Ethics statement

All genetic manipulations and containment work were approved by the UVA Biosafety Committee and are in compliance with the section III-D-1-a of the National Institutes of Health guidelines for research involving recombinant DNA molecules.

Cell lines and bacterial strains

HeLa cells were obtained from ATCC (CCL-2) and cultured at 37°C with 5% CO₂ in DMEM high glucose (Invitrogen) supplemented with 10% heat inactivated FBS (Invitrogen). C. trachomatis Lymphogranuloma venereum, Type II were obtained from ATCC (L2/434/Bu VR-902B). C. muridarum was obtained from Michael Starnbach (Harvard Medical School, Boston, MA). Chlamydia propagation and infection were performed as previously described [48].

Plasmid construction

Restriction enzymes and T4 DNA ligase were obtained from New England Biolabs (Ipswich, MA). PCR was performed using Herculase DNA polymerase (Stratagene). PCR primers were obtained from Integrated DNA Technologies and are listed in S1 Table. All inserts were sequenced using Genscript. Plasmid maps were generated using Serial Cloner (http://serialbasics.free.fr/Serial_Cloner.html).

C. trachomatis and C. muridarum transformation

Our calcium-based transformation protocol was adapted from Wang et al. [14] and modified from Agaisse et al. [15] as follows. For one transformation the reaction was set by combining 500ng of plasmid DNA (extracted from the E.coli GM2163 (dam^- dcm^-) strain), C. trachomatis L2 or C. muridarum (empirically determined so that 100% of the cells were infected), 6.7μl of 5X CaCl₂ Buffer (50mM Tris, 250mM CaCl₂ pH 7.4), and water to a final volume of 33.5μl, added in that order and incubated for 30min at room temperature. Mixing the reaction by pipetting up and down every 5min can be performed but is not necessary. At the end of the 30min, 2ml of media (DMEM high glucose supplemented with 10% FBS) was added to the transformation reaction and transferred to one well of a 6 well plate containing a confluent monolayer of HeLa cells (5x10⁵ cells plated the day before). The plate was centrifuged for 10–15 min at 1,200 rpm and incubated for 36-48h at 37°C in the presence of 5% CO₂. At 12-18h post-infection (p.i.), the media was removed and 2ml of fresh media supplemented with 500μg/ml spectinomycin (Fisher BioRegent) and 1μg/ml of cycloheximide (Sigma) was added. The first passage was performed at 36-48h p.i. (36h p.i. was favored for C. muridarum). The media was removed and the infected cells were lysed by addition of 0.5ml of water/well, scraped, collected and spun for 5min at 1,200rpm. The supernatant was diluted in 2ml of media containing 500μg/ml spectinomycin and 1μg/ml cycloheximide and added to a fresh cell monolayer (seeded the day before) and incubated for 48 h at 37°C in the presence of 5% CO2. Passages were performed until enough infectious particles were recovered to generate a frozen stock and proceed to plaque purification. For both C. trachomatis and C. muridarum, transformants were routinely observed at the second passage and occasionally observed after the first passage.

Immunoblotting

Protein samples were separated by SDS-PAGE and transferred to nitrocellulose membranes. The membranes were blocked for 1h at room temperature in 1xPBS containing 0.05% Tween and 5% fat-free milk. Primary and HRP-conjugated secondary antibodies were diluted in 1xPBS containing 0.05% Tween and 5% fat-free milk and respectively incubated overnight at 4°C and 1h at room temperature. Proteins were detected using the Amersham ECL immunoblotting detection reagent as per manufacturer recommendation and a Bio-Rad ChemiDoc imaging system.

Immunofluorescence and microscopy

At the indicated times, HeLa cells seeded onto glass coverslips were fixed for 30 min in PBS containing 4% paraformaldehyde. Immunostainings were performed at room temperature. Antibodies were diluted in PBS containing 0.1% Triton X-100. Samples were washed with PBS and examined under an epifluorescence or spinning disc confocal microscope. Images were processed using the Imaris software (Bitplane, Belfast, United Kingdom).

Antibodies

The following primary antibodies were used: mouse monoclonal anti-FLAG (1:1,000 (IF), 1:10,000 (WB), Sigma), rabbit polyclonal anti-GFP (1:2,000, Molecular Probes), rabbit polyclonal anti-mCherry (1:2,000, BioVision), and rabbit polyclonal anti-actin (1:10,000, Sigma). The rabbit polyclonal anti-OmcA was generated against a synthetic OmcA peptide (CGGDTHQDAEHGPQARE) by Alpha Diagnostic International and used at 1:10,000. The following secondary antibodies were used: peroxidase-conjugated goat anti-rabbit IgG (1:10,000, Jackson ImmunoResearch), peroxidase-conjugated goat anti-mouse IgG (1:10,000, Jackson ImmunoResearch), and goat anti-mouse AlexaFluor 488 (1:500, Invitrogen).

Time-lapse video microscopy

HeLa cells were seeded in 8 well Lab-Tek II Chambered Coverglass dishes (MatTek, Ashland, MA) and infected with the indicated fluorescent C. trachomatis or C. muridarum strain at a multiplicity of infection of 0.2. At 8h p.i. and for the next 70h, images were captured every 20min on a Leica DMi8 confocal microscope equipped with the Andor iXon ULTRA 888BV EMCCD camera, CSU-W1 confocal scanner unit, and a humidified live cell environmental chamber set at 37°C and 5% CO₂. The Imaris imaging software (Bitplane, Belfast, United Kingdom) was used to analyze and process the data. Videos were converted to mp4 format using the HandBrake compression platform.

