Engineering energetically efficient transport of dicarboxylic acids in yeast Saccharomyces cerevisiae

Abstract

Biobased C4-dicarboxylic acids are attractive sustainable precursors for polymers and other materials. Commercial scale production of these acids at high titers requires efficient secretion by cell factories. In this study, we characterized 7 dicarboxylic acid transporters in Xenopus oocytes and in Saccharomyces cerevisiae engineered for dicarboxylic acid production. Among the tested transporters, the Mae1(p) from Schizosaccharomyces pombe had the highest activity toward succinic, malic, and fumaric acids and resulted in 3-, 8-, and 5-fold titer increases, respectively, in S. cerevisiae, while not affecting growth, which was in contrast to the tested transporters from the tellurite-resistance/dicarboxylate transporter (TDT) family or the Na⁺ coupled divalent anion-sodium symporter family. Similar to SpMae1(p), its homolog in Aspergillus carbonarius, AcDct(p), increased the malate titer 12-fold without affecting the growth. Phylogenetic and protein motif analyses mapped SpMae1(p) and AcDct(p) into the voltage-dependent slow-anion channel transporter (SLAC1) clade of transporters, which also include plant Slac1(p) transporters involved in stomata closure. The conserved phenylalanine residue F329 closing the transport pore of SpMae1(p) is essential for the transporter activity. The voltage-dependent SLAC1 transporters do not use proton or Na⁺ motive force and are, thus, less energetically expensive than the majority of other dicarboxylic acid transporters. Such transporters present a tremendous advantage for organic acid production via fermentation allowing a higher overall product yield.

Keywords: SLAC1; Schizosaccharomyces pombe MAE1; cell factories; dicarboxylic acids; efflux transporters.

Fig. 1.

Carboxylic acid transport assays in Xenopus oocytes and S. cerevisiae. (A) The expression of GFP, ScCtp1-GFP, and their N-terminal fusions with the HsOrai1 leader peptide was examined by confocal microscopy, scanning oocytes on the z axis from the most outer surface toward the inner cytoplasmic space. (B) Efflux of carboxylic acids from oocytes. The bars represent the carboxylic acid contents of the medium (means of 3 to 4 biological replicates each involving 20 oocytes with SDs shown as error bars) 3 h after injecting fumarate and citrate into the control (water injected) oocytes with no heterologous transporter and into the oocytes expressing individual candidate transporters. Asterisks mark significant changes in comparison with the control (**P < 0.01, *P < 0.05). The carboxylic acid concentrations in the medium were also examined before metabolite injection (before injection) and after 3 h incubation of oocytes without injecting metabolites (without injection). (C) Time course of metabolite concentrations in the fermentation broth of the transporter-expressing S. cerevisiae strains and the control strain. Error bars show the SDs of 3 biological replicates.

Fig. 2.

SpMae1(p) harbors a SLAC1 protein domain responsible for voltage-dependent transport. The maximum likelihood phylogenetic tree was built on the Whelan and Goldman substitutional matrix. Bootstraps and branch lengths are shown in blue and green, respectively. Protein domains were predicted with e values below 10⁻⁴⁰. Sequence accession numbers are provided in SI Appendix, Table S6.

Fig. 3.

Structural homology of SLAC1 transporters and functional importance of the conserved regions. (A) Structural homology between HiTehA(p) and SpMae1(p). (B) Cytosolic face representation of the HiTehA(p) and SpMae1(p) transport tunnels with conserved phenylalanine residues blocking the substrate transport channel in SLAC1 transporters. (C) The 2 phenylalanine residues residing with their phenyl ring inside the transport channel (F107_SpMae1/F96_HiTehA[p] and F329_SpMae1/F276_HiTehA[p]) are conserved among the SLAC1 transporters. (D) Structural changes in SpMae1 after replacement of the conserved phenylalanines with alanine. (E) Malate concentration in the fermentation broth of yeast expressing wild-type SpMae1(p), SpMae1(p)^F329A, SpMae1(p)^{F107A, F329A}, and SpMae1(p)^{∆C terminus}. Yeast strains were grown with 50 or 100 g/L of initial glucose concentration. Error bars represent SDs of 3 replicates. Statistical significant changes (P < 0.01) are highlighted by double asterisks. (F) The C-terminal differences of SpMae1(p) and its fungal homologs.

Fig. 4.

Effect of expression of native and F/A mutated transporter variants in yeast on the production of malate and on growth. (A and B) Malate titers in a mineral medium with calcium carbonate for buffering. (C and D) Malic acid titers in a mineral medium without pH buffering. (E) Normalized fluorescence of yeast cells expressing native and F/A mutated transporter variants fused with GFP at the C terminus. (A–E) Error bars represent SDs of 3 biological replicates. Statistical significant changes (P < 0.01) are highlighted by double asterisks.