Research ArticleDrug Discovery

High-throughput metabolomic analysis predicts mode of action of uncharacterized antimicrobial compounds

See allHide authors and affiliations

Science Translational Medicine  21 Feb 2018:
Vol. 10, Issue 429, eaal3973
DOI: 10.1126/scitranslmed.aal3973
  • Fig. 1 Antibiotic-induced metabolome responses in M. smegmatis.

    (A) Metabolomics workflow. Cells were grown in 700-μl volumes in 96-well plates to an OD595 of about 0.4, before addition of 10 μl of the antimicrobial compound. Cell culture (80 μl) was withdrawn from each well at each sampling time. Forty microliters was used to determine cell density, and the remaining 40 μl was added to cold extraction buffer. Supernatant was directly injected into a time-of-flight mass spectrometer, and relative changes in metabolite intensities were extrapolated from processing of the metabolome data. (B) Compounds tested. Almost half of the compounds tested included different concentrations of reference antimicrobials (yellow) and chemical stress agents (green) with known MoAs; the remainder were compounds from a GSK library used at 10 μM concentration (blue). (C) Distribution of MoAs for the 62 reference compounds. ATP, adenosine 5′-triphosphate. (D) Schematic representation of the drug-metabolome response data set. For each antimicrobial compound tested, the dynamic profile of 1006 metabolites was interrogated. As an example, the top graph illustrates the response of the folic acid biosynthesis intermediate 4-aminobenzoic acid to the antimicrobial para-aminosalicylic acid (PAS) (red, 104 μM; gray, 41 μM; blue, 25 μM). The bottom graph shows the response of the bacterial metabolite mycobactin to the known antimicrobial isoniazid (red, 1.5 mM; gray, 0.22 mM; blue, 0.11 mM). Thick lines represent the results from the impulse model fitting analysis for the three drug concentrations. Metabolic profiles of 4-aminobenzoic acid and mycobactin across all conditions are shown in light gray. (E) Distribution of metabolic response onset times for antibiotics belonging to the seven main antibiotic categories tested in this study. The onset time is defined as the time at which metabolite changes reached half of their maximum change after treatment of M. smegmatis with the compounds. For each perturbation (treatment with compound), metabolites with a model fitting R2 ≥ 0.6 and a maximum absolute log2 fold change ≥ 2 were retained.

  • Fig. 2 Commonalities among metabolite changes in response to antimicrobial treatment.

    (A) Correlation between growth rate and metabolite abundance in M. smegmatis after treatment with antimicrobial compounds. Each dot represents a metabolite. The two axes represent the mean R2 across all tested conditions and the mean of maximum Z scores across all tested conditions and time points. Color reflects the degree of Spearman correlation between maximum Z score and growth inhibition across all tested conditions. Metabolites with an average R2 ≥ 0.5 and log2 Z score ≥ 0.5 are shown. UDP, uridine 5′-diphosphate; GDP, guanosine diphosphate; CDP, cytidine 5′-diphosphate. (B) Pathway enrichment for metabolome responses to antibiotics with seven known MoAs: (1) cell wall synthesis inhibitors, (2) DNA cleavage, (3) folic acid biosynthesis inhibitors, (4) quinolones, (5) mycolic acid biosynthesis inhibitors, (6) protein synthesis inhibitors, and (7) RNA synthesis inhibitors. Enrichment was performed with the 50 most frequently identified genes for each antibiotic class. The heat map shows enriched KEGG metabolic pathways with q ≤ 0.01.

  • Fig. 3 Pairwise similarity of antimicrobials with respect to metabolic changes induced in M. smegmatis.

    (A) Similarity heat map for 62 reference antimicrobials. Similarity calculated between each drug-perturbed condition is represented as a symmetric heat map. Diagonal values are not taken into account and are set to not-a-number (gray). Highlighted boxes correspond to the seven main MoAs of the 62 reference compounds. (B) Magnification of panel in (A) showing antimicrobial compounds that blocked gyrase and cell wall synthesis in M. smegmatis. (C) Receiver operating characteristic (ROC) curve measuring the ability of metabolome-based predictions using the iterative hypergeometric test (45) to discriminate antibiotics sharing similar MoAs. Notably, we considered only MoAs that applied to more than one antimicrobial reference compound. CPR, ciprofloxacin; LVX, levofloxacin; MFL, moxifloxacin; NAL, nalidixic acid; NFL, norfloxacin; OFL, ofloxacin; AMX, amoxicillin; AMP, ampicillin; CCL, cefaclor; CTX, ceftriaxone; OCI, oxacillin; BAC, bacitracin; CYC, d-cycloserine; EMB, ethambutol; FOS, fosfomycin; AUC, area under the curve.

