Fig 1.
In situ proteomic labeling of the C. trachomatis inclusion membrane.
(A) Plasmid used to transform C. trachomatis L2 and localize APEX2 to the inclusion membrane. IncB-APEX2 fusion expression was under the control of a tetracycline inducible promoter. Flag epitope tag was added between IncA and APEX2. (B) Schematic for APEX2 localization and biotinylation reaction. Cells were infected with C. trachomatis expressing IncB-APEX2 for 8, 16, or 24 hours, expression was induced by anhydrotetracycline, and incubation with biotin-phenol and hydrogen peroxide for 1-minute catalyzed biotinylation of proteins within 20 nm of APEX2. Biotinylated proteins were enriched using streptavidin coated agarose resin and relative abundance estimated using mass spectrometry-based proteomics. (C) Western blot analysis of cells infected for 16 hours with Chlamydia expressing IncB-APEX2 (first four columns) or untagged APEX2 (last column). Blots were probed with streptavidin-HRP to detect biotinylated proteins. Anti-beta-actin and anti MOMP antibodies were used as human and Chlamydia loading controls, respectively. BP, biotin-phenol; ATc, anhydrotetracycline. (D) Immunofluorescence microscopy of cells infected with C. trachomatis IncB-APEX2 and after 1 min inclusion membrane protein labeling at 8, 16, and 24 hpi. Representative images are shown. Biotin labeled proteins were identified by streptavidin-Alexa 488, Chlamydia were labeled with an anti-MOMP antibody, DNA labeled with DAPI. Single channel images are displayed in inverted grayscale. Merged panels display all three-color channels. Scale bars = 16 μm.
Fig 2.
Global analysis of the inclusion membrane interaction proteome throughout the C. trachomatis developmental cycle.
(A) Heatmap of inclusion membrane interacting proteome identified by APEX2 at 8, 16, and 24 hpi. Data from 6 replicate experiments per time point were averaged and compared against 6 replicate controls at similar times. Colors represent p-values, proteins not detected have no color. Enlarged section of heatmap shows proteins significantly enriched at all time points. Significance determined by t-test or g-test, p < 0.05. (B) C. trachomatis ORFs identified on the inclusion membrane, with p values displayed for their presence in 6 replicate samples for each time point. Locus tags highlighted in green represent annotated Inc proteins. Values highlighted in blue represent p values < 0.05. (C) KEGG pathway overrepresentation analysis of inclusion membrane proteome. Overrepresentation determined by hypergeometric algorithm with Benjamini Hochberg method for multiple test correction. Values highlighted in blue represent p values < 0.05. (D) Subcellular location enrichments of APEX2 identified proteins. Color intensity reflects the number of proteins for each location annotation pulled from the Human Protein Atlas database. Numbers in parentheses indicate the total number of reference proteins contained under that annotation in the Human Protein Atlas database. (E) Spatial distribution of inclusion interacting proteins from 8–24 hpi using Human Protein Atlas annotations and manual entry of Inc proteins. Color intensity indicates the number of proteins assigned to location annotations. A, actin filaments; Cy, cytosol; E, endosomes; ER, endoplasmic reticulum; Ex, extracellular/secreted; G, Golgi apparatus; IM, inclusion membrane; L, lysosomes; LD, lipid droplets; M, mitochondria; MT, microtubules; MTOC, microtubule organizing center; N, nucleus; P, peroxisomes; PM, plasma membrane; V, vesicles.
Fig 3.
Network analysis of C. trachomatis inclusion membrane interacting proteins.
