The Capsule of Acinetobacter baumannii Protects against the Innate Immune Response

Abstract

Acinetobacter baumannii is an opportunistic pathogen that has recently emerged as a global threat associated with high morbidity, mortality, and antibiotic resistance. We determined the role of type I interferon (IFN) signaling in A. baumannii infection. We report that A. baumannii can induce a type I IFN response that is dependent upon TLR4-TRIF-IRF3 and phagocytosis of the bacterium. Phase variants of A. baumannii that have a reduced capsule, lead to enhanced TLR4-dependent type I IFN induction. This was also observed in a capsule-deficient strain. However, we did not observe a role for this pathway in vivo. The enhanced signaling could be accounted for by increased phagocytosis in capsule-deficient strains that also lead to enhanced host cell-mediated killing. The increased cytokine response in the absence of the capsule was not exclusive to type I IFN signaling. Several cytokines, including the proinflammatory IL-6, were increased in cells stimulated with the capsule-deficient strain, also observed in vivo. After 4 h in our acute pneumonia model, the burden of a capsule-null strain was significantly reduced, yet we observed increases in innate immune cells and inflammatory markers compared to wild-type A. baumannii. This study underscores the role of phase variation in the modulation of host immune responses and indicates that the capsule of A. baumannii plays an important role in protection against host cell killing and evasion from activation of the innate immune response.

Keywords: Acinetobacter baumannii; Capsule; Host-pathogen interactions; Innate response; Lung; Pneumonia; Type I interferon; Type III interferon.

Fig. 1

A. baumannii activates type I IFN signaling via TLR4 and IRF3/5. aHeat map of type I IFN-related genes from RNA-seq of murine lungs infected with A. baumannii and uninfected controls. FC-fold change in A. baumannii infected samples compared to uninfected controls. b, cBone marrow-derived macrophages were stimulated overnight with PBS or A. baumannii for ELISA quantitation. d–gBone marrow macrophages were stimulated with A. baumannii for 2 h before qRT-PCR analysis. dWT and HK A. baumannii were added to macrophages in the presence/absence of cytoD and CQ. eRole of TLR receptors using knockout macrophages. fInhibition of IFN signaling using R. sphaeroides LPS (Rs-LPS). gRole of IRF was determined using knockout macrophages. hBMDCs from WT and knockout mice were quantified for Ifnb. N ≥3 from at least two independent experiments. ***p < 0.001, **p < 0.01, and *p < 0.05 relative to untreated or WT controls using a Student's t test (b, c, e, f) or ANOVA (d, g, h). Ab, A. baumannii; ANOVA, one-way analysis of variance; HK, heat-killed; cytoD, cytochalasin D.

Fig. 2

Type I IFN expression is increased in the absence of the capsule in a TLR4-dependent manner. aBMDCs were stimulated for 24 h with A. baumannii and supernatants quantified for IFN-β. b, cBMDCs were stimulated for 2 h before Ifnb and Tnf were quantified by qRT-PCR. dAlcian blue-stained SDS-PAGE gel of capsule extraction from strains. eWT and Tlr4^−/− BMDCs were infected with WT or translucent variants (WT-T) of A. baumannii for 2 h and Ifnb expression was assessed by qRT-PCR. Values were normalized to PBS controls. fBMDCs were stimulated with A. baumannii WT or translucent (WT-T) strains for 2 h with or without R. sphaeroides LPS (Rs-LPS), before RNA extraction and qRT-PCR analysis. N ≥ 6 from at least two independent experiments. ****p < 0.0001, ***p < 0.001, and **p < 0.01 compared to WT controls using an ANOVA with Dunnett's post test. ANOVA, one-way analysis of variance.

Fig. 3

A. baumannii capsule-mutant and translucent strains have increased susceptibility to host cell phagocytosis and killing. BMDCs were infected with FITC labeled A. baumannii WT, capsule-mutant (wzc), or translucent variants (WT-T) for 5 or 30 min. Cells were harvested and subjected to flow cytometry. aRepresentative flow cytometry plots of gated cells and positive populations at 30 min. bQuantification of phagocytosis. BMDCs (c), neutrophils (d), or peritoneal macrophages (e) were infected with A. baumannii strains for 2 h before bacteria were quantified. fGrowth of WT, wzc, and WT-T strains. Graphs display means with standard deviation. N ≥ 6 for all experiments. ***p < 0.001, **p < 0.01, and *p < 0.05 using a Student's t test (d, e) or ANOVA (b, c). ANOVA, one-way analysis of variance.

Fig. 4

Role of type I and III IFN signaling in A. baumannii infection. aMice were intranasally infected with 10⁷A. baumannii 5075 for 24 h before BALF, lung homogenates, and spleen homogenates were enumerated for bacteria. bWT and DKO mice were intranasally infected with 10⁷A. baumannii 5075 for 24 h before BALF, lung homogenates and spleen homogenates were quantified for bacteria. WT and DKO mice were infected as before and samples were collected 48 h later for bacterial burdens (c) and survival (d). eMice were infected intraperitoneally with 10⁵A. baumannii 5075 and bacterial counts enumerated 24 h later. Each point represents a mouse. Line display median. Data were assessed using a nonparametric Mann-Whitney test. DKO, double knockout Ifnar^−/−/Ifnlr^−/−.

Fig. 5

Absence of the capsule enhances cytokine production. BMDCs were stimulated with WT A. baumannii 5075 or a wzc mutant for 24 h. Supernatants were collected for cytokine quantification. aCytokines with increased expression in the absence of the capsule. Decreased cytokines in the absence of the capsule (b) and unaltered cytokines (c). Gray, PBS control; black, WT A. baumannii; white, wzc mutant. Graphs display means with standard deviation. N = 9 for infected samples and 2 for PBS controls. ****p < 0.0001, ***p < 0.001, **p < 0.01, and *p < 0.05 using a Student's t test.

Fig. 6

Capsule of A. baumannii protects against the innate immune response during infection. C57BL/6J mice were intranasally infected with 10⁷ CFU of A. baumannii 5075 for 4 h. aBacterial counts in BALF and lung homogenates. Each point represents a mouse. Lines display median. bFlow cytometric analysis of innate immune cells in BALF. cQuantification of cytokines in BALF using ELISA. N = UN-2, WT-8, wzc-7. Graphs show means with standard deviation. UN, uninfected; ns, not significant. ***p < 0.001, **p < 0.01, and *p < 0.05 using a nonparametric Mann-Whitney test (a, b) and Student's t test (c).