  • Fig. 4 Metabolome-based predictions of MoAs for 212 GSK compounds.

    (A) Grouping of metabolome similarity–based predictions for the 212 GSK compounds into known MoAs. (B) Impact of antimicrobial drugs on normalized FolA in vitro activity. The measured dihydrofolate conversion rate was normalized to the activity measured with DMSO vehicle only. All tested compounds were dissolved in DMSO: 40 μM streptomycin (STR), 40 and 1333 μM trimethoprim (TRM, TRM-H), 40 μM PAS, and 40 μM of six GSK compounds. (C) RecA promoter activity in exponentially growing E. coli treated with the following: four GSK compounds predicted to be quinolone-like agents (BRL-7940SA, BRL-10988SA, GSK1066288A, and GSK695914A), norfloxacin, ampicillin, DMSO, and three GSK compounds (GSK2534991A, GSK1826825A, and GSK1518999A) predicted to be a protein synthesis inhibitor, a folic acid biosynthesis inhibitor, or with an unknown MoA, respectively. (D) Gyrase activity of M. tuberculosis measured using an in vitro supercoiling assay at different concentrations of moxifloxacin or GSK1066288A. (E) Gyrase activity of E. coli in the presence of DMSO, moxifloxacin, streptomycin, or GSK1066288A. Activity of denatured gyrase was used as a negative control (ø).

  • Fig. 5 Analysis of metabolite changes in M. smegmatis after treatment with the GSK compound GSK2623870A.

    (A) Pairwise similarity between M. smegmatis metabolome response profiles to 16 GSK compounds with no similarity to known MoAs. (B) Schematic representation of trehalose monomycolate exporter protein MmpL3. Transmembrane segments are represented in violet. Circles indicate the locations of amino acid changes associated with resistance of M. tuberculosis to antimycobacterial lead compounds previously found to select for resistance mutations in MmpL3. These compounds include SQ109 (blue), THPP (orange), and SPIROS (purple). The black star indicates the amino acid change associated with resistance to the GSK compound GSK2623870A. The genomes of the M. tuberculosis H37Rv and Beijing GC1237 mutant strains contain an A to G SNP at position 755, which resulted in a tyrosine to cysteine missense mutation at position 252 of the MmpL3 protein. (C) Results from limited proteolysis analysis. Each dot in the volcano plot represents the relative difference in peptide abundance between the treated and untreated proteome extracts. Proteins highlighted in red are known to physically interact with fatty acid synthase FAS-I (figs. S17 and S18). For each protein, the size of the dot reflects the number of interacting partners (80) with significant conformational changes. (D) Rapid metabolic changes induced by the GSK antimicrobial compound GSK2623870A. Each dot corresponds to the R2 and Z-score values of the metabolite 5 min after exposure of M. smegmatis to the antimicrobial compound. Metabolites highlighted in red are involved in fatty acid metabolism.

Supplementary Materials

  • www.sciencetranslationalmedicine.org/cgi/content/full/10/429/eaal3973/DC1

    Materials and Methods

    Fig. S1. Schematic description of MS data normalization and analysis.

    Fig. S2. Analysis of correlations across biological replicates.

    Fig. S3. Distribution of responsive metabolites.

    Fig. S4. Number of affected metabolites per MoA.

    Fig. S5. Distribution of growth inhibitory activities.

    Fig. S6. Pathway enrichment for metabolome responses to antibiotics with seven major MoA.

    Fig. S7. Pairwise drug similarity.

    Fig. S8. Metabolome-based similarity.

    Fig. S9. Similarity between compounds with equivalent MoAs as a function of the difference in growth inhibition.

    Fig. S10. Distribution of predicted MoAs.

    Fig. S11. Pairwise compound chemical distance.

    Fig. S12. M. tuberculosis gyrase assay.

    Fig. S13. M. tuberculosis and E. coli gyrase assay.

    Fig. S14. Distribution of growth inhibitory activities for GSK compounds with classified and unclassified MoAs.

