542 C. Santos-Buelga and V. de Freitas
Vivar-Quintana et al. 2002). Using thiolysis, mean polymerization degrees
between 2.9 and 10.7 have been obtained for the compounds present in those humps
(Remy et al 2000; Salas et al. 2005), although not only pigments but also proantho-
cyanidins and colorless compounds present in the same extracts must account for
those results. According to our knowledge, the largest anthocyanin-tannin oligomers
detected in red wines correspond to colorless adducts consisting of one anthocyanin
moiety linked to a procyanidin trimer (Salas et al. 2005), although the possibility of
larger adducts, either colored or not, is hypothesized by the same group. Nonethe-
less, as far as we know, no actual polymericpigments have been characterized in red
wines, which does not mean that their presence should be discarded, as they may
instable and/or escape the usual techniques for separation and analysis. On the other
hand, the existence of various centres of asymmetry in the derived pigments that
contain tannin moieties makes possible the occurrence of many isomers whose dif-
ficult chromatographic separation could also be accounting for the humps observed
in the HPLC profiles of matured and aged red wines.
In assays carried out by our group, it was found that when ‘hump’ fractions
previously separated from red wines were further analyzed by RP-HPLC after
changing their pH to neutrality or by simple diluting them, they suffered a notably
decrease in their size at the same time that peaks of monomeric anthocyanins
(mostly malvidin 3-glucoside) not observed in the original hump fraction appeared
in the chromatograms (non-published results). This observation allows speculating
that, although the existence of large pigment structures cannot be ruled out, the
‘polymeric pigment material’ of red wines could be constituted, at least in part, by
anthocyanins retained physically rather than chemically into tannin clusters. This
had already been postulated by Somers (1966), who found that wine tannin material
separated by gel filtration on Sephadex gels contained pigment molecules that were
easily released by mild acid treatment, indicating that weak linkages were involved
in the retention of anthocyanins into the tannin matrix.
The pyranoanthocyanins in their diverse forms are the derived pigments more
usually found in red wines. They have been estimated to constitute about 70% of the
total derived pigments present in a two-year-old red wine (Alcalde-Eon et al. 2006)
and account for up to 50% of overall pigment material in a five-year-old wine
(Boido et al. 2006). Anthocyanin-ethyl-flavanol derivatives and, more recently, the
products of the direct condensation between anthocyanins and tannins are also
commonly detected in red wines, but their concentrations are lower than those of
pyranoanthocyanins and their relative contribution decrease in older wines (Alcalde-
Eon et al. 2006). When evaluated by the peaks observed in the HPLC chro-
matograms, the levels of all these pigments are apparently too low to explain the
color of the wines. The fact that they are totally or partially resistant to the dis-
coloring effect of the lowacidity and the presence SO 2 and, therefore, able to
express their color in wine conditions, contrary to the anthocyanins, might in part
explain this contradiction (Vivar-Quintana et al. 2002). In addition, the sum of the
contribution of the minor amounts of the diverse existing pigments and possible syn-
ergisms among them might also account for some of the color expression. Despite
this, it is assumed that a significant part of color expressed by aged red wines is still