The Long Pentraxin PTX3 Controls Klebsiella Pneumoniae Severe Infection

Abstract

Klebsiella pneumoniae is a common pathogen in human sepsis. The emergence of multidrug-resistant K. pneumoniae strains represents a major clinical challenge in nosocomial and community acquired infections. The long pentraxin PTX3, a key component of humoral innate immunity, is involved in resistance to selected pathogens by promoting opsonophagocytosis. We investigated the relevance of PTX3 in innate immunity against K. pneumoniae infections using Ptx3^-/- mice and mouse models of severe K. pneumoniae infections. Local and systemic PTX3 expression was induced following K. pneumoniae pulmonary infection, in association with the up-regulation of TNF-α and IL-1β. PTX3 deficiency in mice was associated with higher bacterial burden and mortality, release of pro-inflammatory cytokines as well as IL-10 in the lung and systemically. The analysis of the mechanisms responsible of PTX3-dependent control of K. pneumoniae infection revealed that PTX3 did not interact with K. pneumoniae, or promote opsonophagocytosis. The comparison of susceptibility of wild-type, Ptx3^-/-, C3^-/- and Ptx3^-/- /C3^-/- mice to the infection showed that PTX3 acted in a complement-independent manner. Lung histopathological analysis showed more severe lesions in Ptx3^-/- mice with fibrinosuppurative, necrotizing and haemorrhagic bronchopneumonia, associated with increased fibrin deposition in the lung and circulating fibrinogen consumption. These findings indicate that PTX3 contributes to the control of K. pneumoniae infection by modulating inflammatory responses and tissue damage. Thus, this study emphasizes the relevance of the role of PTX3 as regulator of inflammation and orchestrator of tissue repair in innate responses to infections.

Keywords: Klebsiella pneumoniae; Pentraxin 3 (PTX3); complement – immunological term; inflammation; innate immunity; sepsis.

Figure 1

Induction of PTX3 expression during Klebsiella pneumoniae infection. PTX3 concentration in lung homogenate (A) and serum (B) of wild-type mice infected intranasally with 10⁴ CFUs K. pneumoniae serotype 2 (ATCC 43816). Each dot is mean ± SEM (n = 4 mice). PTX3 concentration at 6, 24 and 48 hours after infection were compared with basal levels using ANOVA multiple analysis. *6h vs 48h, **0h vs 48h. (C–E) Correlation between the concentration of PTX3 and IL-1β (C), TNF-α (D) or MPO (E) in the lung during the course of the infection. Linear best-fit curve.

Figure 2

Susceptibility of Ptx3^-/- mice to K. pneumoniae infection. (A–E) Bacterial load in the lung (A) and blood (B) at 18 hours after intranasal infection (n = 6 mice) and in the lung (C), blood (D), and liver (E), at 44 hours after intranasal infection (n = 15-18). The median of CFU count is shown. Two-tailed Mann-Whitney test. (F, G) Kaplan-Meier survival curves of wild-type and Ptx3^−/− mice inoculated either intranasally with 10⁴ CFUs (n = 6 mice) (F) or intraperitoneally with 500 CFUs (n = 7 mice) K. pneumoniae (G). Survival curves were compared with the Log-Rank test. (H, I) CFU count in the spleen (H) and lung (I) at 44 hours after intranasal infection in mice treated with mPTX3 (10 μg/mouse at 0 and 24 hours after infection) (n = 7-8 mice). The median of CFU count is shown. Two-tailed Mann-Whitney test. (A, F) show one representative experiment out of two (A, F). (C–E) are the pool of two separate experiments. (B, G–I); one experiment performed.

Figure 3

Analysis of opsonophagocytosis and complement in PTX3-dependent resistance to K. pneumoniae. (A) Immunofluorescence analysis by wild-field microscopy and (B) flow cytometry analysis of CFSE-labelled K. pneumoniae (upper panels in (A), left zebra plot panels in (B) and FITC- labeled A. fumigatus (lower panels in (A) and right zebra plot panels in (B) incubated with or without PTX3, followed by anti-hPTX3 biotinylated antibody and streptavidin-Alexa Fluor 647 (AF647). Quantification of percentage of binding of hPTX3 (20 μg/ml) to K. pneumoniae (Kp) and A. fumigatus (AF) of the three experiments performed is shown in the right panel in (B) FITC-labelled A. fumigatus was used as positive control. (C) Left panel: Representative histogram showing phagocytosis (CFSE fluorescent intensity in CD11b⁺Ly6G⁺ neutrophils) in wild-type or Ptx3^-/- blood. Basal = blood without K. pneumoniae. Right panel: Phagocytosis of CFSE-labelled (K) pneumoniae by wild-type or Ptx3^-/- neutrophils with or without PTX3 (25 μg/ml). Results are reported as a percentage of wild-type neutrophil phagocytosis (n = 8 mice, from two separate experiments). Graphs show Mean ± SEM. Two-tailed Mann Withney test and ANOVA. (D) Kaplan-Meier survival curve of wild-type, Ptx3 ^-/-, C3^-/- and Ptx3/C3^-/- mice inoculated intranasally with K. pneumoniae. N= 3 wild-type, 5 Ptx3^-/-, 4 C3^-/- and 5 Ptx3/C3^-/- mice. Survival curves were compared with the Log-Rank test.

