Sorting Out Antibiotics' Mechanisms of Action: a Double Fluorescent Protein Reporter for High-Throughput Screening of Ribosome and DNA Biosynthesis Inhibitors

Abstract

In order to accelerate drug discovery, a simple, reliable, and cost-effective system for high-throughput identification of a potential antibiotic mechanism of action is required. To facilitate such screening of new antibiotics, we created a double-reporter system for not only antimicrobial activity detection but also simultaneous sorting of potential antimicrobials into those that cause ribosome stalling and those that induce the SOS response due to DNA damage. In this reporter system, the red fluorescent protein gene rfp was placed under the control of the SOS-inducible sulA promoter. The gene of the far-red fluorescent protein, katushka2S, was inserted downstream of the tryptophan attenuator in which two tryptophan codons were replaced by alanine codons, with simultaneous replacement of the complementary part of the attenuator to preserve the ability to form secondary structures that influence transcription termination. This genetically modified attenuator makes possible Katushka2S expression only upon exposure to ribosome-stalling compounds. The application of red and far-red fluorescent proteins provides a high signal-to-background ratio without any need of enzymatic substrates for detection of the reporter activity. This reporter was shown to be efficient in high-throughput screening of both synthetic and natural chemicals.

FIG 1

(A) Map of the pDualrep2 reporter plasmid. SulA, promoter of the sulA gene; TrpL2A, trpL leader coding sequence carrying W10A and W11A substitutions. Genes coding for RFP and Katushka2S are shown. (B) Agar plate covered with the E. coli BW25113 strain transformed with the pDualrep2 plasmid and spotted with levofloxacin (left) and erythromycin (right). The plate was scanned in the 605-nm channel to detect RFP fluorescence. (C) The same plate was scanned in the 647-nm channel to detect Katushka2S fluorescence. (D) Overlay of images from panels B and C as a pseudocolor image. RFP is shown in a green bit plane; Katushka2S is shown in a red bit plane.

FIG 2

(A) Induction of Katushka2S fluorescence by translation-stalling antibiotics on petri plates. The lawn of E. coli BW25113 transformed with pDualrep2 was grown after the addition of antibiotic disks (indicated). The antibiotics were erythromycin (ERY), roxithromycin (ROX), azithromycin (AZI), chloramphenicol (CM), kanamycin (KAN), tetracycline (TET), streptomycin (STR), lincomycin (LIN), clindamycin (CLI), levofloxacin (Lev), tobramycin (Tob), neomycin (NEO), spectinomycin (SPM), etamycin A (ETA A), hygromycin B (HYG B), griseoviridin (GRS), tylosin (TYL), amicoumacin A (AMI), fusidic acid (FUS), puromycin (PUR), and gentamicin (GEN). (B) Induction of RFP fluorescence by DNA biosynthesis inhibitors on petri plates. The lawn of E. coli BW25113 transformed with pDualrep2 was grown after the addition of antibiotic disks (indicated). The antibiotics were erythromycin (ERY), sulfanilamide (SUL), polymyxin (POL), rifampin (RIF), nalidixic acid (NAL), levofloxacin (LEV), ciprofloxacin (CIP), etoposide (ETO), furagin (FUR), and microcin B17 (MB17).

FIG 3

Determination of minimal concentrations of the antibiotics that could be detected with pDualrep2. (A) Analysis on the petri plates. Indicated antibiotic solutions (2 μl) were spotted onto the lawn of the E. coli BW25113 strain transformed with pDualrep2. The plate was scanned in RFP and Katushka2S channels, shown as green and red pseudocolors, respectively. (B) Analysis in liquid medium. The part of the 96-well plate containing liquid culture of E. coli BW25113 transformed with pDualrep2 is presented as scanned in the RFP and Katushka2S channels, shown as green and red pseudocolors, respectively. Indicated antibiotics were added to the wells in concentrations decreasing from bottom to top as follows: 60, 30, 15, 7.5, 3.75, and 1.85 ng/ml of levofloxacin (LEV); 200, 100, 50, 25, 12.5, and 6.25 μg/ml of erythromycin (ERY); 5, 2.5, 1.25, 0.612, 0.3, and 0.15 μg/ml of chloramphenicol (CM); 1.2, 0.6, 0.3, 0.15, 0.075, and 0.0035 μg/ml of tetracycline (TET); and 5, 2.5, 1.25, 0.625, 0.33, and 0.15 μg/ml of kanamycin (KAN).

FIG 4

HTS based on the pDualrep2 reporter. (A) An agar plate with the lawn of the E. coli ΔtolC strain transformed with the pDualrep2 plasmid and 2-μl spots of compounds tested at 0.17 mg/ml concentrations. Erythromycin (1 mg/ml) and levofloxacin (1.5 μg/ml) 2-μl spots were used as controls. The locations of F2 and G6 spots demonstrating the induction of Katushka2S and RFP expression, respectively, are marked. (B) The same experiment as that shown in panel A but without the pDualrep2 plasmid. (C) The structure of the compound located in the F2 spot. (D) The structure of the compound located in the G6 spot. (E) Inhibition of in vitro synthesis of catalytically active luciferase by the compound located in the F2 spot (circles) and G6 spot (squares). Approximation curves are marked correspondingly. Relative luciferase activity of synthesized luciferase is shown as the measure of protein synthesis efficiency.