NMR spectroscopy and chemometrics-based analysis of grapevine
Ali, K.
Citation
CHAPTER 11
General discussion, conclusions, and perspectives
Kashif Ali
Natural Products Laboratory, Institute of Biology, Leiden University,
The Netherlands
General discussion and conclusions
Enormous progress has been observed in biological sciences in the past decade. Many new tools have been developed to study biological systems, e.g. ‘omics’. This has resulted into new approaches to study complex biological systems: ‘systems biology’.
This new field is generally defined as the study of the interactions between the components of different biological systems. Different omics approaches, including genomics, transcriptomics, proteomics, and metabolomics, are considered as the building blocks of systems biology as the comprehensive knowledge about the interaction of these building blocks can lead scientists to eventually completely define a living system. Considering the belief of genes to metabolites, metabolomics is closest to the organisms’ phenotype hence the analytical coverage of all metabolites can provide a better understanding of complex biochemical systems.
In metabolomics, the ultimate goal is to get the complete overview of the entire metabolic contents of the system. So far, despite all the technical and analytical advancements, this is not possible for many reasons. Firstly, plant metabolomics studies deal with a very large number of metabolites which estimated in the range of 30,000 (Oksman-Caldentey and Inzé 2004; Verpoorte et al. 2008). Secondly, they occur in quite different concentrations, with different stability profiles. Thirdly, plant metabolites are very diverse in their chemical structures and properties and thus polarities. Hence the qualitative and quantitative analysis of all metabolites in a given sample in a single analysis is not really feasible.
Analytical techniques applied for metabolomics should be unbiased, rapid, reproducible, and stable over time, while requiring only simple sample preparation (Verpoorte et al.
2010). Many platforms are available for metabolome analysis and among them mass
General discussion and conclusions
analytical method for grape metabolomics, monitoring berry ripening, characterization of type, vintage effects and sensory attributes in wine, and analysis of bioactivity in grape and wine and its correlation with NMR data using different multivariate data analyses methods. Grapevine leaves are also studied with the aim to highlight the metabolic differences among the resistant and susceptible cultivars. Furthermore the metabolic response of a resistant and a susceptible grapevine cultivar upon inoculation with pathogenic fungi has also been studied.
The use of NMR in metabolomics studies has many advantages but the overlapping of signals in the spectra represents the major obstacle for compound identification. This problem is usually overcome by the use of different 2D techniques, i.e. J-resolved, 1H-
1H COSY, HMBC and HSQC, which provide additional information regarding compound structures. Among the above mentioned techniques J-resolved and 1H-1H COSY are widely used due to short measuring time with good quantitative features.
They showed to be quite effective in the confirmation of metabolites like phenylpropanoids and flavonoids (Liang et al. 2006). Different metabolites have been identified in grapes, wine, and grapevine leaves using 1H NMR with the help of 2D- NMR techniques. The metabolites identified cover a wide diversity and include amino acids, organic acids, carbohydrates, hydroxycinnamates, hydroxybenzoates, stilbenoids, flavanoids, and flavonoids.
Multivariate data analysis methods are considered as the integral part of any metabolomics based research. In this thesis, several multivariate data analysis methods are used including principal component analysis (PCA), projections to latent structures- discriminant analysis (PLS-DA), hierarchical cluster analysis (HCA), projections to latent structures (PLS), bidirectional orthogonal PLS (O2PLS), and bidirectional orthogonal PLS-DA (O2PLS-DA). Principal component analysis (PCA) is considered as a primary tool in metabolomics used to reduce the dimensionality of a multivariate dataset, and thus helping to better understand possible differences among samples. It is an unsupervised method hence the clustering or separation is purely due to similarities or differences, respectively, among all the samples. Similar to PCA, Hierarchical cluster analysis (HCA) is an unsupervised method. In HCA, based on samples similarity or distance, progressive pair-wise grouping of samples is ocurred. The HCA results can be
seen as a dendrogram in which branch lengths reflect the differences among the groups and thus provide an easy visualization of the similarities of samples.
The application of supervised analyses, like PLS-DA and O2PLS-DA, is considered to be the next step in multivariate data analysis. These analyses, unlike the unbiased system used for PCA, are performed with pre-input information regarding the data. In other words, the samples are classified into different groups based on similar characteristics. Methods like PLS and O2PLS are used to correlate two different types of data, as in the case of this study, NMR data is correlated with the bioactivity data of grapes and wine against TNFα production.
