C. trachomatis harbors the homolog of a Bacillus NlpC/P60 protein

Searching the genome of C. trachomatis we found the pathogen to encode an NlpC/P60 protein, that shows 29% sequence identity with Bacillus cereus YkfC (YkfC_Bce) and is hereinafter referred to as YkfC_Ctr (CT127). NlpC/P60 enzymes, including YkfC_Bce, are cysteine peptidases that have a catalytic triad comprising Cys, His, and a polar residue [31]. Besides the C-terminal NlpC/P60 domain, YkfC_Bce contains two SH3b domains. The active site is located at the SH3b1-NlpC/P60 interface, whereas the SH3b2 domain is distal to the active site. Residues from the SH3b1 and NlpC/P60 domain form the so-called S2-S1 binding site cavity that determines the enzyme’s substrate specificity [31]. Our in silico sequence and structure analyses for YkfC_Ctr revealed that all catalytic residues and the S2-S1 binding cavity are conserved (A and L Figs in S1 Text). Phyre² predicts that these active site residues also form a catalytic center closely resembling the B. cereus YkfC one (Fig 1A). The chlamydial homolog lacks membrane-spanning or lipoprotein-associated domains and a typical signal sequence for transport across the cytoplasmic membrane, suggesting a cytosolic localization of the enzyme. This may be supported by the identification of oligopeptide transporter system OppABCDF in C. trachomatis [30] which potentially transports PGN degradation-derived peptides into the chlamydial cell cytoplasm for further recycling as described for free-living bacteria.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. In silico analysis and activity of chlamydial YkfC in E. coli.

⁠ (a) Comparison of the 3D in silico models of the active site from YkfC_Ctr and YkfC_Bce as characterized by Xu et al., 2010 [31]. The Cys-His-His catalytic triad is shown in red, residues contributing to the S1 and S2 sites essential for substrate recognition in YkfC_Bce are highlighted in yellow and blue, respectively. Residues contributing to both sites are highlighted in green. Complementation of E. coli ΔmepS and E. coli ΔdapDΔmpl double mutant by chlamydial YkfC. (b) Spot plate assay of E. coli carrying a deletion for the gene encoding periplasmic NlpC/P60 endopeptidase MepS. The mutant was transformed with different pBAD24 constructs and grown at 42°C on low osmolarity medium containing 0.04% L-arabinose. Complementation of the osmo- and temperature-sensitive ΔmepS could only be observed for pBAD24 expressing E. coli mepS, but neither for ykfC_Ctr nor its active site mutant ykfC_Ctr (C172A). An empty vector control served as a negative control. (c) Growth effect in the E. coli ΔdapDΔmpl double mutant containing chlamydial ykfC constructs. Cells were grown in nutrient-rich medium containing 40 μM PGN tripeptide and protein production was induced with 0.4% L-arabinose. Cytoplasmic production of wild type YkfC_Ctr, but not its active site mutant, restored growth of the E. coli double mutant, which is auxotrophic for mDAP and unable to utilize PGN tripeptide under nutrient rich conditions. Error bars indicate ± s.d. (n = 3).

https://doi.org/10.1371/journal.ppat.1011047.g001

YkfC from C. trachomatis functions as a PGN tripeptide peptidase in E. coli

A number of the so far characterized cysteine-dependent NlpC/P60 proteins from free-living bacteria act as cell wall endopeptidases or peptide recycling enzymes [32]. To address these two possibilities for the C. trachomatis homolog, we tested if plasmid encoded YkfC_Ctr was able to restore growth in two different E. coli surrogate systems.

First, we investigated whether YkfC_Ctr functions as a cell wall endopeptidase that cleaves cross-bridges between glycan chains within PGN as has been shown for other members of the NlpC/P60 family [31,32]. In E. coli, NlpC/P60 protein MepS (formerly Spr) is a DD-endopeptidase which specifically cleaves D-Ala-mDAP cross-links within PGN, thereby allowing incorporation of new glycan strands during cell growth [33]⁠. E. coli mutants lacking MepS show a thermosensitive growth defect at low osmolarity [34]. We took advantage of this phenotype to test for possible PGN endopeptidase activity of YkfC_Ctr. Expression of neither wild type YkfC_Ctr nor the active site mutant YkfC_CtrC172A, in which the conserved Cys172 residue within the Cys-His-His triad was replaced by Ala, restored growth of an E. coli MepS mutant strain when grown at a non-permissive temperature and low osmolarity (Fig 1B). These results make a PGN endopeptidase activity unlikely for YkfC_Ctr and suggest a role in recycling of PGN-derived peptides.

To test for recycling activity of the chlamydial NlpC/P60 protein we used an E. coli ΔdapDΔmpl double mutant strain that reports on degradation of the PGN-derived tripeptide L-Ala-γ-D-Glu-mDAP through specific cleavage of the non-canonical γ-D-Glu-mDAP bond which is absent in proteins [35]. This strain is auxotrophic for mDAP and deficient in utilizing the PGN tripeptide. Auxotrophy for mDAP was achieved by deletion of dapD, which codes for a critical enzyme within the mDAP/L-Lys synthesis pathway. Direct incorporation of the tripeptide into newly synthesized PGN precursors was halted by deleting the gene mpl which encodes the ligase linking the tripeptide to the activated N-acetylamino sugar UDP-MurNAc, forming precursor UDP-MurNAc-L-Ala-γ-D-Glu-mDAP. We hindered the ΔdapDΔmpl double mutant from hydrolyzing tripeptide through its peptidase MpaA that releases mDAP from the peptide by growing the cells under nutrient rich growth conditions in our test system. Under these conditions, the E. coli ΔdapDΔmpl double mutant requires supplementation with mDAP for vital growth as expression of mpaA is diminished by the transcriptional regulator PgrR, affecting E. coli in hydrolyzing PGN tripeptide and thus in using it as a source of mDAP [35].

When incubated under nutrient rich growth conditions, the double mutant was severely impaired in using PGN tripeptide resulting in defective growth (Fig 1C, empty vector control). The expression of YkfC_Ctr restored growth, suggesting that YkfC_Ctr functions as a PGN tripeptide peptidase in E. coli and released sufficient mDAP from the peptide to promote vital growth (Figs 1C, B in S1 Text). To further investigate whether the observed growth promotion was connected to the catalytic activity of YkfC_Ctr, we expressed the active site mutant YkfC_CtrC172A under the same conditions. The active site mutant YkfC_CtrC172A was unable to restore growth in the surrogate system, indicating that the observed growth effect depended on the presence of catalytically active YkfC_Ctr (Fig 1C).

In sum, our observations suggest that YkfC_Ctr is a functional γ-D-Glu-mDAP peptidase when expressed in E. coli and that it operates in recycling of tripeptides released during PGN degradation.

