Development of oriC-plasmids for use in Mycoplasma hyorhinis

Mycoplasma hyorhinis (M. hyorhinis) is an opportunistic pig pathogen, belonging to the class Mollicutes. It causes polyserositis, arthritis and cancers in vitro, increasing attention of the researchers. Currently, there is no available genetic tool to manipulate its genome. This study describes a development of oriC-plasmids harboring either large (pGEMT-LoriC) or minimum (pGEMT-MoriC) origin of replication (oriC) of M. hyorhinis along with tetracycline resistance marker.These plasmids were successfully transformed into M. hyorhinis with average transformation frequency of 1.5 × 10−4 and 2.0 × 10−5 transformants/CFU for pGEMT-LoriC and pGEMT-MoriC respectively, and were integrated at the chromosomal oriC as well as remained freely replicating. We also constructed a Mini-oriC-HT1 targeting plasmid by inclusion of hlyC arms and was used to inactivate hlyC at average frequency of 50%. The efficiency of hlyC inactivation was further improved (by 90%) when Mini-oriC-HT2 that contains E. coli recA was used. In both cases, hemolysin mutant bacteria diminished the ability to lyse mouse RBCs compared to wild-type (P < 0.001). OriC-plasmids described in this study may, therefore open the way for functional genomics in M. hyorhinis. Furthermore, this is a first study demonstrated the gene associated with a hemolytic phenotype in mycoplasmas.

At present, the oriC-plasmids have not been studied in M. hyorhinis. The aim of this study was to construct and evaluate the usefulness of these plasmids as genetic tools in M. hyorhinis. Here, the oriC-region of M. hyorhinis was characterized and used for the development of oriC-plasmids that replicates in M. hyorhinis.
In many cases, the oriC-plasmids harboring large oriC-fragments tend to integrate into the oriC-region of the genomic DNA by homologous recombination between the gene fragment cloned on the plasmid and the corresponding site on the chromosome due to the sequence homology. The stability of the oriC-plasmids can be improved by utilizing a minimum oriC that reduces the possibility of the integration of the plasmid into the host chromosome 14,18,22 . Herein, a mini-oriC plasmid that could replicate and maintained stably extra-chromosomally has been developed and evaluated. This study also provides evidence of homologous recombination occurring in M. hyorhinis.
The hemolytic activity has been described in other mycoplasmas 23 . Since the HUB-1 genome sequence has a hlyC (NCBI Reference Sequence: WP_014582557.1) related sequence identified by automated prediction, it is hypothesized that this sequence is involved in hemolysis. Therefore, hlyC was chosen as the target gene for these experiments. We initially confirmed that, the supernatant of wild-type M. hyorhinis can lyse mouse RBCs ( Supplementary Fig. S1). When the mini-oriC plasmid that contained relatively small regions (~500 bp) of homology with chromosomal DNA of hlyC flanking tetM used to transform M. hyorhinis, the tetM inserted into the chromosome and disrupted the hlyC. The supernatant of the hemolysin mutant strain diminished the ability to lyse mouse RBCs. Therefore, this manuscript demonstrates that hlyC may be involved in mouse RBC lysis.
RecA plays an important role in DNA recombination and repair 24 . In M. mycoides subsp. capri, the efficiency of homologous recombination using suicide plasmids was enhanced by the inclusion of a recA (from E. coli) into the constructs 7,25 . Therefore, in this study, we further augmented the occurrence of the homologous recombination at the hlyC site by introducing recA into our targeting oriC-plasmids.
The ability to transform M. hyorhinis has been long described. However, the transformation procedure and the oriC-plasmid systems developed in this study, have the potential to be used in targeted gene disruption, and may also in gene complementation and expression studies in this organism and possibly other mycoplasmas.

Results
Vectors construction. No functional oriC had previously been isolated from M. hyorhinis. The large oriC region (LoriC) of M. hyorhinis (1935-bp) which consisted of the complete dnaA and intergenic regions upstream and downstream of the dnaA (NCBI Reference Sequence: WP_020104468.1) were identified (Fig. 1). The genomic regions upstream of the dnaA contained a single putative DnaA box and the AT-rich region is about 83% (Fig. 1). Likewise, the genomic regions downstream of the dnaA contained a single putative DnaA box and the AT-rich region is about 85% (Fig. 1). The minimum oriC region (MoriC) of about 876-bp, only contains the regions upstream and downstream the dnaA (Fig. 1).
The PCR products of tetM 1917-bp (without promoter), spiralin gene promoter 313-bp, large oriC (LoriC) 1935-bp, mini-oriC (MoriC) 876-bp, recA 1062-bp, hemolysin left arm (LA) 437-bp and hemolysin right arm (RA) 405-bp were prepared (Fig. 2) and separately cloned into pGEM ® -T vector. To determine whether the LoriC region is able to function as an autonomously replicating sequence, we first constructed pGEMT-LoriC plasmid, harboring LoriC, and later reduced the oriC to a minimum and cloned to generate pGEMT-MoriC. To target hlyC, both hemolysin arms were cloned flanking the tetM to yield Mini-oriC-HT1 plasmid. Finally, the efficiency of hemolysin targeting had improved by introducing recA expressed by spiralin gene promoter to yield Mini-oriC-HT2 plasmid. The general strategy used to construct the oriC-plasmids was shown in (Fig. 3).

