Development of a Plasmid Shuttle Vector System for Genetic Manipulation of Chlamydia psittaci

Psittacosis, caused by avian C. psittaci, has a major economic impact in the poultry industry worldwide and represents a significant risk for zoonotic transmission to humans. In the past decade, the tools of genetic manipulation have been improved for chlamydial molecular studies. While several genetic tools have been mainly developed in Chlamydia trachomatis, a stable gene-inducible shuttle vector system has not to date been available for C. psittaci. In this study, we adapted a C. trachomatis plasmid shuttle vector system to C. psittaci. We constructed a C. psittaci plasmid backbone shuttle vector called pCps-Tet-mCherry. The construct expresses GFP in C. psittaci. Importantly, exogeneous genes can be inserted at an MCS and are regulated by a tet promoter. The application of the pCps-Tet-mCherry shuttle vector system enables a promising new approach to investigate unknown gene functions of this pathogen.

T he obligate intracellular Gram-negative bacterium Chlamydia psittaci (C. psittaci) is an important zoonotic pathogen that can be encountered in more than 400 bird species, as well as in sheep, cattle, swine, horses, goats, and cats (1)(2)(3). When avian C. psittaci strains infect humans, they can cause severe atypical pneumonia, with even a fatal outcome in some cases. In contrast, nonavian C. psittaci strains might be excluded as potential zoonotic risk factors since the case reports of human psittacosis usually include an avian source (1)(2)(3).
In host cells, chlamydiae undergo a unique developmental cycle that alternates between two distinct bacterial forms, infectious elementary bodies (EBs) and the replicating reticulate bodies (RBs). The developmental cycle is completed within the confines of a membrane-bound vacuole termed the inclusion (4). During infection, chlamydiae interact with various host cell organelles to acquire host-derived nutrients for their survival (5)(6)(7).
The lack of tools for genetic manipulation of chlamydiae hampered molecular research progress for many years. However, Binet and Maurelli first demonstrated transformation of C. psittaci, showing successful allelic exchange on the bacterial chromosome (11). Afterward, Wang et al. demonstrated a stable targeted genetic modification method of C. trachomatis using a plasmid shuttle vector (12). Their system has been modified for other Chlamydia species (spp.) including C. pneumoniae and C. muridarum, but not C. psittaci (12)(13)(14). To replicate efficiently, the plasmid shuttle vector usually has to match the chlamydial plasmid with the target host species due to the presence of replications barriers (13)(14)(15). In this study, we established a stable targeted genetic modification system for C. psittaci infections.

RESULTS
Construction of a plasmid shuttle vector for transformation of C. psittaci. Plasmid DNA sequence comparison of a nonavian isolate C. psittaci 01DC12 (GenBank: HF545615.1) shares 99.97% identity to the avian isolate C. psittaci 6BC (GenBank: CP002587.1) (Fig. S1 in the supplemental material). In the amino acid sequence of protein coding regions, only one amino acid is different between pCps6BC and p01DC12 (Fig. S1). Due to the sequence similarity and biosafety considerations, the p01DC12 plasmid derived from C. psittaci 01DC12 was selected for construction of the shuttle vector.
We constructed a 12,008-bp pCps-Tet-mCherry shuttle vector using a fragment of pBOMB4-Tet-mCherry, including genes for the green fluorescent protein (GFP), mCherry, and ampicillin resistance (AmpR) (16), along with the full sequence of the p01DC12 plasmid derived from C. psittaci 01DC12 (Fig. 1A). mCherry can be replaced with various target genes at a multiple cloning site (MCS) using restriction digestion. Target genes can be regulated by a tetracycline-inducible (tet) promoter (Fig. 1B).