YkfC_Ctr is active in vitro

To further characterize the chlamydial NlpC/P60 homolog at a molecular level we proceeded with biochemical analyses using the heterologously overproduced and purified enzyme (CA Fig in S1 Text). Consistent with our results from the surrogate test system, the purified enzyme used PGN tripeptide as a substrate. Our thin-layer chromatography (TLC) analysis showed that wildtype YkfC_Ctr was able to release mDAP from the PGN tripeptide L-Ala-γ-D-Glu-mDAP by hydrolyzing the γ-D-Glu-mDAP bond (Fig 2A and 2B). This non-protein bond, which is found in Gram-negative PGN and its degradation products, binds the lateral (γ)-carboxyl group of a D-amino acid (D-Glu) to the α-amino group of the non-proteinogenic amino acid mDAP [27]. The enzyme’s activity was time-dependent (CB Fig in S1 Text) and showed a pH optimum of 7.5–8.5 (Fig 2C). Consistent with our observation that the expression of YkfC_CtrC172A failed to rescue growth in the E. coli ΔdapDΔmpl surrogate system, the purified protein variant was inactive against L-Ala-γ-D-Glu-mDAP (Fig 2B). PGN recycling needs to be tightly controlled and product inhibition could provide a negative feedback. To investigate possible product inhibition, we also tested if mDAP inhibits the enzyme in concentrations comparable to the substrate concentration. mDAP did not show any inhibitory effect on the enzyme, suggesting negative feedback through this reaction product does not occur (CC Fig in S1 Text). As we could not completely rule out that the heterologous YkfC_Ctr produced in E. coli was contaminated by the zinc dependent host peptidase MpaA which is also capable of releasing mDAP from the PGN tripeptide, we investigated whether the metal chelator EDTA had any effect in our assay. In accordance with our active site mutant experiments that did not reveal background activity, we could not detect any inhibition of substrate turnover under these conditions, further attesting for high purity and the absence of interfering host peptidases in our protein purification (CD Fig in S1 Text).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. In vitro activity of YkfC_Ctr.

(a) Schematic representation of the proposed peptidase activity of YkfC_Ctr releasing mDAP from L-Ala-γ-D-Glu-mDAP tripeptide. (b) In vitro peptidase activity was analyzed by TLC using L-Ala-γ-D-Glu-mDAP tripeptide as a substrate. YkfC_Ctr, but not its active site mutant YkfC_CtrC172A, released mDAP from the tripeptide as shown by comparing migration of the reaction product and an mDAP standard. (c) pH dependence of YkfC_Ctr activity. The chlamydial YkfC homolog showed highest activity between pH 7.5 and 8.5 and was inactive at acidic pH 5.5.

https://doi.org/10.1371/journal.ppat.1011047.g002

Our results identified Ykfc_Ctr as a peptidase that cleaves mDAP from PGN tripeptide thus corroborating its role in recycling PGN-derived degradation products.

YkfC_Ctr is specific for PGN-derived peptides

To substantiate that YkfC_Ctr is dedicated to recycling of PGN degradation products, we screened different PGN-derived peptides and intermediates from the de novo PGN precursor biosynthesis pathway as potential substrates (Fig 3). Testing PGN-derived peptides, we found the enzyme to be active not only against the tripeptide L-Ala-γ-D-Glu-mDAP, but also against the tetrapeptide L-Ala-γ-D-Glu-mDAP-D-Ala, another typical degradation product released by PGN amidases such as chlamydial AmiA [20]. As evidenced by our MS analyses, cleavage in the L-Ala-γ-D-Glu-mDAP tripeptide and the L-Ala-γ-D-Glu-mDAP-D-Ala tetrapeptide only occurred between D-Glu and mDAP and neither between L-Ala and D-Glu nor between mDAP and D-Ala, when the tetrapeptide was used as a substrate. These data further supported that the enzyme specifically cleaves γ-D-Glu-mDAP bonds. Modification of the tripeptide through mDAP amidation at the free carboxyl group as in Bacillus PGN did not affect the enzyme (D Fig in S1 Text). However, removal of L-Ala in position 1, resulting in dipeptide γ-D-Glu-mDAP, abolished activity (Fig 3A). NlpC/P60 recycling enzymes from free-living bacteria were described to depend on a free L-Ala terminus for proper substrate recognition [32]. To test whether this requirement applies to chlamydial YkfC, we next investigated PGN intermediates where L-Ala in position 1 of the peptide is linked to sugar moieties. Indeed, YkfC_Ctr showed no activity on the final PGN building block lipid II, the nucleotide-activated precursor UDP-MurNAc-tripeptide or the PGN recycling intermediate MurNAc-tripeptide (Fig 3A and 3C). To further elaborate on the enzyme’s specificity, we additionally tested a set of 10 (non)-PGN peptides which did not fulfill the above determined requirement combining the presence of a γ-D-Glu-mDAP bond and a free L-Ala terminus. None of the peptides was used as a substrate by YkfC_Ctr (E Fig in S1 Text). YkfC_Ctr had also no activity on high molecular weigth PGN in remazol dye release assays, which is consistent with the E. coli complementation experiments and further supports a function in recycling PGN-derived peptides in the cytoplasm (Fig 3D).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Substrate specificity of YkfC_Ctr.

YkfC_Ctr hydrolyzes the bond between D-Glu and mDAP in the PGN tri- and tetrapeptide. (a) In vitro activity of YkfC_Ctr on soluble precursors and recycling products. TLC analysis shows that YkfC_Ctr uses PGN-derived tripeptide as a substrate but neither the lipid II precursor UDP-MurNAc-tripeptide nor the PGN recycling intermediate MurNAc-tripeptide. PGN tripeptide is the minimal substrate as the dipeptide γ-D-Glu-mDAP was not hydrolyzed under our test conditions. (b) LC-MS analysis of YkfC_Ctr activity on CwlC amidase derived peptides from E. coli PGN. YkfC_Ctr, but not its active-site mutant, cleaves the tri- and tetrapeptide as shown by the disappearance of these substrates and the production of mDAP and mDAP-D-Ala products, respectively. Data are presented as extracted ion chromatograms applying the theoretical masses of tripeptide (3P, [M−H]⁻ = 389.167), tetrapeptide (4P, [M−H]⁻ = 460.204), DAP ([M−H]⁻ = 189.087) and DAP-Ala ([M−H]⁻ = 260.124) within an error range of ± 0.02. (c) In vitro activity of YkfC_Ctr on the membrane bound PGN precursor lipid II. Neither YkfC_Ctr nor its active site mutant YkfC_CtrC172A exhibit detectable activity on PGN precursor lipid II as shown by the unaltered migration distance on the lipid TLC and the absence of amino acid or peptide reaction products on the peptide TLC. DD-carboxypeptidase PBP6_Cpn releases the terminal D-Ala from lipid II and was included as a positive control to show that the employed TLC method is suitable to detect changes within the structure of lipid II (lipid TLC) or the release of amino acids and peptides from lipid II (peptide TLC). (d) Dye-release assay and photometric analysis of the reaction products. YkfC_Ctr and its active-site mutant YkfC_CtrC172A did not exhibit hydrolytic activity when incubated with remazol-brilliant-blue labelled PGN. Lysozyme served as a positive control releasing the labelled muropeptides into the supernatant. Unpaired t-test revealed statistical significance in comparison with the lysozyme control, two-tailed ***P-value ≤0.0001. Error bars indicate ± s.d. (n = 3). (lipid II penta: undecaprenyl-pyrophosphoryl-MurNAc-(GlcNAc)-L-Ala-γ-D-Glu-mDAP-D-Ala-D-Ala; lipid II tetra: undecaprenyl-pyrophosphoryl-MurNAc-(GlcNAc)-L-Ala-γ-D-Glu-mDAP-D-Ala).

https://doi.org/10.1371/journal.ppat.1011047.g003

Taken together, we found evidence that YkfC from C. trachomatis is tailored for recycling of PGN peptides that are released from the PGN glycan chains by amidases. Provided that the peptides meet the requirement of both a γ-D-Glu-mDAP bond and a free L-Ala terminus, the enzyme appeared to show a rather relaxed substrate spectrum as attested by use of an amidated mDAP variant. Precluding hydrolysis of PGN biosynthesis intermediates such as lipid II and its precursor UDP-MurNAc-tripeptide, where the N-terminus is occupied by an N-acetylmuramyl residue, guarantees that cytosolic YkfC_Ctr does not interfere with intracellular lipid II biosynthesis as otherwise, the tightly orchestrated processes of PGN ring biosynthesis and cell division would collapse.

