The Antibiotic Negamycin Crosses the Bacterial Cytoplasmic Membrane by Multiple Routes

Abstract

Negamycin is a natural pseudodipeptide antibiotic with promising activity against Gram-negative and Gram-positive bacteria, including Enterobacteriaceae, Pseudomonas aeruginosa, and Staphylococcus aureus, and good efficacy in infection models. It binds to ribosomes with a novel binding mode, stimulating miscoding and inhibiting ribosome translocation. We were particularly interested in studying how the small, positively charged natural product reaches its cytoplasmic target in Escherichia coli Negamycin crosses the cytoplasmic membrane by multiple routes depending on environmental conditions. In a peptide-free medium, negamycin uses endogenous peptide transporters for active translocation, preferentially the dipeptide permease Dpp. However, in the absence of functional Dpp or in the presence of outcompeting nutrient peptides, negamycin can still enter the cytoplasm. We observed a contribution of the DppA homologs SapA and OppA, as well as of the proton-dependent oligopeptide transporter DtpD. Calcium strongly improves the activity of negamycin against both Gram-negative and Gram-positive bacteria, especially at concentrations around 2.5 mM, reflecting human blood levels. Calcium forms a complex with negamycin and facilitates its interaction with negatively charged phospholipids in bacterial membranes. Moreover, decreased activity at acidic pH and under anaerobic conditions points to a role of the membrane potential in negamycin uptake. Accordingly, improved activity at alkaline pH could be linked to increased uptake of [³H]negamycin. The diversity of options for membrane translocation is reflected by low resistance rates. The example of negamycin demonstrates that membrane passage of antibiotics can be multifaceted and that for cytoplasmic anti-Gram-negative drugs, understanding of permeation and target interaction are equally important.

Keywords: Dpp; DtpD; Escherichia coli; antibiotic; calcium; membrane; natural product; negamycin; peptide transporters; uptake.

FIG 1

Negamycin structure and mode of action. (A) Structure of negamycin with pKa values (8). (B) Effect of negamycin on coupled in vitro transcription-translation using an E. coli S30 extract and plasmid-based Photinus pyralis luciferase as reporter. Error bars showing standard deviation (SD) of five independent experiments. (C) Effect of negamycin, streptomycin (positive control), or tetracycline (negative control) in a whole-cell miscoding assay demonstrating the readthrough of a stop codon within the luciferase gene. Error bars indicating SD of two independent experiments. RLU, relative luminescence units.

FIG 2

Negamycin uses multiple peptide transporters and competes with peptides for uptake, whereas entry options for bialaphos are limited to Opp and Dpp transporters. (A and B) Impact of single or multiple ABC peptide transporter deletions in the E. coli BW25113 background on negamycin (A) or bialaphos (B) susceptibility in M9 medium. (C) Impact of peptide addition on negamycin MICs in M9. Negamycin susceptibility of E. coli BW25113 in M9 decreases with supplemented polypeptone (PP) in a concentration-dependent manner. (D) Uptake of [³H]negamycin into E. coli BW25113 wild type or its isogenic ΔdppA mutant in M9. [³H]negamycin (specific activity: 0.052 Ci/mmol) was added to the cells at a concentration of 32 μg/ml. One sample of the wild type was supplemented with 0.05% PP concurrently with negamycin addition. Samples were taken after 15 (left) or 60 (right) min of incubation at 37°C with shaking. Negamycin uptake is significantly reduced in M9 with the addition of 0.05% PP or the deletion of the peptide transporter gene dppA. Each diamond represents an independent MIC determination (A to C) or [³H]negamycin uptake measurement (D). Statistical significance was determined using unpaired Student’s t test with Holm-Bonferroni correction. ns, P > 0.05; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001.

FIG 3

Schematic overview of the dpp operon in E. coli and spontaneous mutations induced by negamycin exposure. Indicated deletions, IS element insertions, and a point mutation were detected in spontaneous negamycin resistant mutants isolated from M9 agar plates containing 2× MIC negamycin. Locus numbers based on E. coli K-12 MG1655 (GenBank access no. U00096.3). SNP, single nucleotide polymorphism.

