Rosemary Anne Cripwell

Increasing interest in the bioeconomy has spurred the development of integrated methods to convert organic waste streams, particularly starch-rich substrates, into bioethanol. However, starch-based ethanol fermentations are vulnerable to bacterial contamination, particularly by lactic acid bacteria (LAB). Severe contamination can cause significant economic losses due to stuck fermentations and ethanol plant shutdowns. Although bacterial contamination can be managed with antibiotics, this approach is not cost-effective at an industrial scale and may increase the risk of selecting for antibiotic-resistant strains. Natural antimicrobial peptides (AMPs) can inhibit LAB contaminants in yeast fermentations, but commercial applications are limited by their low abundance and high production costs. Engineering Saccharomyces cerevisiae to produce recombinant AMPs might provide a cost-effective strategy to control LAB, thereby boosting ethanol yields during fermentation. Despite a comprehensive toolkit for gene expression in S. cerevisiae, only a few successful cases of bacteriocin expression have been reported. Since starch-to-ethanol fermentation is a key application for recombinant AMPs, this review explores strategies to optimize the expression of bacteriocin-encoding genes in S. cerevisiae. The ideal scenario would be a single yeast strain capable of producing amylases for starch hydrolysis, fermenting glucose to ethanol, and expressing bacteriocins to inhibit LAB contaminants. One-sentence summary: Yeast strains can produce heterologous antimicrobial peptides that help prevent contaminating bacteria from interfering with the starch-to-ethanol fermentation process.

Recombinant expression in the yeast Saccharomyces cerevisiae offers an alternative approach to developing large-scale production systems for class II bacteriocins from lactic acid bacteria, such as enterocin A, mundticin ST4SA and plantaricin 423. An important consideration for bacteriocin activity is disulphide bond formation: mature mundticin ST4SA has one, and plantaricin 423 and enterocin A each have two disulphide bonds. The native bacteriocin operon typically includes accessory proteins that facilitate disulphide bond formation, but this gene is absent in the enterocin A operon. In this study, the recombinant expression of a codon-optimised gene for enterocin A in S. cerevisiae , was compared to that for a codon-optimised plantaricin 423 and mundticin ST4SA, previously successfully expressed in S. cerevisiae . Shake flasks delivered more than twofold higher peptide EntA_Opt levels than PlaX_Opt and MunX_Opt, with even higher EntA_Opt expression levels in batch fermentations. However, the bacteriocin activity of plantaricin 423 was considerably lower than that of enterocin A and mundticin ST4SA. It is postulated that this could be a result of incorrect disulphide bond conformation due to the absence of the plantaricin accessory protein, PlaC, in S. cerevisiae . Nano-LC-MS/MS analysis showed various post-translational modifications for the peptides, with a greater proportion of EntA_Opt peptides with the correct disulphide bond conformation than for PlaX_Opt. This study demonstrated that S. cerevisiae is a promising host to produce recombinant class IIa bacteriocins, particularly enterocin A. However, the co-expression of accessory proteins should be investigated to improve the activity of recombinant plantaricin 423.

Antimicrobial peptides or bacteriocins are excellent candidates for alternative antimicrobials, but high manufacturing costs limit their applications. Recombinant gene expression offers the potential to produce these peptides more cost-effectively at a larger scale. Saccharomyces cerevisiae is a popular host for recombinant protein production, but with limited success reported on antimicrobial peptides. Individual recombinant S. cerevisiae strains were constructed to secrete two class IIa bacteriocins, plantaricin 423 (PlaX) and mundticin ST4SA (MunX). The native and codon-optimised variants of the plaA and munST4SA genes were cloned into episomal expression vectors containing either the S. cerevisiae alpha mating factor (MFα1) or the Trichoderma reesei xylanase 2 (XYNSEC) secretion signal sequences. The recombinant peptides retained their activity and stability, with the MFα1 secretion signal superior to the XYNSEC secretion signal for both bacteriocins. An eight-fold increase in activity against Listeria monocytogenes was observed for MunX after codon optimisation, but not for PlaX-producing strains. After HPLC-purification, the codon-optimised genes yielded 20.9 mg/L of MunX and 18.4 mg/L of PlaX, which displayed minimum inhibitory concentrations (MICs) of 108.52 nM and 1.18 µM, respectively, against L. monocytogenes. The yields represent a marked improvement relative to an Escherichia coli expression system previously reported for PlaX and MunX. The results demonstrated that S. cerevisiae is a promising host for recombinant bacteriocin production that requires a simple purification process, but the efficacy is sensitive to codon usage and secretion signals.

Consolidated bioprocessing (CBP) combines enzyme production, saccharification and fermentation into a one-step process. This strategy represents a promising alternative for economic ethanol production from starchy biomass with the use of amylolytic industrial yeast strains. Recombinant Saccharomyces cerevisiae Y294 laboratory strains simultaneously expressing an α-amylase and glucoamylase gene were screened to identify the best enzyme combination for raw starch hydrolysis. The codon optimised Talaromyces emersonii glucoamylase encoding gene (temG_Opt) and the native T. emersonii α-amylase encoding gene (temA) were selected for expression in two industrial S. cerevisiae yeast strains, namely Ethanol Red™ (hereafter referred to as the ER) and M2n. Two δ-integration gene cassettes were constructed to allow for the simultaneous multiple integrations of the temG_Opt and temA genes into the yeasts’ genomes. During the fermentation of 200 g l−1 raw corn starch, the amylolytic industrial strains were able to ferment raw corn starch to ethanol in a single step with high ethanol yields. After 192 h at 30 °C, the S. cerevisiae ER T12 and M2n T1 strains (containing integrated temA and temG_Opt gene cassettes) produced 89.35 and 98.13 g l−1 ethanol, respectively, corresponding to estimated carbon conversions of 87 and 94%, respectively. The addition of a commercial granular starch enzyme cocktail in combination with the amylolytic yeast allowed for a 90% reduction in exogenous enzyme dosage, compared to the conventional simultaneous saccharification and fermentation (SSF) control experiment with the parental industrial host strains. A novel amylolytic enzyme combination has been produced by two industrial S. cerevisiae strains. These recombinant strains represent potential drop-in CBP yeast substitutes for the existing conventional and raw starch fermentation processes.