Spatiotemporal analysis of RB to EB transition in C. trachomatis

Confocal time lapse videos were analyzed using the Imaris imaging software (Bitplane, Belfast, United Kingdom) by tracking individual inclusions in space and time. Quantification was performed on three independent confocal time lapse imaging experiments. A total of 46 primary inclusions and 29 secondary inclusions were analyzed. Secondary inclusions were distinguished from primary inclusions based on time of inclusion appearance and the timing of primary inclusion lysis. The graphs were generated using GraphPad Prism. A Student’s t test was performed and statistical significance was set to P < 0.0001.

Quantification of time to GFP Expression: The frame in which an inclusion containing 4–6 mCherry-positive RBs was clearly identifiable and the frame in which GFP expression is first visible within that same inclusion were identified. The 20-minute frame rate was then used to calculate the time elapsed between 4–6 mCherry-positive RBs and the start of GFP expression.

Quantification of inclusion volume at time of GFP expression: For each inclusion, at the time of first GFP expression, a single three-dimensional object corresponding to the raw mCherry signal of the bacteria was generated. The volume of the inclusion was calculated as the sum of the pixels within the 3D object.

Tracking of the spatial distribution of the GFP signal over time: Individual inclusions were tracked for 120min after GFP expression was first visible. At each 20min time mark, the spatial distribution of the GFP signal within individual inclusions was determined by analyzing individual Z stacks of covering the entire inclusion and recording if the signal was at the edge of the inclusion (edge), within the inclusion lumen (internal), or at both the edge and lumen (edge and internal). The mCherry signal of the bacteria was used to determine the boundaries of the inclusion.

A new set of spectinomycin resistant cloning vectors for Chlamydia trachomatis

Following the development of a transformation system for Chlamydia [14], we described a series of p2TK2_Amp-SW2 E. coli-C. trachomatis shuttle cloning vectors [15, 17]. These plasmids encode an ampicillin resistance cassette. Based on a report by Lowden et al. that spectinomycin could also be used to select C. trachomatis transformants [49], we generated a new series of p2TK2-SW2 plasmids. The simplest vector of this series is referred to as p2TK2_Spec-SW2. It is very similar to p2TK2_Amp-SW2 [15], except that it encodes the aminoglycoside 3′ adenyltransferase gene (aadA), which confers resistance to spectinomycin. The incDEFG promoter, which is expressed throughout the developmental cycle [40], was chosen to drive the expression of the aadA gene to ensure the constitutive expression of the antibiotic resistance. The presence of a multiple cloning site (MCS) facilitates the cloning of any transgene of interest.

p2TK2_Spec-SW2 was further improved by the addition of a constitutively expressed mCherry fluorescent reporter driven by the groESL promoter, as previously described for our p2TK2_Amp-SW2 derivatives [17]. The map and the sequence of the corresponding vector, which is referred to as p2TK2_Spec-SW2 mCh(Gro_L2), are respectively shown in Fig 1A and S1 Fig. We have also adapted p2TK2_Spec-SW2 mCh(Gro_L2), to include a 3xFLAG tag downstream from the NotI restriction site and upstream from the incDEFG terminator, to allow for convenient cloning and expression of translational fusion between C. trachomatis ORFs of interest and the 3xFLAG tag. The map and sequence of the corresponding vector, p2TK2_Spec-SW2 mCh(Gro_L2) MCS-3xFLAG, are presented in Fig 1B and S2 Fig, respectively.

Download:

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

Fig 1. A new set of versatile spectinomycin resistant expression vectors for C. trachomatis.

Vectors maps of p2TK2_Spec-SW2 mCh(Gro_L2) (A), p2TK2_Spec-SW2 mCh(Gro_L2)-MCS-3xFLAG (B), p2TK2_Spec-SW2 mCh(Gro_L2) Tet-IncV-3xFLAG (C), and p2TK2_Spec-SW2 mCh(Gro_L2) Tet-KpnI-IncV-3xFLAG (D). The pSW2 plasmid (SW2) is shown in black; the incDEFG operon promoter in dark blue; the aadA ORF that confers spectinomycin resistance (Spec) in blue; aadA terminator light blue; the E.coli origin of replication (Ori) in grey; the groESL operon promoter in dark red; the mCherry ORF in red; the groESL operon terminator in light red; unique restriction sites in violet; the tet repressor in orange; the tetA promoter in light orange; the incV ORF in yellow; the 3xFLAG tag in teal; the incDEFG operon terminator in light green. (E, G) Immuno-blots of cell lysates from HeLa cells infected with strains of C. trachomatis transformed with p2TK2_Spec-SW2 mCh(Gro_L2) Tet-IncV-3xFLAG (E) or p2TK2_Spec-SW2 mCh(Gro_L2) Tet-KpnI-IncV-3xFLAG (G) expressing mCherry constitutively and IncV-3xFLAG under the control of the aTc inducible promoter. The cells were infected for 24h in the absence (-aTc) or in the presence (+aTc) of 10ng/ml aTc added at 20h p.i. for 4h. Immuno-blots of the corresponding lysates were probed using antibodies against FLAG, mCherry, and actin. (F, H) Confocal micrographs of inclusions of the C. trachomatis strains as described in E and G, respectively. The cells were fixed 24h p.i., immunostained with anti-FLAG antibodies and imaged using a confocal microscope. A single plane crossing the middle of the inclusion is shown. The left panels correspond to the bacteria (Ct mCherry, red) and the middle panels to the 3xFLAG signal (IncV-3xF, green). The merge is shown on the right. Scale bar: 5μm.

