Improving Industrially Relevant Phenotypic Traits by Engineering Chromosome Copy Number in Saccharomyces pastorianus

Abstract

The lager-brewing yeast Saccharomyces pastorianus is a hybrid between S. cerevisiae and S. eubayanus with an exceptional degree of aneuploidy. While chromosome copy number variation (CCNV) is present in many industrial Saccharomyces strains and has been linked to various industrially-relevant traits, its impact on the brewing performance of S. pastorianus remains elusive. Here we attempt to delete single copies of chromosomes which are relevant for the production of off-flavor compound diacetyl by centromere silencing. However, the engineered strains display CNV of multiple non-targeted chromosomes. We attribute this unintended CCNV to inherent instability and to a mutagenic effect of electroporation and of centromere-silencing. Regardless, the resulting strains displayed large phenotypic diversity. By growing centromere-silenced cells in repeated sequential batches in medium containing 10% ethanol, mutants with increased ethanol tolerance were obtained. By using CCNV mutagenesis by exposure to the mitotic inhibitor MBC, selection in the same set-up yielded even more tolerant mutants that would not classify as genetically modified organisms. These results show that CCNV of alloaneuploid S. pastorianus genomes is highly unstable, and that CCNV mutagenesis can generate broad diversity. Coupled to effective selection or screening, CCNV mutagenesis presents a potent tool for strain improvement.

Keywords: Saccharomyces pastorianus; chromosome copy number stability; chromosome missegregation; lager beer brewing; strain engineering.

FIGURE 1

Centromere silencing strain construction. (A) The plasmid pART series illustrated by pART008 is designed to introduce the amdS selection marker and the constitutive strong promoter GAL1 (GAL1_p) upstream a centromere (ScCEN6 in the case of pART008). In addition pART plasmids contain an E. coli replication origin from pBR322 and the bla gene conferring resistance to β-lactam antibiotics. (B) Integration of the centromere silencing construct at the ScCEN6 locus in S. pastorianus CBS 1483. Schematic representation of the integration of amdS selection marker and the GAL1 promoter containing linear fragment amplified with primers 8451 and 8454. The integration is directed by homology regions of ∼500 bp to complete the double cross over integration.

FIGURE 2

Vicinal diketones formation in S. pastorianus. (A) Schematic illustration of the diacetyl, pentanedione and branched chain amino acid biosynthesis pathways. (B) Valine, isoleucine, pentanedione and diacetyl biosynthetic pathways and chromosome copy number of the genes involved in the metabolic routes in strains CBS 1384 (blue bar), IMI373 (ScCHRX-SeCHRX^{– 1} SeCHRIX^–1) (green bar) that was targeted for loss of ScCHRX, IMI375 (ScCHRII^–1 ScCHRV^–1 ScCHRVIII^–3 ScCHRXIII^–1 ScCHRXIV^–1 SeCHRI^–1 SeCHRIII-ScCHRIII^–1 SeCHRVII-ScCHRVII^–1 SeCHRVIII-XV⁺¹) (orange bar) that was targeted for loss of ScCHRXIV and IMI381 (ScCHRIV⁺¹ ScCHRVIII^–1 ScCHRIX⁺¹ ScCHRXI^–1 ScCHRXII^–1 SeCHRX-SeCHRX^–1 SeCHRII-IV^–1 SeCHRXI⁺¹) (yellow bar) that was targeted for the loss of SeCHRX. The chromosome copy number was predicted using Magnolya (Nijkamp et al., 2012).

FIGURE 3

Chromosome copy number in various S. pastorianus strains. (a) Chromosome copy number of the parental S. pastorianus strain CBS 1483. (b) Chromosome copy number of strains obtained after centromere silencing of SeCEN6 (in IMI350), ScCEN9 (in IMI352), ScCEN10 (in IMI359), ScCEN12 (in IMI360), ScCEN14 (in IMI361), SeCEN3 (in IMI363), SeCEN8 (in IMI366), SeCEN10 (in IMI367), SeCEN12 (in IMI368) and SeCEN14 (in IMI369). (c) Chromosome copy number of a strain engineered for targeted loss of ScCEN14 before centromere silencing induction. (d) Chromosome copy number of strains after electroporation without DNA. (e) Chromosome copy number of single isolates of the S. pastorianus strain CBS 1483. (f) Chromosome copy number of isolates obtained of treatment with methyl benzimidazole 2 yl carbamate (MBC). Chromosome copy number was predicted using Magnolya (Nijkamp et al., 2012).

