MiR-155 induction by F. novicida but not the virulent F. tularensis results in SHIP down-regulation and enhanced pro-inflammatory cytokine response

Abstract

The intracellular gram-negative bacterium Francisella tularensis causes the disease tularemia and is known for its ability to subvert host immune responses. Previous work from our laboratory identified the PI3K/Akt pathway and SHIP as critical modulators of host resistance to Francisella. Here, we show that SHIP expression is strongly down-regulated in monocytes and macrophages following infection with F. tularensis novicida (F.n.). To account for this negative regulation we explored the possibility that microRNAs (miRs) that target SHIP may be induced during infection. There is one miR that is predicted to target SHIP, miR-155. We tested for induction and found that F.n. induced miR-155 both in primary monocytes/macrophages and in vivo. Using luciferase reporter assays we confirmed that miR-155 led to down-regulation of SHIP, showing that it specifically targets the SHIP 3'UTR. Further experiments showed that miR-155 and BIC, the gene that encodes miR-155, were induced as early as four hours post-infection in primary human monocytes. This expression was dependent on TLR2/MyD88 and did not require inflammasome activation. Importantly, miR-155 positively regulated pro-inflammatory cytokine release in human monocytes infected with Francisella. In sharp contrast, we found that the highly virulent type A SCHU S4 strain of Francisella tularensis (F.t.) led to a significantly lower miR-155 response than the less virulent F.n. Hence, F.n. induces miR-155 expression and leads to down-regulation of SHIP, resulting in enhanced pro-inflammatory responses. However, impaired miR-155 induction by SCHU S4 may help explain the lack of both SHIP down-regulation and pro-inflammatory response and may account for the virulence of Type A Francisella.

Figure 1. SHIP expression is down-regulated in response to F.n. infection.

(A) PBM were infected with F.n. at an MOI of 50 for 24 hours. Cell lysates were analyzed by Western blotting use an anti-SHIP antibody in the top panel. ‘R’ designates resting/uninfected cells and ‘24 h’ designates infected with F.n. The lower panel is a reprobe of the same membrane with an anti-actin antibody. (B) THP-1 were infected with F.n. at an MOI of 100 for 24 hours. Cell lysates were resolved as in Figure 1A. (C) Western blots for SHIP in murine bone marrow-derived macrophages infected with F.n. at an MOI of 50 for 24 hours. Cell lysates were resolved as in Figure 1A. (D) Predicted interaction between miR-155 and the 3′UTR of SHIP (INPP5D) mRNA. (E) Normalized luciferase activity in cells transfected with the 3′UTR of SHIP (psiCHECK_INPP5D) or with vector alone (psiCHECK), and cotransfected with a control Renilla luciferase vector. Synthetic miR-155 or non-specific (scrambled) miRs were subsequently transfected at concentrations of 0, 10, 25 and 50 nM. Luminometer readings were taken 48 hours post-transfection. The graph represents f-luc expression normalized to r-luc expression, then normalized to percent maximal response.

Figure 2. miR-155 is induced in vitro and in vivo in response to F.n. infection.

(A–B) PBM were infected with F.n. for 24 hours at an MOI of 1, 10, or 50. Relative miR-155 (A) and BIC (B) expression were measured by qRT-PCR and then converted to fold change over uninfected. (C–D) PBM were infected with F.n. at an MOI of 50 for 4, 8, or 24 hours. Relative miR-155 (C) and BIC (D) expression were measured by qRT-PCR and then converted to fold change over uninfected. Graphs represent the mean and standard deviation of samples from three independent infections. (E) Mice were injected peritoneally with 200 CFU of F.n. or PBS, then euthanized at 24, 48, and 72 hours post-infection. Relative miR-155 expression was measured by qRT-PCR for the liver, lung and spleen and then converted to fold change over uninfected. An n of 6 was used for each timepoint. Graphs represent the mean and standard deviation. Data were analyzed by a paired Student t test. An asterisk (*) indicates a p-value<0.05.

Figure 3. Bacterial viability contributes to miR-155 induction through the TLR pathway.