https://doi.org/10.1371/journal.pone.0217753.g001

We have also engineered the p2TK2_Spec-SW2 mCh(Gro_L2) vector to include the TetR repressor and the tetA promoter (tetA^P) to express C. trachomatis ORFs from the anhydrotetracycline (aTc)-inducible promoter. A first vector, p2TK2_Spec-SW2 mCh(Gro_L2) Tet-IncV-3xFLAG was constructed by cloning a DNA fragment generated by overlapping PCR and comprising of TetR-tetA^P-IncV into the KpnI and NotI restriction sites of p2TK2_Spec-SW2 mCh(Gro_L2) MCS-3xFLAG. The map and sequence of the resulting plasmid are presented in Fig 1C and S3 Fig, respectively. This configuration allows for a seamless junction between tetA^P and the start codon of IncV, which has been our preferred arrangement to avoid a potential disruption of aTc regulation [17, 27, 29, 51].

In an attempt to streamline the cloning process and avoid multiple rounds of overlap PCR, a second vector, p2TK2_Spec-SW2 mCh(Gro_L2) Tet-KpnI-IncV-3xFLAG was also generated (Fig 1D and S4 Fig for map and sequence respectively). With this configuration, the KpnI and NotI restriction sites present downstream of the TetR-tetA^P regulatory element and upstream of the 3xFLAG allow for a single step cloning of any gene of interest. The potential downside of this configuration is the introduction of a KpnI restriction site between tetA^P and the start codon of the downstream ORF, which could negatively affect the aTc dependent expression (see below).

Altogether, we present a new set of versatile, spectinomycin resistant expression vectors for C. trachomatis.

Addition of nucleotides between tetA^P and the downstream ORF can affect aTc induction

We next tested if introducing a KpnI restriction site between tetA^P and the start codon of the downstream ORF would affect aTc induction. For this purpose, we used the T3SS effector IncV. We generated and introduced in C. trachomatis p2TK2_Spec-SW2 mCh(Gro_L2) Tet-IncV-3xFLAG and p2TK2_Spec-SW2 mCh(Gro_L2) Tet-KpnI-IncV-3xFLAG. The transformants were plaque purified and IncV expression was determined by immunoblot or immunofluorescence in the absence or presence of aTc. As previously observed with our Amp^R vector [29], the Tet-IncV-3xFLAG configuration led to the tight repression of IncV expression in the absence of aTc, with no IncV-3xFLAG protein detectable by immunoblot or immunofluorescence (Fig 1E left lane and 1F top panels). Addition of aTc, led to IncV-3xFLAG expression and localization to the inclusion membrane (Fig 1E right lane and 1F bottom panels). Similar levels of IncV-3xFLAG expression were obtained with aTc induction of the Tet-KpnI-IncV-NotI-3xFLAG construct (Fig 1G, right lane and 1H bottom panels). However, and although to a lesser extent, the IncV-3xFLAG protein was also expressed in the absence of aTc, as determined both by immunoblot and immunofluorescence (Fig 1G left lane and 1H top panels).

Altogether, these results suggest that introducing a KpnI restriction site between tetA^P and the start codon of IncV affected the aTc dependent repression of the promoter leading to weak but constitutive expression of the construct in the absence of inducer.

A spectinomycin resistant cloning vectors for Chlamydia muridarum

Skilton et al. have recently described, pGFP-Nigg, an E. coli-C. muridarum shuttle vector encoding CDS2-8 from the C. muridarum Nigg plasmid and CDS1 from the C. trachomatis SW2 plasmid, an ampicillin resistance cassette, and the green fluorescent protein (GFP) [39]. We have generated a different version of this plasmid, referred to as p2TK2_Spec-Nigg, which includes a spectinomycin selectable marker (aadA under the C. trachomatis incDEFG promoter) and a MCS (see material and methods for details on cloning strategy) (Fig 2A and S5 Fig for map and sequence respectively). In addition, we have introduced the mCherry reporter under the control of the C. trachomatis groESL promoter to generate p2TK2_Spec-Nigg mCh(Gro_L2) (Fig 2B and S6 Fig for map and sequence respectively). The latter was transformed into C. muridarum following the same protocol used for C. trachomatis and led to mCherry expressing C. muridarum at the second passage (Cm mCherry). mCherry expression was stable after plaque purification of the strain (Fig 2C).

Download:

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

Fig 2. C. muridarum strain expressing mCherry from the groESL promoter and TC0273-3xFLAG under the control of the aTc inducible promoter.

