The Legionella pneumophila effector DrrA is sufficient to stimulate SNARE-dependent membrane fusion

Abstract

The intracellular bacterial pathogen Legionella pneumophila subverts host membrane transport pathways to promote fusion of vesicles exiting the endoplasmic reticulum (ER) with the pathogen-containing vacuole. During infection there is noncanonical pairing of the SNARE protein Sec22b on ER-derived vesicles with plasma membrane (PM)-localized syntaxin proteins on the vacuole. We show that the L. pneumophila Rab1-targeting effector DrrA is sufficient to stimulate this noncanonical SNARE association and promote membrane fusion. DrrA activation of the Rab1 GTPase on PM-derived organelles stimulated the tethering of ER-derived vesicles with the PM-derived organelle, resulting in vesicle fusion through the pairing of Sec22b with the PM syntaxin proteins. Thus, the effector protein DrrA stimulates a host membrane transport pathway that enables ER-derived vesicles to remodel a PM-derived organelle, suggesting that Rab1 activation at the PM is sufficient to promote the recruitment and fusion of ER-derived vesicles.

Figure 1. DrrA associates with PM syntaxins

GFP-tagged DrrA proteins and FLAG-tagged SNARE proteins were produced in HEK293-FcγRII cells and interactions were analyzed after the SNARE proteins from cells extracts were precipitated using anti-FLAG agarose. Immunoblot analysis using the antibodies indicated to the right of each blot show protein levels in the blots of the lysate (3% of input) and blots of the immunoprecipitate (IP). (A) Linear diagrams of the DrrA and syntaxin proteins with the bar on top indicating the position of amino acid residues that encompass structural domains in these proteins. The DrrA protein has an adenylyl transferase domain (ATase), a guanine nucleotide exchange factor domain (GEF) and a PM-localization domain containing a PI4P-binding region (PML). The syntaxin proteins have a Habc regulatory domain (H), a SNARE motif (S) and a transmembrane domain (TM). (B) The binding of GFP-DrrA_61-647 with the precipitated FLAG-tagged SNARE indicated on top of the panel was assessed by immunoblot analysis. (C) The binding of GFP-DrrA_451-647 with the precipitated FLAG-tagged SNARE indicated on top of the panel was assessed by immunoblot analysis. (D) The binding of GFP-DrrA_61-647 (left panels) and GFP-DrrA_451-647 (right panels) with the precipitated FLAG-tagged syntaxin proteins deleted of their transmembrane domain (ΔTMD) indicated on top of each panel was assessed by immunoblot analysis.

Figure 2. DrrA translocated into host cells during infection associates with PM syntaxins

FLAG-tagged syntaxin proteins produced in HEK293-FcγRII were infected with Legionella and precipitated using anti-FLAG agarose. Immunoblot analysis was used to determine the amount of DrrA translocated into cells during infection that coprecipitated with the FLAG-tagged syntaxins. (A) HEK293-FcγRII cells producing 3x-FLAG-Stx3 were infected with wild type Legionella (WT), a Dot/Icm-deficient mutant (ΔdotA) or a mutant deficient in the DrrA protein (ΔdrrA) as indicated above each lane. Intact intracellular bacteria were collected in the pellet after centrifugation of the cell lysate and the 3x-FLAG-Stx3 was precipitated from the cleared lysate using anti-FLAG agarose. Immunoblots show the amount of DrrA protein inside the intracellular bacteria (left panel). The top right anti-DrrA blot and the bottom right anti-FLAG blot show the amount of DrrA and 3xFLAG-Stx3 in the cleared host cell lysate (3% of input) and in the 3x-FLAG-Stx3 immunoprecipitate (IP), respectively. (B) HEK293-FcγRII cells producing the indicated 3x-FLAG-tagged SNARE proteins were infected with wild type Legionella and proteins were precipitated from a lysate containing host cytosolic proteins using anti-FLAG agarose. The amount of translocated DrrA in the cleared lysate (3% of input) and the amount of the 3x-FLAG-tagged protein and DrrA protein in the immunoprecipitate (IP) was determined by immunoblot analysis.