YkfC_Ctr is inhibited by the alkylating agent chloroacetone and the peptidomimetic epoxide E-64

Both our in vivo experiments in the E. coli surrogate system and biochemical analyses with the purified protein pointed to a crucial role of the conserved cysteine C172 within the NlpC/P60 active site. We used a set of cysteine protease inhibitors as chemical tools to further explore the role of cysteine C172 in the catalytic activity of YkfC_Ctr. Chlamydial proteins typically contain a high number of cysteines which are otherwise relatively rare amino acids [18,36]. Besides the cysteine present in the predicted active site, YkfC_Ctr contains 7 additional cysteines of unknown redox state that could be targeted by alkylating compounds which bind to all accessible sulfhydryl-groups. For this reason, we tested inhibitor concentrations for these compounds in relation to the total number of all theoretically available cysteine residues. For the cysteine protease inhibitor E-64 this adaptation was not necessary as the compound selectively binds to the active site of cysteine proteases [37]. Among the two tested alkylating compounds chloroacetone and iodoacetamide, only chloroacetone inhibited the enzyme at a molar inhibitor:protein ratio of 8:1 (Fig 4A). Iodoacetamide did not inhibit YkfC_Ctr even at molar ratios of 80:1 (Fig 4B). The same specificity has been described for the YkfC homolog from Bacteroides ovatus [32]⁠ and the lack of inhibition by iodoacetamide is likely attributed to the significantly lower electronegativity of the iodine compared to the chlorine atom in chloroacetone. The irreversible cysteine protease specific inhibitor E-64 abolished YkfC_Ctr activity at a molar inhibitor:protein ratio of 10:1 (Fig 4D). These data suggest an essential role of the active site cysteine. Future crystallization studies might reveal the overall architecture of the enzyme’s active site and provide detailed information on substrate and inhibitor specificity and recognition.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Effect of different inhibitors on the enzymatic activity of YkfC_Ctr.

TLC analysis of the inhibiting effect of chloroacetone, iodoacetamide, and E-64 on YkfC_Ctr activity. (a) The alkylating chemical chloroacetone proved to be the most potent inhibitor as shown by the absence of the reaction product mDAP at a molar inhibitor:protein ratio of 8:1. YkfC_Ctr retained activity at lower molar ratios (F Fig in S1 Text). (b) The second alkylating compound iodoacetamide did not show a significant inhibitory effect on YkfC_Ctr activity. (c) The cysteine protease inhibitor E-64 carrying a trans-epoxysuccinic acid group coupled to a dipeptide was able to inhibit YkfC_Ctr at a molar ratio of 10:1.

https://doi.org/10.1371/journal.ppat.1011047.g004

Chloroacetone and E-64d have anti-chlamydial activity in cell culture infection experiments

Our in vitro results on an inhibitory effect of chloroacetone and natural compound E-64 on purified YkfC_Ctr prompted us to test for in vivo activity of these cysteine protease inhibitors against C. trachomatis in cell culture experiments. Experiments were performed in the range of inhibitor concentrations and treatment durations that did not exert cytotoxicity towards the HEp-2 host cells in resazurin-based viability assays (G Fig in S1 Text). We propagated C. trachomatis D in human HEp-2 cells, added inhibitors at 6, 10 and 24 hpi and analyzed chlamydial infection at 30 hpi using immunofluorescence microscopy (leading to treatment durations of 24, 20 and 6 h, respectively, Fig 5). E-64 did not show antichlamydial activity up to the highest concentration tested (256 μg/mL) irrespective of treatment duration, probably due to a poor penetration of the lead compound into chlamydial cells. Also cytotoxicity could not be observed for E-64 up to the highest concentration (256 μg/mL) and treatment duration (24h) tested. The more membrane-permeable variant E-64d [38] had antichlamydial activity, reflected by a minimal inhibitory concentration (MIC) of 256 μg/mL after treatment for 6 h (corresponding to an addition of the compound at 24 hpi). Under this treatment regimen, the compound did not exert cytotoxicity towards the host cells but had a detrimental effect on chlamydial cells culmulating in the disruption of chlamydial inclusions and cells—possibly due to its concerted interference with several cellular processes. Analyses with extended E-64d treatment durations were limited by first cytotoxic effects of the variant at a concentration of 256 μg/mL. For E-64d treatment durations of 20 and 24 hpi (corresponding to an addition of the compound at 10 and 6 hpi), we observed no significant reduction in the inclusion number at 128 μg/mL, the highest concentration without cytotoxicity after these prolonged treatment durations. Chloroacetone had the most pronounced antichlamydial activity resulting in a MIC of 4 μg/mL for all tested treatment durations and without detectable impairment of host cell viability in the range of MIC (G Fig in S1 Text).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Determination of minimal inhibitory concentrations of YkfC inhibitors towards C. trachomatis.

Overview of the inhibitor susceptibility analyses (a). HEp-2 cells were infected with C. trachomatis serovar D strain UW-3/CX, E-64, its variant E-64d, and chloroacetone were added at 6, 10 and 24 hpi and the antichlamydial activity was analyzed at 30 hpi (resulting in total treatment durations of 24, 20 and 6 h, respectively). Experiments were done in triplicate and analyzed in the range of concentrations and treatment durations that had no cytotoxicity towards the HEp-2 host cells (G Fig in S1 Text). Exemplarily pictures of untreated and DMSO-treated (b) as well as inhibitor-treated cells at the minimal inhibitory concentration (MIC) or the highest concentration tested (c) were included in the figure. Scale bar: 10 μm. The MIC was the lowest concentration that reduced the number of chlamydial inclusions by more than 90% as defined previously [55]. E-64 had neither antichlamydial nor cytotoxic activity up to the highest concentration and treatment duration tested (256 μg/mL for 24 h). In contrast, E-64d treatment for 6h severely impaired chlamydial infection without cytotoxicity resulting in a MIC of 256 μg/mL. Extended E-64d treatment with 256 μg/mL for 20 and 24 h (corresponding to an addition at 6 and 10 hpi) lead to reduced host cell viability. Treatment with 128 μg/mL, the highest concentration without cytotoxicity under these treatment durations, had only a minor non-significant antichlamydial effect. With an MIC of 4 μg/mL determined for all tested treatment durations, chloroacetone had the highest antichlamydial activity without cytotoxicity towards host cells in the range of MIC.