    Fig. S15. Pathway enrichment analysis for compounds with potential unconventional MoAs.

    Fig. S16. Common metabolic responses across GSK compounds with potential new MoAs.

    Fig. S17. Protein-protein interactions among proteins with significant conformational changes detected by limited proteolysis analysis.

    Fig. S18. Robustness of results from limited proteolysis.

    Fig. S19. Similarity between compounds with equivalent MoAs as a function of growth inhibition.

    Fig. S20. Data normalization.

    Fig. S21. Schematic representation of the procedure used to estimate pairwise similarity among tested compounds.

    Table S1. Antibiotic perturbation list.

    Table S2. Metabolome data (perturbation name versus metabolites data matrix).

    Table S3. Impulse model fitting results (maximum fold change and R2 matrices).

    Table S4. MoA predictions (list of top predicted MoAs and complete matrix of enrichment q values).

    Table S5. Gene-drug assignments with P values coming from the network locality analysis.

    Table S6. List of metabolites used to annotate peaks in the mass spectra.

    Table S7. Comparison of similarity metrics.

    Table S8. Results from analysis of limited proteolysis data.

    References (8186)

  • Supplementary Material for:

    High-throughput metabolomic analysis predicts mode of action of uncharacterized antimicrobial compounds

    Mattia Zampieri,* Balazs Szappanos, Maria Virginia Buchieri, Andrej Trauner, Ilaria Piazza, Paola Picotti, Sébastien Gagneux, Sonia Borrell, Brigitte Gicquel, Joel Lelievre, Balazs Papp, Uwe Sauer

    *Corresponding author. Email: zampieri{at}imsb.biol.ethz.ch

    Published 21 February 2018, Sci. Transl. Med. 10, eaal3973 (2018)
    DOI: 10.1126/scitranslmed.aal3973

    This PDF file includes:

    • Materials and Methods
    • Fig. S1. Schematic description of MS data normalization and analysis.
    • Fig. S2. Analysis of correlations across biological replicates.
    • Fig. S3. Distribution of responsive metabolites.
    • Fig. S4. Number of affected metabolites per MoA.
    • Fig. S5. Distribution of growth inhibitory activities.
    • Fig. S6. Pathway enrichment for metabolome responses to antibiotics with seven major MoA.
    • Fig. S7. Pairwise drug similarity.
    • Fig. S8. Metabolome-based similarity.
    • Fig. S9. Similarity between compounds with equivalent MoAs as a function of the difference in growth inhibition.
    • Fig. S10. Distribution of predicted MoAs.
    • Fig. S11. Pairwise compound chemical distance.
    • Fig. S12. M. tuberculosis gyrase assay.
    • Fig. S13. M. tuberculosis and E. coli gyrase assay.
    • Fig. S14. Distribution of growth inhibitory activities for GSK compounds with classified and unclassified MoAs.
    • Fig. S15. Pathway enrichment analysis for compounds with potential unconventional MoAs.
    • Fig. S16. Common metabolic responses across GSK compounds with potential new MoAs.
    • Fig. S17. Protein-protein interactions among proteins with significant conformational changes detected by limited proteolysis analysis.
    • Fig. S18. Robustness of results from limited proteolysis.
    • Fig. S19. Similarity between compounds with equivalent MoAs as a function of growth inhibition.
    • Fig. S20. Data normalization.
    • Fig. S21. Schematic representation of the procedure used to estimate pairwise similarity among tested compounds.
    • References (8186)

    [Download PDF]

    Other Supplementary Material for this manuscript includes the following:

    • Table S1 (Microsoft Excel format). Antibiotic perturbation list.
    • Table S2 (Microsoft Excel format). Metabolome data (perturbation name versus metabolites data matrix).
    • Table S3 (Microsoft Excel format). Impulse model fitting results (maximum fold change and R2 matrices).
    • Table S4 (Microsoft Excel format). MoA predictions (list of top predicted MoAs and complete matrix of enrichment q values).
    • Table S5 (Microsoft Excel format). Gene-drug assignments with P values coming from the network locality analysis.
    • Table S6 (Microsoft Excel format). List of metabolites used to annotate peaks in the mass spectra.
    • Table S7 (Microsoft Excel format). Comparison of similarity metrics.
    • Table S8 (Microsoft Excel format). Results from analysis of limited proteolysis data.

Stay Connected to Science Translational Medicine

Navigate This Article