The expected plasmid size of 12,008 bp was confirmed by digestion of pCps-Tet-mCherry with NotI (Fig. S2). Sequencing of the pCps-Tet-mCherry plasmid shuttle vector revealed that the full sequence of p01DC12 was 100% identical to the original sequence (GenBank: HF545615.1). The origin of replication in pCps-Tet-mCherry was identical to pGFP::SW2 (12). The amino acid sequence of mCherry was identical to that of the pMCherry-C1 vector (TaKaRa Bio, Saint-Germain-en-Laye, France). Another region of the pBOMB4-Tet-mCherry fragment was 100% identical to the literature data (Gen-Bank: KF790910.1) (16).
Since our plasmid shuttle vector was constructed from a plasmid of C. psittaci 01DC12, we selected the same strain as a control to perform our initial transformation. After transformation of C. psittaci 01DC12 with pCps-Tet-mCherry and infection in epithelial cells, a strong GFP signal was detected in C. psittaci 01DC12 (C. psittaci 01DC12-pCps-Tet-mCherry) inclusions at 24 and 48 h postinfection (hpi) (Fig. S3A). Furthermore, mCherry was successfully induced by both 10 and 100 ng/ml anhydro-tetracycline hydrochloride (aTC) treatment, and we did not detect mCherry expression in the absence of aTC ( Fig. S3A and S3B).
Due to their genome sequence similarity and biosafety considerations, we next investigated whether C. psittaci strain 02DC15 could be transformed with pCps-Tet-mCherry. Following transformation, we observed not only expression of GFP, but also GFP, mCherry, and the Tet repressor are shown in green, red, and light blue, respectively. AmpR, pUC ori, and MCIP are shown in light gray. The full sequence of p01DC12 was amplified from C. psittaci 01DC12. Another fragment was amplified from pBOMB4-tet-mCherry. mCherry can be induced by a tetracyclineinducible promoter. (B) An MCS containing NotI, PstI, KpnI, and SalI restriction sites in pCps-Tet-mCherry.
Impact of pCps-Tet-mCherry on the growth of C. psittaci. We investigated whether the transformation of C. psittaci using pCps-Tet-mCherry affected chlamydial growth and morphological characteristics ( Fig. 3 and Fig. S4). While penicillin (PEN) treatment slightly reduced the growth of C. psittaci 02DC15-pCps-Tet-mCherry compared to wild-type C. psittaci 02DC15, C. psittaci 02DC15-pCps-Tet-mCherry showed similar growth characteristics compared to wild-type C. psittaci 02DC15 in the absence of PEN (Fig. 3A). Moreover, we confirmed both 10 and 100 ng/ml of aTC did not inhibit the growth of C. psittaci 02DC15-pCps-Tet-mCherry (Fig. 3A). The same trend was observed using the control strain C. psittaci 01DC12 (Fig. S4A). Immunofluorescence analysis revealed similar inclusion morphology between wild-type and transformed C. psittaci regardless of the presence or absence of aTC ( Fig. 3B and Fig. S4B). These data indicate that pCps-Tet-mCherry itself and aTC treatment at concentrations used in this study do not interfere with chlamydial growth characteristics.
Plasmids are generally prone to being lost without selection pressure. We next investigated the shuttle vector/plasmid stability in transformed C. psittaci. Plasmid shuttle vector pCps-Tet-mCherry was stably maintained in both C. psittaci 02DC15-pCps-Tet-mCherry and the control strain C. psittaci 01DC12-pCps-Tet-mCherry in the presence of PEN ( Fig. 4A and Fig. S6A). Although copy numbers were similar in the presence or absence of PEN at initial culture, pCps-Tet-mCherry was significantly reduced in the absence of PEN from passage 1 ( Fig. 4A and B, Fig. S6A and B). This indicates that the pCps-Tet-mCherry plasmid could be stably retained with selection pressure for long-term maintenance.