https://doi.org/10.1371/journal.ppat.1011047.g005

In a complementary approach, we employed a dCas12 CRISPR interference-based inducible knockdown system [39]. We verified knockdown of ykfC_Ctr in C. trachomatis L2 by measuring the transcriptional levels of ykfC_Ctr in comparison to euo and omcB, as controls for developmental cycle progression [40] (HA Fig in S1 Text). We also assessed genomic DNA levels and did not observe significant differences after induction of dCas12 expression (HB Fig in S1 Text). Expression of the dCas12 in the non-targeting background resulted in an approximately 50% reduction in infection forming units, suggesting there is a metabolic burden to overexpressing dCas12 that does not impact replication rate (HB and HC Fig in S1 Text). Indirect immunofluorescence assay (IFA) analysis did not reveal a meaningful change in inclusion and chlamydial morphology (HD and HE Fig in S1 Text). Nonetheless, after knocking down ykfC_Ctr transcripts, there was a further, statistically significant 50% reduction in production of infection forming units in the ykfC knockdown strain as compared with the non-targeting control (HC Fig in S1 Text). The less natural environment in the permissive mouse fibroblast McCoy cells and residual expression and enzymatic activity of YkfC_Ctr might have contributed to the less intense effects in the ykfC knockdown experiments.

All in all, we used YkfC_Ctr inhibitors as tools in in vivo cell culture studies and provided further evidence that the NlpC/P60 protein is important for chlamydial biology. In particular, the inhibitory effect of the more membrane-permeable variant of natural lead compound E-64, although less pronounced compared to chloroacetone, may open the way for further studies on YkfC_Ctr as a target for antichlamydial drug development.

Localization of YkfC in C. trachomatis cells

To further explore the biological function of YkfC_Ctr we performed localization analyses in C. trachomatis (Fig 6). To this end, we made an anhydrotetracycline (aTc)-inducible construct encoding YkfC_Ctr tagged with six histidines at the C-terminus (YkfC_Ctr_6xH). This construct was transformed into C. trachomatis serovar L2 lacking its endogenous plasmid (-pL2). Using indirect immunofluorescence assays (IFA), we detected YkfC_Ctr_6xH as puncta at the periphery of non-dividing C. trachomatis cells (Fig 6B). We did not observe any gross morphological effects of YkfC_Ctr_6xH overexpression on inclusion size or bacterial numbers, as confirmed by inclusion forming unit assays where the YkfC_Ctr_6xH expressing strain generated 96 ± 20% of infection forming units compared to the uninduced control (I Fig in S1 Text). These data indicate that Ykf_Ctr_6xH overexpression does not impact chlamydial developmental cycle progression.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Localization of YkfC_Ctr_6xH in C. trachomatis.

McCoy cells were infected with the transformant encoding YkfC_Ctr_6xH. YkfC_Ctr_6xH expression was induced at 4 hpi with 10 nM aTc. At 10.5 (a) or 24 hpi (b), the infected cells were fixed with aldehydes to preserve the budding morphology of dividing cell (10.5 hpi) or 100% MeOH (24 hpi) and stained with anti-6xH (red) and anti-MOMP (green) antibodies. The images were acquired on a Zeiss AxioImager.Z2 equipped with an Apotome2 using a 100X objective. Scale bar: 0.5 μm (a) or 2 μm (b). YkfC_Ctr_6xH showed as puncta at the periphery of cells. As exemplarily shown in (a), assessment of dividing cells revelaed that 90% of cells in the early stages of cell division (19 out of 21 analyzed cells in the pre-dividing and following budding stage) did not feature YkfC_Ctr_6xH signals at the (future) site of division (i.e. at the MOMP-enriched site in pre-dividing cells or at the site of daughter cell formation in the budding stage). In contrast, 60% of cells in the late dividing stage (15 out of 25 analyzed late diving cells) had YkfC_Ctr_6xH puncta near the septum. UI: uninfected; I: infected.

https://doi.org/10.1371/journal.ppat.1011047.g006

When analyzing dividing cells, we found evidence for an overall enrichment of the Ykf_Ctr_6xH signal near the septum as cell division progresses (Fig 6A). Only 10% of cells in the early stages of cell division featured Ykf_Ctr_6xH signals proximal to the site of division (i.e. at the MOMP (major outer membrane protein)-enriched site which determines the future site of daughter cell formation [22,41] in the pre-dividing stage or at the site of polarized asymmetric daughter cell formation in the following budding stage). In contrast, Ykf_Ctr_6xH puncta near the septum could be detected in 60% of cells in the late cell-division stage. These observations suggest that YkfC_Ctr might be recruited adjacent to the cell division plane later in the process, consistent with the function of an enzyme which is involved in recycling of PG material, a process that is likely coordinated with constriction and degradation of the transient PG ring during separation of daughter cells.

NOD1 activation by C. trachomatis is increased under YkfC inhibitor treatment

PGN-derived peptides are known to activate innate immunity factor NOD1 that binds to the γ-D-Glu-mDAP dipeptide as a minimal ligand [42]. We wanted to learn whether YkfC_Ctr is active in vivo and contributes to immune evasion through degrading the recognition motif of NOD1 in PGN-derived peptides. To this end, we performed NOD1 activation experiments with C. trachomatis infected human NOD1 reporter cells in the presence of an YkfC inhibitor. Indeed, NOD1 activation was significantly increased under treatment with the YkfC inhibiting compound chloroacetone as compared with the untreated infected vehicle control (Fig 7). The increase of NOD1 activation was observed at concentrations (0.016 and 0.031 μg/mL) that did not impact on inclusion size (JA Fig in S1 Text) and were below both the MIC of chloroacetone towards C. trachomatis (4 μg/mL) and cytotoxicity towards the infected reporter cell line (IC₅₀ 15 μg/mL, JB Fig in S1 Text). In control experiments with uninfected NOD1 reporter cells, chloroacetone did not modulate the level of NOD1 activation (JC Fig in S1 Text). These results provide evidence that C. trachomatis YkfC degrades NOD1 stimulating PGN-derived peptides not only in vitro but also in infected host cells supporting a role in immune evasion.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. NOD1 signaling in C. trachomatis infected reporter cells is increased under YkfC inhibitor treatment.