FIG 4

Basic pH and calcium are additive in improving negamycin activity. Negamycin MICs of P. aeruginosa PAO1 and S. aureus ATCC 29213 (A) and of E. coli BW25113 (B) in 0.5% PP medium adjusted to different pH values and in combination with CaCl₂. (C) Influence of pH and CaCl₂ on [³H]negamycin uptake in E. coli BW25113 in PP. A total amount of 32 μg/ml of [³H]negamycin (specific activity: 0.052 Ci/mmol) was added to the cells. CaCl₂ and [³H]negamycin were added to the cultures in parallel. Samples were taken after 15 (left) or 60 (right) min of incubation with [³H]negamycin at 37°C with shaking. Alkaline pH increased [³H]negamycin uptake, while we could not detect such an effect after the addition of CaCl₂. Each diamond represents an independent MIC determination (A and B) or [³H]negamycin uptake measurement (C). Statistical significance was determined using unpaired Student’s t test with Holm-Bonferroni correction. ns, P > 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001.

FIG 5

Calcium supports the binding of negamycin to phospholipid membranes. Surface acoustic wave (SAW) sensorgrams comparing the interaction of negamycin with a POPC membrane (A), a POPC/DOPG membrane (B), or a POPC/DOPG membrane in the presence of 2.5 mM CaCl₂ (C) or with a POPC/DOPG membrane in the presence of 2.5 mM MgCl₂ (D). The empty triangles at the x axis indicate the injection steps of negamycin. The total concentration of negamycin ranges from 7.5 × 10⁻⁸ M (outermost left triangle) to 1 × 10⁻⁵ M (outermost right triangle). Filled arrows at the curves indicate binding events. The large phase shifts at high negamycin concentrations which drop back after injection are a consequence of increased viscosity and do not indicate real binding events.

FIG 6

Negamycin forms a complex with calcium. (A) Thin-layer chromatography (TLC) analysis of negamycin with CaCl₂, MgCl₂, or NaCl at different molar ratios. (B and C) Isothermal titration calorimetry (ITC) of negamycin titrated with CaCl₂. (B) Heat differences monitored by differential power (DP) measurements upon two consecutive series of 19 injections of CaCl₂ after baseline correction and subtraction of control experiments of CaCl₂ titration into buffer, buffer titration into buffer, and negamycin titration into buffer. (C) Binding enthalpies (ΔH) against the Ca²⁺/negamycin molar ratio. ITC data were fitted to the one-site binding model. Due to the low binding affinity, fitting required to preset stoichiometry manually, which we set to 1:1. With these settings, a K_d of 7.98 mM was obtained.

FIG 7

Mutations affecting the respiratory chain reduce negamycin sensitivity in peptide-rich media. Negamycin MICs of different energy mutants of E. coli BW25113 in M9 (A), 0.5% PP (B), or MHBII (C). Each diamond represents an independent MIC determination. Statistical significance was determined using unpaired Student’s t test with Holm-Bonferroni correction, comparing the various deletion strains to the wild type (WT). ns, P > 0.05; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001; n. g., no growth under this condition.

FIG 8

Anaerobic growth conditions reduce negamycin susceptibility in peptide-containing media and in peptide transporter mutants. (A) Negamycin MICs of E. coli BW25113 wild type in different media under aerobic and anaerobic conditions. Statistical significance was determined using Student’s t test comparing anaerobic to aerobic growth conditions. (B) Effect of peptide transporter deletions in the E. coli BW25113 background on negamycin susceptibility under aerobic versus anaerobic growth conditions in M9 medium. Each diamond represents an independent MIC determination. Statistical significance was determined using unpaired Student's t test with Holm-Bonferroni correction. ns, P > 0.05; *, P ≤ 0.05; ***, P ≤ 0.001; ****, P ≤ 0.0001.