium), the lowest values of lightness were observed in those added with the enological tannins. The enological tannins employed in the present study contained, among other compounds, proanthocyanidins and hydroxybenzoic acids, which may act as co-pigments of the anthocyanins [
35]. Consequently, the lightness of the samples could be reduced due to the hyperchromic effect induced by co-pigmentation [
35]. No differences were observed between the lightness values of the model systems depending on the type of enological tannin added and, in both cases, the model system were darker than the corresponding model system without enological tannin.
The evolution of chroma (C*
ab) was similar for all the model systems studied (
Table 2). In all cases a decrease in the values of chroma was observed over time, which can be attributed to the loss of pigments [
34]. However, as in the case of L*, the change observed in the values of C*
ab of the model systems prepared in the fermentative medium was more important than that observed for the model systems prepared with standard wine. It can be related to the lower loss of anthocyanins in the latter ones. The model systems treated with enological tannins were those that showed the lowest decreases in the value of chroma and, consequently, were those that showed the highest values at the end of the study. As in the case of lightness, the difference in the chroma value between model systems
A and
AC and
AS might be related to differences on the co-pigmentation phenomenon [
35].
The hue was affected by the addition of the enological tannins in the same way in both types of model systems (
Table 3). The model systems to which the enological tannin was not added (
A and
AF) showed lower values of hue than those to which either the enological tannin
C (
AC and
ACF) or the enological tannin
S (
AS and
ASF) was added. In the case of model systems prepared in the fermentative medium, it could be explained by the higher formation of anthocyanin-derived pigments such as A-type and B-type vitisins, which show an orange-red color [
34], i.e., higher hue values than the corresponding anthocyanins. In the case of model systems prepared with standard wine, the lower hue values observed in model system
A could be attributed to the higher levels of anthocyanins in this media in relation to model systems
AC and
AS. Moreover, it has to be pointed out that enological tannins showed, when solubilized, a light yellow-brown color that could affect, mainly, the hue of the model systems. Thus, the highest values of hue observed in the model systems added with enological tannins might also be explained by the contribution to color of the enological tannins.