Development of a novel heterologous β-lactam-specific whole-cell biosensor in Bacillus subtilis

Background Whole-cell biosensors are a powerful and easy-to-use screening tool for the fast and sensitive detection of chemical compounds, such as antibiotics. β-Lactams still represent one of the most important antibiotic groups in therapeutic use. They interfere with late stages of the bacterial cell wall biosynthesis and result in irreversible perturbations of cell division and growth, ultimately leading to cell lysis. In order to simplify the detection of these antibiotics from solutions, solid media or directly from producing organisms, we aimed at developing a novel heterologous whole-cell biosensor in Bacillus subtilis, based on the β-lactam-induced regulatory system BlaR1/BlaI from Staphylococcus aureus. Results The BlaR1/BlaI system was heterologously expressed in B. subtilis and combined with the luxABCDE operon of Photorhabdus luminescens under control of the BlaR1/BlaI target promoter to measure the output of the biosensor. A combination of codon adaptation, constitutive expression of blaR1 and blaI and the allelic replacement of penP increased the inducer spectrum and dynamic range of the biosensor. β-Lactams from all four classes induced the target promoter PblaZ in a concentration-dependent manner, with a dynamic range of 7- to 53-fold. We applied our biosensor to a set of Streptomycetes soil isolates and demonstrated its potential to screen for the production of β-lactams. In addition to the successful implementation of a highly sensitive β-lactam biosensor, our results also provide the first experimental evidence to support previous suggestions that PenP functions as a β-lactamase in B. subtilis. Conclusion We have successfully established a novel heterologous whole-cell biosensor in B. subtilis that is highly sensitive for a broad spectrum of β-lactams from all four chemical classes. Therefore, it increases the detectable spectrum of compounds with respect to previous biosensor designs. Our biosensor can readily be applied for identifying β-lactams in liquid or on solid media, as well as for identifying potential β-lactam producers.


Figure S2a
: Disk diffusion assay of the biosensors and controls with additional b-lactams, bacitracin and water. The four b-lactam antibiotics shown here are cefotaxime (CTX, 200 µg/ml), cephalosporin C (CFC C, 500 µg/ml), aztreonam (ATM, 2000 µg/ml) and meropenem (MEM, 10 µg/ml). White light pictures indicate the positions of the disks on the plate. The corresponding images from luminescence detection are displayed underneath. Top row from left to right: (1) W168 wild type (Control 1), (2) a strain that constitutively expresses the lux operon (Control 2), (3) a strain carrying the lux operon without any promoter (TMB2841, Control 3) and (4) detection of b-lactams by Biosensor 1 (TMB3641). Bottom row from left to right: (1) detection of b-lactams by the Biosensor 1 in D penP (TMB3713), (2) a control strain carrying all biosensor parts but the regulator blaI (Control 4), (3) detection of b-lactams by Biosensor 2 (TMB5608) as well as (4) another control strain carrying all biosensor parts despite the receptor blaR1 (Control 5) (see Table 1). Representative images of triplicates are shown.   Here, the inhibitory effect of the four β-lactams cefalexin, cefoxitin, cephalosporin C and cefotaxime are shown as well as for bacitracin, a non-β-lactam. The MICs for B. subtilis W168 wild type and strains missing either ybxI (TMB3668) or penP (TMB3667) as well as both genes (TMB3675) coding for potential β-lactamases are shown in comparison. Note that the concentrations tested differ among the antibiotic compounds as demonstrated by the changing the x-axis. Experiments have been performed in triplicates.

Figure S5 (A) and (B): Growth (OD600nm) and luminescence signal (RLU/OD600nm) of Biosensor 1 in the presence of different b-lactams.
The upper row displays the growth measured as OD600nm, while the row immediately below shows the luminescence intensity in relative luminescence units normalized over OD600nm (y-axis). Both the y-axis for the OD600nm measurements (growth) and the y-axis presenting RLU/OD600nm are valid for all the compounds presented in the respective row. The x-axis displays the time in hours. Induction with the antibiotics occurred after 1 hour of growth as indicated by the black vertical dotted line. The assayed strains and their corresponding colors are indicated in the legend on the upper right. Continuous lines and round symbols always represent growth, while dotted lines with triangular symbols represent the luminescence signal over time. (A) Growth and luminescence of Control 3 (see Table 1, TMB2841), Biosensor 1 and Biosensor 1 in ΔpenP after induction with β-lactams belonging to the group of penicillins. (B) Growth and luminescence of Control 3, Biosensor 1 and Biosensor 1 in ΔpenP after induction with β-lactams belonging to the group of cephalosporins. Experiments have been performed in triplicates.

