Transportome-wide engineering of Saccharomyces cerevisiae

Abstract

Synthetic biology enables the production of small molecules by recombinant microbes for pharma, food, and materials applications. The secretion of products reduces the cost of separation and purification, but it is challenging to engineer due to the limited understanding of the transporter proteins' functions. Here we describe a method for genome-wide transporter disruption that, in combination with a metabolite biosensor, enables the identification of transporters impacting the production of a given target metabolite in yeast Saccharomyces cerevisiae. We applied the method to study the transport of xenobiotic compounds, cis,cis-muconic acid (CCM), protocatechuic acid (PCA), and betaxanthins. We found 22 transporters that influenced the production of CCM or PCA. The transporter of the 12-spanner drug:H(+) antiporter (DHA1) family Tpo2p was further confirmed to import CCM and PCA in Xenopus expression assays. We also identified three transporter proteins (Qdr1p, Qdr2p, and Apl1p) involved in betaxanthins transport. In summary, the described method enables high-throughput transporter identification for small molecules in cell factories.

Keywords: Betaxanthins; Cell factory; Metabolic engineering; Muconic acid; Protocatechuic acid; Transporter protein.

Fig. 1

Overview of the workflow for transporter identification on the example of cis, cis-muconic acid. (a) Double-stranded DNA fragments were designed and synthesized for Cas9-mediated disruption of 361 S. cerevisiae putative non-essential transporters. The fragments comprise the sgRNA and 165 bp repair donor that introduces premature termination codons via 1 bp frameshift. (b) Detection of in vivo cis, cis-muconic acid (CCM) production using BenM biosensor. A transportome-wide mutant library was generated and sorted by FACS. The high- and low-fluorescent clone pools were sequenced to identify the transporters that influence CCM production.

Fig. 2

Optimization of CRISPR-Cas9-mediated gene disruption in S. cerevisiae. (a) Overview of the plasmid and donor fragments design. (b) The efficiency of ADE2 disruption in a strain carrying a genome integrated Cas9 (ST8420) transformed with 1 μg of plasmid or plasmid pool. The plasmids were constructed with double-stranded DNA fragments consisting of the sgRNA-scaffold RNA-SUP4t-donor construct. Data shown are mean values ± SDs of triplicates for transformant number and ADE2 disruption ratio, while one representative plate is shown for each transformation.

Fig. 3

Biosensor setup and validation for the high-throughput screen. (a) The metabolic pathway for PCA and CCM biosynthesis. (b) Fluorescence output of CCM-responsive BenM biosensor in CCM producing strains. (c) Flow cytometric analysis of the CCM producing strains harboring BenM biosensor. (d) Fluorescence output of PCA-responsive PcaQ biosensor in PCA producing strains. MM, mineral medium, SD-CSM, synthetic defined medium with complete supplement mixture. Data shown in panel b and d are mean values ± SDs of biological triplicates, while a representative sample is shown for flow cytometric analysis in panel c and d.

Fig. 4

Frequency change-based transporter identification for altered CCM and PCA production. (a, b) Frequency change of transporter genes for reverse validation. The frequency of each transporter was defined as the ratio of the read count to the total experimental counts for each sample. Frequency change for each gene was compared with the corresponding control samples. (c, d) The CCM or PCA production of strains reverse engineered with the transporter disruption. ST8424 and ST8859 were used as the parental strain, respectively, and metabolite was quantified on 72 h culture. Data shown are from biological replicates in panel a and b, and data shown in panel c and d are mean values ± SDs of biological triplicates. Statistical differences between the control and test strains were determined by two-tailed Student's t-test. *P < 0.05; **P < 0.01.

Fig. 5

Characterization of transport activity of Tpo2p. (a) Intracellular and extracellular concentrations of CCM in the yeast strains with disrupted Tpo2p. EV – ST8424 with an empty vector (negative control). EV+2^nd AroY - ST8424 expressing an additional copy of AroY (positive control). TPO2-dis – ST8424 with a disrupted TPO2 gene. The strains were cultivated for 72 h in MM pH 6.0. (b–g) The transporter activity of Tpo2p on CCM and PCA. Oocyte cells expressing TPO2 were injected with CCM and PCA solution (b, c, and d) or immersed into CCM and PCA solution (e, f, and g). Cellular CCM and PCA were extracted and measured using LC-MS. Data shown are mean values ± SDs of biological triplicates. Statistical difference between control (Con) and indicated samples was determined by two-tailed Student's t-test. *P < 0.05; **P < 0.01.

Fig. 6

Transporter identification for betaxanthins (a) Tyrosine hydroxylase (TYH) and DOPA dioxygenase (DOD) were expressed in yeast to produce betalamic acid (BA). BA spontaneously reacts with intracellular amino acids, generating yellow to orange fluorescent pigments - betaxanthins. The betaxanthins producing yeast is transformed with plasmid library for transporter disruption, and the resulting clones are tested for increased intracellular betaxanthins accumulation (Btx_intra). The causal transporter disruption are identified by sgRNA-donor sequencing. Betaxanthins production (b) and distribution (c) by variant strains pre-selected in the screening, based on increased color intensity in colonies and higher cellular betaxanthins retention. Data shown in panel b and c are mean values ± SDs of biological triplicates. Statistical difference between control (ST10367) and indicated strains was determined by two-tailed Student's t-test. *P < 0.05; **P < 0.01.