ZBP1 senses Brucella abortus DNA triggering type I interferon signaling pathway and unfolded protein response activation

Abstract

The innate immune system promptly detects and responds to invading pathogens, with a key role played by the recognition of bacterial-derived DNA through pattern recognition receptors. The Z-DNA binding protein 1 (ZBP1) functions as a DNA sensor inducing type I interferon (IFN) production, innate immune responses and also inflammatory cell death. ZBP1 interacts with cytosolic DNA via its DNA-binding domains, crucial for its activation. Brucella abortus is the etiologic agent of brucellosis in livestock and humans, leading to significant economic losses and public health impact. Despite other innate immune sensors that recognize B. abortus DNA, including Toll-like receptor 9 and the Stimulator of interferon genes (STING), here we evaluated the ZBP1 participation as a cytosolic receptor sensing Brucella infection. Using macrophages derived from ZBP1 knockout (KO) mice we demonstrated that ZBP1 partially contributes to IFN-β expression upon B. abortus infection or Brucella DNA transfection. The knockdown of STING by siRNA decreased the residual IFN-β signal elicited by B. abortus infection, demonstrating the presence of a redundant cytosolic DNA-sensing mechanism driving type I IFN production. Furthermore, ZBP1 is involved in type I IFN signaling inducing IRF-1 expression. Additionally, ZBP1 also contributes to Unfolded Protein Response (UPR) activation during infection. However, ZBP1 does not influence the production of proinflammatory mediators, inflammasome activation and it is dispensable to control bacterial infection in mice or replication in macrophages. This study highlights the complex interactions of Brucella components with innate immune receptors and identifies ZBP1 as a sensor for B. abortus DNA-induced IFN-β response.

Keywords: Brucella abortus; ZBP1; macrophage; type I interferon; upr.

Figure 1

B. abortus induces ZBP1 activation and type I interferon expression. Macrophages from wild-type (WT) or ZBP1 KO mice were infected with B. abortus (Ba) for 16 h and the IFN-β (A) and IFN-α (B) expression levels were determined by real-time RT-PCR. Non-infected cells (NI, control) were incubated under the same experimental conditions without bacteria. (C) Macrophages from wild-type (WT) or ZBP1 KO mice were stimulated with transfected B. abortus DNA for 16 h and the IFN-β expression levels were determined by real-time RT-PCR. Fugene alone (FUG) was used as control. (D) Macrophages from wild-type (WT) or ZBP1 KO mice were transfected with non specific siRNA (si CONTROL) or STING siRNA (si STING) for 2 days. Then, cells were infected with B. abortus for 16 h and the IFN-β expression levels were determined by real-time RT-PCR. The data (A-D) are presented as mean ± SD. (A-C), * (comparison between WT and KO), p < 0.05, two-way ANOVA. (D), * (comparison between WT and KO) or & (comparison between si CONTROL-treated and si STING-treated), p < 0.05, two-way ANOVA.

Figure 2

ZBP1 promotes the Brucella-induced UPR response. Macrophages from wild-type (WT) or ZBP1 KO mice were infected with B. abortus (Ba) for 16 h and the BiP (A) and XBP1(s) (B) expression levels were determined by real-time RT-PCR. Non-infected cells (NI, control) were incubated under the same experimental conditions without bacteria. The data (A, B) are presented as mean ± SD. (A, B), * (comparison between WT and KO), p < 0.05, two-way ANOVA.

Figure 3

ZBP1 participates in the type I interferon signaling elicited by Brucella. (A) Western blot analysis of IRF-1, BiP and ZBP1 in wild-type (WT) or ZBP1 KO macrophages lysates, non-infected (NI) or infected with B. abortus at 8 h or 16 h Equal loading was verified by measuring β-actin levels in the corresponding cell lysates. The densitometry analysis of Western blot of IRF-1 (B), BiP (C) and ZBP1 (D) were performed relative to β-actin. The data (B-D) are presented as mean ± SD. (B-D) * (comparison between WT and KO), p < 0.05, two-way ANOVA.

Figure 4

ZBP1 is not necessary for the production of proinflammatory cytokines and NO during B. abortus infection. The cytokines IL-12 (A), IL-6 (B) and TNF-α (C) produced by wild-type (WT) or ZBP1 KO macrophages, non-infected (NI) or infected with B. abortus at 8 h or 16 h, were detected in cell supernatants using ELISA. (D) NO₂ ⁻ (nitrite) accumulation in cell supernatants from wild-type (WT) or ZBP1 KO macrophages, non-infected (NI) or infected with B. abortus at 8 h or 16 h, were measured by Griess reaction. The data (A-D) are presented as mean ± SD. No statistical difference was observed (comparison between WT and KO), p < 0.05, two-way ANOVA.

Figure 5

ZBP1 is dispensable for inflammasome activation and cell death during B. abortus infection. (A) The cytokine IL-1β released by wild-type (WT) or ZBP1 KO macrophages, non-infected (NI) or infected with B. abortus at 8 h or 16 h, were detected in cell supernatants using ELISA. (B) Western blot analysis of wild-type (WT) or ZBP1 KO macrophages, non-infected (NI) or infected with B. abortus at 8 h or 16 h The protein pro-IL-1β was detected in cell lysates, and the active form of caspase-1 (p20 subunit) in supernatants. Equal loading was verified by measuring β-actin levels in the corresponding cell lysates. (C) Cell death was assessed by measuring LDH release in the supernatant of wild-type (WT) or ZBP1 KO macrophages, non-infected (NI) or infected with B. abortus (Ba) at the indicated time points. The data (A, C) are presented as mean ± SD. No statistical difference was observed (comparison between WT and KO), p < 0.05, two-way ANOVA.

Figure 6

ZBP1 is not required for the control of B. abortus infection. (A) Residual B. abortus CFU in the spleen of wild-type (WT) or ZBP1 KO mice infected intraperitoneally with B. abortus were determined at 2 and 4 weeks post-infection (wpi). Splenocytes from 4-week-infected mice were stimulated with B. abortus, 5 μg/ml ConA, 1 μg/ml LPS, or medium as a negative control. Supernatants from the splenocytes were harvested 48 or 72 hours after stimulation and analyzed by ELISA for TNF-α (B) or IFN-γ (C), respectively. (D) Macrophages from wild-type (WT) or ZBP1 KO mice were infected with B. abortus for 8 h or 16 h and the CFU assessed in cell lysates. The data (A-D) are presented as mean ± SD. No statistical difference was observed (comparison between WT and KO), p < 0.05, two-way ANOVA.