Fig 1.
Bac7 dual killing mechanism and K. pneumoniae molecular response to elevated ATP.
The figure shows the mechanisms of K. pneumoniae entry and killing by bac7 next to the bacterial response to combat bac7 dual membrane and ribosomal stress. A shows the two mechanisms of cell entry by bac7 where black arrows indicate 1) transporter mediated uptake and 2) membrane translocation into the cytosol. Red arrows indicate the metabolite changes upon bac7 treatment and blue directional arrows show the causative relationship towards the dual killing mechanisms. B shows that in response to bac7, K. pneumoniae activates MgtA to transport extracellular magnesium ions into the cell (black arrows), and MgtC to inhibit uptake of phosphate ions (Pi) through a transporter yet to be defined in S. enterica. These actions decrease intracellular Pi and ATP, which consequently increases transcription of rRNA and intracellular magnesium concentrations (red arrows). Pi limitation may be the connection to the decreases in oxidative phosphorylation, the tricarboxylic acid cycle, and c-di-GMP regulated fimbriae and cellulose operons in our transcriptional profiling (dashed blunt arrow).
Table 1.
Expression changes with bac7 treatment.
Fig 2.
Transcriptional insight reveals complex mechanism of killing and biofilm disruption.
Volcano plot of hvKp NTUH-K2044 most significantly differential expressed genes (DEGs) (-log10 FDR > 10) following 7.5 µmol L-1 bac7 for 30 minutes (A). Red dots show gene changes relevant to the mechanism of killing and blue dots show biofilm important gene changes. RT-qPCR validation of phoP and mgtC (B), and c-di-GMP regulated bcsA and fimH (C) with 7.5 µmol L-1 bac7 for 30 minutes. Intracellular ATP assessment following 7.5 µmol L-1 bac7 for 30 minutes measuring relative luminescence units (RLU) using a CellTiter-Glo kit and normalized to the respective OD600 (RLU/OD600) (D). Intracellular c-di-GMP levels following 7.5 µmol L-1 bac7 for 30 minutes using a Cyclic di-GMP ELISA kit (E). The RT-qPCR was normalized using a ftsZ housekeeping gene. One-way ANOVA was used to determine significance compared to the no treatment groups for B-E with Dunnett’s correction for multiple comparisons with adjusted p-values shown (asterisks indicate p-values **** < 0.0001, ** < 0.01, and ns > 0.1) and error shown reported as ±SEM. Immunoblot of hvKp NTUH-K2044 with 7.5 µmol L-1 bac7 treatment using cultures normalized by optical density showing FimA antibody staining next to RNA polymerase alpha subunit control (F). The relative abundance of FimA from n = 11 western blots was quantified using ImageJ with error shown as ±SD and significance determined by unpaired t-test (p-value 0.0076) (G). Transmission electron microscopy (TEM) of hvKp NTUH-K2044 with or without 7.5 µmol L-1 bac7 using antibody labeling of FimA fimbriae (H). Percent piliation was determined by manually counting at least 50 cells/sample and dividing the number of cells with extruding pili by the total number of cells counted in the TEM images for I.
Fig 3.
Bac7 transient membrane localization induces hvKp membrane depolarization but not leakage.
Membrane depolarization of hvKp NTUH-K2044, hvKp KPPR1S, KPPR1S ∆wcaJ, and cKp MKP103 using a DiSC3 membrane potential dye (A-D). The dye was quenched into the membrane for 30 minutes before adding bac7 2-fold dilutions and the arrow indicates the peptide addition. Single cell fluorescent microscopy with FM 4-64 membrane dye (red), DAPI DNA stain (blue), after 15 minutes and 45 minutes incubation with 7.5 µmol L-1 FITC-bac7 (green) (E). Transmission microscopy of hvKp NTUH-K2044 with a 0.1% phosphotungstic negative stain with or without 7.5 µmol L-1 bac7 (F). Black arrow indicates membrane blebbing locations with treatment and red arrows show capsule deformation. Error for A-D are reported as ±SEM.
Table 2.
MIC and MBEC90 values of clinical antibiotics and bac7 towards K. pneumoniae isolates.
Fig 4.
Bac7 displays differential membrane depolarization across K. pneumoniae clinical isolates.
The figures show the membrane depolarization and FITC-bac7 uptake using the MRSN clinical diversity panel isolates. DiSC3 fluorescent readings after addition of 0.95 µmol L-1 bac7 for 30 minutes with the isolates ordered from low to high colistin resistance (left to right), respectively based on MICs determined in our previous work (A). Mucoid isolates that harbor the rmpADC locus are marked with two asterisks (**) and isolates with red text were chosen for the single cell fluorescent microscopy. Single cell fluorescent microscopy with FM 4-64 membrane dye (red) after 15 minutes incubation with 1.9 µmol L-1 FITC-bac7 (green) (B). The number of cells (n) displaying high levels of FITC-bac7 uptake are shown for each isolate. Flow cytometry analysis with BacLight red cell stain following 1.9 µmol L-1 FITC-bac7 treatment of MRSN 365679 and 607210 for 30 minutes using FITC and APC lasers for green and red fluorescence, respectively (C). Gating was performed to generate 4 quadrants (Q1: FITC + /APC-; Q2: FITC + /APC + Q3: FITC-/APC + ; Q4: FITC-/APC-) with a high bar set to quantify super FITC fluorescent cells in Q2. Quantification of the cells in quadrant 2 (FITC + /APC+) and quadrant 3 (FITC-/APC+) with triplicate flow cytometry samples (D). Single flow cytometry figures shown as representative of triplicate experiments in C. Error for A and D are reported as ±SEM.
