Controlling heterologous gene expression in yeast cell factories on different carbon substrates and across the diauxic shift: a comparison of yeast promoter activities

Abstract

Background: Predictable control of gene expression is necessary for the rational design and optimization of cell factories. In the yeast Saccharomyces cerevisiae, the promoter is one of the most important tools available for controlling gene expression. However, the complex expression patterns of yeast promoters have not been fully characterised and compared on different carbon sources (glucose, sucrose, galactose and ethanol) and across the diauxic shift in glucose batch cultivation. These conditions are of importance to yeast cell factory design because they are commonly used and encountered in industrial processes. Here, the activities of a series of "constitutive" and inducible promoters were characterised in single cells throughout the fermentation using green fluorescent protein (GFP) as a reporter.

Results: The "constitutive" promoters, including glycolytic promoters, transcription elongation factor promoters and ribosomal promoters, differed in their response patterns to different carbon sources; however, in glucose batch cultivation, expression driven by these promoters decreased sharply as glucose was depleted and cells moved towards the diauxic shift. Promoters induced at low-glucose levels (P(HXT7), P(SSA1) and P(ADH2)) varied in induction strength on non-glucose carbon sources (sucrose, galactose and ethanol); in contrast to the "constitutive" promoters, GFP expression increased as glucose decreased and cells moved towards the diauxic shift. While lower than several "constitutive" promoters during the exponential phase, expression from the SSA1 promoter was higher in the post-diauxic phase than the commonly-used TEF1 promoter. The galactose-inducible GAL1 promoter provided the highest GFP expression on galactose, and the copper-inducible CUP1 promoter provided the highest induced GFP expression following the diauxic shift.

Conclusions: The data provides a foundation for predictable and optimised control of gene expression levels on different carbon sources and throughout batch fermentation, including during and after the diauxic shift. This information can be applied for designing expression approaches to improve yields, rates and titres in yeast cell factories.

Figure 1

GFP activity driven by different promoters on different carbon sources. GFP fluorescence, culture pH and biomass accumulation (OD₆₀₀) of the P _TEF1-yEGFP strain (a) and the P _TEF1-yEGFP-CLN2 _PEST strain (b) in flask batch cultivation in YNB broth with 20 g L⁻¹ glucose as the carbon source are shown. GFP fluorescence of various promoter-yEGFP-CLN2 _PEST strains on 20 g L⁻¹ glucose in microtitre plate culture (c) and of various promoter-yEGFP strains on various carbon sources in microtitre plate culture (d) are also shown. The GFP fluorescence in (c) was ranged using Tukey’s test: three levels were identified (dashed lines a > b > c) within which the difference between group members (bold lines) was insignificant (p > 0.05). In TEF1-M (d) construction, A XhoI site plus triple “A” was inserted between TEF1 promoter and the start codon of yEGFP. The insert in (d) shows a zoomed-in GFP fluorescence scale for the weaker promoters, P _PDA1, P _CYC1, P _TPS1 and P _CUP1. The analysis of variance for the fluorescence levels in (d) is shown in Additional file 1: Figure S3. The auto-fluorescence was determined from the reference strains (ILHA GH4 for the yEGFP strains and ILHA GFP3 for the yEGFP-CLN2 _PEST strains) in parallel. Symbol Asterisk represents that the value is <50 and not significantly different from auto-fluorescence (t test, p > 0.05). Mean values ± standard deviations are shown from replicate cultivations.

Figure 2

Promoter performance over the diauxic shift: a cell growth and pH of the reference strain ILHA GH4; b the profile of extracellular metabolites. c the fluorescence of GFP controlled by promoters classically considered to be “constitutive”; d the fluorescence of GFP controlled by low-glucose-inducible promoters. The vertical dash lines are at 8, 12 and 16 h. The auto-fluorescence was determined from the reference strain in parallel. The medium was buffered with 100 mM MES. Mean values ± standard deviations are shown from duplicate cultivations.

Figure 3

Copper induction of the CUP1 promoter: a population growth of the reference strain ILHA GH4; b CUP1 promoter-regulated GFP activity. 1 M copper sulphate was added to final concentrations of 0, 100, 200 or 300 μM at 5 h (vertical dash line). The auto-fluorescence was determined from the reference strain in parallel. The medium was buffered with 100 mM MES. Mean values ± standard deviations from duplicate cultivations are shown.

Figure 4

Physical maps of plasmids pITGFP3 (a) and pILGFP3 (b): rep pUC19 replicon in E. coli, bla ampicillin resistant gene in E. coli, P _URA3 URA3 promoter of S. cerevisiae, KlURA3 Kluyveromyces lactis URA3 gene, yEGFP yeast enhanced green fluorescence gene, CLN2 _PEST encoding the protein-destabilizing peptide from cyclin 1 of S. cerevisiae, T _URA3 URA3 terminator of S. cerevisiae.