PMC

(A) The efficiency of Rig-i knockdown was determined by quantitative RT-PCR under uninfected, viral and bacterial infected, and transfected conditions. Differences in Rig-i transcript levels were statistically significant (*, p<0.05, Students t-test) under resting, infected, and transfected conditions. (B) Rig-i is involved in the host type I interferon response to L. pneumophila RNA. Rig-i knockdown leads to reduced Ifnb expression upon transfection with DNase-treated L. pneumophila RNA. Quantitative RT-PCR was carried out 4 hours post stimulation. Control knockdown macrophages induced a statistically significant (*, p<0.05) higher level of Ifnb transcript in ΔflaA L. pneumophila and Sendai virus infected macrophages. No significant difference was found in response to untreated and RNase-treated L. pneumophila nucleic acids.

(A) The type I interferon response to L. pneumophila involves Ips-1 in vivo. Ips-1 ^−/− and heterozygous littermate mice were infected intranasally with 2.5×10⁶ LP01ΔflaA. Bronchoalveolar lavage with PBS was performed 20 hours post infection. Type I interferon levels in the bronchoalveolar lavage fluid (BALF) were analyzed by bioassay and recombinant IFNβ was used to determine a standard curve. A two-tailed t-test determined the differences in IFNβ levels were statistically significant (*, p<0.01, Student's t-test) upon comparison of Ips-1 ^+/− and Ips-1 ^−/− mice. (B) L. pneumophila colony forming units are not significantly different in Ips-1 ^+/− and Ips-1 ^−/−. Bronchoalveolar lavage fluid from infected Ips-1 ^+/− and Ips-1 ^−/− mice was centrifuged to isolate cells. Hypotonic lysis of cells was performed and CFU were plated on buffered yeast extract charcoal plates with antibiotic selection for L. pneumophila. A two-tailed t-test determined that CFU in Ips-1 ^+/− and Ips-1 ^−/− mice 20 hours post infection were not statistically significantly different (ns, p>0.5, Student's t-test).

(A) Inhibition of Pol III had no effect on Ips-1-dependent Ifnb induction by L. pneumophila. Ips-1 ^+/− and Ips-1 ^−/− macrophages were pretreated (controls were untreated) with 20 µM ML-60218 10 hours before infection with ΔflaA and ΔflaAΔsdhA L. pneumophila at an MOI of 1. Ifnb induction was analyzed by quantitative RT-PCR 4 hours post infection. Ifnb message was normalized to ribosomal protein rps17 levels. (B) L. pneumophila genomic DNA does not induce IFNB in HEK293T cells. HEK293T cells were pretreated, or left untreated, with 20 µM ML-60218 10 hours before transfection with 1.0 µg/ml pA:T, L. pneumophila genomic DNA, or salmon sperm DNA. Ifnb induction was analyzed by quantitative RT-PCR 4 hours post infection. Ifnb message was normalized to S9 levels. (C) WT (C57BL/6) macrophages were infected at an MOI of 0.01 in the presence or absence of 20 µM ML-60218 and growth of luminescent L. pneumophila strains was determined by RLU at 0, 24, 48, and 72 hours post infection. For inhibitor conditions, macrophages were pretreated with 20 µM ML-60218 10 hours before infection. Macrophages were infected with WT (LP02) L. pneumophila or with isogenic Δdot L. pneumophila (D) ΔflaA L. pneumophila (E).

(A) Type I interferon receptor signaling and caspase1-dependent pyroptotic cell death are not required for superinduction of Ifnb by the sdhA mutant. Bone marrow derived C57BL/6, Ifnar ^−/− and Casp1 ^−/− macrophages were infected with wild type, Δdot, and ΔsdhA L. pneumophila at an MOI of 1. Ifnb message was analyzed by qPCR from macrophage RNA harvested 4 hours post infection. (B) Caspase1-dependent pyroptotic cell death occurs independently of the type I interferon receptor. Bone marrow derived C57BL/6, Ifnar ^−/− and Casp1 ^−/− macrophages were infected with wild type, Δdot, and ΔsdhA L. pneumophila at an MOI of 1 and release of lactate dehydrogenase (LDH) in cell supernatants was measured 4 hours post infection. Specific cell lysis was calculated as a percentage of detergent lysed cells with spontaneous LDH release subtracted. No statistically significant difference was found between B6 and Ifnar ^−/− macrophages infected with ΔsdhA L. pneumophila (p>0.1, Student's t-test). ND, not detected.

