Bactericidal, anti-biofilm, and anti-virulence activity of vitamin C against carbapenem-resistant hypervirulent Klebsiella pneumoniae

Abstract

The emergence of carbapenem-resistant hypervirulent Klebsiella pneumoniae (CR-hvKP) causing high mortality in clinical patients infers the urgent need for developing therapeutic agents. Here, we demonstrated vitamin C (VC) exhibited strong bactericidal, anti-biofilm, and virulence-suppressing effects on CR-hvKP. Our results showed such a bactericidal effect is dose-dependent both in vitro and in the mouse infection model and is associated with induction of reactive oxygen species (ROS) generation. In addition, VC inhibited biofilm formation of CR-hvKP through suppressing the production of exopolysaccharide (EPS). In addition, VC acted as an efflux pump inhibitor at subminimum inhibitory concentration (sub-MIC) to disrupt transportation of EPS and capsular polysaccharide to bacterial cell surface, thereby further inhibiting biofilm and capsule formation. Furthermore, virulence-associated genes in CR-hvKP exposed to sub-MIC of VC were downregulated. Our findings indicated VC could be an effective and safe therapeutic agent to treat CR-hvKP infections in urgent cases when all current treatment options fail.

Keywords: Microbiofilms; Microbiology; Molecular biology.

Graphical abstract

Figure 1

Time-kill curve and development of resistance of CR-hvKP strains KP1088 and HvKP3 treated with VC (A) KP1088 in the early exponential phase treated with VC. (B) KP1088 in the early exponential phase treated with VC. (C) HvKP3 in the early exponential phase treated with VC. (D) KP1088 in the late exponential phase treated with VC. The bacterial CFUs per mL at different time points during 24 h were determined. All experiments were performed in triplicate, and the mean ± SD is shown. (E) Development of resistance to VC in KP1088 and HvKP3. Strain KP1088 and HvKP3 were subjected to serial passages in the presence of different concentrations of VC. The y axis is the highest concentration of VC the cells could grow during the serial passages. KP1088 could only grow in VC at 0.5∗MIC and the highest concentration HvKP3 could grow was 0.25∗MIC after incubation for three days

Figure 2

Biofilm formation in CR-hvKP strains upon treatment with VC (A and B) Biofilm formation in KP1088 (A) and HvKP3 (B) upon treatment with different concentrations of VC for 24 h. Data are representative of their experiments performed in triplicate. (C and D) Viable cell counting of KP1088 (C) and HvKP3 (D) biofilm treated with various concentrations of VC for 24 h. (E) CLSM imaging of biofilm formation in strain KP1088 and HvKP3 upon treatment with VC. The left column is images of biofilm of KP1088 without treatment and after treatment with 8 mg/mL and 16 mg/mL of VC. The right column is images of HvKP3 biofilm without treatment and after treatment with 8 mg/mL and 16 mg/mL of VC. (F) COMSTAT analysis of biovolumes from KP1088 biofilm treated with VC. (G) COMSTAT analysis of biovolumes from HvKP3 biofilm treated with VC. ∗, p < 0.05, ∗∗, p < 0.005, ∗∗∗p < 0.0005, ∗∗∗∗p < 0.00005 (independent t-test).

Figure 3

Scanning electron microscopic (SEM) images of HvKP3 biofilm The upper panel is SEM images of HvKP3 biofilm without treatment and visualized by 1000-fold (left) 10,000-fold (middle) and 20,000-fold (right) magnification. The lower panel is SEM images of HvKP3 biofilm grown in the presence of 4 mg/mL of VC and visualized by 1000-fold (left),10,000-fold (middle) and 20,000-fold (right) magnification

Figure 4

Antibacterial and anti-virulence effect of VC on K. pneumoniae strain KP1088 and HvKP3 (A) ROS generation in K. pneumoniae strain KP1088 and HvKP3 after 4 h exposure to VC of different concentrations. ROS production is expressed as the fluorescence intensity of ROS upon normalization to viable-cell counts. Anti-virulence effect of VC on K. pneumoniae strain KP1088 and HvKP3. (B) EPS production in K. pneumoniae KP1088 and HvKP3 upon incubation overnight in the presence of different concentrations of VC. The production of EPS was determined and expressed as glucose equivalents by matching against a glucose standard curve. (C) CPS production of KP1088 and HvKP3 incubated overnight in the presence of different concentrations of VC. CPS production was determined by matching against a uronic acid standard curve and normalized to the total viable bacterial cell counts. Results were expressed as uronic acid equivalents per 10⁶ CFU. All experiments were performed in triplicate, and the mean ± SD is shown

Figure 5

Membrane potential determination in CR-hvKP (A and B) Determination of effect of different concentrations of VC on bacterial cell membrane potential of strain KP1088 (A) and HvKP3 (B). The K⁺ transporter valinomycin (5μM) was used as positive control. The assay was conducted with 100mM of KCl. All experiments were performed in triplicate, and the mean ± SD is shown

Figure 6

Inhibition of Nile Red efflux by VC in a dose-dependent manner (A–D) Intensity of Nile Red fluorescence in strain KP1088 after treatment with different concentrations of VC; (B) Nile red fluorescence intensity of KP1088 after treatment of VC, followed by washout of VC after staining; (C) Nile red fluorescence of HvKP3 treated with different concentrations of VC; (D) Nile red fluorescence of HvKP3 upon treatment of VC and washout of VC after staining. 50 mM of Glucose was added to the assay at 2.8 min; the fluorescence was monitored for another 20 min. All experiments were performed in triplicate, and the mean ± SD is shown

Figure 7

qRT-PCR analysis of CR-hvKP treated with VC Fold expression of the mucoid phenotype-encoding genes, fimbriae genes, and biofilm-related genes in K. pneumoniae strain KP1088 and HvKP3 cells with or without treatment with VC at concentrations of 0.5 × and 0.25 × MIC, respectively. Mean ± SD from three replicates are depicted

Figure 8

Mouse infection model (A and B) Efficacy of VC in mouse infection model with strain KP1088 (A) and HvKP3 (B) as infecting agent. Neutropenic NIH mice were infected intraperitoneally with 1.0 × 10⁴ CFU of KP1088 and 6.4 × 10⁴ CFU of HvKP3, respectively, followed by treatment with saline (control), VC at 20 mg/kg and VC at 80 mg/kg. Each mouse was treated five times, with 12 h interval between each treatment. The number of dead mice was recorded at each treatment

Figure 9

Schematic representation of the antimicrobial and anti-virulence mechanisms of VC ROS (OH·) was generated in Fenton's reaction induced by intracellular VC①. ROS storm causes damages to bacterial cells including lipid peroxidation②, membrane disruption③, protein denaturation④ and DNA breakage⑤, resulting in collapse of the cell structure. Extracellular VC also causes dissipation in bacterial membrane proton motive force ⑥, which drives efflux activities. As a result, transportation of EPS was hampered ⑦, so that the biofilm formation was inhibited. In addition, export of CPS to the bacterial cell surface was also disrupted when efflux activities were inhibited ⑧. VC also inhibits expression of genes that encode various virulence factors ⑨, including rmpA and rmpA2 (capsule formation) ⑩, fimB, mrkJ, and ecpA (fimbriae formation) ⑪ and luxS, galF, fimB, mrkJ, ecpA, rmpA, and rmpA2 (biofilm formation) ⑫. Graphical abstract was created with BioRender.com (license number GP23G531KW).