conference schedule and program with abstracts - Horticulture ...

O-19
Integrated analysis of transcriptome and metabolome to understand grapevine
development and its relation to wine attributes
H. Peña-Cortés* 1,2 , A. Cuadros-Inostroza 2 , I. Ramírez 1 , E. Gonzalez 3 , S. Ruiz 3 , C. Caldana 2 , P.
Giavalisco 2 , L. Willmitzer 2
1 Biotechnology Center, Technical Federico Santa Maria University, Valparaiso, Chile; 2 Max-
Planck Institute for Plant Molecular Physiology, 14476 Potsdam-Golm, Germany; 3 Talca
University, Talca, Chile
*Corresponding author : hugo.pena@usm.cl
Different physiological and biochemical processes are involved in fruit setting, development and
ripening of grapevine berries. Diverse efforts using varied technologies are being applied in
order to explain certain key biological stages of this fruit which could allow, in a near future, to
improve the quality of grapevine berries and consequently the quality of wine. To gain
information concerning the genes and metabolites involved in such processes, we measured
primary and secondary metabolites together with transcript levels of three Vitis vinifera cultivar
berries during their growth period. Samples of cultivars Carmenère, Merlot and Cabernet
Sauvignon were collected every three days and during two growth seasons, starting with flowers
and finishing with mature berries. We established a qRT-PCR platform to investigate the
transcriptional changes during growth development and setting that allows us to determine the
expression levels of around 800 grapevine genes. We measured metabolite levels by using GC-
TOF-MS and UPLC/FT-MS for primary and secondary metabolites respectively. Since huge
amount of data is generated, we use a network analysis approach based on correlations as a tool
to investigate the relationship between transcripts and metabolite changes and their role in
regulation of metabolic pathways during ripening. Understanding the processes involved in grape
ripening will contribute to obtain better fruits quality which would help to improve wine quality.
Considering the advantages provided by using LC-MS to analyse secondary metabolites, we are
applying this technology to investigate whether a non-targeted wine chemical composition
analysis could provide suitable information on the relationship between the quality scores -given
to Chilean commercial red wines by experienced winemakers- and wine matabolome data and in
that way provide an objective method to discriminate wine quality scores and other wine
attributes like variety, origin or vintage. The analysis of the metabolic profiles of the wine
samples by unsupervised multivariate techniques and in house developed bioinformatics tools
recognize signatures”(Biomarkers”) in wine metabolome which allow to differentiate attributes
like variety, vintage, origin and wine quality scores.
38