(A) The 11C11 mutant harbors a transposon insertion in sdhA. The transposon insertion site is in the 3′ end of the open reading frame of the sdhA locus at nucleotide position 3421. (B) A clean deletion mutant of sdhA recapitulates the 11C11 transposon mutant and hyperinduces type I interferon. Bone marrow derived Myd88 ^−/− Trif ^−/− macrophages were infected with stationary phase L. pneumophila strains at a MOI of 1. Cell supernatants were harvested 8 hours post infection and assayed for type I interferon induction by an L929-ISRE luciferase bioassay. Type I interferon levels were determined by generating a standard curve with recombinant IFNβ An unmarked clean deletion of sdhA was compared to wild type, Δdot, the transposon mutant 11C11, and a triple deletion of sdhA and the two L. pneumophila paralogs, sidH and sdhB. Differences in IFNβ induction were statistically significant between WT L. pneumophila and the transposon mutant 11C11 (***, p<0.0005, Student's t-test). Differences between 11C11, ΔsdhA and ΔsdhAΔsdhBΔsidH were not statistically significant (ns, p>0.05, Student's t-test). (C) A clean deletion mutant of sdhA recapitulates the 11C11 transposon mutant and hyperinduces transcriptional activation of Ifnb. Bone marrow derived Myd88 ^−/− Trif ^−/− macrophages were infected with wild type, Δdot, ΔsdhA, 11C11, ΔsdhAΔsdhBΔsidH stationary phase L. pneumophila and transcriptional induction of Ifnb was analyzed by quantitative RT-PCR. (D) Complementation of the sdhA mutant results in loss of the Ifnb hyperinduction phenotype. MyD88 ^−/− Trif ^−/− BMDM were infected at an MOI of 1 with ΔflaA, Δdot, ΔflaAΔsdhA and ΔflaAΔsdhA L. pneumophila carrying vector or a plasmid expressing full length SdhA. Expression of Ifnb message was assessed by quantitative RT-PCR 4 hours post infection.

