Fig 1.
(A) Crystallographic structure of the CTD of SidD, residues 350–507, shown in cartoon ribbon format in three orthogonal orientations. Nomenclature for helices continues the layout previously described for the N-terminal domain of SidD [25]. The zinc and magnesium ions are represented as grey and green spheres, respectively. (B) Final electron density map (2Fobs-Fcalc contoured at 1.5σ, blue mesh) within the area of the two metal ions showing the coordination sphere of the magnesium ion. (C) Crystallographic structure of SidD, residues 37–507, shown in cartoon ribbon format in two orientations. The N-terminal domain (NTD) is colored in slate and the CTD in yellow.
Fig 2.
SAXS-derived conformational assemblies of SidD37-507.
(A) Comparison of the experimental SAXS profile (blue circles) along the computed scattering from the crystallographic structure (grey line) and the profile calculated from the multi-state model obtained by the program MultiFox (orange line). The SAXS patterns are displayed as the logarithm of the scattering intensity (I) versus the momentum transfer (s). Inset: plot of the pair distance distribution function P(r). (B) Overlay of the DAMMIF-derived ab initio shape envelope with the crystal structure of SidD37-507. (C) MultiFoXS analyses of SidD37-507 SAXS data shows that two conformation ensembles are present in solution as seen from the radius of gyration (Rg) distribution. Cartoons represent the two selected conformers by MultiFoXS and their contribution percentage. (D) Superposition of the two MultiFoXS-selected conformers of SidD37-507 through their C-terminal domains (yellow) showing a large interdomain movement. The N-terminal domains are colored in green and slate for the conformers with 24% and 76% contribution, respectively.
Fig 3.
A hydrophobic loop within the CTD is required for SidD localization.
(A) Schematic representation of the position and composition of the loopCTD and alignment with SidD homologs from L. rowbothamii (Lr) and L. belliardensis (Lb). Hydrophobic residues within the loop region are colored in red. NTD, N-terminal domain; CTD, C-terminal domain. Numbers indicate amino acid positions. (B) Intracellular localization of CTD variants. Transiently transfected COS-1 cells producing either GFP (control) or GFP-CTD variants were fixed and stained using an antibody directed against the Golgi marker protein giantin (middle). Merged images show SidD proteins in green and giantin in red. Scale bar, 10 μm. Line scans (right) indicate pixel intensity of the green (GFP) and red (giantin) fluorescent signals along the dashed lines (distance in μm). (C) Quantification of (B) showing percentage of cells with SidD enrichment at the Golgi compartment. Numbers are results from at least 100 cells per sample and experiment. The graph represents the average of three biological replicates. (D) Effect of aromatic residue substitutions on intracellular localization of the CTD. Transiently transfected COS-1 cells producing either GFP (control, see panel B) or the indicated GFP-CTD mutants were chemically fixed, and stained for giantin to label the Golgi. The localization of CTD was evaluated by fluorescence microscopy. Amino acid substitutions at position F370, Y374, F376, and F377 were as follows: alanine (A), serine (S), tyrosine (Y), phenylalanine (F). Scale bar, 10 μm. (E) Quantification of (D) scoring cells with colocalization of GFP-CTD and giantin. The graph represents the average of at least 100 transfected cells from three biological replicates.
Fig 4.
The hydrophobic loop is crucial for liposome binding of SidD during liposome flotation.
(A) Illustration of the liposome flotation assay (left). (B) Coomassie blue-stained SDS-PAGE gel showing the binding of recombinant SidD37-507, but not SidDΔloop, to free liposomes. (C) Representative cryo-EM images of a control liposome (left), a liposome incubated with SidD37-507 (middle), and a liposome incubated with SidDΔloop (right). Scale bar, 50 nm. The plot below each image represents the corresponding cross-sectional electron density profile along the perimeter of the liposome.
Fig 5.
The carboxy-terminal helix bundle determines localization specificity.
(A) Schematic representation of CTD and its variants. Numbers indicate amino acid positions; asterisks represent residues altered by site-directed mutagenesis. The hydrophobic loop is shown in grey, and the region required for specific localization of CTD to Golgi membranes is highlighted in green. The intracellular localization of each CTD mutant (as shown in (C)) is summarized on the right. (B) Ribbon diagram of SidD-CTD (aa 322–450) colored in orange and the C-terminal helix-turn-helix bundle is colored in green. The relative position of the deAMPylation domain is shown in transparent slate. (C) Intracellular localization of CTD variants. Transiently transfected COS-1 cells producing the indicated GFP-CTD proteins (left) were chemically fixed and stained for giantin (middle). The localization of SidD relative to giantin is shown on the right. Scale bar, 10 μm. (D) Localization of CTD(322–450) to mitochondria membranes. Transiently transfected COS-1 cells coproducing GFP-CTD(322–450) or GFP-CTD(322–450; F370S) and Mito-RFP (a mitochondria marker) were chemically fixed, and the fluorescence signal was examined by confocal microscopy. Arrowheads indicate the position of membranes magnified in the insets. Scale bar, 10 μm.
Fig 6.
SidD localizes to LCVs via its loopCTD.
(A) GFP-CTD localizes to the surface of LCVs. Transiently transfected COS-1 cells producing indicated GFP-CTD variants were challenged with L. pneumophila for 2 hours. Intracellular bacteria were labeled using anti-Legionella-specific antibody followed by TexRed-conjugated secondary antibody. GFP-CTD localization was examined by fluorescence microscopy. White arrowheads indicate the position of bacteria magnified in the insets. Scale bar, 10 μm. (B) Quantification of (A) scoring SidD-decorated LCVs. Values are an average of at least 50 LCV compartments from three experimental replicates.
Fig 7.
Membrane association of SidD is critical for Rab1 deAMPylation.
(A) Rab1 dynamics on the LCV surface. Bone marrow macrophages challenged with the indicated L. pneumophila strains were chemically fixed at the indicated time points. Rab1 was detected by indirect immunolabeling using a Rab1B-specific antibody followed by secondary Alexa Fluor 488-conjugated antibody. L. pneumophila was stained with anti-Legionella-specific antibody and TexRed-conjugated secondary antibody. Cells were examined by confocal microscopy showing L. pneumophila in red and Rab1 in green. Scale bar, 1 μm. (B) Quantification of LCVs decorated with Rab1 analyzed under (A). At least 100 LCVs were counted per sample to determine the percentage of LCVs decorated with Rab1. The graph shows the average of three independent replications. ***P <0.001, **P <0.01.