Neutrophil extracellular traps (NETs)-mediated killing of carbapenem-resistant hypervirulent Klebsiella pneumoniae (CR-hvKP) are impaired in patients with diabetes mellitus

Abstract

Carbapenem-resistant hypervirulent Klebsiella pneumoniae (CR-hvKP) have been reported in recent years across Asian countries and pose a serious threat to public health. Neutrophils represent the first line of defense against numerous infectious pathogens, such as CR-hvKP. Neutrophil extracellular traps (NETs) constitute one of the major antimicrobial defense mechanisms in neutrophils against invading pathogens, especially against hvKP. Interestingly, previous studies have demonstrated that patients with type 2 diabetes mellitus (T2D) display elevated levels of NETosis but are vulnerable to infections caused by hvKP. The discrepancy propels us to investigate the role of NETs in hvKP infections in the context of T2D. By utilizing a clinical-derived CR-hvKP strain and a combination of NETs complex detection, phagocytosis testing, NETs killing assay and immunofluorescence, and scanning electron microscope assays, we identified defective NETs-mediated killing of CR-hvKP strain in patients with T2D. Specifically, we show that the impaired NETs-mediated killing in T2D is not due to the decreased NETs formation, as the neutrophils isolated from T2D patients exhibited enhanced NETs formation compared to healthy controls. Further, we demonstrate that the reduced NETs activity does not result from the trapping failure of CR-hvKP, but likely associated with the deficient surface damage conferred by the NETs of T2D patients. Our data provide a novel insight into the defective innate immune response against CR-hvKP in T2D.

Keywords: Neutrophil extracellular traps (NETs); antimicrobial activity; carbapenem-resistant hypervirulent Klebsiella pneumoniae (CR-hvKP); innate immunity; type 2 diabetes mellitus (T2D).

Figure 1.

Representative immunofluorescence images (a) and quantification of the percentages of neutrophils undergoing NETosis (b) in the presence of PMA (20 nM) and A23187 (1 μM). Type 2 diabetes mellitus patients (T2D) neutrophil form more extracellular traps than healthy controls (HC) with the same stimulation. For the NETs induction, samples from five diabetics and five healthy volunteers were included. Neutrophils from each sample were induced by PMA or A23187 on three separate coverslips and six random fields were acquired per coverslip. On average, each field approximately contains 81 cells and each sample can count about 1400 cells. (c) Quantification of MPO-DNA complexes in serum samples of HC and T2D patients. (d-g) CR-hvKP C2166 showed highly resistant to neutrophilic phagocytosis using different assay methods. (d) E. coli ATCC 25922 and K. pneumoniae C2166 have a comparable efficiency of FITC labeling and more than 95% of the population can be stained effectively. (e) CR-hvKp C2166 displayed similar levels of resistance to phagocytosis by healthy and diabetic neutrophils after 40-min incubation. E. coli strain ATCC 25922 is susceptible to neutrophil phagocytosis and used as neutrophil phagocytosis function control in this study. (f) FITC-labeled bacteria detection based on flow cytometry; (g) CFUs recovery percentage based on DNase I treated neutrophil killing assays. ns, no significance; *p < 0.05: SSC, side scatter.

Figure 1.

Representative immunofluorescence images (a) and quantification of the percentages of neutrophils undergoing NETosis (b) in the presence of PMA (20 nM) and A23187 (1 μM). Type 2 diabetes mellitus patients (T2D) neutrophil form more extracellular traps than healthy controls (HC) with the same stimulation. For the NETs induction, samples from five diabetics and five healthy volunteers were included. Neutrophils from each sample were induced by PMA or A23187 on three separate coverslips and six random fields were acquired per coverslip. On average, each field approximately contains 81 cells and each sample can count about 1400 cells. (c) Quantification of MPO-DNA complexes in serum samples of HC and T2D patients. (d-g) CR-hvKP C2166 showed highly resistant to neutrophilic phagocytosis using different assay methods. (d) E. coli ATCC 25922 and K. pneumoniae C2166 have a comparable efficiency of FITC labeling and more than 95% of the population can be stained effectively. (e) CR-hvKp C2166 displayed similar levels of resistance to phagocytosis by healthy and diabetic neutrophils after 40-min incubation. E. coli strain ATCC 25922 is susceptible to neutrophil phagocytosis and used as neutrophil phagocytosis function control in this study. (f) FITC-labeled bacteria detection based on flow cytometry; (g) CFUs recovery percentage based on DNase I treated neutrophil killing assays. ns, no significance; *p < 0.05: SSC, side scatter.

Figure 2.

(a) Healthy controls (HC) and type 2 diabetes mellitus (T2D) patients NETs in vitro killing of CR-hvKP C2166. NETs were induced by 100 nM PMA for 3 h, then C2166 (1 × 10⁶ CFUs) were added and co-incubated for another hour. After cells lysing, C2166 were plated on LB agar and enumerated the next day. HC and T2D patients NETs were combined with C2166 and bacterial surface were analyzed under scanning electron microscope. The mean percentage of surface-damaged bacteria by HC and T2D NETs was compared and shown in (b). Representative images were presented in (c) and (d). **p < 0.01, ns: no significance.

Figure 2.

(a) Healthy controls (HC) and type 2 diabetes mellitus (T2D) patients NETs in vitro killing of CR-hvKP C2166. NETs were induced by 100 nM PMA for 3 h, then C2166 (1 × 10⁶ CFUs) were added and co-incubated for another hour. After cells lysing, C2166 were plated on LB agar and enumerated the next day. HC and T2D patients NETs were combined with C2166 and bacterial surface were analyzed under scanning electron microscope. The mean percentage of surface-damaged bacteria by HC and T2D NETs was compared and shown in (b). Representative images were presented in (c) and (d). **p < 0.01, ns: no significance.

Figure 3.

Healthy controls (HC) and type 2 diabetes mellitus patients (T2D) NETs can entrap a similar amount of C2166 strain. Green arrows indicate the ensnared bacteria within the NETs structure.

Figure 3.

Healthy controls (HC) and type 2 diabetes mellitus patients (T2D) NETs can entrap a similar amount of C2166 strain. Green arrows indicate the ensnared bacteria within the NETs structure.