Targeting eEF1A by a Legionella pneumophila effector leads to inhibition of protein synthesis and induction of host stress response

Abstract

The Legionella pneumophila Dot/Icm type IV secretion system is essential for the biogenesis of a phagosome that supports bacterial multiplication, most likely via the functions of its protein substrates. Recent studies indicate that fundamental cellular processes, such as vesicle trafficking, stress response, autophagy and cell death, are modulated by these effectors. However, how each translocated protein contributes to the modulation of these pathways is largely unknown. In a screen to search substrates of the Dot/Icm transporter that can cause host cell death, we identified a gene whose product is lethal to yeast and mammalian cells. We demonstrate that this protein, called SidI, is a substrate of the Dot/Icm type IV protein transporter that targets the host protein translation process. Our results indicate that SidI specifically interacts with eEF1A and eEF1Bgamma, two components of the eukaryotic protein translation elongation machinery and such interactions leads to inhibition of host protein synthesis. Furthermore, we have isolated two SidI substitution mutants that retain the target binding activity but have lost toxicity to eukaryotic cells, suggesting potential biochemical effect of SidI on eEF1A and eEF1Bgamma. We also show that infection by L. pneumophila leads to eEF1A-mediated activation of the heat shock regulatory protein HSF1 in a virulence-dependent manner and deletion of sidI affects such activation. Moreover, similar response occurred in cells transiently transfected to express SidI. Thus, inhibition of host protein synthesis by specific effectors contributes to the induction of stress response in L. pneumophila-infected cells.

Fig. 1. Toxicity of SidI to eukaryotic cells

A. SidI-mediated toxicity to yeast. Growth of yeast cells expressing sidI, sidI*, or sidI** from the galactose inducible promoter on glucose medium (upper panel) or on galactose medium (lower panel). A yeast strain harboring an empty vector was used as a control (vector). Serially diluted yeast cultures grown in glucose medium were spotted followed by incubation at 30 °C for 2 days (glucose) or 3 days (galactose). B. Untagged sidI* and sidI** was expressed in yeast. Total lysates of yeast cells grown in 2% galactose medium were probed with an anti-SidI antibody. C. Inhibition of mammalian cell proliferation by SidI. 293T cells on coverslips with about 80% confluence were transfected to express SidI (left panel) or the SidI* mutant (right panel) (transfection efficiency was about 80%); images were acquired from fixed samples 24 hours posttransfection. Bar: 50 μm. D. Relative viable cell number of samples transfected to express different alleles of sidI. 293T cells transfected to express the indicated proteins for 24 hours were lifted and viable cells were enumerated. Data were expressed in relative to samples transfected with the vector set at 100%. Similar results were obtained from at least three independent experiments each done in triplicate. **, p<0.004.

Fig. 2. Dot/Icm-mediated translocation and growth phase-dependent expression of SidI