Grape berries are undoubtedly among the most important fruit species because of their use in wine making. Grape biochemistry shows a great diversity in terms of structure and function ranging from simple amino acids and sugars to highly complex polymers of condensed tannins. Extracting metabolites from a biological matrix is very important step in metabolomics studies. While comparing different extraction procedures for grape metabolomics, two potential candidates emerged i.e. SPE and direct extraction using deuterated solvents. SPE was found most effective in the case of less concentrated phenolics while the direct extraction give good results in the case of metabolites like organic acids, amino acids, and sugars. It was shown that distribution of metabolites in the extract was highly influenced by the solvent(s) used for extraction or the solubility or polarity of the metabolites. This information clearly shows the importance of such a comparative analysis before starting a metabolomics study. In drawing conclusions of a single extraction solvent metabolomics analysis, one should always be careful as differences can be observed for certain compounds, but that are not necessarily the only compounds that differ among the samples. For comprehensive profile of differences in levels of metabolites, the use of more than one extraction method would be helpful.
General discussion and conclusions
data analyses in combination with 1H NMR data, some genuine differences among the different development stages and among the cultivars used in this study have been detected.
The different stages of grape development mainly differ in their phenolic profile along with significant fluctuations in organic acid and sugar contents. In the light of the results obtained from this research it can be concluded that the initial stages, green and veraison, are metabolically very different from ripe and harvest. Organic acids with the phenolics were highly accumulated in berries during early stages followed by an increase in sugars during the later stages. Veraison was proved to be the key stage since during this stage the grape berries undergo some dramatic metabolic changes. The technique applied here is highly reproducible and effective in analyzing a wide range of compounds of the grape metabolome.
Wine is considered as one of the most complex foods to define on chemical basis as many factors like grape cultivar, climate, cultural practices, fermentation process, yeast strain, and storage etc., influence the chemistry of wine. In this project, 1H NMR based metabolic profiling was shown to be effective in terms of identifying diverse metabolites present in wine. In combination with 1H NMR, PLS and O2PLS were found very effective in highlighting the differences among the wines based on quality scores.
Different metabolites are found positively and negatively correlated with wine taste. It should be mentioned that in terms of quality the interaction among the various components of wine are highly complex and knowledge regarding individual components might not be that much useful as such, increasing our insight in their interactions should be of interest. Metabolomics does not only offer the visualization of the complexity of these specific interactions and provides better understanding of wines but more importantly generate information needed to improve wine quality.
Considering the previous reports on wine analysis, the extraction procedure followed in this study was very effective as compared to direct analysis of wine (Lopez-Rituerto et al. 2009), lyophilization (Baderschneider and Winterhalter 2001), and the use of nitrogen-flow (Amral and Caro 2005). Many problems like low reproducibility, time consuming, and signal overlapping by dominant compounds in wine, and shifting of signals due to differences in pH of wine are associated with these methods. The ethyl acetate extraction used here was fast, does not require a pH control, and has less signal
intensity for wine dominating metabolites like ethanol and lactic acid. In general, this study is a classic example to represent the enormous potential of NMR for the chemical analyses of complex food items. NMR spectroscopy has been already used for wine (Son et al. 2009b; Pereira et al. 2005) classification and characterization but very few reports are available for the quantification of wine components using intensities from 1H NMR spectra (Larsen et al. 2006; Viggiani and Morelli 2008). Reports on wine analysis either involve some chromatography for the identification of phenolics or simply no phenolics identification. Here, 1H NMR with different 2D NMR techniques resulted in successful identification of nearly all the major classes of phenolics (cis- and trans- cinnamic acids, benzoic acid, stilbenoids, flavonols, and flavane-3-ols), along with amino acids, carbohydrates, and organic acids. The recent trend in wine metabolomics is the classification of different wines based on sensory attributes (Cuadros-Inostroza et al. 2010; Rochfort et al. 2010) but non of them use supervised regression models like PLS or O2PLS. Our study is the successful attempt to demonstrate the enormous potential of these chemometrics methods for the correlation of two different data sets.