(A) Interactions between host proteins identified by APEX2 proteomic labeling were obtained from StringDB [89], and visualization map was generated using R with the tidygraph package [86]. Colors represent more interconnected protein communities within the network. The interaction map in A represents the global extent of inclusion membrane interactions identified over the chlamydial developmental cycle. Edges between protein nodes represent interactions that were curated in StringDB either from published experiments or other databases. Proteins which do not have characterized interaction partners (i.e. single nodes) and translation related proteins were omitted for clarity. (B) Temporal dynamics of inclusion membrane protein interaction networks over the developmental cycle. Protein networks present in 8, 16, and 24 hpi proteomic data sets are represented by colored nodes. Gray nodes depict proteins present in the global interaction map (A) but absent from a specific stage of infection.
Fig 4.
RNAi validation of inclusion interacting proteins and comparison to previous data sets.
(A) IFU determination following RNAi depletion of 64 proteins identified from early (8 hpi) inclusions. Cells were transfected with siRNA corresponding to targets shown on the y-axis, infected with C. trachomatis L2, and harvested for IFU determination at 48 hpi. Bars mean(SD); n = 3; IFU shown relative to mean IFU of plate. Black squares next to RNAi targets (y-axis) indicate which time points the protein was enriched in APEX2 mass spectrometry data. Teal bars correspond to at least 1.5-fold increase/decrease compared to the mean, blue bars are over 2 fold increase/decrease compared to the mean. Significance was determined by one-way ANOVA with Dunnett’s multiple comparisons test, comparing to mean infectivity of plate; *, p < 0.05; **, p < 0.01; ****, p < 0.0001. (B) Summary of all proteins shared between APEX2 data and two previous inclusion mass spectrometry data sets [15,16]. Lower table describes all proteins that overlap between APEX2 and AP-MS Inc-specific interactions [15]. Shaded boxes represent significant enrichment found by APEX2. (C) Summary of APEX2 identified proteins with reported microscopy-based associations with Chlamydia inclusions. PMID entries refer to the publications used to provide this evidence.
Fig 5.
ERES proteins Sec16A and Sec31A are recruited to the C. trachomatis inclusion.
(A) Immunofluorescence microscopy of cells showing cellular distribution of Sec16A (anti-Sec16A, first column, green in merge), Chlamydia IncA (anti-IncA, second column, magenta in merge), and DNA (DAPI, third column, blue in merge). (B) Distribution of the COPII outer coat protein Sec31A (anti-Sec31A, first column, green in merge), IncA, and nuclei. Single channel images are displayed in inverted grayscale. Merged panels display all three-color channels. Protein distribution in uninfected HeLa cells are shown in the top row. Deconvolved images from cells infected with C. trachomatis L2 at 24 hpi are shown as a summation of z-series images (merged planes) or a single xy plane. Enlargements (far right) represent the regions marked with white boxes. Scale bars = 16 μm.
Fig 6.
Inhibition of ERES cargo loading abrogates ERES recruitment to the inclusion.
(A) Treatment of C. trachomatis infected cells with FLI-06 disrupted the recruitment of COPII coat protein Sec31A to inclusion membranes. Immunofluorescence microscopy of cells showing cellular distribution of Sec31A (anti-Sec31A, first column, green in merges), inclusion membrane protein IncA (anti-IncA, second column, magenta in merges), and DNA (DAPI, third column, blue in merges). Representative deconvolved merged z-series images of uninfected cells are shown in upper panels, and cells infected with C. trachomatis L2 at 24 hpi are shown in lower panels. 10 μM FLI-06 treatment for 4 h, from 20–24 hpi, resulted in Sec31A distribution similar to that of uninfected cells, and away from inclusion membranes. Scale bars = 16 μm. (B) Quantification of Sec31 in cells infected and treated with FLI-06 as described in A. Sec31 punctae that were touching or overlapping with IncA in untreated or FLI-06 treated cells were counted using Volocity to assess the number of overlapping spots per inclusion. At least 20 inclusions were analyzed per condition, for two independent experiments. Each dot represents an inclusion, lines represent mean (SD). Significance determined by unpaired t-test with Welch’s correction, ****, p < 0.0001.
Fig 7.
