The Proteome of the Isolated Chlamydia trachomatis Containing Vacuole Reveals a Complex Trafficking Platform Enriched for Retromer Components

Abstract

Chlamydia trachomatis is an important human pathogen that replicates inside the infected host cell in a unique vacuole, the inclusion. The formation of this intracellular bacterial niche is essential for productive Chlamydia infections. Despite its importance for Chlamydia biology, a holistic view on the protein composition of the inclusion, including its membrane, is currently missing. Here we describe the host cell-derived proteome of isolated C. trachomatis inclusions by quantitative proteomics. Computational analysis indicated that the inclusion is a complex intracellular trafficking platform that interacts with host cells' antero- and retrograde trafficking pathways. Furthermore, the inclusion is highly enriched for sorting nexins of the SNX-BAR retromer, a complex essential for retrograde trafficking. Functional studies showed that in particular, SNX5 controls the C. trachomatis infection and that retrograde trafficking is essential for infectious progeny formation. In summary, these findings suggest that C. trachomatis hijacks retrograde pathways for effective infection.

Fig 1. Isolation of C. trachomatis inclusions from HeLa cells at 24 h p.i.

A) Workflow of the MACS purification procedure. Inclusions were isolated from C. trachomatis L2 infected cells. Different fractions of the purification procedure were analyzed by phase contrast microscopy. B) Visually intact inclusions were counted in different fractions (n = 3; error bars, SD). C) Immunoblot analysis of different steps of the inclusion purification procedure. In the first three fractions, equal volumes of approximately equal protein amounts were loaded, the MACS 5X lane contains 5X the number of inclusions contained in the infected cell lysate lane. The indicated proteins were detected with specific antibodies. The subcellular compartments indicate the primary location of the probed proteins. D) TEM of gradient purified and washed inclusions. IM, inclusion membrane; EB, elementary body; RB, reticulate body. E) Inclusions were gradient purified from C. trachomatis L2 infected HeLa cells transiently expressing Rab11A-eGFP, fixed, immunostained with IncA antibodies (red) and analyzed using LSCM. The DNA was stained with DAPI (blue). Scale bar, 2 μm; n = 2. See also S1 Fig.

Fig 2. Quantitative proteomics of isolated C. trachomatis inclusions.

A) Workflow of inclusion host proteome analysis. The scheme of three different peptides (1–3) measured by LC-MS/MS shows idealized SILAC ratios for peptides derived from different classes of proteins. (1) Ideal inclusion associated protein. Contaminations are expected to be identical for H and L labeling, corresponding to peptide (2). Proteins that are in part contaminations as well as inclusion associated have a lower L/H ratio (3). The enrichment of proteins in the inclusion fraction compared to the total lysate is calculated based on intensity based absolute quantification (iBAQ). B) The iBAQ enrichment score was calculated by dividing the proportional amount of protein specifically associated with the inclusion by the amount of this protein in the total lysate fraction. For more details see main text and S1 Text. C) Each protein identified and quantified in triplicate was plotted with its SILAC ratio (grey lines, lin/log plot). Known (green lines) and newly validated proteins (blue, yellow and red lines) are indicated. D) Distribution of annotated organellar proteins along the SILAC ratios. Mean SILAC ratios of proteins were pooled into bins of 0.125 (n = 1400). The total number of proteins mapping to the subcellular localization term was determined and the percentage in each bin plotted along against the SILAC ratios. Data points were connected for better visibility. SILAC ratios of 3 and above were pooled in the 3+ bin. Dashed line, approximate cutoff for enrichment. E) Validation of inclusion associated proteins using fluorescent fusion proteins. Confocal IF images showing HeLa cells expressing the indicated fluorescent fusion proteins (green), infected with C. trachomatis L2 (MOI 2). Cells were fixed 24 h p.i. and stained for inclusion membrane (IncA red) and DNA (DAPI, blue). Scale bar, 20 μm. F) Validation by purified inclusions. Inclusions were gradient purified from cells expressing the indicated fusion protein using a small scale protocol and analyzed by LSCM, DNA was stained with DAPI. Scale bar, 5 μm. For supplemental data see S2–S4 Figs.

Fig 3. Global analysis of the host cell derived inclusion proteome.

A) Proteins that were reliably found and quantified in the inclusion proteome or the total cell lysate were annotated with subcellular localization data from UniprotKB. Proteins were quantified according to their iBAQ intensity and the abundance of proteins annotated with the indicated term was summed. One protein can have annotations for several categories and organelles. B) Protein-protein interactions of inclusion associated proteins annotated with the highly enriched GO term `establishment of protein localization´. Connecting lines indicate interactions as reported by STRING database in standard settings. Color of the node represents the enrichment score, the color of border of the nodes are colored according to the SILAC ratio. Main clusters of interacting proteins are encircled with a green line and labeled I-IV. C) Proteins annotated with the GO term `vesicle mediated transport´ were further classified as involved in retrograde or anterograde transport (n = 35) and respective subcategories. Three proteins with incomplete GO annotation were added manually (Rab1B, Rab12, VPS29). Five proteins were annotated with both retrograde and anterograde transport pathways (LMAN1, Rab11B, Rab14, TMED10 and VAMP3). Two additional proteins were annotated with two subcategories (Stx7, VAMP8, both anterograde). E = endosome, ER = endoplasmic reticulum, GA = Golgi apparatus, EX = exocytosis, PM = plasma membrane, LY = lysosome, RE = recycling endosome. D) Enrichment of proteins at the inclusion calculated using iBAQ. Each protein is represented by a line colored according to its enrichment score. Black lines indicate the ranges of log2 transformed fold enrichment. Proteins of interest, white lines. For supplemental data see S15 Fig.