(A-B) Vector maps of p2TK2_Spec-Nigg (A) and p2TK2_Spec-Nigg mCh(Gro_L2) (B). The pNigg plasmid is shown in black; incDEFG promoter in dark blue; the aadA ORF that confers spectinomycin resistance (Spec) in blue; the aadA terminator light blue; the E. coli origin of replication (Ori) in grey; the groESL promoter in dark red; the mCherry ORF in red; the groESL terminator in light red; unique restriction sites in purple. (C) Immunofluorescence images of HeLa cells infected for 18h with a C. muridarum strain transformed with p2TK2_Spec-Nigg mCh(Gro_L2) stained with the Hoechst DNA dye. The left panel corresponds to the DNA (blue) with the nucleus labeled with “N” and the inclusion annotated with an asterisk. The middle panel corresponds to the bacteria (Cm mCherry, red), and the merge is shown on the right. Scale bar: 10μm. (D) Vector map of p2TK2_Spec-Nigg mCh(Gro_L2) TetTC0273-3xFLAG. The pNigg plasmid is shown in black; the incDEFG promoter in dark blue; the aadA ORF that confers spectinomycin resistance (Spec) in blue; the aadA terminator light blue; the E. coli origin of replication (Ori) in grey; the groESL promoter in dark red; mCherry ORF in red; the groESL terminator in light red; the Tet repressor in orange; the tetA promoter in light orange; the tc0273 ORF in yellow; the 3xFLAG in teal; the incDEFG terminator in light green; unique restriction sites in purple. (E) Immuno-blot of cell lysates from HeLa cells infected with a strain of C. muridarum transformed with p2TK2_Spec-Nigg mCh(Gro_L2) TetTC0273-3xFLAG vector expressing mCherry constitutively and TC0273-3xFLAG under the control of the aTc inducible promoter. The cells were infected for 24h in the absence (-aTc) or presence (+aTc) of 10ng/ml aTc added at 2h p.i.. Immuno-blots of the corresponding lysates were probed using antibodies against FLAG, mCherry, and actin. (F) Confocal micrographs of inclusions of the C. muridarum strain as described in E. Cells were fixed 18h p.i., immunostained with anti-FLAG antibody and imaged using a confocal microscope. A single plane crossing the middle of the inclusion is shown. The left panels correspond to the bacteria (Cm mCherry, red), and the middle panels correspond to the 3xFLAG signal (TC0273-3xF, green). The merge is shown on the right. Scale bar: 5μm.

https://doi.org/10.1371/journal.pone.0217753.g002

Altogether, these results indicate that spectinomycin can be used as a selectable marker for C. muridarum and that C. trachomatis promoters (i.e. incDEFG and groESL operon) and mCherry are expressed in C. muridarum.

aTc inducible gene expression in C. muridarum

We next tested if the Tet system could be used for inducible gene expression in C. muridarum. For this purpose, the tc0273 ORF, which encodes the C. muridarum IncV homolog, was cloned as a 3xFLAG translational fusion under the control of the tetA inducible promoter into the KpnI and NotI restriction sites of p2TK2_Spec-Nigg mCh(Gro_L2). Based on our results with C. trachomatis (Fig 1), we chose the configuration in which the tc0273 ORF is seamlessly attached to tetA^P to ensure tight repression. The map and sequence of the resulting plasmid are presented in Fig 2D and S7 Fig, respectively. C. muridarum transformation resulted in mCherry expressing transformants that were plaque purified. In the absence of aTc, the resulting strain expressed mCherry only, as assayed by immunoblot and immunofluorescence (Fig 2E left lane and 2F top panels). In the presence of aTc, TC0273-3xFLAG expression was induced and detectable by immunoblot (Fig 2E right lane). TC0273-3xFLAG expression was confirmed by immunofluorescence microscopy, which also revealed that, as observed for its C. trachomatis homologue, TC0273 localized to the inclusion membrane (Fig 2F bottom panels).

Altogether, these results demonstrated that the Tet system can be used for inducible gene expression in C. muridarum and for studying the cellular localization of C. muridarum effectors, such as Inc proteins.

A transcriptional fluorescent reporter to monitor RB to EB transition

Constitutive expression of a fluorescent reporter such as mCherry is a useful tool to monitor inclusion expansion, however it does not inform if and when the bacteria have undergone RB or EB conversion. For this purpose, we took advantage of the fact that a subset of C. trachomatis genes are expressed late in the developmental cycle, as the bacteria transition from RB to EB [40, 41]. One such gene is omcA. GFP was cloned under the control of the C. trachomatis omcA promoter and introduced between the KnpI and NotI restriction sites of p2TK2_Spec-SW2 mCh(Gro_L2). The resulting plasmid, p2TK2_Spec-SW2 mCh(Gro_L2) GFP(OmcA_L2) (Fig 3A and S8 Fig for map and sequence, respectively) was transformed into C. trachomatis and the resulting strain, referred to as Ct mCh(Gro_L2) GFP(OmcA_L2), was plaque purified. Expression of mCherry or mCherry and GFP did not affect bacterial replication (not shown).

Download:

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

Fig 3. Transcriptional fluorescent reporters to monitor RB to EB transition in C. trachomatis and C. muridarum.

