Genomic adaptation of Burkholderia anthina to glyphosate uncovers a novel herbicide resistance mechanism

Abstract

Glyphosate (GS) specifically inhibits the 5-enolpyruvyl-shikimate-3-phosphate (EPSP) synthase that converts phosphoenolpyruvate (PEP) and shikimate-3-phosphate to EPSP in the shikimate pathway of bacteria and other organisms. The inhibition of the EPSP synthase depletes the cell of the EPSP-derived aromatic amino acids as well as of folate and quinones. A variety of mechanisms (e.g., EPSP synthase modification) has been described that confer GS resistance to bacteria. Here, we show that the Burkholderia anthina strain DSM 16086 quickly evolves GS resistance by the acquisition of mutations in the ppsR gene. ppsR codes for the pyruvate/ortho-P_i dikinase PpsR that physically interacts and regulates the activity of the PEP synthetase PpsA. The mutational inactivation of ppsR causes an increase in the cellular PEP concentration, thereby abolishing the inhibition of the EPSP synthase by GS that competes with PEP for binding to the enzyme. Since the overexpression of the Escherichia coli ppsA gene in Bacillus subtilis and E. coli did not increase GS resistance in these organisms, the mutational inactivation of the ppsR gene resulting in PpsA overactivity is a GS resistance mechanism that is probably unique to B. anthina.

FIGURE 1

Effect of GS on growth of Burkholderia anthina, emergence of GS‐resistant mutants and their characterization. (A) GS inhibits the enolpyruvylshikimate 3‐phosphate (EPSP) synthase AroA. Dashes files indicate that multiple reactions are involved in the biosynthesis of aromatic amino acids, quinones and folates. (B) Emergence of GS‐resistant B. anthina mutants. B. anthina was grown in C‐Glc medium at 37°C until an OD₆₀₀ of about 2.0. 100 μL containing 10⁷ colony forming units (CFU) were propagated on C‐Glc minimal medium plates that were incubated for 24 h at 37°C. (C) Time‐dependent emergence of GS‐resistant B. anthina mutants. B. anthina was grown in C‐Glc medium at 37°C until an OD₆₀₀ of about 2.0. 10⁷ CFU were propagated on C‐Glc plates supplemented with 10 mM GS. The plates were incubated for up to 3 days at 37°C and the emerging mutants were counted. Dots indicate biologically independent replicates and bars indicate mean values. (D) Evaluation of growth of the parental strain and three isolated GS‐resistant mutants on C‐Glc plates in the absence and in the presence of the herbicide. The plates were incubated for 24 h at 37°C. (E) Growth of the parental strain and three isolated GS‐resistant mutants in C‐Glc medium supplemented with increasing amounts of GS in a microplate reader at 37°C.

FIGURE 2

Effect of ppsR mutations identified in the GS‐resistant mutants on the cellular PEP level. (A) Schematic illustration of the genomic organization of the ppsR and ppsA genes and the effect of the PEP synthetase regulatory protein PpsR on the activity of the PEP synthetase PpsA. (B) Venn diagram illustrating the overlap of the genes mutated in Burkholderia anthina during growth under selection with GS (see also Table 1). (C) Localization of the amino substitutions that likely affect the function of the PEP synthetase regulatory protein PpsR. The structure model was generated using the SWISS‐MODEL server for homology modelling of protein structures (Waterhouse et al., 2018) and a model of the dimer structure of the maize pyruvate orthophosphate dikinase regulatory protein (PDBid: 5D0N) (Jiang et al., 2016). (D) Sequence alignment of PpsR homologues from B. anthina (UniProt code A0A103TDR4), Trinickia fusca (UniProt code A0A494XQF5), Caballeronia mineralivorans (UniProt code A0A0J1CLS7), Pseudomonas nitroreducens (UniProt code A0A246FAH5), Neisseria menigitidis (UniProt code Q9K0I1), Buttiauxella agrestisz (UniProt code A0A08GFB9), Klebsiella pneumoniae (UniProt code B5XQE5), Enterobacter roggenkampii (UniProt code A0A167SP00), Escherichia coli (UniProt code B1XG10), Salmonella typhimurium (UniProt code P67197), Shigella flexneri (UniProt code Q0T4R2), Yokenella regensburgei (UniProt code A0A6H0K776), Acinetobacter baumannii (UniProt code B0V7F2), Vibrio vulnificus (UniProt code Q7MF05), and Zea mays (UniProt code A0A1D6IDV8). The amino substitutions that likely affect function of the PEP synthetase regulatory protein PpsR are indicated by red triangles.