(A) PBM were infected with live, heat-killed (HK), or paraformaldehyde-fixed (PFA) F.n. at an MOI of 50 for 24 hours, then miR-155 measured by qRT-PCR. (B) PBM were infected with live F.n. or heat-killed F.n. at an MOI of 50, then with or without gentamicin after 2 hours, then miR-155 was assayed by qRT-PCR. (C) Wild-type (WT) or MyD88^−/− bone marrow-derived macrophages were infected with F.n. at an MOI of 50 for 24 hours, then relative miR-155 expression measured by qRT-PCR and then converted to fold change over uninfected. (D) ELISA for TNFα secretion in supernatants from the samples in Figure 3C. (E) Wild-type (WT) and TLR2 (TLR2 Mut) signaling mutant bone marrow-derived macrophages were infected with F.n. at an MOI of 50 for eight hours and miR-155 was assayed by qRT-PCR. (F) WT and TLR2 Mut macrophages were infected at an MOI of 50 of F.n. for 5, 20, 40, and 60 minutes. Cell lysates were analyzed by Western blotting for pIKKα, pNFκBp65, and Actin. (G) PBM were pretreated for 30 minutes with inhibitors of PI3K (LY), ERK (UO), JNK (SP), p38 (SB), or with DMSO vehicle control, infected with F.n. at an MOI of 50 for 6 hours, then miR-155 was assayed by qRT-PCR. (H) PBM were pretreated with the IKK inhibitor BAY7085 (BAY) or DMSO vehicle control for 90 minutes, infected with F.n. at an MOI of 50 for 6 hours, then miR-155 was assayed by qRT-PCR. (I) ELISA for TNFα in supernatants from the samples in Figure 3F. All experiments were performed in triplicate, and each experiment performed three times. Graphs represent mean and standard deviation. Data were analyzed by a paired Student t test. An asterisk (*) indicates a p-value<0.05.

Figure 4. Effects of host-cell entry and inflammasome activation on miR-155 induction.

(A) PBM were pretreated for 30 minutes with 5 µg/ml cytochalasin D (CytoD), infected with F.n. at an MOI of 50 for 6 hours, then relative miR-155 expression measured by qRT-PCR. (B) CFU assays were conducted in parallel with the samples from Figure 4A. (C) Wild-type (WT) and caspase-1^−/− bone marrow-derived macrophages were infected with F.n. at an MOI of 50 for 8 hours and miR-155 was assayed by qRT-PCR. (D) ELISAs were done to measure IL-1β in supernatants from the samples in Figure 4C. Graphs represent the mean and standard deviation from three independent infections. Data were analyzed by a paired Student t test. An asterisk (*) indicates a p-value<0.05.

Figure 5. miR-155 promotes pro-inflammatory cytokine production during F.n. infection.

(A) PBM were transfected with vector control or miR-155 overexpression construct. 22 hours post-transfection, RNA was extracted and miR-155 over-expression was verified by qRT-PCR. (B) SHIP expression was assayed from the samples in Figure 5A. (C–D) 14 hours post-transfection PBM were infected with F.n. at an MOI of 50 for 8 hours and (C) TNFα and (D) IL-6 measured in supernatants. (E) Wild-type and miR-155^−/− macrophages were assayed for miR-155 expression. (F) SHIP expression was assayed by qRT-PCR from the samples in Figure 5E. (G) Wild-type and miR-155^−/− macrophages were infected with F.n. at an MOI of 50 for eight hours. IL-6 in the supernatant was measured by ELISA. All experiments were performed in triplicate. Graphs represent the mean and standard deviation of samples from three independent experiments. Data were analyzed by a paired Student t test. An asterisk (*) indicates a p-value<0.05.

Figure 6. Virulent F.t. elicits suboptimal miR-155 and pro-inflammatory cytokine responses.

(A–B) PBM (n = 7) were infected with F.n. or F.t. at an MOI of 100 for 24 hours. Relative miR-155 (A) and BIC (B) expression were measured by qRT-PCR. (C) PBM were infected with F.n. or F.t. at an MOI of 50 for 24 hours, then cell lysates probed for SHIP by Western blotting (top panel). The lower panel is a reprobe of the same membrane with anti-actin antibody. (D–E) PBM were infected with F.n. or F.t. at an MOI of 50 for 24 hours. TNFα (D) and (E) IL-6 in supernatants were measured by ELISA. Graphs represent the mean and standard deviation of samples from three independent infections. Data were analyzed by a paired Student t test. An asterisk (*) indicates a p-value<0.05.

Figure 7. Model of miR-155 induction and function during Francisella infection.

Francisella is recognized on the host cell surface by TLR2. The signal is transmitted through the adaptor protein MyD88. Subsequently MAPKs, PI3K and Akt are activated, which leads to enhanced NFκB activity, inflammatory cytokine production, and effective host response. SHIP negatively regulates the activation of Akt to prevent effective host response. During Francisella infection miR-155 is induced through the TLR signaling pathway, PI3K/Akt, ERK, JNK, and NFκB. MiR-155 induction in turn down-regulates SHIP to promote the activation of the PI3K/Akt pathway and inflammatory cytokine production. The highly virulent F. tularensis suppresses or subverts the induction of miR-155 in human monocytes, while the relatively avirulent F. novicida does not.