The Type III Secretion Effector CteG Mediates Host Cell Lytic Exit of Chlamydia trachomatis

Abstract

Chlamydia trachomatis is an obligate intracellular bacterium causing ocular and urogenital infections in humans that are a significant burden worldwide. The completion of its characteristic infectious cycle relies on the manipulation of several host cell processes by numerous chlamydial type III secretion effector proteins. We previously identified the C. trachomatis CteG effector and showed it localizes at the host cell plasma membrane at late stages of infection. Here, we showed that, from 48 h post-infection, mammalian cells infected by wild-type C. trachomatis contained more infectious chlamydiae in the culture supernatant than cells infected by a CteG-deficient strain. This phenotype was CteG-dependent as it could be complemented in cells infected by the CteG-deficient strain carrying a plasmid encoding CteG. Furthermore, we detected a CteG-dependent defect on host cell cytotoxicity, indicating that CteG mediates chlamydial lytic exit. Previous studies showed that Pgp4, a global regulator of transcription encoded in the C. trachomatis virulence plasmid, also mediates chlamydial lytic exit. However, by using C. trachomatis strains encoding or lacking Pgp4, we showed that production and localization of CteG are not regulated by Pgp4. A C. trachomatis strain lacking both CteG and Pgp4 was as defective in promoting host cell cytotoxicity as mutant strains lacking only CteG or Pgp4. Furthermore, CteG overproduction in a plasmid suppressed the host cell cytotoxic defect of CteG- and Pgp4-deficient chlamydiae. Overall, we revealed the first chlamydial type III secretion effector involved in host cell lytic exit. Our data indicates that CteG and Pgp4 participate in a single cascade of events, but involving multiple layers of regulation, leading to lysis of host cells and release of the infectious chlamydiae.

Keywords: Chlamydia trachomatis; effectors; host-pathogen interactions; pathogen egress; type III secretion.

Figure 1

A C. trachomatis cteG::aadA insertional mutant is defective in progeny generation. (A) Schematic representation of the genomic region of ctl0360/cteG (light blue), which was disrupted by insertion of a modified group II intron (grey) carrying a spectinomycin-resistance gene, aadA (yellow) to generate a C. trachomatis cteG mutant strain (cteG::aadA) (Pais et al., 2019). The cteG::aadA mutant strain was transformed with plasmids encoding CteG fused to a double hemagglutinin tag (2HA; red; pCteG-2HA; also named pCteG-2HA[Pgp4⁺] in Table 1 and in Figures 4 , 5 ), native CteG (pCteG) ( Supplementary Figure S1 ), or CteG and two (ctl0359/fabI and ctl0361; pFabI-CteG-CTL0361) or one (ctl0359/fabI; pFabI-CteG) of its flanking genes (dark blue). (B) Two identical tissue culture plates seeded with HeLa cells were infected with C. trachomatis L2/434, cteG::aadA mutant and complemented (cteG::aadA harboring a plasmid encoding CteG-2HA, also named pCteG-2HA[Pgp4⁺] in Table 1 and in Figures 4 , 5 ) strains at a MOI of 0.06. In one plate (input), the IFUs obtained in a primary infection were quantified at 24 h p.i. by immunofluorescence microscopy after fixation and immunolabelling of the chlamydiae; in the second plate (output), cells were lysed at 40 h p.i. and the number of released infectious particles were quantified after infecting for 24 h a new plate seeded with HeLa cells followed by fixation, immunolabelling of the chlamydiae, and immunofluorescence microscopy. For each strain, the relative progeny generation was obtained by dividing the number of IFUs in the output by those in the input. See Materials and Methods for a detailed description of the procedure. (C) cteG::aadA mutant strains harboring pFabI-CteG or pFabI-CteG-CTL0361 (see Panel A) were assessed in terms of infectious progeny generation as in (B) by comparison with the parental (L2/434) and mutant (cteG::aadA) strains. Data in (B, C) correspond to the mean ± standard error of the mean (n=3). Statistical significance was determined by using ordinary one-way ANOVA and Dunnett post-test analysis relative to the L2/434 strain (*p<0.05; **p<0.01; ***p<0.001).

Figure 2

C. trachomatis displays a CteG-dependent defect in egress from infected host cells. HeLa 229 cells were infected with C. trachomatis parental (L2/434), mutant (cteG::aadA), and complemented (cteG::aadA harboring a plasmid encoding CteG-2HA; also named pCteG-2HA[Pgp4⁺] in Table 1 and in Figures 4 , 5 ) strains at an MOI of 0.06 for 48, 72 or 96 h. At each time post-infection (p.i.), cell supernatants were collected (supernatant fraction) and adherent cells were lysed by osmotic shock to recover intracellular chlamydiae (lysate fraction). Fresh layers of HeLa cells were infected with serial dilutions of both supernatant (A) and lysate (B) fractions to quantify the number of recoverable inclusion-forming units (IFUs/mL). Data correspond to the mean ± standard error of the mean (n≥3). For each time point, statistical significance was determined by using ordinary one-way ANOVA and Dunnett post-test analysis relative to the L2/434 parental strain (ns, non-significant; *p<0.5; **p<0.01, ***p<0.001). For statistical analysis, natural logarithm was applied to data to ensure normality of the populations. (C) HeLa cells were left non-infected (N.I.), or were infected for 48 h with C. trachomatis L2/434, cteG::aadA or cteG::aadA harboring pCteG-2HA (also named pCteG-2HA[Pgp4⁺] in Table 1 and in Figures 4 , 5 ) at a MOI of 0.06. The proteins in the supernatant fraction (containing extracellular bacteria) were analyzed by immunoblotting with an antibody against C. trachomatis Hsp60 and the lysate fraction (intracellular bacteria) was analyzed by immunoblotting with antibodies against C. trachomatis Hsp60 and human α-tubulin (cell loading control), and using SuperSignal West Pico detection kit (Thermo Fisher Scientific) to detect proteins in the lysate fraction or SuperSignal West Femto detection kit (Thermo Fisher Scientific) to detect proteins in the supernatant fraction. Bands were quantified using Fiji software, and the Hsp60 signal in the supernatant fraction (Sup) was normalized to that in the lysate fraction (Lys) and to tubulin signal (Tub). Bars correspond to mean ± standard error of the mean (n=3).