Figure S5 (C)-(E): Growth (OD600nm) and luminescence signal (RLU/OD600nm) of Biosensor 1 in the presence of different b-lactams.
The upper row displays the growth measured as OD600nm, while the row immediately below shows the luminescence intensity in relative luminescence units normalized over OD600nm (y-axis). Both the y-axis for the OD600nm measurements (growth) and the y-axis presenting RLU/OD600nm are valid for all the compounds presented in the respective row. The x-axis displays the time in hours. Induction with the antibiotics occurred after 1 hour of growth as indicated by the black vertical dotted line. The assayed strains and their corresponding colors are indicated in the legend on the upper right. Continuous lines and round symbols always represent growth, while dotted lines with triangular symbols represent the luminescence signal over time. (C) Growth and luminescence of Control 3 (see Table 1 Table 1), Biosensor 2 and Biosensor 2 in ΔpenP after induction with β-lactams belonging to the group of penicillins. (B) Growth and luminescence of Control 5, Biosensor 2 and Biosensor 2 in ΔpenP after induction with β-lactams belonging to the group of cephalosporins. Experiments have been performed in triplicates. The upper row displays the growth measured as OD600nm, while the row immediately below shows the luminescence intensity in relative luminescence units normalized over OD600nm (y-axis). Both the y-axis for the OD600nm measurements (growth) and the y-axis presenting RLU/OD600nm are valid for all the compounds presented in the respective row. The x-axis displays the time in hours. Induction with the antibiotics occurred after 1 hour of growth as indicated by the black vertical dotted line. The assayed strains and their corresponding colors are indicated in the legend on the upper right. Continuous lines and round symbols always represent growth, while dotted lines with triangular symbols represent the luminescence signal over time. (C) Growth and luminescence of Control 5 (see Table 1), Biosensor 2 and Biosensor 2 in ΔpenP after induction with a β-lactam belonging to the group of carbapenems. (D) Growth and luminescence of Control 5, Biosensor 2 and Biosensor 2 in ΔpenP after induction with a β-lactam belonging to the group of monobactams. (E) Growth and luminescence of Control 5, Biosensor 2 and Biosensor 2 in ΔpenP after induction with the controls bacitracin and water. Experiments have been performed in triplicates.

Figure S7 (A) and (B): Growth (OD600nm) and luminescence signal (RLU/OD600nm) of control strains in response to different b-lactams. (A)
The upper row displays the growth measured as OD600nm, while the row immediately below shows the luminescence intensity in relative luminescence units normalized over OD600nm (y-axis). Both the y-axis for the OD600nm measurements (growth) and the y-axis presenting RLU/OD600nm are valid for all the compounds presented in the respective row. The x-axis displays the time in hours. Induction with the antibiotics occurred after 1 hour of growth as indicated by the black vertical dotted line. The assayed strains and their corresponding colors are indicated in the legend on the upper right. Continuous lines and round symbols always represent growth, while dotted lines with triangular symbols represent the luminescence signal over time. (A) Growth and luminescence of Control 1 (wild type, see Table 1), Control 2 (TMB3090, constitutively expressed lux operon) and Control 4 (Biosensor 2 in ΔpenP missing blaI repressor construct) after induction with β-lactams belonging to the group of penicillins. (B) Growth and luminescence of Control 1 (see Table 1, wild type), Control 2 (see Table 1, TMB3090, constitutively expressed lux operon) and Control 4 (see Table 1, Biosensor 2 in ΔpenP missing blaI repressor construct) after induction with β-lactams belonging to the group of cephalosporins. Experiments have been performed in triplicates.

Figure S7 (C)-(E): Growth (OD600nm) and luminescence signal (RLU/OD600nm) of control strains in response to different b-lactams.
The upper row displays the growth measured as OD600nm, while the row immediately below shows the luminescence intensity in relative luminescence units normalized over OD600nm (y-axis). Both the y-axis for the OD600nm measurements (growth) and the y-axis presenting RLU/OD600nm are valid for all the compounds presented in the respective row. The x-axis displays the time in hours. Induction with the antibiotics occurred after 1 hour of growth as indicated by the black vertical dotted line. The assayed strains and their corresponding colors are indicated in the legend on the upper right. Continuous lines and round symbols always represent growth, while dotted lines with triangular symbols represent the luminescence signal over time. (C) Growth and luminescence of Control 1 (wild type, see Table 1), Control 2 (TMB3090, constitutively expressed lux operon) and Control 4 (Biosensor 2 in ΔpenP missing blaI repressor construct) after induction with a β-lactam belonging to the group of carbapenems. (D) Growth and luminescence of Control 1 (wild type, see Table 1), Control 2 (TMB3090, constitutively expressed lux operon) and Control 4 (Biosensor 2 in ΔpenP missing blaI repressor construct) after induction with a β-lactam belonging to the group of monobactams. (E) Growth and luminescence of Control 1 (wild type, see Table 1), Control 2 (TMB3090, constitutively expressed lux operon) and Control 4 (Biosensor 2 in ΔpenP missing blaI repressor construct) after induction with the controls bacitracin and water. Experiments have been performed in triplicates.    The y-axis shows the luminescence intensity in relative luminescence units normalized over OD600nm (RLU/OD600nm). The x-axis shows the concentration range of the respective antibiotic that varies depending on the compounds MIC. Experiments have been performed in triplicates. Figure S11: Results from the dose response assay with Biosensor 2 (TMB5608). The y-axis shows the luminescence intensity in relative luminescence units normalized over OD600nm (RLU/OD600nm). The x-axis shows the concentration range of the respective antibiotic that varies depending on the compounds MIC. Experiments have been performed in triplicates.
Figure S12: Growth and luminescence signal of Biosensor 2 DpenP in response to other cell wall antibiotics. The left y-axis displays the growth (OD600nm), while the right y-axis indicates the luminescence intensity in relative luminescence units normalized over OD600nm (RLU/OD600nm). The xaxis shows the time in hours. Induction with the antibiotics occurred after 1 hour of growth as indicated by the black vertical dotted line. The strains and their corresponding colors are indicated in the legend to the right of the graphs. Continuous lines and round symbols always represent growth, while dotted lines with triangular symbols represent the luminescence signal over time. Concentrations tested were beneath the MIC values found in literature, thus final concentrations tested were: daptomycin, 0.05 µg/ml; D-cycloserine, 5 µg/ml; phosphomycin, 1.25 µg/ml; polymyxin, 0.25 µg/ml; tunicamycin, 0.015 µg/ml; vancomycin, 0.125 µg/ml. Experiments have been performed in triplicates.