Evaluating Different Virulence Traits of Klebsiella pneumoniae Using Dictyostelium discoideum and Zebrafish Larvae as Host Models

Abstract

Multiresistant and invasive hypervirulent Klebsiella pneumoniae strains have become one of the most urgent bacterial pathogen threats. Recent analyses revealed a high genomic plasticity of this species, harboring a variety of mobile genetic elements associated with virulent strains, encoding proteins of unknown function whose possible role in pathogenesis have not been addressed. K. pneumoniae virulence has been studied mainly in animal models such as mice and pigs, however, practical, financial, ethical and methodological issues limit the use of mammal hosts. Consequently, the development of simple and cost-effective experimental approaches with alternative host models is needed. In this work we described the use of both, the social amoeba and professional phagocyte Dictyostelium discoideum and the fish Danio rerio (zebrafish) as surrogate host models to study K. pneumoniae virulence. We compared three K. pneumoniae clinical isolates evaluating their resistance to phagocytosis, intracellular survival, lethality, intestinal colonization, and innate immune cells recruitment. Optical transparency of both host models permitted studying the infective process in vivo, following the Klebsiella-host interactions through live-cell imaging. We demonstrated that K. pneumoniae RYC492, but not the multiresistant strains 700603 and BAA-1705, is virulent to both host models and elicits a strong immune response. Moreover, this strain showed a high resistance to phagocytosis by D. discoideum, an increased ability to form biofilms and a more prominent and irregular capsule. Besides, the strain 700603 showed the unique ability to replicate inside amoeba cells. Genomic comparison of the K. pneumoniae strains showed that the RYC492 strain has a higher overall content of virulence factors although no specific genes could be linked to its phagocytosis resistance, nor to the intracellular survival observed for the 700603 strain. Our results indicate that both zebrafish and D. discoideum are advantageous host models to study different traits of K. pneumoniae that are associated with virulence.

Keywords: Danio rerio; Dictyostelium discoideum; host-pathogen interactions; hypervirulent Klebsiella pneumoniae; intracellular survival; resistance to phagocytosis.

Figure 1

Use of the D. discoideum social development assay to evaluate K. pneumoniae virulence. (A) Main stages of the D. discoideum social cycle, which culminates with the formation of the fruiting bodies. The scheme was adapted from Fey et al. (2007). (B) Micrographs of the social cycle progression when feeding amoeba with different K. pneumoniae strains or with K. aerogenes DBS0305928 (control). Scale bar, 100 μm. (C) Semi-quantitative assessment of the social cycle progression in presence of different Klebsiella strains using several replicates. The symbols represent the mean of six independent experiments and the error bars show the standard deviation. The details on scoring criteria can be found in the Materials and Methods section.

Figure 2

Bacterial phagocytosis assays revealed differences in invasiveness and intracellular survival among different K. pneumoniae strains. (A) Percentage of bacteria that were internalized after co-incubation of amoebae-Klebsiella for 1 h. (B) Total intracellular bacteria were titrated at the indicated times (T_x), and the results were expressed as the fraction respect the survival at time zero (T₀). In both (A,B), the symbols represent the mean of three independent experiments and the standard deviation. ^*p < 0.05, ^****p < 0.0001. (C) Live-cell imaging of D. discoideum cells (green) interacting with Klebsiella (red), representative of the experiments showed in (A,B). Bright field and fluorescence images were blended in order to better appreciate the boundaries of the amoeba cells. At time 24 h, two pictures of amoebae infected with Kp 700603 are shown in order to illustrate that both extracellular (upper picture) and intracellular bacteria (lower picture) could be observed. Scale bar, 10 μm.

Figure 3

Evaluation of the resistance to phagocytosis of K. pneumoniae using D. discoideum as host cell model. (A) Extracellular viable bacteria count after different times of Klebsiella-amoeba co-incubation. At each indicated time, the mixed cultures were centrifuged at low rpm in order to sediment D. discoideum cells and intracellular bacteria, and the CFU remaining in the supernatant were titrated. The symbols represent the mean of three independent experiments and the error bars correspond to the standard deviation. At times 3 and 24 h, Kp RYC492 and Kp 700603 showed a significantly higher number of extracellular bacteria, compared with K. aerogenes (Two-way ANOVA, ^*p < 0.05, ^****p < 0.0001). (B) Confocal microscopy images representative of the experiment showed in (A). At each indicated time, an aliquot of the mixed cultures was directly mounted for microscopic observation (without centrifuging). D. discoideum cells are green and bacterial cells are red. Bright field and fluorescence images were blended in order to better appreciate the boundaries of the amoeba cells. Scale bar, 10 μm.

