Evolving dual-trait EPSP synthase variants using a synthetic yeast selection system

Abstract

The enzyme 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS) functions in the shikimate pathway which is responsible for the production of aromatic amino acids and precursors of other essential secondary metabolites in all plant species. EPSPS is also the molecular target of the herbicide glyphosate. While some plant EPSPS variants have been characterized with reduced glyphosate sensitivity and have been used in biotechnology, the glyphosate insensitivity typically comes with a cost to catalytic efficiency. Thus, there exists a need to generate additional EPSPS variants that maintain both high catalytic efficiency and high glyphosate tolerance. Here, we create a synthetic yeast system to rapidly study and evolve heterologous EPSP synthases for these dual traits. Using known EPSPS variants, we first validate that our synthetic yeast system is capable of recapitulating growth characteristics observed in plants grown in varying levels of glyphosate. Next, we demonstrate that variants from mutagenesis libraries with distinct phenotypic traits can be isolated depending on the selection criteria applied. By applying strong dual-trait selection pressure, we identify a notable EPSPS mutant after just a single round of evolution that displays robust glyphosate tolerance (K_i of nearly 1 mM) and improved enzymatic efficiency over the starting point (~2.5 fold). Finally, we show the crystal structure of corn EPSPS and the top resulting mutants and demonstrate that certain mutants have the potential to outperform previously reported glyphosate-resistant EPSPS mutants, such as T102I and P106S (denoted as TIPS), in whole-plant testing. Altogether, this platform helps explore the trade-off between glyphosate resistance and enzymatic efficiency.

Keywords: EPSPS; directed evolution; glyphosate; synthetic biology; yeast.

Fig. 1.

Generation of a tetrafunctional ScARO1 by removing EPSPS function and development of a synthetic yeast model enables growth dependence of a heterologous EPSPS. (A) Overview of the shikimate pathway in plants, beginning with PEP and E4P converted into DAHP by DAHP synthase. Glyphosate competitively inhibits EPSP synthase to prevent the creation of aromatic amino acids. (B) Overview of the wild-type pentafunctional ARO1 in yeast. Mutation D731A successfully abolished catalytic function of EPSPS portion while keeping other domains intact and functional. (C) Schematic of synthetic yeast model with two plasmids. (D) Synthetic yeast host either expressing ZmEPSPS (green curve) or lacking ZmEPSPS (gray curve). Data represent the average of three biological triplicates picked from individual colonies at random, and the shaded error bar region represents the SE.

Fig. 2.

Tuning the selection pressure of synthetic yeast host via expression optimization. (A) Specific Growth Rate for EPSPS and TIPS when expressed under varying promoter strengths (promoter series) in 0 mM glyphosate. (B) Resulting growth rates in 5 mM glyphosate. Differences in growth rate between EPSPS and TIPS enable a selection window. Core1p promoter was selected due to more consistent growth characteristics and differences between wild-type and mutant EPSP synthases. (C) Growth curves of well-studied mutants P106S and TIPS compared to wild-type EPSPS Variants grown in 0.5 mM glyphosate in the synthetic yeast model, near the maximum value of what is seen in plant tissue during field application. Glyphosate tolerance is highest in TIPS. (D) Variants grown in 0 mM glyphosate in the synthetic yeast model. The growth deficit caused by TIPS is evident. Absorbance at 600 nm was measured on a Tecan plate reader over time. Each column represents the average of three biological replicates, where specific growth rate was calculated individually for each replicate. Error bars represent the SD. Statistical analysis was performed using ANOVA with Tukey’s test; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns = no significance.

Fig. 3.

A comparison of example variants resulting from either mono- or dual-trait selection pressure. (A) EPSPS evolution entry point mutation pools. Each point is the average growth rate of biological triplicates grown in either 0 mM or 0.25 mM glyphosate, the level at which wild-type EPSPS cannot grow. (B) TIPS evolution entry point mutation pools. Each point is the average growth rate of biological triplicates grown in either 0 mM or 5 mM glyphosate, a level at which high glyphosate tolerance can be evaluated.

Fig. 4.

Specific growth rate of each EPSPS variant in increasing levels of glyphosate. Each heatmap square represents the average of three biological replicates.

Fig. 5.

Crystal structure of ZmEPSPS TIPS (PSKR) with mutations highlighted. The original TIPS mutations are circled (dotted yellow). Complete X-ray crystallography data collection and refinement statistics are provided in SI Appendix, Table S1.

Fig. 6.

Maize plants expressing selected EPSPS variants demonstrated tolerance to glyphosate. (A) Wild type maize plants were transformed with the indicated variants and with TIPS as a positive control. Plants were regenerated on media which included glyphosate to select only those plants that were transformed with the indicated EPSPS variant and demonstrated glyphosate tolerance. Following regeneration of plants, the plants were allowed to recover and were screened for molecular quality and plant health. The resulting advanced plants were moved to the greenhouse for seed production and surplus plants were spray challenged with glyphosate. (B) Only two EPSPS variants resulted in enough plants for a glyphosate spray challenge, the TIPS positive control and TIPS P126S/K296R. Plants were challenged with a glyphosate spray and injury observed 1 wk following treatment. Three of the 10 TIPS control plants were injured by the glyphosate spray while none of the TIPS P126S/K296R expressing plants showed visual injury. (C) Plants were assayed in a hybrid background in the following generation in a typical greenhouse spray assay and monitored for glyphosate injury using visual injury ratings (plant stunting, chlorosis, necrosis, malformation, and death) at 18 d posttreatment.