Vector maps of p2TK2-SW2 mCh(Gro_L2)GFP(OmcA_L2) (A) and p2TK2-Nigg mCh(Gro_L2)GFP(OmcA_Cm) (B). The pSW2 and pNigg plasmids are shown in black; the incDEFG promoter in dark blue; the spectinomycin resistance gene (aadA) and ampicillin resistance gene (bla) in blue; the terminator in light blue; the E. coli origin of replication (Ori) in grey; the groESL promoter in dark red; mCherry ORF in red; the groESL terminator in light red; the omcA promoter in dark green; GFP in green; the incDEFG terminator in light green; unique restriction sites in purple. (C) Immuno-blot of cell lysates from HeLa cells infected with a strain of C. trachomatis transformed with p2TK2_Spec-SW2 mCh(Gro_L2) or p2TK2_Spec-SW2 mCh(Gro_L2) GFP(OmcA_L2). Lysates were collected at the indicated time and probed using antibodies against GFP, OmcA, mCherry, and actin. (D-E) Quantification of the signal intensity of the indicated marker from immunoblots as shown in (C). The average signal intensity from three independent experiments is shown. mCherry: red, OmcA: black and GFP: green. (F-G) Selected 3-dimensional merged frames from Video S1 (Top panels, (F)), Video S2 (Bottom panels, (F)), Video S3 (Top panels, (G)) and Video S4 (Bottom panels, (G)) acquired by time-lapse video confocal microscopy of HeLa cells infected with C. trachomatis (F) or C. muridarum (G) expressing mCherry (red) from the groESL promoter (Top panels) or mCherry (red) and GFP (green) under the control of the C. trachomatis groESL and omcA promoter, respectively (Bottom panels). Frames were selected to provide representative examples of primary (left) and secondary (right) inclusions. The time p.i. is indicated above each frame. Scale Bar: 10μm.

https://doi.org/10.1371/journal.pone.0217753.g003

HeLa cells were infected with this strain or with C. trachomatis expressing mCherry only (Ct mCh(Gro_L2)). At various times from 18h p.i., when the inclusions mostly harbor RBs, to 44h p.i., when C. trachomatis completes its developmental cycle and most of the bacteria have converted to EB, samples were collected and analyzed by immunoblot using antibodies against GFP, OmcA, mCherry and actin (Fig 3C). Quantification of three independent experimental replicates are presented in Fig 3D and 3E. As expected for a constitutively expressed reporter, both strains were positive for mCherry at all time points analyzed (Fig 3C mCherry and 3D and 3E red line). Also as expected, the overall intensity of the mCherry signal increased over time, especially during the first 30h, reflecting RB multiplication. For both strains, a weak OmcA signal was detectable starting at 28h p.i., when asynchronous RB to EB conversion is occurring. The OmcA signal intensified over time as the developmental cycle progressed and more RBs transitioned into EBs (Fig 3C OmcA and 3D and 3E black line). Additionally, GFP was only detected in the Ct mCh(Gro_L2) GFP(omcA_L2) strain, and most importantly for the validation of our approach, the expression of GFP mirrored the expression of endogenous OmcA (Fig 3C GFP and 3D and 3E green line).

Altogether these results indicate that, at the population level, GFP expression from the C. trachomatis omcA promoter correlates with the expression of endogenous OmcA and therefore constitute an adequate reporter of the RB to EB conversion during C. trachomatis developmental cycle.

Visualizing C. trachomatis RB to EB transition by time lapse-microscopy

We next performed time-lapse confocal microscopy to visualize the expression of the mCherry and GFP reporters at a single inclusion level. For this purpose, HeLa cells infected with the Ct mCh(Gro_L2) or Ct mCh(Gro_L2) GFP(OmcA_L2) strain were imaged starting 8h p.i.. Z stacks were acquired every 20min for 70h to capture several rounds of infections. Selected frames and the corresponding movies are presented in Fig 3F and S1 and S2 Movies, respectively. With the Ct mCh(Gro_L2) strain, mCherry expression was detected early when the inclusion harbored only a few bacteria (Fig 3F, top panels, Primary inclusions, 16h). Moreover, the bacteria remained fluorescent during the entire duration of the movie, which included the formation of primary and secondary inclusions (Fig 3F, top panels, Secondary inclusions, 56h). The Ct mCh(Gro_L2) GFP(omcA_L2) strain presented similar timing of mCherry expression to the Ct mCh(Gro_L2) strain (Fig 3F, bottom panels, Primary inclusions). In addition, in primary inclusions, a few bacteria started to express GFP at 26h p.i. (Fig 3F, bottom panels, Primary inclusions, 26h). The number of GFP-positive bacteria per inclusion progressively increased over time, until the inclusion was eventually filled with GFP-expressing bacteria (Fig 3F, bottom panels, Primary inclusions). Importantly, the same pattern of mCherry and GFP expression was observed during secondary infections. Early on, secondary inclusions were mCherry-positive only and displayed increasing GFP expression as time progressed (Fig 3F, bottom panels, Secondary inclusions, compare 56h and 60h), further confirming the temporal regulation of the GFP reporter.

Altogether, these results indicate that our fluorescent reporters allow for live imaging of early to late events of C. trachomatis developmental cycle. Moreover, we showed that, in C. trachomatis, the omcA promoter-dependent expression of GFP leads to a temporal expression, which correlates with the late developmental stages of the bacteria, therefore constituting an adequate reporter for monitoring C. trachomatis RB to EB conversion.