Figure 4

K. pneumoniae strains show different virulence behavior in the zebrafish larvae as revealed by survival curves and neutrophil recruitment. (A) Scheme of a 5-dpf zebrafish larva. Bacteria were injected in the dorsal caudal artery (red arrow) for establishing a rapid systemic infection, or in the otic vesicle (blue circle) for a localized infection. (B) Kaplan-Meier survival curves following injection of 3 dpf larvae with various strains of K. pneumoniae or with E. coli DH5α as a control. Larvae survival was monitored for 3 days after injection. The data are representative of three independent experiments, with a total of 57 larvae per bacterial strain. (C) Representative picture of a larva infected for 24 h with K. pneumoniae RYC492 strain, after acute bacteremia and death, and an unaffected larva injected with E. coli as a control. Bacteria expressing mCherry fluorescent protein is shown red. Scale bar, 500 μm. (D) Neutrophils attracted to the otic vesicle of Tg(BACmpo:gfp) zebrafish larvae injected with K. pneumoniae strains or E. coli DH5α after 24 h post-injection (hpi). Bacteria are red and neutrophils are green. The white dashed lines encircle the otic vesicle. Scale bar, 100 μm. (E) Bacterial burden in the otic vesicle of zebrafish larvae after 24 hpi. Measurements of red fluorescence from mCherry-labeled bacteria are plotted as arbitrary units (AU) for each K. pneumoniae strain or for E. coli DH5α. The data are representative of three replicates, with a total of 21 larvae per condition. (F) Neutrophils infiltrated in the otic vesicle of Tg(BACmpo:gfp) zebrafish larvae after 24 hpi. Measurements of green fluorescence from EGFP-labeled neutrophils are plotted as arbitrary units (AU) for larvae injected with K. pneumoniae strains, E. coli DH5α, or PBS. The data are representative of three independent experiments, with a total of 21 larvae per bacterial strain. Error bars represent the standard deviation among the replicates. Statistical differences compared to E. coli DH5α were assessed using a Kruskal-Wallis analysis followed by Dunn's multiple comparisons test. ^****p < 0.0001.

Figure 5

K. pneumoniae can infect zebrafish larvae by immersion and colonize its gastrointestinal tract. (A) Scheme of the gastrointestinal tract of a 5-dpf zebrafish larva, showing three arbitrarily defined zones (anterior, middle and posterior). (B) Semi-quantitative assessment of the intestinal colonization upon immersing zebrafish larvae in a suspension containing different K. pneumoniae strains or with E. coli DH5α. At the indicated time, larvae were anesthetized, observed under a fluorescence stereomicroscope, and the total intestinal colonization was scored in a scale from 1 to 9, following the criteria described in the Materials and Methods section. The data presented correspond to six replicates with a total of 75 larvae per bacterial strain. (C) Fluorescence microphotographs of zebrafish larvae representative of the data showed in (B). Bright field images of representative non-infected larvae at 48 and 96 h post-immersion are presented for more clarity. The larvae vasculature is shown in green, while infecting bacteria are shown in red. Scale bar, 500 μm. ^****p < 0.0001.

Figure 6

Klebsiella pneumoniae RYC492 virulence correlates with a more prominent and irregular capsule and with increased biofilm formation ability. (A) K. pneumoniae cells were subjected to classical capsule staining with India ink, and observed and photographed using confocal microscopy. Pictures of 3 representative cells of each bacterial strain are presented. The capsule corresponds to the white area between the bacillary cell and the black background. Scale bar, 2 μm. (B) Biofilm formation assessment of K. pneumoniae strains or E. coli DH5α upon growth in a 96-well plate during 18 h. Bacterial cells adhered to the wells were stained with crystal violet, where biofilm formation is proportional to the absorbance of the bound stain (measured at 590 nm). The values shown correspond to the mean of three independent experiments and the error bars represent the standard deviation. RYC492 strain showed a significantly higher ability to form biofilms than Kp 700603 (one-way ANOVA, ^****p < 0.0001), and then the rest of the bacterial strains tested.