NOD1 reporter cells were infected with C. trachomatis D and used to analyze abundance of NOD1 activating PGN-derived material shed by the intracellular pathogen in the presence of YkfC_Ctr inhibitor chloroacetone. Infected NOD1 reporter cells were treated 6 hpi with chloroacetone and analyzed 18 hpi (a) and 24 hpi (b) resulting in treatment durations of 12 h and 18 h, respectively. An increase of NOD1 activation was observed at concentrations (0.016 and 0.031 μg/mL) below both cytotoxicity towards the reporter cell line (IC₅₀ 15 μg/mL after 24 h treatment, JA Fig in S1 Text) and the MIC of chloroacetone towards C. trachomatis (4 μg/mL, observed after treatment durations of 6, 20 and 24 hpi, Fig 5). Addition of TriDAP 6 hpi was used as a positive control for NOD1 activation. Error bars indicate ± s.d. (n = 3). Ns: not significant. ***P-value ≤0.0001.

https://doi.org/10.1371/journal.ppat.1011047.g007

NlpC/P60 proteins are widely conserved within the order Chlamydiales

Our discovery of a functional YkfC homolog in C. trachomatis prompted us to analyze the evolutionary conservation of NplC/P60 proteins among the Chlamydiales. Using sequences of YkfC from C. trachomatis and B. cereus as queries for BLAST alignments, we found genes coding for putative NlpC/P60 proteins in all members of the Chlamydiaceae and the genomically less reduced Chlamydia-related bacteria belonging to the Simkaniaceae, Parachlamydiaceae, and Criblamydiaceae families, but not within the family Waddliaceae [1,43]. To gain insight into the function of these putative NlpC/P60 proteins, we selected three candidates for initial analyses in our two E. coli surrogate systems reporting on possible PGN endopeptidase or peptide recycling activity.

Consistent with our results for C. trachomatis YkfC, neither the homolog from the other established human pathogen C. pneumoniae nor the putative NlpC/P60 proteins from the Chlamydia-related bacteria Simkania negevensis and Estrella lausannensis rescued growth in the endopeptidase MepS defective E. coli strain (Fig 8A), making a role in periplasmic PGN turnover unlikely for all three candidates. The homolog from C. pneumoniae did not restore growth of the E. coli ΔdapDΔmpl double mutant strain under nutrient-rich growth conditions (Figs 8B, K in S1 Text). Both NlpC/P60 proteins from the Chlamydia-related organisms lack a typical signal sequence for transport across the cytoplasmic membrane and were capable of restoring growth in the E. coli surrogate test strain suggesting their ability to cleave PGN-derived tripeptide (Figs 8B, K in S1 Text). The predicted structure of S. negevensis NlpC/P60 (NlpC/P60_Sne) contrasts with all other chlamydial proteins. NlpC/P60_Sne lacks both SH3b domains including the S2 site residues (corresponding to Glu83, Thr84, and Tyr118 in YkfC_Bce) important for interactions with L-Ala in position 1 of PGN peptides in other NlpC/P60 PGN peptidases. The structure predictions show highest matches with an NlpC/P60 PGN hydrolase from the protozoan Trichomonas vaginalis (L and M Figs in S1 Text) [44]. These data indicate that our understanding of substrate binding in prokaryotic NlpC/P60 proteins is not yet fully established. Moreover, our results suggest that NlpC/P60 proteins are likely conserved as peptide recycling enzymes not only in C. trachomatis but also the less related, deeply-rooting S. negevensis and E. lausannensis.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 8. In vivo complementation of the E. coli ΔdapDΔmpl and E. coli ΔmepS surrogate systems with different chlamydial NlpC/P60 homologs.

(a) Spot plate assay of E. coli carrying a deletion for the gene encoding periplasmic NlpC/P60 endopeptidase MepS. The mutant was transformed with different pBAD24 constructs encoding for an NlpC/P60 homolog from C. pneumonia, S. negevensis or E. lausannensis and their cognate active-site mutants. Plates with low osmolarity media containing 0.04% L-arabinose were incubated at 42°C. Complementation of the osmo- and temperature-sensitive ΔmepS could only be observed for pBAD24 producing E. coli MepS, but not for any chlamydial NlpC/P60 homolog. An empty vector control served as a negative control. (b) Growth effect in the E. coli ΔdapDΔmpl double mutant containing different chlamydial NlpC/P60 constructs. Cells were grown in nutrient-rich medium containing 40 μM PGN tripeptide and protein production was induced with 0.4% arabinose. NlpC/P60 from S. negevensis and E. lausannensis, but not from C. pneumoniae, restored growth of the E. coli double mutant, which is auxotrophic for mDAP and unable to utilize PGN tripeptide under nutrient rich conditions. None of the active site mutants was able to complement the growth defect. Error bars indicate ± s.d. (n = 3).

https://doi.org/10.1371/journal.ppat.1011047.g008

To further explore the function of NlpC/P60 proteins in the biology of both Chlamydia-related genera, we analyzed gene expression and protein localization of of NlpC/P60 homologs in S. negevensis and E. lausannensis (Fig 9A). Transcription of nlpC/p60 could be detected in both species (Fig 9A) and was normalized to the number of bacteria, as quantified by qPCR (NA Fig in S1 Text). In E. lausannensis the expression level of nlpC/p60 was similar to mreB, and both transcript levels peaked synchronously during cell divsion, as expected. To date, nothing is known about cell division in S. negevensis. As described above, our E. coli complementation experiments indicate that S. negevensis NlpC/P60 is capable of recycling PGN-derived peptides like the other tested proteins but the predicted structure contrasted with the other peptidases. In S. negevensis, mreB showed a less pronounced increase in expression during the replication phase as compared with E. lausannensis, and also the nlpC/p60 gene was more constantly expressed throughout the developmental cycle. This might imply functional differences between NlpC/P60 in these two chlamydial species. Further studies are required to decipher these differences and understand the evolution of PGN-remodelling and division mechanisms in the Chlamydiales order.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 9. NlpC/P60 proteins of Chlamydia-related bacteria are expressed during the developmental cycle and show a peripheral localization.

Vero cells infected with S. negevensis or E. lausannensis were harvested at the indicated times post-infection. (a) RNA levels of nlpC and mreB transcripts were quantified by qRT-PCR and normalized to the number of bacteria, which was estimated by qPCR (NA Fig in S1 Text). RNA level is shown as foldchange compared to time 72 h. Error bars indicate ± s.d. (n = 4). NlpC/P60_Ela (b) and NlpC/P60_Sne (c) show a peripheral localization throughout the developmental cycle of E. lausannensis and S. negevensis, respectively. Infected cells were harvested and labelled with anti-NlpC/P60 antibodies (green), species-specific antibodies (red) and DAPI (blue). Localization was investigated by confocal microscopy. Treatment with penicillin of E. lausannensis causes formation of enlarged aberrant bodies, but does not influence NlpC/P60 localization. Bars = 2 μm.

https://doi.org/10.1371/journal.ppat.1011047.g009

Next, we raised antibodies against S. negevensis and E. lausannensis NlpC/P60 (NB and NC Figs in S1 Text) and investigated the subcellular localization of the proteins using confocal microscopy. In line with our observations in C. trachomatis, a peripheral cell envelope associated localization was observed in both Chlamydia-related genera throughout the developmental cycle (Fig 9B and 9C). Contrasting with C. trachomatis where the YkfC protein was detected as punctae, the S. negevensis and E. lausannensis NlpC/P60 proteins were more evenly distributed over the envelope. These divergent localization patterns may reflect potential differences in the organization and regulation of PGN machineries between Chlamydia and the evolutionary less reduced Chlamydia-related bacteria. Previous studies in chlamydiae indicated that persistence-inducing β-lactam antibiotics trigger delocalization of PGN biosynthesis enzymes [10]⁠. To test whether this effect also applies for PGN recycling enzymes, we treated E. lausannensis with penicillin but did not detect changes in the localization of the NlpC/P60 enzyme (Fig 9B). S. negevensis were not investigated under penicillin treatment, as these chlamydiae are resistant to penicillin [45,46]. Taken together, we found NlpC/P60 proteins to be widespread among Chlamydiales and provide evidence that the homologs in the Chlamydia-related S. negevensis and E. lausannensisare were expressed throughout the developmental cycle and act as peripherally located PGN peptide recycling enzymes.

