Relationship between polyphenolic composition and some sensory properties in red wines using multiway analyses

Abstract

Polyphenolic compounds are responsible for important sensory properties of red wines (especially colour, astringency and, to a lesser extent, bitterness). Various model solution studies have investigated the relationships between specific phenolic compounds and sensory perception. The purpose of the present study was to relate polyphenolic composition to sensory data for a very large number of different commercial wines, using multiway analyses. Two homogeneous populations of commercial red wines (61 French and 60 German wines) were analyzed. Thirty simple polyphenolic compounds (anthocyanins, flavonols and phenolic acids) and some red pigment derivatives have been quantified using direct injection by liquid chromatography (LC) coupled to diode array detection and mass spectrometry. Condensed tannins were analyzed using LC after thiolysis. Sensory perception was assessed using descriptive analysis by four trained panels (2 countries × 2 years). We first checked the consistency of the sensory data both through the 2 assessment years, and through the two countries. Although reproducibility was high, especially through countries, slight scale factor differences and scale shifts were detected. A chemical consensus was then built using common components and specific weights analysis for both wine subsets, where the sensory variables of each panel were projected separately. Although data structure was very different between the two wine subsets, common main chemical-sensory relationships were confirmed. The hypothesis of a relationship between flavonol aglycones and bitterness was raised. A partial least squares regression was performed and predicted linearly astringency with a R² value of 0.80, but not colour intensity and hue, due to non-linear relationship and saturation of visual perception.

Introduction

Polyphenols make up a large and complex family of molecules, with diverse structures, properties and sizes (monomers to polymers). Further complexity results from reactions of native phenolic compounds originating from grapes during winemaking and ageing. Some of these compounds are responsible for important sensory properties of red wines (especially colour and taste, i.e. astringency and, to a lesser extent, bitterness). Besides, some of these perceptions may be modified by other sensory characteristics (e.g. sourness, sweetness) related to other wine components.

Numerous studies have been performed in attempts to analyze phenolic compounds in wines. Simple polyphenolic compounds (anthocyanins, flavonols, phenolic acids and flavanol monomers and some oligomers) and some red pigment derivatives have been identified and sometimes quantified in wines, rarely at the same time, by liquid chromatography (LC) on reversed-phase columns, often coupled to diode array detection (DAD) and/or mass spectrometry (MS) [1], [2], [3], [4], [5], [6]. Global methods, such as Folin-Ciocalteu [7], have been used to quantify total polyphenol indices, including simple polyphenolic compounds and condensed tannins, i.e. proanthocyanidins, which are oligomers to polymers of flavanols. Tannins have been quantified by colorimetric methods, such as reaction with vanillin [8], [9] or dimethylaminocinnamaldehyde (DMCA) [10], and by depolymerization in acidic media at high temperature using the Bate–Smith reaction [11], [12]. But more complete and valuable information about tannins (e.g. mean degree of polymerization (mDP), percentage of galloylation) has been acquired using depolymerization by acido-catalysis in the presence of a strong nucleophile (e.g. thiolysis [13], phloroglucinolysis [14]).

Sensory analysis of food products is the central link between chemical composition and consumer preference. Many studies now involve some sensory assessments, performed using either descriptive analysis (DA), delivering sensory profiles for visual, olfactory or gustatory attributes or time–intensity assessment, providing information on perceived intensity of a given taste character over time [15].

Native anthocyanins and pigments derived from them play the main role in colour perception The role of flavonols, as yellow pigments, and the complex phenomenon of copigmentation [16] are still under investigation. Besides, various relationships between phenolic compounds and taste have been emphasized in model solutions. Astringency is mainly a tactile sensation [10], due to tannin interaction with salivary glycoproteins, generating a loss of lubrication or with oral epithelium [17]. Vidal et al. [18] showed in model solutions that astringency perception of tannins increased with their mDP and their percentage of galloylation [19]. Tannin derivatives were as astringent as tannins themselves, contrary to the pigment-tannin derivatives, which were less astringent. Rubico and McDaniel reported that acids could also play a role in astringency perception by interacting with salivary proteins [20], rather in relation with their acidity than because of their structure [21]. However, phenolic acids probably do not play any role due to their weak acidity and the strongly bulked wine matrix. It was also shown that gallic acid was not astringent in water at 10 g/l [22]. Robichaud and Noble studied the bitterness of some phenolic compounds [23]. They found that catechin and gallic acid were more bitter than astringent. However, no general rule has been reported to relate the structure of the molecules and bitterness [24]. Some flavonols were reported to be bitter in 5% aqueous ethanol [25], [26]. Verette et al. investigated the addition of caffeic acid and its esters in wine with a concentration of 150 mg/l, and did not observe any modification in acidity perception [27]. Wine phenolic compounds do not exhibit any sweetness. However, sweetness, acidity, viscosity and ethanol content, were reported to affect astringency and bitterness perception [28], [29].

Many studies have reported works crossing different data tables, outcoming from different instrumental measurements, or from instrumental and sensory assessment in either commercial or experimental food products. Different statistical and chemometric tools have been used for exploratory or predictive goals. Principal component analysis (PCA) was the first multivariate tool used for describing and visualizing the relationships between several groups of variables, also when crossing instrumental and sensory data [30]. However, PCA does not take into account the initial grouping of the variables. Interpretation may therefore be difficult. In order to cope with such difficulties, multiway techniques have been developed. General procrustes analysis (GPA) was used to correlate sensory and chemical data sets for ice creams [31]. Brossaud used multiple factor analysis (MFA) for studying the chemical-sensory relationship for wines produced in the Loire valley [32]. GPA and MFA probably build a consensus giving more importance to the information, which is common to all processed data tables than to the specific information derived from each of them, at least for the first dimensions. Some other multiway techniques allow the synthesis of both common and specific information. One such method is common components and specific weights analysis (CCSWA), which has been used to cross different instrumental data tables [33], [34]. Multiple linear regression (MLR) has been used for modelling or predicting the sensory properties of wines [35]. However, this was done from polyphenol indices rather than from detailed polyphenolic composition. When the explanatory variables are linearly dependent, for example when coming from spectroscopic methods, other chemometric techniques, such as stepwise MLR, stepwise principal component regression (SPCR) or partial least squares regression (PLS) are needed [7], [36], [37], [38].

The whole polyphenolic information, i.e. native and derived anthocyanins, phenolic acids, flavonols and tannins (quantity and quality) has rarely been gathered at the same time and tentatively crossed with sensory data for a very large number of different commercial wines.

The purpose of the present study was first to completely characterize and analyze the phenolic compounds of a large number of red wines from two very different types. The same wines were then assessed using DA. These two kinds of information (analytical and sensory) were then processed using chemometric tools and multiway analyses, namely CCSWA and PLS.