Figure 4

Analysis of inflammation in (K.) pneumoniae infection. (A-C) Cytokine concentration in lung homogenates at 18 hours (n = 6 mice) (A) and 44 hours (n = 18 mice, pool of two separate experiments) post-intranasal infection (B) or in blood 44 hours post-intranasal infection (n = 5-6 mice) (C). (D) IL-10 concentration in lung homogenate (n = 18 mice, pool of two separate experiments) and blood (n = 5-6 mice) 44 hours post-intranasal infection. (E) MPO activity in lung homogenates (n = 7 mice). (F) Flow cytometry analysis of neutrophils and monocytes/macrophages in the lung (n = 5 mice). (G) White blood cell count (n = 7-8 mice). (H) Percentage of neutrophils in blood (n = 5 mice). Mean ± SEM is shown. (A–D, F, H) Mann Whitney test. (G) Multiple t-test. The outliers were removed through ROUT.

Figure 5

Analysis of lung damage and fibrin deposition of K. pneumoniae intranasally infected mice. (A) Macroscopic aspect of lungs of wild-type and Ptx3^-/- mice infected with K. pneumoniae. A representative picture out of five is shown. (B–E): Histological analysis of wild-type mice. (B) Multiple foci of suppurative bronchopneumonia (arrow line). (C, D) Lesions centered on bronchiolar structures. (E) Bacteria in the bronchiolar and alveolar spaces or within foamy macrophages (arrowhead). (F–I): Histological analysis of Ptx3^-/- mice. (F) Multifocal areas of fibrinosuppurative and necrotizing and/or hemorrhagic bronchopneumonia (arrow left-up). (G) Intra alveolar and intrabronchial hemorrhages (arrow). (H, I) Perivascular aggregates of bacteria associated with edema (*). B, Bronchiole; BV, blood vessel. (J) Quantification of the hemorrhage grade in the lung (n = 4 mice). (K) Blood fibrinogen concentration (n = 7-8 mice) at 18 and 44 hours after infection and (L) TAT levels (n = 5 mice) at 44 hours after infection. (M) Fibrin deposition in the lung by confocal microscopy. Left panels: representative images of wild-type and Ptx3^-/- mice. Right panel: Quantitation of fibrin/ogen deposition. Each dot represents the mean percentage of the immunoreactive area (IRA) of four fields per mice (n = 4 mice). (J–M) Mean ± SEM is shown. Groups were compared using two-tailed Mann-Whitney test.

Figure 6

Analysis of the cellular source of PTX3 (A) PTX3 protein concentration in serum of chimeric mice 44 hours after infection. Mean ± SEM is shown (n = 3-5 mice). (B) Bacterial load in the lung of Ptx3 ^fl/fl Cdh5-cre/ERT2^+/- and Ptx3 ^fl/fl Cdh5-cre/ERT2^-/- mice 44 hours after infection. The median of CFU count is shown. (C) PTX3 concentration in lung homogenate and blood of Ptx3 ^fl/fl Cdh5-cre/ERT2^+/- and Ptx3 ^fl/fl Cdh5-cre/ERT2^-/- mice. Mean ± SEM is shown. n = 5 male and female mice (shown in black and blue, respectively). Groups were compared using two-tailed Mann-Whitney test. (D) Immunofluorescence confocal microscopy images of PTX3 (magenta), PDGFRα (blue), CD31 (red) and α-SMA (green) staining in the lung of wild-type mice 44 hours after infection. Representative merged image is shown (n = 3 mice; 3 fields were observed for each mouse). Scale bar = 250μm. (E) PTX3 production by embryonic fibroblasts after stimulation with heat killed K. pneumoniae at different bacteria:cells ratio for 8 and 24 h. TNF and LPS were used as positive control.