Inhibition of ERES cargo loading or specificity reduces C. trachomatis developmental growth.
(A) Experimental design used to test the impact of ERES disruption on Chlamydia developmental growth. Colored bars mark the times when FLI-06 or 4PBA was applied to infected cells. All cells were harvested for IFU determination at 48 hpi. The effects of ERES inhibition were determined for (B,D) primary infection, through measuring the diameters of inclusions at 48 hpi, or (C,E) IFU production at 48 hpi. Bars denote the mean (n = 3; SD); white bars correspond to untreated controls; gray bars are inhibitor treated in decreasing concentration, 10 μM, 5 μM, or 1 μM FLI-06 and 5 mM, 2 mM, 1 mM 4PBA. Colored bars on x-axis correspond to the inhibitor application key in A. Significance determined by one-way ANOVA with Dunnett’s multiple comparisons test, comparing to untreated control. *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. (F) Infected cells were treated from 24–48 hpi with 10 μM FLI-06, 5 mM 4PBA, or 0.5 μg/mL chloramphenicol (CAM). Genomic DNA was extracted, and genome copy number was determined by using quantitative PCR for the GroEL2 gene. Significance was determined by one-way ANOVA with Dunnett’s multiple comparisons test, comparing to control. **, p < 0.01; ***, p < 0.001. (G) Relative expression of genes upregulated during ER stress in infected or mock infected cells after treatment with 10 μM FLI-06, 5 mM 4PBA, or 1 μM thapsigargin from 20–24 hpi. Gray bars are infected, black are uninfected. Expression was determined using quantitative PCR. Significance was determined using a two-way ANOVA with Dunnett’s multiple comparisons test. There was no significant difference between infected and uninfected cells, thapsigargin treatment was significantly different from the control. Uninfected treatments were compared to uninfected control, infected treatments compared to infected control. **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.
Fig 8.
Membrane contact sites and ceramide uptake are maintained during inhibition of ERES.
Cells were infected with C. trachomatis, then transfected with a plasmid expressing pmScarlet-CERT (A, purple in merge) or pmClover3-PDI (B, green in merge). At 20 hpi, DMSO, 10 μM FLI-06, 5 mM 4PBA, or 3 μg/mL BFA were added to cells. At 24 hpi, cells were fixed and stained with anti-Sec31A (A, green in merge) or anti-IncA (B, purple in merge). DNA was stained with DAPI (blue in merge). Images are single planes from deconvolved z-series. Scale = 16 μm. (B) Inset shows regions along the inclusion membrane where PDI was closely apposed or overlapping IncA. (C) Average inclusion fluorescence after incubation with NBD-C6-Ceramide. Representative images shown on right, NBD-ceramide in green, DAPI in blue. Significance determined by one-way ANOVA with Tukey’s multiple comparison test. Any comparisons not shown were not significant. **, p < 0.01; ***, p < 0.001.
Fig 9.
RNAi depletion of ERES regulatory proteins Sec12 and cTAGE5 disrupt Chlamydia growth.
(A) Cells were treated with siRNA oligonucleotides and incubated for 48 hours, then infected with C. trachomatis L2. At 48 hpi, cells were lysed and infectious Chlamydia EB from each sample group were tested for IFU by infecting new cells. IFU values were compared to scramble siRNA treated, infected cells. Significance determined by one-way ANOVA with Dunnett’s multiple comparisons test, compared to control IFU. Bars, mean (SD);***, p < 0.001; ****, p < 0.0001; n ≥ 3. (B) Sec12 (anti-Sec12, green in merge) colocalizes with Sec31 (anti-Sec31A, purple in merge) in infected cells. Representative deconvolved merged z-series image. Scale bar = 16 μm. (C) Sec12 (anti-Sec12, green in merge) overlaps with IncA (anti-IncA, purple in merge) in a similar manner to Sec16 or Sec31. Single plane from deconvolved z-series image. Scale bar = 16 μm.