(A) Induction of interferon beta (Ifnb) by L. pneumophila is largely dependent on Ips-1. Bone marrow derived Ips-1^+/+, Ips-1 ^−/−, and Irf3 ^−/− macrophages were infected with wild type and Δdot L. pneumophila at a multiplicity of infection (MOI) of 1. Ifnb induction was analyzed by quantitative RT-PCR 4 hours post infection. Ifnb message was normalized to ribosomal protein rps17 levels. Differences in Ifnb transcript induction were statistically significant in Ips-1^+/+versus Ips-1 ^−/− macrophages (*, p<0.05) and Ips-1 ^−/− versus Irf3 ^−/− (*, p<0.05, Student's t-test) when infected with wild type L. pneumophila. (B) Induction of Ifnb by L. pneumophila is partially dependent on Mda5. Bone marrow derived Mda5 ^+/+ and Mda5 ^−/− macrophages were infected with wild type and Δdot L. pneumophila at a multiplicity of infection (MOI) of 1. Ifnb induction was assessed by quantitative RT-PCR as in (A). Differences in Ifnb induction were statistically significant (*, p<0.05, Student's t-test) between Mda5 ^+/+ and Mda5 ^−/− infected with wild type L. pneumophila. (C) Retroviral transduction of a Rig-i shRNA, but not the control shRNA, knocks down expression of Rig-i in MyD88 ^−/− Trif ^−/− immortalized macrophages. Stable transduction of MyD88 ^−/− Trif ^−/− immortalized macrophages was performed with a retroviral vector containing a control and Rig-i shRNA. Level of Rig-i knockdown was determined by quantitative RT-PCR under uninfected, infected, and poly I:C stimulation conditions. Differences in Rig-i transcript levels were statistically significant (*, p<0.05, Students t-test) under resting, infected, and ligand-stimulated conditions. (D) Rig-i is involved in the host type I interferon response to infection with L. pneumophila. Rig-i knockdown leads to reduced Ifnb expression in response to infection with ΔflaA L. pneumophila, as well as stimulation with poly I:C. Quantitative RT-PCR was carried out 4 hours post infection. Control knockdown macrophages induced a statistically significant (*, p<0.05) higher level of Ifnb transcript in response to ΔflaA L. pneumophila and poly I:C. No significant difference was found in uninfected or Δdot L. pneumophila infected macrophages. (E) Induction of Ifnb by L. pneumophila is independent of Zbp-1 (Dai). Bone marrow derived Zbp-1^+/+ and Zbp-1 ^−/− macrophages were infected with L. pneumophila strains and analyzed for Ifnb induction as in (A) and (B). Differences in Ifnb transcript levels between Zbp-1^+/+ and Zbp-1 ^−/− macrophages infected with L. pneumophila were not statistically significant (ns, p>0.1, Student's t-test).

(A) Hyperinduction of Ifnb by ΔsdhA L. pneumophila is largely dependent on Ips-1. Bone marrow derived Ips-1^+/+ and Ips-1 ^−/− macrophages were infected with wild type, Δdot, and ΔsdhA L. pneumophila at an MOI of 1. Ips-1^+/+ and Ips-1 ^−/− macrophages were transfected with 1.0 µg/ml poly I:C. 4 hours post infection and stimulation, macrophages were harvested and assessed for Ifnb induction as in . Ips-1^+/+ infected with WT L. pneumophila induced statistically significant higher levels of Ifnb transcript than Ips-1 ^−/− (**, p<0.005, Student's t-test). The same phenotype was seen in Ips-1^+/+ infected with ΔsdhA L. pneumophila (**, p<0.005) and transfected with poly I:C (*, p<0.05) when compared to Ips-1 ^−/−. (B) Hyperinduction of Ifnb by ΔsdhA L. pneumophila is partially dependent on Mda5. Bone marrow derived Mda5^+/+ and Mda5 ^−/− macrophages were infected with ΔflaA, Δdot, and ΔflaAΔsdhA L. pneumophila at an MOI of 1. Theiler's virus (TMEV) was overlaid onto Mda5^+/+ and Mda5 ^−/− macrophages. 4 hours post bacterial and viral infection, macrophages were harvested and assessed for Ifnb induction by qRT-PCR as in . Ifnb message was induced statistically significantly in Mda5^+/+ macrophages infected with ΔflaA L. pneumophila versus Mda5 ^−/− (*, p<0.05, Student's t-test). Mda5^+/+ also responded statistically significantly to ΔflaAΔsdhA L. pneumophila over Mda5 ^−/− (*, p<0.05, Student's t-test), while Theiler's virus elicited a robust Ifnb response from Mda5^+/+ not seen in Mda5 ^−/− (***, p<0.005, Student's t-test). (C) Retroviral transduction of a Rig-i shRNA, but not the control shRNA, knocks down expression of Rig-i. MyD88 ^−/− Trif ^−/− immortalized macrophages were stably transduced with retroviral vector containing a control and Rig-i shRNA. Level of Rig-i knockdown was determined by quantitative RT-PCR under uninfected and infected conditions. Differences in Rig-i transcript levels were statistically significant (*, p<0.05, Students t-test) under resting and infected conditions. (D) Rig-i is involved in the hyperinduction of type I interferon by ΔsdhA L. pneumophila. Rig-i knockdown leads to reduced Ifnb expression in response to infection with WT and ΔsdhA L. pneumophila, as well as Sendai virus. Quantitative RT-PCR was carried out 4 hours post infection. Control knockdown macrophages induced a statistically significant (*, p<0.05) higher level of Ifnb transcript in response to WT and ΔsdhA L. pneumophila and Sendai virus. No significant difference was found in uninfected or Δdot L. pneumophila infected macrophages.