Visualizing C. muridarum developmental cycle, including RB to EB transition, by time lapse-microscopy

We next investigated if the RB to EB transition omcA reporter could be used in C. muridarum. For this purpose, GFP was cloned under the control of the C. muridarum omcA promoter and introduced between the KpnI and NotI restriction sites of p2TK2_Spec-Nigg mCh(Gro_L2). The resulting plasmid, p2TK2_Spec-Nigg mCh(Gro_L2) GFP(OmcA_Cm) (Fig 3B and S9 Fig for map and sequence, respectively) was transformed into C. muridarum and the derived strain, Cm mCh(Gro_L2) GFP(OmcA_Cm) was plaque purified. HeLa cells infected with Cm mCh(Gro_L2) or Cm mCh(Gro_L2) GFP(OmcA_Cm) were analyzed by time-lapse confocal microscopy under the same conditions described above for C. trachomatis. Selected frames and the corresponding movies are presented in Fig 3G and S3 and S4 Movies, respectively. Live imaging of the Cm mCh(Gro_L2) strain revealed that the mCherry reporter was expressed early and throughout the duration of the movie which included secondary infections (Fig 3G Top panels). In addition, comparison of the kinetics of the developmental cycle of the mCherry expressing C. trachomatis and C. muridarum strains, indicated that C. muridarum developmental cycle appeared to progress faster than C. trachomatis developmental cycle (Compare time points of the selected frames in Fig 3F and 3G). Live imaging of the Cm mCh(Gro_L2) GFP(OmcA_Cm) strain revealed that, early on, inclusions were expressing mCherry only, but progressive and increasing appearance of the GFP signal occurred as the developmental cycle progressed (Fig 3G Bottom panels, Primary inclusions). Moreover, nascent secondary inclusions were at first mCherry-positive but GFP-negative (Fig 3G bottom panels, Secondary inclusions, compare 45h and 50h).

Altogether, these results indicated that our mCherry reporter allows for live imaging of early and late events of C. muridarum developmental cycle. They also revealed that in our experimental set up C. muridarum appeared to go through one round of infection at a faster pace than C. trachomatis. Moreover, our data indicated that GFP expression from the C. muridarum omcA promoter was temporally regulated and coincided with late stages of the developmental cycle, supporting the use of this reporter for monitoring RB to EB conversion.

Spatiotemporal analysis of RB to EB transition in C. trachomatis

To gain insight into the spatiotemporal dynamics of the RB to EB transition, we analyzed when and where the GFP signal was first observed. Time-lapse confocal microscopy was performed on HeLa cells infected with the Ct mCh(Gro_L2) GFP(OmcA_L2) strain, as described in Fig 3, and individual inclusions were tracked over time. We first quantified the time required for inclusions containing 4 to 6 mCherry-positive bacteria, a stage at which we could confidently identify inclusions, to display the first GFP-positive bacteria (Fig 4A). The fastest and slowest primary inclusions displayed GFP-positive bacteria after 240min and 440min, respectively. Overall, the average time for the initiation of RB to EB conversion in primary inclusion was 365min +/-53min, once the inclusion contains 4–6 bacteria. This number was significantly reduced in secondary inclusions (268min +/-64min). We also determined that, when the first GFP-positive bacteria were observed, the average volume of the inclusions was 423μm³ +/-116 μm³ and 139 μm³ +/-45 μm³ in primary and secondary inclusions, respectively (Fig 4B).

Download:

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

Fig 4. Spatiotemporal analysis of RB to EB transition in C. trachomatis.

(A-B) Quantification of time from 4–6 RBs per inclusion to first GFP expression in minutes (A) and of the volume of the inclusions at the time of first GFP expression in μm³ (B). Measurements were performed on primary and secondary inclusions harboring the C. trachomatis mCh(Gro_L2) GFP(OmcA_L2) strain and imaged by time-lapse video confocal microscopy. Each circle represents one inclusion. Error bars are Standard Deviation. ****P < 0.0001. (C) Representative 3-dimensional merged frames acquired by time-lapse video confocal microscopy of C. trachomatis mCh(Gro_L2) GFP(OmcA_CL2) inclusions exhibiting GFP-positive bacteria at the edge only (top panel, light green), both edge and internal (middle panel, dark green), and internal only (bottom panel, orange). Red: mCherry, Green: GFP. Scale Bar: 5 μm. (D) Tracking of the spatiotemporal distribution of the GFP signal over time in primary inclusions using categories defined in (C). Time prior to GFP expression is represented in red. Inclusions were grouped in three classes (I, II, III) based on spatiotemporal pattern of GFP expression. (E) Selected 3-dimensional merged frames acquired by time-lapse video confocal microscopy and representative of Class I inclusions, where GFP localization change from edge only to edge and internal over time. Red: mCherry, Green: GFP. Scale Bar: 5μm.

https://doi.org/10.1371/journal.pone.0217753.g004

We next determined if the GFP signal was occurring randomly or at a specific location within the inclusion. For this purpose, primary inclusions were monitored until GFP expression was first detected and the spatial distribution of the GFP signal was tracked and recorded over time. Three categories of inclusions were observed. The GFP signal was at the edge only, at the edge and internally, or internally only (Fig 4C, light green, dark green and orange respectively). In all cases, at the time of conversion, the GFP signal was restricted to one to three bacteria. Tracking of multiple primary inclusions revealed that for 85% of the inclusions, an example of which is provided in Fig 4E, the GFP signal appeared first in one to three bacteria localized at the edge of the inclusion only (Fig 4C, top panel, light green and 4D, Class I, light green). As time progressed, the GFP signal was observed in bacteria localizing at the edge of the inclusion but also in the center of the inclusion (Fig 4C, middle panel, dark green and 4D, Class I, dark green). The time to transition from inclusions containing GFP-positive bacteria at the edge only, to inclusions with GFP-positive bacteria at the edge and internally, ranged from 20min to 120min (Fig 4D, Class I). Two inclusions, annotated with an asterisk, required longer (240min and 160min). For about 6% of the primary inclusions, the GFP signal was observed at the edge and internally at all time points analyzed (Fig 4C, middle panel, dark green and 4D, Class II, dark green). Because frames were acquired every 20min, it is possible that we were not able to capture these inclusions with GFP-positive bacterium at the edge only, despite its occurrence. If so, it would suggest that transition from edge only to edge and internal could occur in less that 20min. Finally, 9% of the primary inclusions displayed the GFP signal internally at first and at the edge and internally later on (Fig 4C, bottom panel, orange and 4D, Class III). It is also possible that the initial GFP expressing bacteria at the edge of the inclusion was missed because of the timing of the acquisition. Secondary inclusions followed the same trend (not shown).