FIGURE 4

Characterisation of CBS 1483, and IMI373, IMI375 and IMI381 under brewing conditions. The S. pastorianus strains CBS 1483 (black circle), IMI373 (ScCHRX-SeCHRX^–1 SeCHRIX^–1) (white circle), IMI375 (ScCHRII^–1 ScCHRV^–1 ScCHRVIII^–3 ScCHRXIII^–1 ScCHRXIV^–1 SeCHRI^–1 SeCHRIII-ScCHRIII^–1 SeCHRVII-ScCHRVII^–1 SeCHRVIII-XV⁺¹) (red circle) and IMI381 (ScCHRIV⁺¹ ScCHRVIII^–1 ScCHRIX⁺¹ ScCHRXI^–1 ScCHRXII^–1 SeCHRX-SeCHRX^–1 SeCHRII-IV^–1 SeCHRXI⁺¹) (green circle) were grown in air-capped 100 ml serum bottles in undiluted industrial wort at 12 °C. Average and standard deviation from duplicates (IMI373 and IMI375) or triplicates (IMI381 and CBS 1483) are shown. (A) displays glucose, (B) maltose, (C) maltotriose, (D) ethanol determined by liquid chromatography; (E) diacetyl, (F) pentanedione concentrations produced in industrial wort measured using static headspace gas chromatography and (G) optical density measured at 660 nm (OD₆₆₀ is directly related with biomass concentration in suspension).

FIGURE 5

Characterisation of mutants obtained by centromere-silencing mutagenesis in medium containing 10% ethanol. Strains IMX1882 (blue squares), IMX1886 (blue triangles), IMX1891 (blue diamonds), IMX1893 (blue circles), CBS 1483 (red squares) and IMI361 (red triangles) were grown in triplicate in micro-aerobic bottle fermentations on SMD medium containing 10% ethanol at 20°C during 9 days. During the experiment OD₆₆₀ (A) and extracellular glucose concentrations (B) were monitored. The growth rates were calculated from at least 6 measurements with an R² superior to 0.95 (C).

FIGURE 6

Characterisation of mutants obtained by MBC mutagenesis in medium containing 10% ethanol. Strains IMS0687 (blue squares), IMS0698 (blue triangles), IMS0703 (blue diamonds), IMS0716 (blue circles) and CBS 1483 (red squares) were grown in triplicate in micro-aerobic bottle fermentations on SMG medium containing 10% ethanol at 20°C during 9 days. During the experiment OD₆₆₀ (A) and extracellular glucose concentrations (B) were monitored. The growth rates were calculated from at least 6 measurements with an R² superior to 0.95 (C).

FIGURE 7

SNV on monosomic ScCHRXIII after MCB treatment predominantly originate from Loss of heterozygosity. (A) Venn Diagram of Single Nucleotide Variations (SNV) found in coding sequences relative to the parental strain CBS 1483 identified in S. pastorianus strains IMS0687 (blue), IMS0703 (red) and IMS0716 (green) obtained after MBC treatment. Genes found in IMS0687-IMS0703 intersection and denoted in red were located on ScCHRXIII whose copy number decreased from two to one copy. (B) Schematic representation of the SNV derived from LOH on ScCHRXIII of IMS0687 and IMS0703 MBC treated mutants.

FIGURE 8

Bar plot of chromosome copy number changes observed in strains generated in this study. Chromosomes were not lost equally in the sequenced strains, untargeted chromosome alterations have been summed over all the analyzed strains for each chromosome and shown in order of observed copy number changes. Instances of loss are shown in green and gain in red.