(A) Purified genomic DNA and RNA from L. pneumophila induces type I interferon independently of MyD88 and Trif. Bone marrow derived Myd88 ^−/− Trif ^−/− macrophages were stimulated by transfection of 3.3 µg/ml purified L. pneumophila DNA, L. pneumophila RNA, pA:T (DNA), and pI:C (RNA). Nucleic acids were treated with DNase and/or RNase A before transfection. Macrophage supernatants were harvested 8 hours post stimulation and analyzed for IFNβ levels by L929-ISRE luciferase bioassay. IFNβ production by DNase-treated L. pneumophila RNA was statistically significantly higher compared to DNase-treated L. pneumophila DNA (**, p<0.005). In addition, RNase A-treated L. pneumophila RNA produced statistically significant lower levels of IFNβ (***, p<0.0001, Student's t-test) than L. pneumophila RNA and RNase-treated L. pneumophila DNA. (B) Genomic L. pneumophila DNA does not induce type I interferon in a Ips-1-dependent manner. Ips-1 ^−/− and heterozygous littermate bone marrow derived macrophages were stimulated by transfection of 1.0 µg/ml pA:T and purified genomic L. pneumophila DNA. Macrophages were infected with ΔflaA L. pneumophila at an MOI of 1. Sendai virus (SeV) was overlaid onto Ips-1 ^−/− and heterozygous littermate macrophages. Transcriptional activation of Ifnb was determined by quantitative RT-PCR as described in . (C) The viral RNA sensor Mda5 is not required for induction of type I interferon by L. pneumophila DNA. WT (C57BL/6) and Mda5 ^−/− bone marrow derived macrophages were stimulated by transfection of 1.0 µg/ml pA:T and purified genomic L. pneumophila DNA. Sendai virus (SeV) and Theiler's virus (TMEV) were overlaid onto WT and Mda5 ^−/− macrophages. Quantitative RT-PCR was used to determine Ifnb gene expression. (D) Non-CpG containing DNA (pA:T) does not induce Ips-1-dependent Ifnb at all concentrations tested. Ips-1 ^−/− and heterozygous littermate bone marrow derived macrophages were stimulated with a titration of pA:T by transfection of 10, 1.0, 0.1, 0.01 µg/ml pA:T. The difference between Ips-1 ^+/− and Ips-1 ^−/− macrophages transfected with pA:T was not statistically significant (ns, p>0.1, Student's t-test). Sendai virus (SeV) was overlaid onto Ips-1 ^−/− and heterozygous littermate macrophages (**, p<0.005). Cell supernatants were collected 8 hours post stimulation/infection. Induction of type I interferon was determined by L929-ISRE luc bioassay. Units are relative light units (RLU). (E) Genomic L. pneumophila DNA induces type I interferon independently of Ips-1 at all concentrations tested. Ips-1 ^−/− and heterozygous littermate bone marrow derived macrophages were stimulated with a titration of purified genomic L. pneumophila DNA by transfection of 10, 1.0, 0.1, 0.01 µg/ml L. pneumophila DNA. No statistically significant difference was found between Ips-1 ^+/− and Ips-1 ^−/− macrophages transfected with genomic L. pneumophila DNA (ns, p>0.1, Student's t-test). Sendai virus (SeV) was overlaid onto Ips-1 ^−/− and heterozygous littermate controls (**, p<0.005). Macrophage supernatants were collected 8 hours post stimulation/infection. Type I interferon levels were determined by L929-ISRE luc bioassay, units are relative light units (RLU).