Modular Metabolic Engineering and Synthetic Coculture Strategies for the Production of Aromatic Compounds in Yeast

Abstract

Microbial-derived aromatics provide a sustainable and renewable alternative to petroleum-derived chemicals. In this study, we used the model yeast Saccharomyces cerevisiae to produce aromatic molecules by exploiting the concept of modularity in synthetic biology. Three different modular approaches were investigated for the production of the valuable fragrance raspberry ketone (RK), found in raspberry fruits and mostly produced from petrochemicals. The first strategy used was modular cloning, which enabled the generation of combinatorial libraries of promoters to optimize the expression level of the genes involved in the synthesis pathway of RK. The second strategy was modular pathway engineering and involved the creation of four modules, one for product formation: RK synthesis module (Mod. RK); and three for precursor synthesis: aromatic amino acid synthesis module (Mod. Aro), p-coumaric acid synthesis module (Mod. p-CA), and malonyl-CoA synthesis module (Mod. M-CoA). The production of RK by combinations of the expression of these modules was studied, and the best engineered strain produced 63.5 mg/L RK from glucose, which is the highest production described in yeast, and 2.1 mg RK/g glucose, which is the highest yield reported in any organism without p-coumaric acid supplementation. The third strategy was the use of modular cocultures to explore the effects of division of labor on RK production. Two two-member communities and one three-member community were created, and their production capacity was highly dependent on the structure of the synthetic community, the inoculation ratio, and the culture media. In certain conditions, the cocultures outperformed their monoculture controls for RK production, although this was not the norm. Interestingly, the cocultures showed up to 7.5-fold increase and 308.4 mg/L of 4-hydroxy benzalacetone, the direct precursor of RK, which can be used for the semi-synthesis of RK. This study illustrates the utility of modularity in synthetic biology tools and their applications to the synthesis of products of industrial interest.

Keywords: combinatorial engineering; division of labor; microbial communities; p-coumaric acid; raspberry ketone; synthetic biology.

Figure 1

Raspberry ketone synthesis pathway refactoring in Saccharomyces cerevisiae. (a) RK synthesis pathway in S. cerevisiae, which begins with the conversion of L-tyrosine to p-coumaric acid by tyrosine ammonia lyase (TAL). Conversion of p-coumaric acid is then converted into RK in three additional enzymatic steps, by p-coumaroyl-CoA synthetase (4CL), benzylacetone synthase (BAS), and raspberry ketone reductase (RKS). (b) Promoter-ORF combinations used in the combinatorial promoter library for tuning the expression of the RK synthesis pathway enzymes. (c) RK titers of 96 randomly selected strains after 3 days of growth in SC minus uracil media. Experimental measurements are RK amounts as determined by LC–MS from spent media and shown as individual values from a single replicate. (d) Relative promoter strengths of the RK synthesis pathway genes from the five strains identified in the RK producer screen. Promoter strengths are rank order (1–15) corresponding to their relative promoter strength as characterized in the Yeast MoClo Toolkit.

Figure 2

Modular metabolic engineering strategy to improve raspberry ketone synthesis and production in S. cerevisiae. (a) Four modules are introduced for the RK synthesis pathway, and the enzymes of each module are grouped by colors. Mod. RK (red) is described in Figure 1, Mod. Aro (yellow): overexpression of ScARO3^K222L, ScARO4^K229L: DAHP synthase; ScARO7^G141S: chorismate mutase; Mod. p-CA (green): overexpression of VvPAL: phenylalanine ammonia-lyase from Vitis vinifera; AtC4H: cinnamate-4-hydroxylase from Arabidopsis thaliana; FjTAL: tyrosine ammonia-lyase from Flavobacterium johnsoniae; Mod. M-CoA (blue): overexpression of ScALD6: aldehyde dehydrogenase from S. cerevisiae; SeACS1^L641p: acetyl-CoA synthetase with mutate site L641P from Salmonella enterica; ScACC1^S659A,S1157A: acetyl-CoA carboxylase with two mutation sites: S659A, S1157A from S. cerevisiae. (b) Description of gene combinations of each module and modular engineered strains. (c–f) Engineering strains are cultured at 30 °C and 250 rpm in 96 deep well plates in synthetic minimal medium (SM) and synthetic complete medium (SC) for 72 h. The OD values from plate reader are shown in (c) and (e). The precursors p-coumaric acid (p-CA), 4-hydroxy benzalacetone (HBA) and product RK are shown in (d) and (f). These results show the average of three replicates and the SD.

Figure 3

Modular coculture strategy via division of labor to explore the production of RK. (a) Diagram of modular coculture strategies: p-CA producing strain RLAs1654 was used as the sender strain (S1) and two receiver strains (S2, S3) including RLAs1642 and RLAs1643 based on their promising RK titers from monocultures. Two pairs of two-member cocultures (CL_RK1, CL_RK2) and one pair of three-member coculture (CL_RK3) were designed, and each member within these cocultures was communicated by p-CA diffusion. (b) Different initial inoculation ratios were selected to study the production of RK, p-CA, and HBA. The ratios for CL_RK1 were 20:1, 6:1, 1:1, 1:6, 1:20, and 1:100; for CL_RK2, they were 1:1, 1:20, and 1:100; and for CL_RK3, they were 1:1:1, 1:20:1, and 10:20:1. These three pairs of cocultures were cultured using 96-well deep plates in SM and SC for 72 h. (c, d) The OD values of cocultures in both SM and SC at 72 h from a microplate reader. (e, f) Titers of precursor p-CA and product HBA and RK at 72 h in both SM and SC. The values presented here are average of triplicates and SD.