Fig 4. C. trachomatis infection shows SNX recruitment to the inclusion and leads to partial separation of the retromer subcomplexes.

A) Confocal IF images showing localization of retromer components during C. trachomatis L2 infection (MOI 1). HeLa cells were fixed 24 h p.i. and stained with indicated antibodies; DNA was stained with DAPI (blue). Scale bar, 5 μm; n = 3. B) Confocal IF images showing localization of eGFP fusion proteins of SNX5 and SNX6 recruited to the inclusion. HeLa cells were infected with C. trachomatis L2 (MOI 2) 4 h prior to transfection, fixed at 24 h p.i. and stained with indicated antibodies; DNA was stained with DAPI (blue). Scale bar, 10 μm; n = 2. C) Confocal IF images showing fibers positive for SNX1 and IncA in C. trachomatis L2 infected HeLa cells (MOI 2). Cells were fixed at 24 h p.i. and stained with indicated antibodies; DNA was stained with DAPI (blue). Images show a maximum intensity projection of a z-stack, arrows indicate SNX1/IncA fibers. Scale bar, 10 μm; n = 3. D) Confocal IF images showing co-localization of eGFP-SNX2 fusion protein with endogenous VPS35 in C. trachomatis L2 infected (Ctr L2, MOI 2) and uninfected (NI) HeLa cells. HeLa cells were infected 4 h prior to transfection, fixed at 24 h p.i. and stained with indicated antibodies; DNA was stained with DAPI (blue). Scale bar, 10 μm; n = 2. E) Immunoblots of retromer components during C. trachomatis L2 infection (MOI 2) at 24 h p.i. n = 2. For supplemental data see S5–S7 Figs.

Fig 5. Silencing of SNX5 promotes C. trachomatis infectious progeny formation.

A) Confocal IF images showing C. trachomatis L2 infected (MOI 0.5), siRNA-treated HeLa cells. Cells were transfected with siRNA pools against indicated genes or control siRNA (AllStars), fixed at 24 h p.i. and stained with indicated antibodies; DNA was stained with DAPI (blue). Scale bar, 50 μm; n = 2. B) Analysis of the primary infection in siRNA-treated HeLa cells from panel A. Inclusions and nuclei were microscopically counted in 5 fields of view and ratio was calculated (n = 2; error bars, SE). C) Reinfection assay assessing the effect of SNXs knockdown on infectious progeny formation 48 h p.i. n = 3; error bars, SE; *** indicates p value < 0.005. D) Graph showing the effect of SNXs knockdown on relative bacterial genome copy number in C. trachomatis L2 infected (Ctr L2, MOI 0.5) HeLa cells at 48 h p.i. Genome copy number was determined by qRT-PCR (n = 2; error bars, SE). For supplementary data see S8 Fig.

Fig 6. Retro-2 reduces infectious progeny formation of C. trachomatis.

A) to C) Reinfection assays assessing the effect of Retro-2 on infectious progeny formation. HeLa cells were infected with C. trachomatis L2 (MOI 2) and, at 8 h p.i., cells were either treated with indicated concentrations of Retro-2 or DMSO. Cells were harvested and inclusion forming units (IFU) per ml were determined A) at 48 h p.i. (n = 3; error bars, SE; *** indicates p value < 0.005) or B) at different time points after infection (n = 3; error bars, SE). C) C. trachomatis EBs were treated with 200 μM Retro-2 or DMSO as solvent control for 30 minutes at room temperature before pelleting, followed by washing and infection of HeLa cells at MOI 2 for 48 h (n = 4; error bars, SE). D) Graph showing the effect of Retro-2 on relative bacterial genome copy number in C. trachomatis L2 infected (MOI 2) HeLa cells at 24 h and 48 h p.i. Cells were infected and, at 8 h p.i., either treated with indicated concentrations of Retro-2 or DMSO as solvent control. Genome copy number was determined by qRT-PCR (n = 4; error bars, SE; *** indicates p value < 0.005). E) TEM of C. trachomatis L2 infected (Ctr L2, MOI 2) HeLa cells at 48 h p.i. treated with indicated concentrations of Retro-2 or DMSO. G = ghost, IB = intermediate body, RB = reticulate body, EB = elementary body. Scale bar, 1 μm (upper micrographs) and 500 nm (lower micrographs). F) Reinfection assay demonstrating the recovery of infectious progeny after treatment with Retro-2 and PenG. HeLa cells were infected and treated with the indicated compound from 8 h p.i (Retro-2) or 24 h (PenG) until 48 h p.i. Cells were washed with fresh infection medium and incubated until 96 h p.i. to allow for recovery of the bacteria (n = 3, error bars = SE). G) Reinfection assay assessing the effect of Retro-2 and SNX knockdown on infectious progeny formation. HeLa cells were transfected with siRNA pools against indicated genes or control siRNA (against Luciferase), infected with C. trachomatis L2 (MOI 2), treated with Retro-2 at 8 h p.i., harvested at 48 h p.i. for reinfection assay (n = 2; error bars, SE).