Role of acinetobactin-mediated iron acquisition functions in the interaction of Acinetobacter baumannii strain ATCC 19606T with human lung epithelial cells, Galleria mellonella caterpillars, and mice

Abstract

Acinetobacter baumannii, which causes serious infections in immunocompromised patients, expresses high-affinity iron acquisition functions needed for growth under iron-limiting laboratory conditions. In this study, we determined that the initial interaction of the ATCC 19606(T) type strain with A549 human alveolar epithelial cells is independent of the production of BasD and BauA, proteins needed for acinetobactin biosynthesis and transport, respectively. In contrast, these proteins are required for this strain to persist within epithelial cells and cause their apoptotic death. Infection assays using Galleria mellonella larvae showed that impairment of acinetobactin biosynthesis and transport functions significantly reduces the ability of ATCC 19606(T) cells to persist and kill this host, a defect that was corrected by adding inorganic iron to the inocula. The results obtained with these ex vivo and in vivo approaches were validated using a mouse sepsis model, which showed that expression of the acinetobactin-mediated iron acquisition system is critical for ATCC 19606(T) to establish an infection and kill this vertebrate host. These observations demonstrate that the virulence of the ATCC 19606(T) strain depends on the expression of a fully active acinetobactin-mediated system. Interestingly, the three models also showed that impairment of BasD production results in an intermediate virulence phenotype compared to those of the parental strain and the BauA mutant. This observation suggests that acinetobactin intermediates or precursors play a virulence role, although their contribution to iron acquisition is less relevant than that of mature acinetobactin.

Fig 1

Interaction of A. baumannii bacteria with A549 alveolar epithelial cells. (A) Sterile medium. (B to D) Adhesion of parental A. baumannii ATCC 19606^T cells (B) or cells of the s1 acinetobactin production (C) or t6 utilization (D) isogenic derivatives examined by SEM after A549 cell monolayers were infected with 2 × 10⁶ bacteria for 1 h. Bars, 1 μm. (E) Surface-attached bacterial counts collected from the SEM micrographs of infected monolayers, examples of which are shown in panels B to D. Data were collected using five A549 cells in different fields from three different biological replicates. Values are means ± 1 standard deviation (error bars). (F) Intracellular CFU counts obtained after lysates of Gm-treated A549 cells were plated on LB agar and incubated overnight at 37°C. Data represent three independent experiments done in duplicate each time. Values are means ± 1 standard deviation (error bars).

Fig 2

Role and expression of the acinetobactin-mediated iron acquisition functions by intracellular bacteria. A549 monolayers were incubated with sterile medium (Sm) or infected for 3 h with 5 × 10³ cells of the ATCC 19606^T parental strain (P) or cells of the s1 or t6 mutant. Monolayers were also coinfected with 5 × 10³ cells of the parental strain mixed with equal number of cells of the s1 or t6 isogenic derivative. CFU counts were obtained after Gm-treated A549 cell lysates were plated on LB agar (La) and LB agar containing 40 μg/ml Km (La-K) to determine total bacterial and mutant cell counts, respectively. (Inset) Detection of the acinetobactin outer membrane receptor protein BauA. A. baumannii ATCC 19606^T whole-cell lysates were prepared from cells cultured in supplemented DMEM (lane 3) or from intracellular bacteria recovered after infection of epithelial cells (lane 4). Protein samples loaded in lane 1 or 2, which were prepared from bacteria cultured in LB broth containing 100 μM FeCl₃ or 100 μM DIP, respectively, were used as a control for differential production of BauA in response to free-iron availability. Horizontal bars with numbers indicate the fold change before the slash and the P value for the compared samples after the slash. Values are means ± 1 standard deviation (error bars).

Fig 3

CLSM analysis of A549 monolayers infected with A. baumannii ATCC 19606^T isogenic strains. (A to D) Monolayers were infected with ATCC 19606^T-GFP (B), s1-GFP (C), or t6-GFP (D) cells, using uninfected monolayers (A) as a negative control. Production of GFP was detected using a green filter. All monolayers were stained with propidium iodide, which was detected using a red filter. All images were taken at a magnification of ×630, and Z stack analysis was performed at 0.63 μM increments. Bar, 20 μm. (E) Graphic representation of intracellular bacterial counts. Horizontal bars with numbers indicate the fold change before the slash and the P value for the compared samples after the slash. Values are means ± 1 standard deviation (error bars).

Fig 4

A549 apoptotic response to bacterial infection determined by TUNEL assays and fluorescence microscopy. A549 monolayers were infected with cells of the ATCC 19606^T parental strain (19606), the s1 acinetobactin synthesis mutant, or the t6 acinetobactin transport mutant. A549 cells incubated in the presence of sterile medium (Sm) and sterile medium supplemented with 2 M HCl were used as negative and positive controls, respectively. Samples were examined by fluorescence microscopy at a magnification of ×400 or ×600. (Right) Bar graph of the number of apoptotic foci determined from the cognate micrographs captured with fluorescence microscopy. Horizontal bars with numbers indicate the fold change before the slash and the P value for the compared samples after the slash. Values are means ± 1 standard deviation (error bars).

Fig 5

G. mellonella infection and killing assays. (A) For infection assays, caterpillars were injected with 1 × 10⁵ bacteria of the ATCC 19606^T parental strain (19606) or the s1 or t6 iron-deficient isogenic derivative and incubated at 37°C in darkness for 18 h. Dilutions of whole-larva lysates were plated on nutrient agar, and colony counts were determined after overnight incubation at 37°C. Horizontal bars with numbers indicate the fold change before the slash and the P value for the compared samples after the slash. Values are means ± 1 standard deviation (error bars). (B to E) For killing assays, caterpillars were infected with high (1 × 10⁵ bacteria [B and C]) or low (1 × 10² bacteria [D and E]) bacterial doses in the absence (B and D) or presence (C and E) of 100 μM Fe₃Cl. For negative controls, caterpillars were injected with comparable volumes of PBS (B and D) or PBS plus 100 μM Fe₃Cl (C and E). The inset in panel B shows the melanization of infected caterpillars (I) and no pigmentation of uninfected (U) caterpillars.

Fig 6

Postinfection tissue bacterial loads in mice infected with the ATCC 19606^T, s1, and t6 strains. C57BL/6 mice (n = 8 mice/group) were infected with ∼1 × 10⁶ CFU of the indicated strain, and bacterial loads were measured in spleen homogenates 16 h postinfection. Each symbol represents the bacterial load from an individual mice, with the horizontal line indicating the median of the group. The asterisk indicates that the values for the ATCC 19606^T parental strain (19606) and the s1 and t6 iron-deficient isogenic derivatives were statistically significantly different (P < 0.001).