Altogether our spatiotemporal analysis of RB to EB conversion suggests that this process starts at a specific time (365min +/-53min and 268min +/-64min in primary and secondary inclusions, respectively), after the bacteria have undergone 2–3 rounds of replication, and correlates with a specific size of the inclusion (423μm³ +/- 116μm³ and 139μm³ +/- 45μm³ in primary and secondary inclusions, respectively). In addition, we provide evidence that the first bacteria to convert localize at the edge of the inclusion and that within 20-120min converting/converted bacteria are observed at the edge and in the lumen of the inclusion.

Spectinomycin resistant vectors for Chlamydia

Beta-lactams, such as ampicillin or penicillin, were the first antibiotics used for the selection of C. trachomatis transformants [14]. Chloramphenicol and spectinomycin selections were described later on [21, 49]. However, most of the C. trachomatis vectors available so far display the beta-lactamase gene and therefore cannot be used for the complementation of TargeTron or FRAEM mutants in which the beta-lactamase gene either interrupts or replaces the targeted gene [22, 52]. To circumvent this problem and complement a ct383::bla mutant, Weber et al. inserted the aadA gene under the control of the incD promoter in their pBOMB4-Tet-CT383 vector, leading to a vector that conferred resistance to spectinomycin, in addition to ampicillin [20]. Our p2TK2-SW2 derivatives offer the advantage of conferring resistance to either ampicillin (p2TK2_Amp-SW2) or spectinomycin (p2TK2_Spec-SW2). Moreover, since their MCS are compatible, transgenes can easily be transferred from one vector to another.

Cloning downstream of the tet^A promoter

We have cloned the C. trachomatis incV ORF downstream of the tet^A promoter, either seamlessly, or by introducing a KpnI restriction site between the promoter and the start codon. Seamless junction led to a tight aTc regulation with the construct being off or on in the absence or presence of aTc, respectively. In addition, when expressed, IncV localized to the inclusion membrane. This is in contrast with the construct that included a KpnI site. This configuration led to leaky expression of IncV in the absence of inducer. In addition, a great proportion of the IncV molecules appeared to be trapped inside the bacteria. The only difference between the 2 constructs is the presence of the KpnI restriction, which would suggest that the addition of the corresponding 6 nucleotides is responsible for affecting tight repression, but also translocation. Potential reasons for the latter are unclear to us. Wickstrum et al. included an AgeI site between the tet^A promoter and start codon of GFP without affecting regulation [16]. Similarly, Bauler et al. used a NotI site [18]. At this point we cannot conclude if our results are specific to IncV and/or KpnI; regardless, they indicate that caution should be taken when designing strategies to clone effectors downstream of the tet^A promoter.

A versatile vector for Chlamydia muridarum

Since the development of genetic tools to study Chlamydia, there has been a bias towards C. trachomatis. Tools for C. muridarum, the mouse adapted strain commonly used to model human genital tract ascending infections in mice, are still limited. Most of the vectors described so far have been used to investigate the role of plasmid encoded ORFs in C. muridarum virulence [34, 35, 37–39]. Here, we present C. muridarum vectors that are as versatile as the one described for C. trachomatis. We have shown that in addition to GFP [39] and luciferase [37], mCherry can be expressed in C. muridarum without affecting its developmental cycle. Our study also reveals that spectinomycin can be used to select for transformants, hereby providing an additional selection marker for C. muridarum. By including a MCS, to facilitate the cloning of any transgene of interest, and an aTc inducible promoter, we have performed a tagged effector localization study. Our results indicate that the Tet inducible system is functional in C. muridarum and confirmed the predicted inclusion localization of TC0273 [11]. Altogether, increasing the versatility of C. muridarum vectors will be instrumental for mutant complementation and subsequent in vivo studies.

We were able to image C. muridarum developmental cycle using a constitutive mCherry reporter or a reporter for RB to EB conversion. Our live imaging analysis indicates that constitutive mCherry expression allows for detection of nascent inclusions containing a few RBs. In addition, we noticed that C. muridarum secondary inclusions are detected earlier than C. trachomatis secondary inclusions and that the developmental cycle progresses almost twice as fast as C. trachomatis developmental cycle, in agreement with a previous report [53].

Advantages of a fluorescent reporter to monitor RB to EB conversion

Inclusion Forming Units (IFUs) assay, where the number of infectious bacteria (i.e. EBs) recovered at a given stage of the developmental cycle is determined by infection of a fresh monolayer of cells with serial dilutions of infected cells lysates, is a standard assay to determine if Chlamydia undergoes a productive developmental cycle. While accurately measuring the amount of EBs, this assay is labor intensive, limited in the number of conditions that can be tested at once, and not easily adaptable to high-throughput screening. Moreover, it does not provide information on potential defects in entry, EB to RB conversion or RB replication.

