The quorum sensing volatile molecule 2-amino acetophenon modulates host immune responses in a manner that promotes life with unwanted guests

Abstract

Increasing evidence indicates that bacterial quorum sensing (QS) signals are important mediators of immunomodulation. However, whether microbes utilize these immunomodulatory signals to maintain infection remain unclear. Here, we show that the Pseudomonas aeruginosa QS-regulated molecule 2-amino acetophenone (2-AA) modulates host immune responses in a manner that increases host ability to cope with this pathogen. Mice treated with 2-AA prior to infection had a 90% survival compared to 10% survival rate observed in the non-pretreated infected mice. Whilst 2-AA stimulation activates key innate immune response pathways involving mitogen-activated protein kinases (MAPKs), nuclear factor (NF)-κB, and pro-inflammatory cytokines, it attenuates immune response activation upon pretreatment, most likely by upregulating anti-inflammatory cytokines. 2-AA host pretreatment is characterized by a transcriptionally regulated block of c-JUN N-terminal kinase (JNK) and NF-κB activation, with relatively preserved activation of extracellular regulated kinase (ERK) 1/2. These kinase changes lead to CCAAT/enhancer-binding protein-β (c/EBPβ) activation and formation of the c/EBPβ-p65 complex that prevents NF-κB activation. 2-AA's aptitude for dampening the inflammatory processes while increasing host survival and pathogen persistence concurs with its ability to signal bacteria to switch to a chronic infection mode. Our results reveal a QS immunomodulatory signal that promotes original aspects of interkingdom communication. We propose that this communication facilitates pathogen persistence, while enabling host tolerance to infection.

Figure 1. 2-AA enhances survival following BI.

(A) Mice were injected with 2-AA (6.75 mg/kg mice) or PBS 6 h (n = 20), 2 d (n = 20), 4 d (n = 20), 8 d (n = 20), or 30 d (n = 20) prior to BI with PA14. The data shown are averages of two independent experiments. Significance of survival rate differences was determined using the Kaplan-Meier method, with a hazard ratio of 1.8932 (95% CI, 1.0664–6.0718). Infection (−) drastically reduced survival relative to (2-AA W/O BI) controls (p = 0.03). Delivery of 2-AA 4 d before BI (red) had a particularly powerful influence on survival versus mice not pretreated with 2-AA (p = 0.03). A less remarkable, but still significant, survival benefit was also observed in BI mice pre-exposed to 2-AA 6 h, 2 d, 8 d, or 30 d before BI (all p = 0.03 vs. non-infected 2-AA exposed controls). (B) Relative to the effects observed with 2-AA (n = 20), 4 d pretreatment with the 2-AA analogs 4-AA (n = 8; p = 0.03), 2-NA (n = 8; p = 0.03), or MA (n = 8; p = 0.03), or the 2-AA metabolite 3OH-2-AA (n = 8; p = 0.03) prior to PA14 infection had weak, though still statistically significant, positive effects on survival after infection. Significance of survival rate differences was calculated as in A. (C) Bacterial loads in the local muscle 7 d post-BI were significantly higher in mice pretreated with 2-AA 4 d before BI (n = 7) than in control mice subjected to BI without 2-AA pretreatment (n = 7; p<0.05, Kruskal-Wallis test). CFU data are presented on a log₁₀ scale. (D) CFU counts at the site of infection in mice 11 d postinfection. The 2-AA treated mice showed proliferation and higher counts than mice that were not treated with 2-AA. (n = 6; p<0.001, Kruskal-Wallis test). CFU data are presented on a log₁₀ scale.

Figure 2. 2-AA pretreatment modulates the pro-inflammatory response in vivo.

Multiplex ELISA showed that BI induced surges in serum levels of IL-1α, IL-1β, IL-4, IL-10, IFN-γ, and TNF-α 24 h post-BI, and that 2-AA pretreatment delivered 4 d before BI attenuated the surges in IL-1α, IL-1β, IL-4, IFN-γ, and TNF-α, while augmenting the surge in IL-10 (n = 4 per group). Mean values calculated from 2–4 replicate experiments are depicted with SD error bars. *p<0.05 vs. naïve; **p<0.05 vs. BI (Student's t test).

Figure 3. Histopathology of lung tissues after 2-AA treatment.

(A) Control healthy (non-infected) lung tissue 4 d after 2-AA treatment. (B) Inflammatory cell infiltration with large areas of consolidation in lung parenchyma 48 h after infection with PA14 (Black arrows indicate the infiltration and necrotic foci). (C) Lack of infiltration 48 h after PA14 infection in the lungs of mice pretreated with 2-AA 4 d prior to BI.

Figure 4. 2-AA pretreatment modulates activation of the NF-κB pathway in mouse macrophages.