A. Transfer of SidI by the Dot/Icm system between bacterial cells. Protein transfer was performed by using either Lp02(Dot/Icm⁺) or Lp03(dotA⁻) expressing a Cre::SidI fusion as the donor strains and a strain containing a floxed reporter as a recipient (Luo and Isberg, 2004). The transfer of Cre mediated by the effector SidF (Banga et al., 2007) was used as a positive control. Solid bars, donor strain L. pneumophila Lp02(Dot/Icm⁺); striped bars, donor strain L. pneumophila Lp03(dotA⁻). B-D. SidI-mediated translocation of the transfer-deficient SidCΔC100 into macrophages. Mouse macrophages were infected with bacteria expressing SidC, SidCΔC100 or SidCΔC100::SidI for one hour; L. pneumophila and SidC was differently labeled with distinctive antibodies followed by appropriate secondary antibodies as described (VanRheenen et al., 2006). At least 150 vacuoles in triplet samples were scored in each experiment. Representative images of vacuoles harboring sidCΔC100 (C) or sidCΔC100::sidI (D); bacteria were labeled in red and SidC was stained in green. E. Dot/Icm-mediated transfer of SidI measured by using the pertusis toxin Cya as a reporter. U937 infected with Lp02(Dot/Icm⁺) or Lp03 (dotA⁻) containing cya-sidI fusion plasmid were analyzed for cAMP production. The translocated effector SidJ (Liu and Luo, 2007) was used as positive control. Solid bars, L. pneumophila Lp02(Dot/Icm⁺); striped bars, L. pneumophila Lp03 (dotA⁻). F. SidI was translocated into host cells by the Dot/Icm system during infection. U937 macrophages were infected with relevant bacterial strains at an MOI of 5 for 6 hours and the collected cells were lysed with 0.2% saponin. Soluble proteins precipitated with methanol were separated by SDS-PAGE, transferred to nitrocellulose membranes. SidI in insoluble fraction was also probed (low panel). The cytosolic protein isocitrate dehydrogenase was probed as a control for bacterial cell lysis. Lanes: 1, wild type strain Lp02(Dot/Icm⁺); 2, Lp03(dotA⁻); 3, Lp02ΔsidI(Dot/Icm⁺); 4, total cell lysate of Lp02(Dot/Icm⁺). G-H. Expression of sidI is induced when L. pneumophila was grown to exponential phase. L. pneumophila cultures established by diluting stationary phase cells at 1:20 (starting OD₆₀₀=0.2) in fresh AYE broth were grown at 37°C with vigorous shaking. The growth of the bacteria was monitored by measuring the values of OD₆₀₀ at a 2-hour interval (G). At indicated time points (hours), equal amounts of cells (estimated by OD₆₀₀ values) were withdrawn and lysates of the samples were separated by SDS-PAGE; proteins transferred to nitrocellulose membranes were probed with anti-SidI antibodies (H); the sidI deletion mutant grown at OD₆₀₀=1.8 was loaded as a control (H last lane). The bacterial isocitrate dehydrogenase (ICDH) was detected as loading controls.

Fig. 3. SidI interacts with components of the mammalian translation elongation machinery

A. Components of the host protein synthesis machinery were retained by SidI. Affigel beads coated with GST (lane 1) or His₆-SidI (lane 3) were incubated with mammalian cell lysates. His₆-SidI coated beads incubated with cell lysis buffer (lane 2) was used as another control. After washing with lysis buffer, proteins separated by SDS-PAGE were visualized by silver staining; bands only retained by the His₆-SidI coated beads were identified by MALDI /mass spectrometry analysis. Relevant protein size markers (in kDa) were indicated. B. Co-purification of eEF1A and eEF1Bγ with GST-SidI from rabbit reticulocyte lysates (RRL). Each GST-tagged protein was incubated with RRL and glutathione coated beads were used to retrieve GST-SidI or GST-SidF. After removing unbound proteins by extensive washing, the retained proteins resolved by SDS-PAGE were detected by Coomassie bright blue staining and identified by mass/spectrometry analysis. Relevant markers (in kDa) were indicated. C. Direct binding of SidI to eEF1A and eEF1Bγ. His₆-SidI was incubated with GST-tagged eEF1A, eEF1Bγ or eEF1Bβ, and the protein complex was captured with glutathione beads, retained SidI was detected by immunoblotting. D. SidI* formed complexes with eEF1A or eEF1Bγ in mammalian cells. Lysates of 293T cells transfected to express GFP-SidI* or GFP-SidF were subjected to immunoprecipitation with an anti-GFP antibody, the precipitates resolved by SDS-PAGE were detected for eEF1A and eEF1Bγ using specific antibodies. 5% (50 μg) of total protein was probed as input controls (lanes 1 and 2). TCL: total cell lysates.

Fig. 4. Inhibition of protein synthesis by SidI

A. SidI abolished in vitro synthesis of SidF from plasmid pcDNA3-sidF. SidI or heat-inactivated SidI was added in an in vitro protein synthesis system (Promega) and the production of SidF was measured by immunoblotting. Lanes: 1, markers (kDa); 2, H₂O; 3, 180 ng SidI; 4, 540 ng SidI; 5, 540 ng heat-inactivated SidI. B. Dose-dependent inhibition of SidF production. Different amounts of SidI were added to RRL containing pcDNA3-sidF and the production of SidF was detected by Western blot. The levels of relevant proteins were monitored by immunoblotting with the appropriate antibodies. Actin was detected as a loading control (lowest panel). C. Inhibition of translation of luciferase mRNA. Fifty ng of GST (1), His₆-SidI (2), heat-inactivated His₆-SidI (3), His₆-SidI** (4) or His₆-SidI* (5) was added to RRL containing luciferase mRNA. Production of light was detected after adding the luciferase assay reagent. D. Dose-dependent inhibition of protein synthesis by SidI and its mutants. His₆-SidI or indicated mutants were added to RRL containing luciferase mRNA and the synthesis of luciferase was monitored by the production of light. E. Inhibition of in vivo protein synthesis by SidI. Seven hours after transfection, ³⁵S-methionine was added to 293T cells for 3 hours and the incorporation of S³⁵ was measured. 1, vector pEGFP-C1; 2, wild type sidI; 3, sidI*. Similar results were obtained from at least three independent experiments and the data shown are from one representative experiments done in triplicate. ***, p<0.0002; **, p<0.007; *, p<0.05.