Figure 3

C. trachomatis displays a CteG-dependent defect in host cell lysis. HeLa 229 cells were infected with C. trachomatis parental (L2/434), mutant (cteG::aadA), and complemented (cteG::aadA harboring a plasmid encoding CteG-2HA; also named pCteG-2HA[Pgp4⁺] in Table 1 and in Figures 4 , 5 ) strains at a MOI of 0.3. (A) At 48, 72 or 96 h post-infection (p.i.), cells were fixed with methanol, immunolabelled with antibodies against C. trachomatis Hsp60 (red) and appropriate fluorophore-conjugated secondary antibodies, and stained with DAPI (host cell nuclei and chlamydial inclusions; blue) and with fluorophore-conjugated phalloidin (host actin cytoskeleton; green). Scale bar, 40 µm. (B) At 48, 72, and 96 h p.i., the release of host lactate dehydrogenase (LDH) into the supernatant of infected HeLa cells was measured using a CytoScan™ LDH Cytotoxicity Assay kit (G-Biosciences). Data are representative of five independent experiments and correspond to the mean ± standard error of the mean of three biological replicates. For each time point, statistical significance was determined by using ordinary one-way ANOVA and Dunnett post-test analysis relative to the L2/434 parental strain (ns, non-significant; *p<0.05; **p<0.01; ***p<0.001).

Figure 4

Pgp4 does not modulate the production or the localization of CteG during C. trachomatis infection. HeLa cells were either left non-infected (N.I.) or infected with C. trachomatis cteG::aadA strains carrying pCteG-2HA[Pgp4⁺] (also named pCteG-2HA in Table 1 and in Figures 1 – 3 ; Pgp4⁺/CteG-2HA⁺) or pCteG-2HA[Pgp4^-] (Pgp4^-/CteG-2HA⁺) (A) At 16, 24, 30 or 40 h post-infection (p.i.), whole cell extracts were prepared and then analyzed by immunoblotting with antibodies against HA (CteG-2HA), C. trachomatis Hsp60 (bacterial loading control) and human α-tubulin (HeLa cell loading control), and using SuperSignal West Pico detection kit (Thermo Fisher Scientific) to detect Hsp60 or α-tubulin, or SuperSignal West Femto detection kit (Thermo Fisher Scientific) to detect CteG-2HA. The band corresponding to full-length CteG-2HA is indicated with an arrow. Zooms of the band pattern of CteG-2HA species at 30 h p.i. in both pgp4⁺ and pgp4^- backgrounds are displayed. The intensity of all bands on each lane was quantified using the software Fiji and summed to obtain the intensity of all CteG-2HA species at a given time point. Each value was normalized to the bacterial and HeLa cell loading controls. Bars correspond to mean ± standard error of the mean (n=3). (B) Cells were fixed with PFA 4% (w/v) at 24 or 40 h p.i. and immunolabelled with antibodies against C. trachomatis major outer membrane protein (MOMP; blue), cis-Golgi network (GM130; green) and HA (CteG-2HA; red), and appropriate fluorophore-conjugated secondary antibodies. Scale bar, 10 µm.

Figure 5

C. trachomatis mediates host cell lysis via a common pathway involving both CteG and Pgp4. (A) HeLa cells were infected with C. trachomatis L2/434 carrying pVector[Pgp4⁺] (CteG⁺/Pgp4⁺), L2/434 carrying pVector[Pgp4^-] (CteG⁺/Pgp4^-), cteG::aadA carrying pVector[Pgp4⁺] (CteG^-/Pgp4⁺), or cteG::aadA carrying pVector[Pgp4^-] (CteG^-/Pgp4^-) for 72 h at an MOI of 0.3, and the LDH released by lysed host cells was measured using a CytoScan™ LDH Cytotoxicity Assay kit (G-Biosciences). (B) As in panel A, but HeLa 229 cells were also infected with cteG::aadA carrying pCteG-2HA[Pgp4⁺] (also named pCteG-2HA in Table 1 and in Figures 1 – 3 ; CteG-2HA⁺/Pgp4⁺), or cteG::aadA carrying pCteG-2HA[Pgp4^-] (CteG-2HA⁺/Pgp4^-). Statistical significance was assessed by using ordinary one-way ANOVA and Tukey post-test analysis. In (A, B), data correspond to mean ± standard error of the mean (n = 4 in panel A and n = 7 in panel B; ns, non-significant; *p<0.01; **p<0.0001). (C) Hypothetical model for the mode of action of CteG and Pgp4 in promoting host cell lysis. After CteG is produced by chamydiae in an inactive form (CteG_I), CteG is activated (CteG_A) in a Pgp4-dependent manner, which could occur within chlamydiae or after delivery of CteG into the host cell cytoplasm. Premature CteG_A-mediated host cell lysis is prevented by the action of an unknown inhibitory factor, which could be another chlamydial effector or a host cell factor. At late stages of infection, the effect of this inhibitor in CteG_A is alleviated by an unknown mechanism and host cell lysis is triggered.