Impact of aTC on host cell metabolism. It is known that a long-term treatment of tetracycline inhibits host cell metabolism, especially mitochondrial function (18). Therefore, we investigated the impact of aTC on mitochondrial activity. aTC (10 ng/ml) at the concentration used in this study attenuated host cell mitochondrial activity, as indicated by basal respiration, ATP production, and maximal respiration (Fig. 5A). This indicates that we have to take into account the effect of aTC on host cell functions. However, when we compared mitochondrial activity in aTCtreated uninfected control cells to aTC-treated C. psittaci 02DC15-pCps-Tet-mCherry infected cells, significantly upregulated mitochondrial activity was observed in aTCtreated C. psittaci 02DC15-pCps-Tet-mCherry infected cells ( Fig. 5B and C). This upregulation was similar to that observed in wild-type C. psittaci 02DC15 inclusion-bound mitochondria ( Fig. S7A and B).

DISCUSSION
C. psittaci has the capability of causing severe disease in animals and humans (1-3). However, due to the lack of genetic manipulation tools for this important pathogen, molecular studies of C. psittaci are lagging far behind that of other human chlamydial pathogens, such as C. trachomatis.
Development of chlamydial shuttle vectors and stable transformation of C. trachomatis, C. muridarum, and C. pneumoniae have been established only in the last decade (12)(13)(14)19). Bauler and Hackstadt further developed a novel plasmid shuttle vector pBOMB4-Tet-mCherry, which encodes the Tet-inducible promoter system for the expression of recombinant target proteins in C. trachomatis (16). Moreover, Weber et al. were able to identify 10 novel inclusion membrane proteins in C. trachomatis infection using this shuttle vector system (20).
Considering there are some barriers to plasmid replication (13)(14)(15) and the advantages of pBOMB4-Tet-mCherry reporter system, the C. psittaci backbone plasmid shuttle vector pCps-Tet-mCherry was constructed from the full sequence of p01DC12 and the fragment of C. trachomatis shuttle vector pBOMB4-Tet-mCherry. We showed that GFP was expressed after transformation of the control strain C. psittaci 01DC12 using pCps-Tet-mCherry. Since avian C. psittaci strains cause severe lung diseases in humans, it is important to assess whether pCps-Tet-mCherry system works in avian C. psittaci strains, including C. psittaci 6BC. However, conducting the experiment using C. psittaci 6BC is generally troublesome due to stricter regulations for the biosafety level. Therefore, we used C. psittaci 02DC15 as a substitute for avian C. psittaci 6BC. While whole-genome analysis using UpSetR (21) revealed a total of 984 CDS in C. psittaci 6BC  (24). Therefore, successful transformation using C. psittaci 02DC15 is particularly meaningful with respect to the biology of C. psittaci infection.
In addition to transformation of C. psittaci 02DC15, the induction of mCherry expression was controlled by aTC treatment. In this study, mCherry was used as an exogenous gene, but it is possible to insert other target genes into the MCS. Therefore, the pCps-Tet-mCherry is a useful plasmid shuttle vector to investigate the function of specific proteins during C. psittaci infection. A tetracycline-inducible system is beneficial for biological research. However, it has been demonstrated that Ͼ100 ng/ml of tetracycline or its derivatives attenuates mitochondrial activity (18,29,30). Since mitochondrial function is essential for chlamydiae to sustain their life (8,31,32), we investigated the impact of aTC on mitochondria in C. psittaci infection.
Oxygen consumption of C. psittaci 02DC15-pCps-Tet-mCherry-infected cells was measured by a Seahorse XF analyzer with solid state sensor probes containing polymer embedded fluorophores (33). The peak absorption and emission of the oxygen sensor are 530 and 650 nm, respectively. Although C. psittaci 02DC15-pCps-Tet-mCherry expresses GFP and mCherry, intracellular dye fluorescence does not interfere with optical sensors given enough distance to the object, as various groups have demonstrated (34,35).
Similar to previous studies (18,29,30), 10 ng/ml of aTC attenuated the mitochondrial oxygen consumption rate in noninfected epithelial cells. However, we could show that mitochondrial activity was significantly upregulated in C. psittaci 02DC15-pCps-Tet-mCherry-infected cells under aTC treatment, similar to that observed in wild-type C. psittaci 02DC15 infection in the absence of aTC. These data indicate that pCps-Tet-mCherry can be a beneficial tool for further biological studies in C. psittaci infection, though we need to take into account the impact of aTC on host cell functions.