Fig. 5. SidI binds eEF1A independent of guanine nucleotide it associates

A. SidI has a lower affinity for nucleotide free eEF1A. GST-eEF1A (lane 1), GST-eEF1A that had been dialyzed against 5 mM EDTA for 20 min. (lane 2) or GST (lane 3) was incubated with His₆-SidI, and the protein complex was captured with glutathione beads. Retained His₆-SidI was detected by immunoblotting. B. SidI binds eEF1A independent of the guanine nucleotides. Nucleotide free GST-eEF1A loaded with nonhydrolyzable GDPβS or (lane 1) or GTPγS (lane 2) was incubated with His₆-SidI and the protein complexes were captured and detected as described in A. The GST protein was used as a control (lane 3). In each case, 5% of the input was detected as controls. C. Mutants SidI*(E482K) and SidI**(R453P) binds eEF1A with affinities comparable to that of wild type protein. The indicated amounts of His₆-SidI or its mutants were added to wells of immulon-2 plates coated with eEF1A, SidI or the mutant proteins retained by eEF1A was detected by ELISA.

Fig. 6. SidI contributes to the activation of HSF1 during L. pneumophila infection

A. L. pneumophila infection-mediated activation of HSF1. Upper panel, cell lysates of U937 macrophages infected with the relevant bacterial strains at an MOI of 5 were subjected to immunoprecipitation with an anti-eEF1A antibody and the precipitated HSF1 was detected after SDS-PAGE. Lanes: 1, uninfected; 2, Lp03(dot/icm⁻); 3, Lp02(dot/icm⁺); 4, Lp02ΔsidI; 5, heat shock; 6, Lp02Δlgt1; 7, Lp02ΔsidI/psidI. Middle two panels: eEF1A and HSF1 was probed respectively to detect cellular levels of these proteins. Lower panel, quantitation of band intensity of upper panel relative to that of the heat shock treated cells set at 100%. B. Activation of HSF1 by SidI but not by the SidI* mutant. Immunoprecipitates were obtained with an eEF1A specific antibody from lysates of 293T cells transfected for 6 hours with the appropriate plasmids. HSF1 was detected as described in A. Lanes: 1, GFP; 2, SidI; 3, SidI*; 4, heat shock. C. Infection of L. pneumophila increases the DNA binding activity of HSF1. Lysates prepared from U937 macrophages infected with relevant bacterial strains were used for assay the binding of HSF1 to HSE. The probe was detected with an antibody specific for digoxigenin used to label the binding target. Lanes: 1, Uninfected; 2, Lp02(dot/icm⁺); 3, Lp03(dot/icm⁻); 4, heat shock; 5, Lp02ΔsidI; 6, Lp02ΔsidI/pSidI. Competition experiments received 5 times more unlabeled probe were performed to determine the specificity of the binding. D. Induction of hsp70. qPCR analysis for hsp70 was performed with cDNA derived from total RNA of U937 cells infected for 8 hours (Infection) or 293T cells transfected to express indicated genes for 6 hours (Transfection). The levels of expression were normalized with mRNA specific for the GAPDH gene. Relative gene expression represents the normalized values of the indicated samples against cDNA of uninfected cells or cells transfected to express GFP. Lanes: 1, uninfected; 2, Lp03(dot/icm⁻); 3, Lp02(dot/icm⁺); 3, Lp02ΔsidI; 5, heat shock; 6, Lp02Δlgt1; 7, Lp02ΔsidI/psidI; 8, vector expressing GFP; 9, GFP::SidI; 10, GFP::SidI*; 10, heat shock. Data shown are representatives of more than three independent experiments.