Fig 5.
Spatial polysaccharide distribution and cellular density generate biofilm mediated resistance in clinical isolates.
The figures show confocal z-stack images of FITC-bac7 treated MRSN 1912 and 564304 biofilms. A and B shows 1 hour treatment of MRSN 1912 and 564304, respectively with 15 μmol L-1 FITC-bac7, stained with BactoView cell stain (red) and calcofluor white polysaccharide stain (blue) to visualize localization of the FITC-bac7 peptide (green). All imaging was performed in triplicate with one representative image shown. BiofilmQ software was used to process the confocal z-stack images and graph the mean relative fluorescence in each z-stack layer (dz (µm)) for red labeled cells and green FITC-bac7 peptide (C and D).
Fig 6.
Mutant analysis reveals both SbmA and MgtC are important for mitigating the biofilm disruption potential of bac7.
The figures show comparisons of the parental MKP103 (WT) and mutants lacking the SbmA polyproline transporter (∆sbmA) and MgtC (∆mgtC). Relative gene expression of phoP and mgtC grown in LB to log phase to analyze baseline expression of these genes in colistin-resistant MKP103 parental isolate compared to hvKp NTUH-K2044 and hvKp KPPR1S using a 16S housekeeping gene for normalization (A). Intracellular ATP levels shown as fold change in RLU/OD600 between no treatment and 1.9 µmol L-1 bac7 treatment for 30 minutes for parental MKP103 (WT), MKP103 ∆sbmA, and MKP103 ∆mgtC (B). Fold change intracellular c-di-GMP levels between no treatment and 7.5 µmol L-1 bac7 for MKP103 parental (WT) and MKP103 ∆mgtC quantified using the Cyclic di-GMP ELISA kit (C). DiSC3 membrane potential dye assessment of inner membrane depolarization with 0, 0.95, and 1.9 µmol L-1 bac7 showing the peak fluorescence observed following 10 minutes treatment for parental MKP103 (WT), MKP103 ∆sbmA, and MKP103 ∆mgtC (D). Biofilm density measurement using crystal violet staining showing normalized (subtract background) biofilm density (OD550) after 24 hours with no treatment or 15 µmol L-1 bac7 (E). 3D-rendering of confocal z-stack images of pre-formed biofilms of the MKP103 parental isolate (WT) and ∆mgtC mutant untreated and treated with 15 µmol L-1 bac7 (F). Cells were stained with SYTO9 (green cells), and matrix polysaccharides were stained with calcofluor white (blue matrix). n = 3 biofilms imaged for each condition with a representative image shown. Two-way ANOVA was used to determine significance for A in comparison to MKP103 with Tukey’s multiple comparison correction, and one-way ANOVA was used to determine significance for C-F comparing the mutants to the WT with Tukey’s multiple comparison correction. Significance is shown with asterisks (adjusted p-values **** < 0.0001, *** < 0.001, ** < 0.01, * < 0.1, and ns > 0.1) and error shown as ±SEM.
Fig 7.
In vivo treatment decreases bacterial burden and colonization of vital organs in a skin abscess biofilm infection model.
Skin abrasions on the ventral side of mice were formed and infected with 1x104 CFU hvKp NTUH-K2044. 5 hours post-infection, a single dose of bac7 and positive control polymyxin B (10X MICs) was administered onto the infection site (A). Four days post-treatment, regions of skin abrasions were excised, and organ systems were harvested and homogenized to quantify the bacteria. Mouse weight assessed at day two and day four post infection and shown as a percentage weight change compared to day 0 (B). Log10 CFU mL-1 of the bacterial load in the skin, kidney, spleen, and liver (C). Data was graphed as box plots to show the median values with error shown as ±SEM. The limit of detection (100 CFU mL-1) is shown as a dotted line on the y-axis. Significance was determined using non-parametric Kruskal-Wallis one-way ANOVA of n = 6 mice with p-values were obtained by comparing untreated groups to the treated groups. BioRender was used to generate the vector images with the figure designed in Adobe Illustrator. Links for the BioRender vectors can be found: Mouse with organs: Created in BioRender. Fleeman, R. (2025) https://BioRender.com/i5yuud0. Mouse vector: Created in BioRender. Fleeman, R. (2025) https://BioRender.com/ackcbgr. Created in BioRender. Fleeman, R. (2025) https://BioRender.com/atzgcky. Petri Dish: Created in BioRender. Fleeman, R. (2025) https://BioRender.com/atzgcky.