Recently, gene knockout techniques such as the TargeTron system have also been established in C. trachomatis (36,37). Our plasmid shuttle vector pCps-Tet-mCherry and the future development of a knockout system are promising new approaches to elucidate chlamydial protein functions during C. psittaci infection.
Taken together, our findings highlight that pCps-Tet-mCherry can be used for C. psittaci transformation and functional studies. This system will enable the identification of novel virulence factors and improvement of treatment strategies for C. psittaci infection.
Genetic transformation. Transformation was performed as described in a previous study (13). pCps-Tet-mCherry was extracted from a Dam-and Dcm-methylase-deficient strain, E. coli GM2163, using a Qiagen plasmid mega kit (Hilden, Germany) (12). PEN (1U/ml) was used for the selection of transformed C. psittaci. A mixture of C. psittaci 01DC12 or 02DC15 1 ϫ 10 8 inclusion forming units (IFU) and 15 g of pCps-Tet-mCherry were incubated in 200 l calcium chloride buffer (10 mM Tris, 50 mM calcium chloride, pH 7.4) for 30 min at room temperature. Then, 200 l of C. psittaci mixed with pCps-Tet-mCherry were incubated with 200 l of trypsinized 8 ϫ 10 6 epithelial cells in calcium chloride buffer for 20 min at room temperature with mild agitation.
The total volume (400 l) of the mixture of epithelial cells, pCps-Tet-mCherry, and C. psittaci 01DC12 or 02DC15 was distributed over 4 wells (each 100 l) containing 2 ml DMEM medium and 1 g/ml cycloheximide in a 6-well plate and incubated at 37°C under 5% CO 2 . At 24 hpi, 1U/ml PEN was added to each well and further incubated for 48 h at 37°C under 5% CO 2 . Afterward, infected epithelial cells were scraped and lysed with glass beads.
Freshly prepared epithelial cells were infected with transformed C. psittaci in 6-well plates containing 5 ml DMEM medium with 1 g/ml cycloheximide and 1U/ml PEN. Passages were performed every 2 to 3 days. GFP expression was observed by passage 2.
Induction of mCherry. A total of 2 ϫ 10 5 epithelial cells in 24-well plates were infected with 2 ϫ 10 5 IFU of transformed C. psittaci in DMEM medium. At 1 hpi, 10 or 100 ng/ml of aTC was added to induce mCherry.
Recovery assay. A total of 5 ϫ 10 4 epithelial cells in 1 ml DMEM medium were seeded into 24-well plates (Greiner bio-one) and cultured overnight at 37˚C under 5% CO 2 . For the infection, 2 ϫ 10 5 IFUs of either untransformed or pCps-Tet-mCherry-transformed C. psittaci strains as well as cycloheximide (1 g/ml) were added to each well. When appropriate, PEN (10 U/ml) and 0, 10, or 100 ng/ml of aTC were added at 1 hpi. The plate was centrifuged at 700 ϫ g for 1 h at 35˚C and incubated for 5, 24, 48, and 72 h. After the indicated time, the cells were subsequently cultured for 48 h for determination of the recoverable C. psittaci.
Real-time quantitative PCR. Determination of plasmid copy number was performed as demonstrated previously (16). Wild type or pCps-Tet-mCherry-transformed C. psittaci were cultured in epithelial cells in the presence or absence of 10 U/ml PEN. Afterward, isolated C. psittaci were boiled in 20 mM dithiothreitol (DTT) for 15 min and the supernatant was collected to obtain the genomic and plasmid DNA mixture. One primer set was designed to detect the hctA gene (GenBank, AEG87858.1; CPS0B_0937) in the genome. Another primer set was designed to detect the GFP gene for quantitation of the pCps-Tet-mCherry plasmid shuttle vector. For the endogenous plasmid, primers were designed to detect CDS5 or the CDS1-CDS2 junction, which is separated by the fragment of pBOMB4-Tet-mCherry. Real-time quantitative PCR was performed using LightCycler 480 SYBR green I Master on LightCycler 480 II (Roche Molecular Biochemicals, Mannheim, Germany). Plasmid copy number was calculated using the threshold cycle (2 ΔΔCT ) method (38) relative to genomic DNA.