Bacterial strains and growth conditions

For propagation of C. trachomatis serovar L2, McCoy (kind gift of Dr. Harlan Caldwell, NIH) host cells were cultured at 37°C with 5% CO₂ in Dulbecco’s Modified Eagle Medium (DMEM; Invitrogen, Waltham, MA) containing 10% (v/v) fetal bovine serum (FBS; Hyclone, Logan, UT) and 10 μg/mL gentamicin (Gibco, Waltham, MA). C. trachomatis serovar L2 (strain 434/Bu) lacking the endogenous plasmid (-pL2; kind gift of Dr. Ian Clarke, Univ. Southampton) was grown in McCoy cells for use in transformations. The infected McCoy cells were incubated in DMEM containing 10% FBS, 10 μg/mL gentamicin, 1 U/mL penicillin G, and 1 μg/mL cycloheximide. All cell cultures and C. trachomatis serovar L2 stocks were routinely tested for Mycoplasma contamination using the Mycoplasma PCR detection kit (Sigma, St. Louis, MO).

For propagation of C. trachomatis serovar D, HEp-2 host cells (ATCC CCL-23, Cell Lines Services) were cultured in DMEM supplemented with 10% (v/v) FBS, 1x MEM non essential amino acids solution, 1x MEM vitamin solution, 2.5 μg/mL amphotericin B and 50 μg/mL gentamicin (Gibco) at 37°C and 5% CO₂. C. trachomatis D/UW-3/CX (ATCC VR-885) was grown in HEp-2 cells for use in MIC experiments. The infected cells were incubated in DMEM containing 10% (v/v) fetal calf serum (FCS), 1x MEM non essential amino acids solution, 1x MEM vitamin solution, 2.5 μg/mL amphotericin B, 50 μg/mL gentamicin and 1.2 μg/mL cycloheximide (Merck, Darmstadt, Germany). Cell culture was tested monthly for Mycoplasma contaminations using the Venor GeM OneStep Mycoplasma Detection Kit (Minerva Biolabs, Berlin, Germany). All MIC experiments were conducted in growth medium without cycloheximide, amphotericin and gentamicin.

All E. coli strains were maintained on Luria Bertani (LB) agar plates, unless stated otherwise. E. coli BL21(DE3) carrying YkfC_Ctr expression plasmids was maintained on agar plates containing 100 μg/mL ampicillin. For maintenance of the mutant strains, that were used in complementation experiments, the agar plates contained 200 μg/mL mDAP and 30 μg/mL chloramphenicol in case of the E. coli ΔdapDΔmpl double mutant strain and 100 μg/mL kanamycin in case of E. coli mepS mutant strain. E. coli NEB10β (New England Biolabs, Ipswich, MA) cells were used for cloning and expansion of the E.coli/C. trachomatis shuttle vector. E. coli NEB10β was grown at 30°C in LB broth. All chemicals and antibiotics were obtained from Sigma unless otherwise noted.

Cloning of C. trachomatis

From C. trachomatis serovar L2 (strain 434/Bu), we amplified the gene ctl0382 (homolog of ct127) with a six histidine tag in the reverse primer using Phusion enzyme (NEB), and the PCR product was linked with pBOMBL::L2 linearized by EagI and KpnI using HiFi assembly system (NEB). The HiFi products were transformed into NEB10β competent cells (NEB) and plated on LB agar containing 100 μg/mL ampicillin.

Immunofluorescence and confocal microscopy of C. trachomatis

McCoy cells were infected with the transformants, and the infected cells were either fixed with aldehydes at 10.5 hpi or 100% MeOH for 10 min at 24 hpi after inducing expression of the construct at 4 hpi with 10nM aTc. The coverslips were stained with primary antibody rabbit anti-6xH (Abcam, Cambridge, MA) and goat anti-major outer membrane protein (MOMP) of C. trachomatis (Meridian, Memphis, TN). Subsequently, the coverslips were stained with secondary antibodies donkey anti-rabbit (594) and donkey anti-goat (488) (Invitrogen).

Transformation into C. trachomatis without endogenous plasmid

A day before beginning the transformation, McCoy cells were plated in a six-well plate (1 × 10⁶ cells/well). C. trachomatis serovar L2 without plasmid (-pL2) (kind gift of Dr. Ian Clarke) was incubated with 2 μg plasmid in Tris-CaCl₂ buffer (10 mM Tris-Cl pH 7.5, 50 mM CaCl₂) for 30 min at room temperature. During this step, the McCoy cells were washed with 2 mL Hank’s Balanced Salt Solution (HBSS) media containing Ca²⁺ and Mg²⁺ (Gibco). After that, cells were infected with the transformants in 2 mL HBSS per well. The plate was centrifuged at 400 × g for 15 min at room temperature and incubated at 37°C for 15 min. The inoculum was aspirated, and DMEM containing 10% FBS and 10 μg/mL gentamicin was added per well. At 8 h post infection (hpi), media containing 1 or 2 U/mL penicillin G and 1 μg/mL cycloheximide was added to the wells after aspirating the initial post-infection media, and the plate was incubated at 37°C until 48 hpi. At 48 hpi, the transformants were harvested and used to infect new McCoy cells. These harvest and infection steps were repeated every 44–48 hpi until mature inclusions were observed (typically collected after two passages).

Cloning of E. coli

For overproduction, the ykfC (ct127) gene from C. trachomatis D/UW-3/CX was amplified by PCR using the primers listed in Table A in S1 Text and cloned into pET52b using the XmaI and EagI restriction sites to generate a protein with an N-terminal Strep-tag II followed by a human rhinovirus 3C protease cleavage site. For complementation assays, the ykfC genes from C. trachomatis D/UW-3/CX (ct127) and C. pneumoniae GiD (cpn0245) as well as the genes encoding the NlpC/P60 proteins of E. lausannensis CRIB 30 (ELAC_2173) and S. negevensis ATCC VR-1471/Z (SNE.A23160) were amplified by PCR using the primers listed in Table A in S1 Text and cloned into pBAD24+.

Site-directed mutagenesis

To generate active site mutants of the respective proteins, the residues C172 in YkfC_Ctr, C173 in YkfC_Cpn, C194 in NlpC/P60_Ela and C153 in in NlpC/P60_Sne were changed to alanine using the QuikChange Lightning Site-Directed Mutagenesis Kit (Agilent Technologies). The respective sense and antisense primers listed in Table A in S1 Text were used according to the manufacturer’s instructions. Correct base changes were confirmed by sequencing.