Figure 3. Rab1 regulates DrrA binding to the SNARE motif in Stx3

In vitro interactions between MBP-tagged syntaxin proteins and purified DrrA proteins were assayed using amylose agarose to pull down MBP protein complexes. Proteins in the reaction (15% of input) and in the MBP pull-down were separated by SDS-PAGE and detected using antibodies specific for either the His, GST or MBP tag as indicated on the right of each blot. (A) Purified Stx3 derivatives fused to MBP indicated on the top were incubated with purified His-tagged DrrA proteins indicated on the right. MBP alone was included as a control for non-specific interactions. His-SNAP23 was included as a positive control for functional SNARE interactions with Stx3. (B) Purified Stx3 and Stx5 proteins fused to MBP are indicated on the top were incubated with His-DrrA_451-647. (C) Purified His-DrrA_1-647, GST-Rab1_S25N or GST-Rab1_Q70L derivatives, and MBP-Stx3 _ΔHab/TMD were included in the in vitro reaction as indicated in the grid on top of the panel. (D) Purified His-DrrA_1-647, His-DrrA_340-647, GST-Rab1, and MBP-Stx3 _ΔHab/TMD were incubated either in the presence or absence of 0.5mM GTP in vitro as indicated in the grid on top of each panel. The MBP pull-down panel (left) shows proteins bound to the MBP-tagged fusions isolated on amylose beads. The GST pull-down panel (right) shows proteins bound to GST-tagged Rab1 isolated on glutathione beads. This control shows that DrrA_1-637 and DrrA_340-647 bound similarly to GST-Rab1 and that complex formation was less efficient in the presence of GTP. (E) Purified His-DrrA_1-647 and MBP-Stx3 _ΔHab/TMD were added to each reaction at an equal molar ratio (0.1 μM) and incubated with the indicated concentrations of purified GST-Rab1.

Figure 4. DrrA and Rab1 coordinate Stx3 interactions with Sec22b

The binding of purified MBP-Stx3 _ΔHab/TMD to vesicle-associated 3x-FLAG-Sec22b was assayed using amylose agarose to pull down MBP protein complexes. (A) Purified MBP-Stx3 _ΔHab/TMD was incubated in vitro with the purified His-DrrA derivatives, GST-Rab1 and nucleotides indicated in the grid above each lane. A PNS fraction containing vesicles displaying 3x-FLAG-Sec22b was incubated with each reaction. Proteins in the reaction (15% or 5% of input) and in the MBP pull-down were separated by SDS-PAGE and detected using antibodies specific for either the His, GST, FLAG or MBP tag as indicated on the right of each blot. (B) Blots from the experiment shown in panel A were scanned to determine relative amounts of 3xFLAG-Sec22b associated with MBP-Stx3 _ΔHab/TMD. Relative amounts of 3x-FLAG-Sec22b were calculated by dividing the signal intensity observed for reaction conditions indicated below each bar by the signal intensity in the control lane where MBP-Stx3 _ΔHab/TMD alone was used to pull-down 3x-FLAG-Sec22b on vesicles. Data represent the mean ± SEM of four independent experiments.

Figure 5. DrrA coordinates the functional pairing of Stx3 and Sec22b in permeabilized cells

Permeabilized HEK293-FcγRII cells producing the indicated GFP-Stx3 and GFP-DrrA proteins were incubated with vesicles displaying 3x-FLAG-Sec22b and protein complexes were precipitated using anti-FLAG agarose. (A) The GFP-DrrA and 3x-FLAG-Stx3 constructs expressed in the permeabilized HEK293-FcγRII cells and the inclusion of GTP or GDP in the reactions are indicated above each lane. Immunoblot analysis using antibodies specific for GFP, SNAP23, and FLAG show protein levels in the reaction (5% or input, left panels) and protein levels in the anti-FLAG immunoprecipitate (IP, right panels). Because SNAP23 has been shown to form a SNARE complex with Sec22b and Sec18, this interaction served as a positive control. (B) As in panel A, HEK293-FcγRII cells producing GFP-Stx3 and GFP-DrrA_61-647 were permeabilized and incubated with vesicles in a PNS fraction obtained from cells producing 3x-FLAG-Sec22b in the presence of GTP. Lysates of these reactions were incubated in the presence or absence of purified NSF and α-SNAP as indicated. Protein levels in the reaction (5% or input) and protein levels in the 3x-FLAG-Sec22b immunoprecipitate (IP) were determined by immunoblot analysis using antibodies specific for GFP, SNAP23, and FLAG.