Grape and wine phenolics have been shown to posses several health promoting activities e.g. for cardiovascular diseases, renal diseases, and cancer (Ali et al. 2010;
Halpern 2008). Nearly all of these beneficial effects associated to wine are due to anti- oxidant and radical scavenging properties of wine phenolics (German and Walzem 2000). Our interest in health effects during this research was to device a method using NMR spectral data of grapes and wine in combination with chemometrics methods to predict the metabolites of pharmacological interests in the crude extract without any chromatographic separation. Using methods like O2PLS and O2PLS-DA, we successfully correlate the NMR data with the bioactivity data from anti-TNFα assay and found several grape and wine phenolics positively correlated with TNFα inhibition.
General discussion and conclusions
Various multivariate data analysis methods were used in combination with NMR spectroscopy in order to correlate the activity data of the extracts with the spectroscopy data of the same. Such analyses of extracts from Hypericum perforatum (Roos et al.
2004), Artemisia annua (Bailey et al. 2004), Citrus grandis (Cho et al. 2009), and Galphimia glauca (Cardoso-Taketa et al. 2008), were successful in linking pharmacological activities with certain compounds. This approach is very effective in the screening of plant extracts in order to identify active compounds without laborious fractionation and chromatographic separation of the crude extract. Fractions from SPE of various red wines from Portugal were analyzed for anti-TNFα activity and the combination of NMR spectroscopy and chemometrics was successfully applied to identify the active ingredients. It is interesting to observe that the wine showed more anti-TNFα activity than the corresponding grape cultivars from the same year. This is mainly due to phenolics in grapes showed very poor systemic absorption and need fermentation to observe any beneficial effect (Coimbra et al. 2005).
Metabolic phenotyping has been widely applied and can be useful to underline some genuine metabolic differences and also to define certain physiological characteristics among different cultivars or even species. Fungal diseases of grapevine are the major cause of huge economics losses, annually. As secondary metabolites (phytoalexins) and pathogen related (PR) proteins are the two basic lines of defense in plants, the understanding of their biosynthesis and plant response towards infection can be of high value. During the course of this study, metabolic characterization of different grapevine species is performed to highlight the major metabolic differences among the resistant and susceptible species. It was discovered that the higher concentrations of phenylpropanoids and flavonoids in resistant grapevine is the major difference from the susceptible grapevine. It was also found that upon inoculation with fungal pathogen, the resistant cultivar is quicker in generating a response than the susceptible cultivar and within 48 hours post inoculation, a significant amount of phenylpropanoids and flavonoids is accumulated in the resistant cultivar.
In connection with studies on other plants, it is evident that phenylpropanoids and flavonoids are involved in resistance against biotic stresses to the plants. Plants like tomato (Lopez-Gresa et al. 2010), tobacco (Choi et al. 2006), different Brassica (Abdel- Farid et al. 2009), and Arabidopsis (Liang et al. 2006) cultivars showed significant
increase in the biosynthesis of phenylpropanoids and flavonoids. It has been shown that grapevine specific phytoalexins, i.e. viniferins, can be produced by both resistant and susceptible cultivars upon infection (Jean-Denis et al. 2006; Slaughter et al. 2008; Pezet et al. 2004). This observation really made our findings significant as it seems the initial hours of infection are very critical in the development of disease. Since both resistant and susceptible cultivars are reported to produce stress related viniferins, the synthesis of higher amounts of phenylpropanoids and flavonoids in resistant cultivar might be the key for resistance.
Future prospects
Although NMR based metabolomics in combination with chemometrics has been proved very efficient and effective in studying different aspects of grapes and wine research and some very interesting data have been generated, still a lot more need to be done in this field. The study of berry development in more cultivars, from different geographic locations, and with more vigorous sampling would lead to provide a deeper insight in various important metabolic events during ripening. The correct identification of the veraison stage is still a very critical question to answer as the berry and resultant wine quality is based on harvesting berries at the right time. The sensory analysis of more wine samples from different parts of the world would certainly help to identify factors responsible for the good taste of wine. This in turn helps the wineries to design efficient fermentation processes, selecting the appropriate yeast strain, and optimize the storage conditions to produce top quality wines.
The more extensive study of grapevine and pathogen interaction, especially at the initial stages, can be of extreme importance. This information can and will provide aid to do
General discussion and conclusions
‘complete’ picture of grape and wine metabolome. Also with the emergence of new analytical tools, with more sensitivity and precision, our understanding regarding the grape and wine chemistry will certainly further increase as there is still a lot to be understood at the level of genomics, proteomics, and especially metabolomics.