YkfC_Ctr overproduction and purification

For purification of YkfC_Ctr, E. coli BL21(DE3) cells were transformed with pET52b-ykfC_Ctr. Cells were grown in 2 liters TB autoinduction medium [56,57] at 37°C for 1 h. Temperature was shifted to 30°C and growth was continued for 20 h. Cells were harvested by centrifugation for 15 min at 7550 × g and 14°C and the resulting cell pellet was resuspended in 40 mL buffer A (25 mM MOPS pH 7.2, 1 M NaCl, 2 mM MgCl₂) supplemented with 2 μg/mL polymyxin B (Carl Roth), 1 mM phenylmethyl sulfonyl fluoride (Merck) and 1 U/ml benzonase (Merck). Cells were broken by sonication, supplemented with 0.25% CHAPS (Carl Roth) and further incubated at 4°C for 1 h with gentle shaking. The suspension was centrifuged for 1 h at 310500 × g at 4°C and the supernatant was poured onto a Strep-tactin column (1 ml bed volume) equilibrated with buffer A. After complete passage the column was washed 5 times with buffer B (25 mM MOPS pH 7.2, 500 mM NaCl, 2 mM MgCl₂, 10% glycerol). The protein was eluted with buffer C (25 mM MOPS pH 7.2, 300 mM NaCl, 2 mM MgCl₂, 10% glycerol and 50 mM biotin) and dialysed against 2 × 1 L dialysis buffer (MOPS pH 7.2, 150 mM NaCl, 2 mM MgCl₂, 10% glycerol and 2 mM DTT) at 4°C overnight. Purity was determined by SDS-PAGE and aliquots stored at -80°C. For inhibitor screens DTT was omitted in all buffers. The active site mutant YkfC_CtrC172A was overproduced and purified in the same manner as wild type YkfC_Ctr.

YkfC_Ctr activity assay

Endopeptidase assays were carried out in a final volume of 25 μl containing 50 mM MOPS pH 7.5, 150 mM NaCl, 2 mM MgCl₂, 0.5 mM substrate and 2 μM enzyme. Each reaction was incubated at 37°C for 2 h and terminated by incubation for 5 min at 100°C. The reactions were centrifuged for 5 min at 17000 × g and 20 μl of the supernatant was analyzed by TLC on a HPTLC alumina silica gel 60 plate (Merck) using 2-butanol-pyridine-ammonia-water (39:34:10:26) as the mobile phase. Separation was visualized by ninhydrin staining. For lipid II as the substrate, reaction products were extracted with 20 μl of n-butanol/pyridine acetate (2:1 v/v, pH 4.2) and analyzed by TLC using chloroform-methanol-water-ammonia (88:48:10:1) as mobile phase and visualized by Hanessian’ s staining.

For PGN peptides derived from E. coli or B. subtilis PGN as the substrate, PGN from each organism was treated with recombinant B. subtilis amidase CwlC as described [58]. CwlC derived peptides from E. coli or B. subtilis PGN were incubated with YkfC_Ctr and the reaction products were analyzed by HPLC-MS using an electrospray ionisation (ESI) time of flight mass spectrometer (MS) (microTOF II, Bruker Daltonics) connected with an UltiMate 3000 HPLC (Dionex). Soluble components in 5 μl samples were separated on a reversed-phase C18 column (Gemini 5 μm C18 110 Å 150 mm × 4.6 mm; Phenomenex) in an UltiMate 3000 HPLC (Dionex) by applying a 30-min linear gradient of 0–40% acetonitrile (buffer A: 100% acetonitrile) at a flow rate of 0.2 ml/min. Before each run, the column was pre-equilibrated with 0.01% formic acid/0.05% ammonium formate, followed by the linear gradient to 100% acetonitrile. After each run, the column was rinsed for 12.45 min with 60% acetonitrile and subsequently re-equibrilated for 12.45 min with 0.01% formic acid/0.05% ammonium formate. The mass-to-charge rations of the separated samples were analysed by MS, operating in negative ion mode with a mass range of 120–1000. Extracted ion chromatograms for the tripeptide (3P; theoretical mass [M−H]⁻ = 389.167), tetrapeptide (4P; [M−H]⁻ = 460.204), DAP ([M−H]⁻ = 189.087) and DAP-Ala ([M−H]⁻ = 260.124) were obtained within an error range of ± 0.02.

For PGN as the substrate, freshly prepared sacculi from E. coli W3110 were stained as described previously [20]⁠. The assay was carried out in a final volume of 100 μl containing 50 mM MOPS pH 7.5, 150 mM NaCl, 2 mM MgCl₂, 1.25 μg sacculi and 4 μM YkfC_Ctr. Each reaction was incubated for 16 h at 37°C. Insoluble material was removed by centrifugation for 10 min at 21000 × g and the supernatant containing soluble reaction products was measured on a spectrophotometer (Nano Photometer, Implen) at 595 nm.

YkfC_Ctr inhibitor assays

All inhibitors were purchased from Merck Millipore (formerly Sigma Aldrich). The assay was carried out as described above with minor modifications. Before adding the substrate, YkfC_Ctr was pre-incubated with the inhibitor for 15 min at room temperature.

E. coli ΔmepS complementation

The E. coli JW2163-3 ΔmepS mutant was received from the Keio E. coli mutant strain collection [59]. Competent cells of this E. coli mutant were transformed with the appropriate vectors (Table A in S1 Text). 5 ml of LB containing 100 μg/ml ampicillin were inoculated with material of a single, freshly transformed colony and incubated for 16 h at 37°C. For each construct material from three different colonies were tested separately in the assay. Subsequently, 500 μl of the culture was diluted 1:100 in nutrient broth (0.5% tryptone, 0.3% yeast extract) containing 100 μg/ml ampicillin. Cultures were incubated at 37°C and harvested at an OD₆₀₀ of 1. Each cell pellet from 1 ml was suspended in 1 ml 0.9% NaCl solution and starting with this suspension a 1:10 dilution series in 0.9% NaCl was performed. The dilutions were plated on nutrient broth agar (NA) plates containing 0.04% L-arabinose using a replica plater (Merck). After drying the plates were incubated for 16 h at 30°C, 37°C and 42°C, respectively.

Construction of the E. coli ΔdapDΔmpl double mutant strain

The AT980 ΔdapD strain requiring mDAP for growth was earlier described [26]⁠. The MLD2502 strain derived from BW25113 that carries a deletion of the chromosomal mpl gene (Δmpl::Cm^R) was described previously [60]. The ΔdapDΔmpl double mutant strain used in this work was constructed by transduction of the Δmpl::Cm^R mutation into the AT980 strain by phage P1 transduction [61]⁠.

E. coli ΔdapD Δmpl complementation

Competent cells of E. coli BW25113 ΔdapD Δmpl were transformed with the appropriate vectors (Table A in S1 Text). 5 ml of LB containing 200 μg/ml mDAP and 100 μg/ml ampicillin were inoculated with material of a single, freshly transformed colony and incubated for 16 h at 37°C. Subsequently, 500 μl of the culture was diluted 1:100 in TB medium [56] containing 200 μg/ml mDAP, 100 μg/ml ampicillin and 0.4% L-arabinose. Cultures were incubated at 37°C and harvested at an OD₆₀₀ of 1. Each cell pellet from 1 ml was washed twice with 1 ml TB. The final suspension was diluted 1:3 in TB and for each well 100 μl were added to 100 μl TB containing 200 μg/ml ampicillin and 0.8% L-arabinose. Wells were further supplemented with 40 μM tripeptide. Samples were incubated for 6 h at 37°C with intermittent shaking every 15 min. Growth was monitored at OD₆₀₀ on a Tecan platereader.

