Yeast systems biology: modelling the winemaker’s art Anthony R. Borneman, Paul J. Chambers and Isak S. Pretorius The Australian Wine Research Institute, PO Box 197, Glen Osmond, Adelaide, SA 5064, Australia Yeast research represents an important nexus between fundamental and applied research. Just as fundamental yeast research transitioned from classical, reductionist strategies to whole-genome techniques, whole-genome studies are advancing to the next level of biological research, referred to as systems biology. Industries that rely on high-performing yeast, such as the wine industry, are therefore poised to reap the many benefits that systems biology can provide. This includes the pro- mise of strain development at speeds and costs which are unobtainable using current techniques. This article reviews the current state of whole-genome techniques available to yeast researchers and outlines how these processes can be used to obtain ‘systems-level’ infor- mation to provide insights into winemaking. The oldest biotechnology meets cutting-edge science The yeast Saccharomyces cerevisiae has been at the forefront of scientific research for decades. Not only is it an excellent model organism for studies in genetics, biochemistry and cell biology, but it also has tremendous economic importance in the food and beverage industries. Nowhere in science is the relationship between applied and fundamental research as evident as in yeast research. In fact, it can be argued that the first biochemists, particu- larly enzymologists, were yeast scientists studying beer and wine fermentation; the word ‘enzyme’ is derived from Greek for ‘in leaven’, referring to leavening from the fermentation of sugars by yeast (for reviews on this history, see Refs [1–3]). Similarly, the field of yeast genetics was pioneered by Øjvinde Winge while working on brewing yeasts in the laboratories of Carlsberg in Denmark [4], and in the forthcoming years, yeast rese- arch, like other biological fields, followed a reductionist approach, characterized by deconstructing complex bio- logical systems into smaller pathways that were amenable to study. However, technological advances have allowed biological research to transition from single-gene, reduc- tionist studies into the ‘whole-genome’ era (Figure 1). Owing to its desirable characteristics, S. cerevisiae has continued to be the system of choice for the development of the majority of these high-throughput ’omic technol- ogies, which are summarized and defined in Box 1. Now, by combining data from multiple whole-genome and classical sources with computational modelling, a new level of biological research, termed systems biology, is emerging (Figure 1). This discipline aims to model cellular functions such that realistic predictions can be made as to how cells will function under given conditions or pertur- bations before experiments are performed. Many whole-genome methodologies have now matured to the point where they have left the research laboratory and are being readily applied in an industrial environ- ment. Likewise, as the field of systems biology matures, data obtained from whole-genome investigation of indus- trial strains will result in improvements to yeast-based processes which would be all but unachievable using cur- rent single-gene techniques. This review seeks to outline the whole-genome methodologies that are available for the study of S. cere- visiae, and how these methods can be integrated to provide a systems-level understanding of yeast growth and metab- olism to be applied in an industrial context. S. cerevisiae as an experimental model for whole genome technology S. cerevisiae was the first eukaryotic organism to have its entire genome sequenced, providing the raw material for true whole-genome technologies [5]. The first of these was published less than a year later, with the expression of all of the genes predicted from the S. cerevisiae genome sequence being monitored simultaneously by the newly-developed technology of DNA microarrays [6]. Since this pioneering study, the use of DNA microarrays to follow global changes in gene expression has blossomed. There are well over 200 microarray datasets concerning S. cerevisiae (each repre- senting multiple array experiments) currently residing in the Gene Expression Omnibus (GEO) public microarray database (http://www.ncbi.nlm.nih.gov/geo/). Since their development for the study of gene expression, the applications for DNA microarrays have diversified enormously such that this platform is now the basis for investigating several discrete biological phenomena (Figure 2a). DNA arrays are now used to deter- mine the location of transcription factor binding sites (ChIP chip) [7,8], to identify protein–RNA interactions (RIP chip) [9] and to detect differences in both DNA content (dupli- cations and deletions) or sequence (nucleotide polymorph- isms) between yeast strains by comparative genome hybridisation (CGH) [10,11]. A modified form of CGH can also be used to identify interstrain polymorphisms and map quantitative trait loci (QTLs), expediting the identification of complex multifactorial loci that influence important phenotypes in yeast [12–14]. Review TRENDS in Biotechnology Vol.25 No.8 Corresponding author: Pretorius, I.S. (sakkie.pretorius@awri.com.au). Available online 27 June 2007. www.sciencedirect.com 0167-7799/$ – see front matter ß 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.tibtech.2007.05.006