(A) Schematic of 2-AA treatment. Macrophages were left untreated (No Pre) or pretreated with 0.8-mM 2-AA or 4-AA for 48 h (2-AA/4-AA Pre). The untreated and 2-AA pretreated cells were then stimulated with 0.2 mM, 0.4 mM, or 2.0 mM 2-AA (for experiment in B) or 4-AA (for experiment in C). (B) Pretreatment with 2-AA blocked NF-κB activation relative to cells not pretreated with 2-AA (0.8 mM). (C) NF-κB was activated by 2-AA analog 4-AA in 4-AA pretreated and not pretreated cells. Mean values calculated from 2–4 replicate experiments are depicted with SD error bars. (D and E) Following stimulation with 2-AA (0.4 mM), cellular extracts prepared from not pretreated and 2-AA pretreated macrophages. Western blots of I-κBα and I-κBβ degradation (D) and phosphorylation of NF-κB subunit p65 (E). Loading was normalized relative to mouse β-actin. (F and G) A TransAM NF-κB assay showed binding of NF-κB p65 and p50 with the NF-κB promoter in not pretreated and 2-AA pretreated cells following stimulation with 2-AA. Mean values calculated from three replicate experiments are depicted with SD error bars. (p<0.05, Student's t test).

Figure 5. 2-AA pretreatment alters the expression of pro- and anti-inflammatory cytokines upon 2-AA stimulation in macrophages.

Levels of TNF-α (A), IFN-γ (B), and TGF-β (C), following 6 h stimulation of 2-AA as measured by ELISA. The experiments were performed in triplicate and the results are expressed as means ± SD. (p<0.05, one-way ANOVA).

Figure 6. 2-AA pretreatment alters activation of the MAPKs and AP-1 in macrophages upon 2-AA stimulation.

Western blotting of cellular extracts with phospho-specific antibodies after 48 h pretreatment with 2-AA (0.8 mM) followed by stimulation with 0.4 mM 2-AA (A) p38 MAPK, (B) JNK1/2 and (C) ERK1/2. One representative experiment (out of three) is shown. Loading was normalized relative to mouse β-actin. A TransAM AP-1 transcription factor assay after 48-h pretreatment with 2-AA (0.8 mM) followed by stimulation with 2-AA, showing the binding of c-Fos (D) and c-Jun (E) to AP-1 promoter. Mean values calculated from three replicate experiments are depicted with SD error bars. (p<0.05, Student's t test).

Figure 7. Inhibition of p65 phosphorylation in 2-AA pretreated cells is accompanied by de novo formation of c/EBPβ-p65 complexes.

(A) Western blots of cellular extracts incubated with c/EBPβ from macrophages that had been incubated for 48 h with 0.8 mM 2-AA (2-AA Pre) or plain medium (No Pre) and subsequently stimulated with 0.4 mM 2-AA for the indicated time periods. (B) Western blot showing inhibition of ERK1/2 and c/EBPβ in 2-AA pretreated cells in the presence of MEK1 inhibitor PD98059 (1 µM, 5 µM, or 10 µM). Loading was normalized relative to mouse β-actin. (C) In cells treated as above, c/EBPβ-p65 complex formation monitored by IP followed by immunoblotting with anti-c/EBPβ or anti-p65 antibodies.

Figure 8. Transcription and NF-κB inhibitors can block the effects of 2-AA pretreatment.

Western blots showing phosphorylation of JNK1/2 in 2-AA pretreated or untreated cells along with CAPE (1.5 µM) (A), MG-132 (1 µM) (B), and actinomycin D (1 µM) (C) following 0.2 mM or 0.4 mM 2-AA stimulation. Loading was normalized relative to mouse β-actin. One representative experiment (of three) is shown for JNK1/2.

Figure 9. Proposed model for 2-AA immunomodulatory mechanisms.

In naïve cells (left), stimulation with 2-AA induces activation of NF-κB, which leads to the phosphorylation and degradation of I-κBα, releasing the NF-κB dimers p65 and p50. 2-AA also induces the p38 MAPK and JNK pathways to stimulate c-Jun and c-Fos. Activation of MAPK and NF-κB pathway upregulates pro-inflammatory genes. In contrast, in 2-AA pretreated cells (right) over-expression of ERK1/2 activates C/EBPβ, which binds directly to p65, resulting in c/EBPβ-p65 complex formation, and preventing 2-AA induced phosphorylation of p65 upon 2-AA stimulation. This interaction inhibits NF-κB mediated transactivation. The activation of JNK and p38 MAPK are repressed in 2-AA pretreated cells. All together, repression of the p38 MAPK, JNK, and NF-κB pathways abrogates the activation of pro-inflammatory mediators.