Plasmid stability. A total of 3 ϫ 10 5 epithelial cells in 1 ml DMEM medium were initially infected with pCps-Tet-mCherry-transformed C. psittaci at 0.5 IFUs/cell in the presence or absence of 10 U/ml PEN. The infected cells were subcultured every 2 days for 5 times. From the second passage, epithelial cells were infected with serial dilutions of C. psittaci at 0.5 to 0.8 IFUs/cell. To determine plasmid copy number, extracted genomic and plasmid DNA mixture was analyzed as described above. At passage 5, the GFP signal was detected by fluorescence microscopy and cells were fixed by methanol. Afterward, chlamydial inclusions were stained by fluorescein isothiocyanate (FITC)-labeled monoclonal chlamydial lipopolysaccharide (LPS) antibodies.
Fluorescence microscopy. A fluorescence microscope Keyence BZ-9000 (Keyence, Osaka, Japan) was used to detect the fluorescence signal of GFP or mCherry expressed in C. psittaci. In addition, untransformed or pCps-Tet-mCherry-transformed C. psittaci-infected cells were analyzed by immunofluorescence staining with mouse anti-chlamydial LPS antibody (green), which stained chlamydial inclusions. Evans blue was used for counterstaining of host cells (red).
Transmission electron microscopy. C. psittaci 02DC15-infected cells were fixed with 2% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M cacodylate buffer for 1 h. Postfixation was performed with 1% OsO 4 in 0.1 M cacodylate buffer for 2 h. Samples were dehydrated with a graded ethanol series and embedded in araldite (Fluka, Buchs, Switzerland). Ultrathin sections were stained with uranyl acetate and lead citrate and were examined with a JEOL 1011 transmission electron microscope (TEM) (JEOL, Tokyo, Japan).
Genome comparion of C. psittaci 02DC15 and C. psittaci 6BC. UpSetR package was used in the genome comparison (21). A total of 13 genomes (12 species of Chlamydia) and their homologous gene groups identified with genome analysis pipeline RIBAP were selected as input for UpSetR.
Metabolic analysis. A total of 2 ϫ 10 4 HeLa cells in RPMI 1640 were treated with 10 ng/ml aTC. In addition, cells were infected with untransformed or pCps-Tet-mCherry-transformed C. psittaci without cycloheximide in 24-well XF plates (Agilent, Santa Clara, CA). Plates were centrifuged at 700 ϫ g for 1 h at 35˚C and incubated for 24 h. Afterward, Mito Stress test kits were used following Seahorse Bioscience manufacturer's instructions with chemical concentrations as follows: oligomycin (0.5 M), FCCP (0.2 M), and antimycin A (1 M) plus rotenone (1 M). Before the assay of pCps-Tet-mCherry transformed-C. psittaci infected cells, expression of GFP and mCherry was detected by fluorescence microscopy BZ-9000 (Keyence).
Statistics. Data are indicated as means Ϯ standard error of the mean (SEM). Statistical analyses were performed by GraphPad Prism 7 statistical software. When three or more groups were compared in the experiment, Sidak's multiple comparison was used in cases where one-way analysis of variance showed statistical significance (P values Յ0.05). Data between two groups were evaluated using Student's t test. In Sidak's multiple comparison and Student's t test, P values of Յ0.05 were considered statistically significant.
Data availability. The sequence of the pCps-Tet-mCherry plasmid shuttle vector reported is available in the DDBJ/EMBL/GenBank databases under accession number LC548057.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.