Figure 6. DrrA alters ER-derived vesicle dynamics at the PM

Cells producing GFP-DrrA_61-647 were permeabilized and incubated with ER-derived vesicles containing RFP-KDEL. TIRF microscopy was used to image ER-derived vesicles associated with the PM in cells producing GFP-DrrA_61-647. (A) A series of time-lapse micrographs showing vesicle dynamics without the addition of fusion buffer was included to demonstrate the stability of vesicle tethering. Upon the addition of fusion buffer vesicle dynamics were imaged in cells that had not been treated with siRNA (no siRNA), in cells where siRNA was used to reduce the levels of the proteins Stx2, Stx3, and Stx4 in the permeabilized cells (acceptor), and in cells where siRNA was used to reduce the levels of Sec22b in the in the (donor) cells producing RFP-KDEL. Squares show the location of randomly chosen vesicles at the beginning and end of each series. Bar, 1μm. (B) Time-lapse video TIRF microscopy images captured as described in panel A were used to follow RFP-KDEL-positive vesicles tethered in the TIRF plane. Vesicles (RFP-KDEL puncta) that were tethered in the TIRF plane upon the addition of fusion buffer (t=0 sec) were followed over time (30 sec). Data show the percent of vesicles present in the TIRF plane at t=0 sec that remained in the TIRF plane at t=30 sec, and represents the mean ± SEM of three independent experiments in which 100 vesicles were scored for each series. p*<0.05, p**<0.03 compared to the no siRNA (mock) control.

Figure 7. SNARE programming by DrrA and Rab1 promotes fusion of ER-derived vesicles with the PM

Tethering and fusion of ER-derived vesicles containing Luciferase-KDEL or ts045-VSVG-GFP with the PM of permeabilized HEK293-FcγRII cells was measured in vitro. (A) Permeabilized HEK293-FcγRII cells were incubated with purified His-DrrA protein either in the presence or absence of nucleotides as indicated. After the addition of a PNS fraction from cells producing Luciferase-KDEL, the tethering of ER-derived vesicles was assessed by determining the Luciferase activity associated with the permeabilized cells, which is presented as the percent of total Luciferase activity added to the reaction. (B) Permeabilized HEK293-FcγRII cells were incubated with purified His-DrrA_1-647 protein and a PNS fraction from cells producing Luciferase-KDEL was added. Tethering and fusion of ER-derived vesicles was assessed independently in parallel wells. The tethering of ER-derived vesicles was assessed by determining the Luciferase activity associated with the permeabilized cells, which is presented as the percent of total Luciferase activity added to the reaction (white bars, cell associated before fusion). Fusion was assessed by comparing the Luciferase activity that remained cell associated after the addition of fusion buffer (black bars) to the Luciferase activity that was liberated into the supernatant (gray bars, secreted). Shown are reactions in which siRNA was used to reduce the levels of the proteins Stx2, Stx3, and Stx4 (Stx2/3/4) or Sec22b in the permeabilized cells (acceptor), and reactions in which siRNA was used to reduce the levels of the proteins Stx2, Stx3, and Stx4 (Stx2/3/4) or Sec22b in the (donor) cells producing Luciferase-KDEL. These data are presented as a ratio of the total Luciferase activity that was associated with cells before fusion. Data are the means ± SEM of three independent experiments. *p<0.05 when compared to the same fraction in no RNAi samples; **p<0.01 when compared to the same fraction in Stx2/3/4 (acceptor) samples; ***p<0.01; compared to the same fraction in Sec22b (donor) samples. (C) Vesicles prepared from cells 15 min after ts045-VSVG-GFP was released from the ER were added to permeabilized control cells (no siRNA) or permeabilized cells in which PM syntaxins had been silenced (Stx2/3/4 KD). TIRF images show tethered ts045-VSVG-GFP-positive vesicles at the PM (before fusion) and the change in ts045-VSVG-GFP fluorescence that occurred after the addition of fusion buffer (after fusion). (D) A model that depicts data on how DrrA could promote the tethering and fusion of ER-derived vesicles with a PM-derived organelle. The first stage shows DrrA recruiting Rab1-GDP to the membrane and associating with Stx3. In the second stage the GEF activity of DrrA activates Rab1 on the PM-derived organelle, and active Rab1 is required for the association of ER-derived vesicles displaying Sec22b. A tethering factor that binds to active Rab1 is predicted to be involved. In the third stage Sec22b and Stx3 interactions result from vesicle tethering. In the final stage the fusion of ER-derived vesicles with the PM-derived organelle membrane is mediated by the assembly of a functional SNARE complex (SNAREpin) consisting of the v-SNARE Sec22b and the t-SNARE comprised of Stx3 and SNAP23.