FIGURE 3

Interaction analysis between PpsR and PpsA homologues from Burkholderia anthina and Escherichia coli. (A) B2H assay to assess the interaction between PpsR and PpsA from B. anthina. (B) B2H assay to assess the interaction between PpsR and PpsA from E. coli. The agar plates were incubated for 36 h at 30°C. (C) B2H assay to assess the interaction between PpsR Q173P and PpsA from E. coli. The ppsR and ppsA alleles were introduced into the plasmids pUT18, pUT18C, pKNT25 and pKT25. Plasmids pUT18 and pUT18C allow the expression of the proteins fused to the N‐ and C‐terminus of the T18 domain of the Bordetella pertussis adenylate cyclase, respectively. Plasmids pKNT25 and pKT25 allow the expression of the proteins fused to the N‐ and C‐terminus of the T25 domain of the adenylate cyclase. The E. coli transformants were spotted onto LB plates supplemented with X‐Gal, IPTG, ampicillin and kanamycin. The agar plate shown in (A) was incubated for 48 h at 30°C, followed by 40 days at 4°C. The agar plates shown in (B) and (C) were incubated for 36 h at 30°C.

FIGURE 4

Relative quantification of the cellular PEP concentrations in Burkholderia anthina and Escherichia coli. (A) Relative quantification of the cellular PEP concentrations in the GS‐resistant B. anthina mutants. The asterisk indicates that the PEP concentration was normalized to the parental strain S2.1. (B) Relative quantification of the cellular PEP concentrations E. coli strain W3110 expressing the native PpsA enzyme from plasmid pBP1224 (+PpsA). The PEP concentration was normalized to the E. coli strain W3110 carrying the empty plasmid pGP380. PEP concentrations in the suppressor mutants are shown as log₂‐fold changes compared to the wild type concentration. Mean value and standard deviation of three replicates are shown.

FIGURE 5

Effect of ppsA overexpression on GS resistance in Escherichia coli. Growth of the E. coli strain W3110 carrying the plasmids pGP380 (−PpsA) and pBP1224 (+PpsA) C‐Glc minimal medium supplemented with the indicated amounts of GS and glutamate in a microplate reader at 37°C.

FIGURE 6

Characterization of the Burkholderia anthina mutant with increased GS resistance. (A) Evaluation of growth of the parental strain and isolated GS‐resistant mutant S2.1 on C‐Glc plates in the absence and in the presence of 35 mM GS. The plates were incubated for 24 h at 37°C. (B) Evaluation of growth of the parental strain and isolated GS‐resistant mutant S2.1 in C‐Glc liquid supplemented in the absence and in the presence of 35 mM GS. (C) The pyruvate kinase catalyses the conversion of ADP/Pi and PEP to ATP and pyruvate in the glycolytic pathway. The PEP synthetase converts pyruvate and ATP to AMP/Pi and PEP that is required for gluconeogenesis. (D) Growth of the B. anthina mutant S2.1 in C medium supplemented with 0.5% (w/v) glucose, 0.6% (w/v) succinate and 0.8% (w/v) glutamate in a microplate reader at 37°C.

FIGURE 7

Mechanisms conferring GS resistance and enzymes that catalyse pyruvate and PEP synthesis. Mechanisms conferring GS resistance in bacteria and fungi. AroA, EPSP synthase; GltT and GltP, Bacillus subtilis glutamate transporters; YhhS and MFS40, major facilitator secondary transporters from Escherichia coli and Aspergillus oryzae, respectively; Pdr5 and Dip5 are amino acid and ABC efflux transporters, respectively, from Saccharomyces cerevisiae; Gat, Bacillus licheniformis GS N‐acetyltransferase; GlpA, Burkholderia pseudomallei hygromycin phosphotransferase; PpsA, PEP synthetase; PpsR, PEP synthetase regulatory protein; PEP, phosphoenolpyruvate; S3P, shikimate‐3‐phosphate; EPSP, enolpyruvylshikimate‐3‐phosphate.