Immunofluorescence and confocal microscopy with Chlamydia-related bacteria

Vero cells (ATCC CCL-81) were cultivated in Dubelcco’s modified Eagle’s medium (DMEM) supplemented with 10% FBS at 37°C and 5% CO₂. Cells were detached and 5 × 10⁵ cells/ml were incubated on a glass coverslip in 24 well plate overnight to allow them to adhere. E. lausannensis or S. negevensis cultivated in Acanthameoba castellanii were filtered with 5 μm filter to remove intact ameobae and diluted 1000 × in DMEM 10% FBS. Infection was synchronized by centrifugation for 15 min at 1790 × g. Infected cells were then incubated 15 min at 37°C, cells were washed and fresh medium was added.

After 24 h of incubation at 37°C 5% CO₂ infected cells were fixed with ice-cold methanol for 5 min, washed 3 times with PBS and incubated overnight in blocking solution (PBS, 1% BSA, 0.1% saponin). Immunofluorescence was performed as previously described [11]⁠. Coverslips were then incubated with antibodies: rabbit anti E. lausannensis, anti S. negevensis [62,63] and antibodies targeting NlpC of each bacterium (produced by immunization of mice with the purified proteins, Eurogentec, Seraing, Belgium). A subsequent incubation with secondary antibodies Alexa Fluor 488 goat anti-mouse and Alexa Fluor 594 donkey anti-rabbit (Thermo FisherScientific, (Waltham, MA) and DAPI was performed. Samples were then imaged by confocal microscopy (LSM 710 confocal microscopy, Zeiss, Oberkochen, Germany) and treated using ImageJ software (U. S. National Institutes of Health, Bethesda, MD).

Transcription levels measurements

Cells were infected with E. lausannensis or with S. negevensis as described above and harvested by scrapping at the different time points post-infection. Samples were split and used for (i) quantification of the number of bacteria by qPCR, (ii) and gene expression measurement by qRT-PCR.

For qPCR, genomic DNA was isolated from 50 μl of cell suspension with the Wizard SV Genomic DNA purification system (Promega) following the manufacturer’s instructions, with a final elution in 200 μl of water. qPCR was performed on 5 μl of DNA and iTaq supermix with ROX (Bio-Rad). Primers SneF and SneR (400 nM each) and 200 nM of probe SneS were used to quantify genome copies of S. negevensis [63]. Primers EstF and EstR (400 nM each) and 200 nM of probe EstS were used to quantify E. lausannensis [62]. qPCR conditions were 5 min at 95°C followed by 40 cycles of 30 s at 95°C and 1 min at 60°C. Amplification and detection of PCR products were performed using a QuantStudio 3 (Applied Biosystems).

For qRT-PCR, RNA was stabilized from 500 μl of cell suspension using RNA Protect (Qiagen, Venlo, Netherlands) by incubation for 5 min at room temperature. Suspension was centrifuged at 5000 x g for 10 minutes, supernatant was removed, and pellet was kept at -20°C. RNA was extracted from the pellet using the RNeasy Plus Kit (Qiagen), following manufacturer’s instructions. DNA was eliminated by selective digestion with DNAse from Ambion DNA-free kit (Thermo Fisher Scientific, Waltham, MA). Reverse transcription was then performed to synthesize cDNA using a Goscript Reverse Transcription System (Promega). qPCR was used to quantify cDNA of mreB and nlpC for both E. lausannensis and S. negevensis. qPCR was performed by mixing 5 μl of 5-times diluted cDNA with 10 μl of iTaq Universal SYBR Green mix (Bio-Rad), 3.6 μl of water and 0.8 μl of forward and reverse primers listed in Table A in S1 Text. qPCR was performed using a QuantStudio 3 (Applied Biosystems, Waltham, MA) with the following conditions: 5 min of denaturation at 95°C followed by 45 cycles of 30 sec denaturation at 95°C and 1 min of annealing/elongation at 60°C.

Minimal inhibitory concentration of cysteine protease inhibitors against C. trachomatis

The effect of cysteine protease inhibitors on active C. trachomatis infection was analysed by fluorescence microscopy. First, 3.5×10⁴ HEp-2 cells were seeded into black 96-well μ-plates (ibidi, Gräfelfing, Germany) and incubated for 3 days at 37°C and 5% CO₂. Cells were washed once with medium and infected with C. trachomatis D/UW-3/CX. After 2 h the EB suspension was exchanged with fresh medium. The effect of E-64, chloroacetone (dissolved in water), and E-64d (dissolved in DMSO) on C. trachomatis was determined by addition of the inhibitors at 6, 10 and 24 hpi. Cells were fixed with 100% (v/v) icecold methanol at 30 hpi and washed once with 1× phosphate-buffered saline (PBS, 4 mM KH₂PO₄ pH 7.4, 16 mM Na₂HPO₄, 115 mM NaCl). Pathfinder chlamydia culture confirmation system (Bio-Rad Laboratories, Feldkirchen, Germany) was used for staining of C. trachomatis and host cell cytoplasm according to the manufacturers’ information. Genomic DNA was labeled with 10 μg/mL DAPI (Thermo Fischer Scientific, Waltham, MA) for 1 min and cells were washed twice with PBS for 10 min at 4°C. Fluorescence microscopy was performed with an Axio observer Z.1 using Zen 2 software (Zeiss, Oberkochen, Germany).

NOD1 reporter assay

Eukaryotic HEK-Blue hNOD1 cells (InvivoGen, Toulouse, France) were cultured as indicated in the manufacturers protocol. Briefly, cells were detached every 2–5 days by incubation with 1x PBS for 5 min at 37°C. For passaging, the cell suspension was diluted with fresh DMEM supplemented with 10% (v/v) FBS, 2 mM L-glutamine, 100 μg/mL Normocin, 30 μg/mL blasticidin and 100 μg/mL Zeocin (InvivoGen, Toulouse, France) and incubated at 37°C and 5% CO₂. The cell line was used to analyze release of NOD1 ligands by C. trachomatis during infection and treatment with cysteine protease inhibitors. 2x10⁵ HEK-Blue hNOD1 cells/mL were seeded into flat TC-96-well plates (Sarstedt, Nümbrecht, Germany) using supplemented DMEM lacking antibiotics and the plates were incubated for 20 h at 37°C and 5% CO₂. Cells were washed once with medium and inoculated with C. trachomatis D/UW-3/CX at an infection rate of 20% for 2 h. Afterwards, the medium was exchanged with reconstituted HEK-Blue detection medium (InvivoGen, Toulouse, France). At 6 hpi, YkfC inhibitor chloroacetone (solved in H₂O) was added in a serial dilution. Positive controls were conducted by the addition of 1.25 μg/mL triDAP 6 hpi (InvivoGen, Toulouse, France). Stimulation of NOD1 was measured at OD_{600 nm} 18 hpi and 24 hpi (corresponding to treatment durations of 12 h and 18 h, respectively) using a Tecan infinite M200 plate reader and Tecan SparkControl software (Tecan Group, Männedorf, Switzerland).