Genome-wide analytical approaches for reverse metabolic engineering of industrially relevant phenotypes in yeast

Abstract

Successful reverse engineering of mutants that have been obtained by nontargeted strain improvement has long presented a major challenge in yeast biotechnology. This paper reviews the use of genome-wide approaches for analysis of Saccharomyces cerevisiae strains originating from evolutionary engineering or random mutagenesis. On the basis of an evaluation of the strengths and weaknesses of different methods, we conclude that for the initial identification of relevant genetic changes, whole genome sequencing is superior to other analytical techniques, such as transcriptome, metabolome, proteome, or array-based genome analysis. Key advantages of this technique over gene expression analysis include the independency of genome sequences on experimental context and the possibility to directly and precisely reproduce the identified changes in naive strains. The predictive value of genome-wide analysis of strains with industrially relevant characteristics can be further improved by classical genetics or simultaneous analysis of strains derived from parallel, independent strain improvement lineages.

Fig. 1

The ‘forward’ metabolic engineering and ‘reverse’ metabolic engineering cycles and their interaction. In forward metabolic engineering, analysis of strains constructed based on rational design often results in scientific questions that need to be addressed by further analysis and consultation of the rapidly expanding knowledge on microbial metabolism and its regulation. In reverse metabolic engineering, generation and analysis of biodiversity – obviously, with special attention for strains that show improved performance – contributes to accelerated strain improvement and knowledge development (After Nielsen, 2001; Bailey et al., 1996; Bro & Nielsen, 2004).

Fig. 2

Transcriptome comparison of a genetically engineered Saccharomyces cerevisiae strain and an arabinose-fermenting derivative strain obtained by laboratory evolution. The dotted lines indicate a twofold difference. The circled dots represent genes located between positions 543 555 and 807 659 with an at least twofold increased transcript level in the evolved strain, indicating a duplication of a 250-kb region on chromosome VII. The increased expression level of YGR043C (diamond), encoding a transaldolase, was subsequently shown to contribute to enhanced arabinose fermentation rates in the evolved strain (reproduced with permission from Wisselink et al., 2010).

Fig. 3

Analyses of evolved Saccharomyces cerevisiae strains adapted for glucose-limited cultivation conditions illustrate the advantages of whole genome sequencing over microarray-based transcriptome analysis in reverse metabolic engineering. Transcriptome analysis of four independently evolved strains consistently yielded more than 180 differentially expressed genes. Genotyping of a single cell line using tiling arrays and whole genome sequencing showed a much smaller number of underlying nonconservative mutations. Some mutations identified by whole genome sequencing went unnoticed by a previous analysis using tiling arrays.