Evidence of Brucella abortus OPS dictating uptake and restricting NF-kappaB activation in murine macrophages

Abstract

Smooth Brucella abortus S2308 is virulent while rough derivatives are attenuated. Intracellular killing is often blamed for these differences. In the studies described, uptake kinetics and interaction of S2308 and S2308 manBA::Tn5 (CA180) rough mutants with macrophages were investigated. The results revealed that smooth B. abortus was rapidly internalized, achieving a maximum level in less than 5 min without additional uptake over the next 30 min. In contrast, continued uptake of the rough mutant was observed and only achieves a maximum level after 30 min. The results were confirmed by the differences in F-actin polymerization, lipid raft staining, early endosome colocalization and electron microscopic observations after smooth and rough Brucella infection. We also demonstrated for the first time that uptake of S2308, but not rough mutant CA180 was PI3-kinase and toll-like receptor 4 (TLR4) dependent. Differences in uptake were associated with differences in macrophage activation with regard to NF-kappaB translocation and cytokine production. These results provide evidence that the presence of B. abortus OPS dictates the interactions between Brucella and specific cell surface receptors minimizing macrophage activation and enhancing Brucella survival and/or persistence.

Fig. 1

B. abortus invasion dynamics and electron micrographs in murine macrophages. The invasion of B. abortus smooth strain S2308 and the rough derivative CA180 was monitored over time using double antibody staining and electron microscopy. Percentage of the cells infected at indicated time points (A). Average internalized bacteria per infected cell (B). A ratio (percent) of internalized bacteria to cell-associated bacteria (C). The results are means ± SD of three independent experiments. A ratio (percent) of internalized bacteria to cell-associated bacteria at early stages (less than 5 minutes) of infection (D). The data is a representative of two experiments with similar results. Scanning electron micrograph of macrophages infected with S2308 and fixed at 5 minutes p.i. (E), or infected with CA180 and fixed at 5 minutes (F) or 20 minutes p.i. (G). Transmission electron micrograph of macrophages infected with S2308 (H, I) and CA180 (J, K) and fixed at 5 min p.i. Bar = 1 μM.

Fig. 2

F-actin polymerization and early endosome marker colocolization during B. abortus invasion. J774.A1 macrophages cultured on coverslips were infected with S2308 and CA180 at an MOI of 10 and fixed with 3% (w/v) paraformaldehyde at selected time points. The bacteria (red) were stained with mouse anti Brucella serum and detected with donkey anti mouse IgG Alexa Fluor 594. F-actin (A) or EEA1 (C) was detected as described in the Materials and Methods. The dynamics of Brucella colocalization with F-actin (B) or EEA1 (D) were revealed in time course. The insert shows the dynamics in early stage of the infection (5 minutes or less). The data shown is a representative of two experiments with similar results.

Fig. 3

Brucella smooth strain invades macrophages using lipid rafts as revealed by GM1 colocalization. J774.A1 macrophages cultured on coverslips were infected with S2308 and CA180 at an MOI of 10 and fixed with 3% (w/v) paraformaldehyde at various time points. (A) The bacteria (red) were stained with mouse anti Brucella serum and detected with donkey anti mouse IgG-Alexa Fluor 594. Ganglioside GM1 (green) was detected by cholera toxin B subunit-FITC staining. (B) Dynamics of Brucella colocalization with GM1. The result is a representative of two experiments with similar results.

Fig. 4

PI3-kinase in the uptake of Brucella in macrophages. J774.A1 was treated with 100 nM wortmannin for 30 minutes before infection and continuously thereafter. Gentamicin protection assay was used to monitor the uptake of S2308 and CA180 in the presence and absence of wortmannin. The data shown represent the means ± SD of three independent experiments. □p<0.05 when compared with untreated control.

Fig. 5

TLR4 is important for smooth Brucella uptake in murine macrophage. BMDMs cultured from (A) TLR4 deficient (HeJ) or (B) TLR4 sufficient control (OuJ) mice were infected with smooth and rough Brucella at an MOI of 100. Uptake of the bacteria was determined using gentamicin protection assay. The results are representatives of four experiments

Fig. 6

Rough Brucella infection induces macrophage activation. J774.A1 macrophages cultured on coverslips were infected with CA180 (A) or S2308 (B) at an MOI of 100. Control cells were treated with 100 ng/ml E. coli LPS (C) or left untreated (D). The cells were fixed with 3% (w/v) paraformaldehyde 1 h p.i. and detected for p65 (red staining) translocation and Brucella (green) using IX70 fluorescence microscopy (Olympus) after immunofluorescence staining. TNF-α (E) and IL1-α (F) were detected using sandwich ELISA after macrophages were infected with smooth and rough B. abortus at an MOI of 100. The supernatants were collected at 8, 24, 48 and 72 h p.i. The results (mean ± SD) are representatives of three experiments with similar results.