Electron microscopy (EM) has been instrumental in investigating RB to EB conversion due to the clear morphological difference of the two developmental forms [5–7]. Although EM offers the unambiguous identification of the different Chlamydia forms, it is labor intensive and the number of inclusions that can be analyzed is limited. In addition, 3D EM revealed that the distribution of EB and RB within the inclusion is heterogeneous, therefore introducing a bias when analyzing inclusion content by conventional 2D EM [7].

Chlamydia intracellular development can also be evaluated by using a fluorescence-based assay where the surface area of the inclusions, as determined per immunostaining of the bacteria or using fluorescent strains, is quantified using computer assisted software. This assay allows for the analysis of a large number of inclusions and is amenable to high-throughput formats. However, it does not provide a full picture of the developmental cycle. Some inclusions could be large and yet mostly contain RBs. On the contrary, small inclusions could be filled with EBs, as shown with Brefeldin A treatment [54].

Here, we have designed a fluorescent reporter to monitor RB to EB conversion. Our rationale stemmed from earlier studies that demonstrated temporal expression of specific Chlamydia genes throughout the developmental cycle [40, 41]. Of particular interest to our design were genes, like omcA, whose transcription initiation corresponded to the time of RB to EB conversion. As anticipated, the timing and asynchronous nature of the expression of the transcriptional omcA::gfp reporter closely mimics the expression of endogenous OmcA. In addition, the timing of GFP expression, as a readout for RB to EB conversion, was consistent with results from previous EM studies. For example, in their 3D EM analysis of the developmental cycle, Lee et al. observed the first intermediate bodies at 24h p.i. [7], which is the same time point at which we observed the first GFP-expressing C. trachomatis. Moreover, Lee et al. also reported that by 40h p.i., 70% of the bacteria were EBs, which is also in agreement with our C. trachomatis imaging results where the inclusions were filled with EBs at that time point.

Altogether, our RB to EB conversion reporter offers an alternative that combines the advantages of the assays listed above, while alleviating some of the limitations. Fluorescence microscopy analysis of cells infected with Chlamydia strains encoding the fluorescent RB to EB conversion reporter allows for quantification, in a single assay, of: 1) the size of the inclusion based on the mCherry signal, 2) the proportion of GFP-positive inclusions in which RB have converted to EB, 3) the proportion of RB that have converted to EB in a given inclusion (i.e. GFP-positive bacteria). Moreover, live imaging can be used to measure the timing of these events. These readouts would be applicable to genome wide siRNA or CRISPR screens or screening of small molecules that inhibit Chlamydia developmental cycle. The reporter could also be introduced in Chlamydia mutants to monitor potential growth defects. Finally, one could also envision that this new tool could be instrumental in visualizing the RB to EB transition in vivo.

Spatiotemporal analysis of RB to EB transition in C. trachomatis

By tracking individual inclusions harboring our C. trachomatis RB to EB reporter strain, we were able to gather spatiotemporal information of RB to EB conversion. In primary inclusions, the timing of the initiation of RB to EB conversion was fairly homogenous and averaged to 365min +/-53min, after the bacteria had undergone 2–3 rounds of replication. It also correlated with similar inclusion volumes (423μm³ +/-116μm³). In secondary inclusions, although the average time to conversion was also consistent from one inclusion to another, it was significantly faster than in primary inclusions (268min +/-64min) and correlated with smaller inclusions (139μm³ +/-45μm³). In our experimental set up, the environment in which primary and secondary inclusions matured was different. Primary inclusions developed in an undisturbed environment, whereas secondary inclusions developed in the presence of lysed infected cells and high numbers of released EB that continue to infect the monolayer. In addition, it is possible that specific nutrients are being depleted, resulting in some level of starvation of the bacteria in secondary inclusions. Altogether our results indicate that the initiation of RB to EB conversion occurs at a specific time after the RBs start to replicate and once the inclusion has reached a certain size. However, the timing and inclusion size appear to vary depending on the environment, suggesting that external signals may also influence this process.

We were also able to visualize where RB to EB conversion is initiated. Our data indicate that the first one to three bacteria to undergo conversion are localized at the edge of the inclusion. Within 20–120 min of the first conversion event, a different set of GFP-expressing bacteria localized at the edge and some GFP-positive bacteria were present in the lumen of the inclusion. These results could fit a model by which RB to EB transition is initiated in bacteria that are in close proximity to the inclusion membrane and is followed by the subsequent accumulation of converting/converted bacteria in the center of the inclusion. This model would follow the one proposed by Wilson et al. in which replicating RBs are attached to the inclusion though their type III secretion system and over time detachment of RBs from the inclusion membrane provides the signal for differentiation into EB which accumulate in the lumen of the inclusion [47]. Because our experimental setup does not allow us to determine if converting RBs are physically attached to the inclusion membrane, we cannot confirm at this time if RB detachment is required to initiate conversion or if completion of the conversion requires detachment. It is also unclear if the size of the RBs controls this process [7]. Regardless, our data clearly indicate that the first event of RB to EB conversion occurs in RB localized at the periphery of the inclusion.