Transformation of M. hyorhinis. Following transformation with pGEMT-LoriC and pGEMT-MoriC
plasmids, tetracycline resistant colonies appeared within 3 to 5 days on KM2-agar plates containing 2.5 µg/ml tetracycline. During the indicated time period, no spontaneous tetracycline-resistant colonies appeared for untransformed controls. The presence of the plasmids in the transformants was confirmed when tetM (part of the plasmid) was detected by tetM specific PCR (with P9 primers, Table 1) using DNA template prepared from hyorhinis strain HUB-1 was predicted around the dnaA. It contains short AT-rich regions of 83% and 85% upstream and downstream dnaA respectively. To generate mini-oriC, the regions upstream and downstream dnaA, were amplified as fragment A (563-bp) and fragment B (313-bp). These two fragments were joined by overlapping PCR to form mini-oriC (MoriC, 876-bp).
Analysis of homologous recombination at the oriC site of M.hyorhinis genome. The oriC-plasmids containing LoriC fragment, have been shown to integrate into the host chromosome via homologous recombination due to the sequence homology with oriC-region. Here, we evaluated whether our pGEMT-LoriC had integrated into the M. hyorhinis genome, and also investigated the site at which the integration had occurred. Briefly, M. hyorhinis DNA extracted from transformed (at 5 th passages) and untransformed control cultures, were subjected to the PCR analysis. Based on M. hyorhinis genomic sequence and pGEMT-LoriC plasmid, we designed integration primers (P10, Table 1) to investigate the predicted integration site of the plasmid at the oriC-region (Fig. 4A). The forward integration primer (P10-F, Table 1) was designed to specifically bind up-stream the LoriC at the M. hyorhinis genome and the reverse integration primer (P10-R, Table 1) was designed to specifically bind the tetM of the pGEMT-LoriC plasmid. Following plasmid integration by single cross-over at the oriC-region of M. hyorhinis, the predicted size of the PCR product was found to be 4037-bp (Fig. 4A). Indeed, the PCR product using integration primers (P10, Table 1) and DNA extracted from pGEMT-LoriC transformants, produced fragments of 4037-bp ( Fig. 4B) indicating the integration of pGEMT-LoriC at the predicted oriC region. Analysis of the DNA extracted from pGEMT-LoriC transformants by tetM specific PCR (with P9 primers, Table 1), had detected the tetM and further confirmed the presence of the plasmid ( Supplementary Fig. S3). In contrast, if the plasmid replicating in an extrachromosomal form (did not integrate), then the PCR fail to generate a band due to the specificity of the specific integration primers (P10, Table 1) (because the forward integration primer binds specifically to M. hyorhinis genome upstream to LoriC, while reverse integration primer binds specifically to tetM of the pGEMT-LoriC plasmid). We also predicted the size of the PCR product when pGEMT-MoriC integrated at the oriC-site (2122-bp), where we get no PCR product ( Supplementary Fig. S4), indicating the absence of plasmid integration at this region by single cross-over.

Inactivation of hlyC using Mini-oriC-HT1.
The hemolytic activity has been described in other mycoplasmas 23 . However, there is no specific gene in mycoplasmas identified to be associated with the hemolysis. In this study, we first confirmed that, the supernatant collected from M. hyorhinis can lyse mouse RBCs ( Supplementary  Fig. S1). Since the HUB-1 genome sequence encodes a hlyC related sequence determined by automated prediction, this sequence is hypothesized to be involved in hemolysis. Therefore, we chose to inactivate the hlyC as it's easy to measure the phenotype of the mutant. In other mycoplasmas, the LoriC-fragment drives the plasmid to integrate at the oriC-region of the host due to sequence homology. Therefore, we constructed an oriC-plasmid that contains minimum oriC-region (876-bp) and was named pGEMT-MoriC. Since we aimed to evaluate whether pGEMT-MoriC plasmid could be used for hlyC inactivation. We cloned hemolysin left and right arms into the pGEMT-MoriC plasmid to flank tetM and the spiralin gene promoter and to yield Mini-oriC-HT1 targeting plasmid. The resultant mutant strain expected to confer resistance to tetracycline when added to the medium at the appropriate concentration. The following transformation with Mini-oriC-HT1 plasmid and plating in KM2-agar containing 2.5 µg/ml tetracycline, we observed the growth of tetracycline resistance colonies. Selected colonies sub-cultured in KM2 medium containing 5.0 µg/ml tetracycline and subjected to PCR analysis using hlyC flanking primers (P11, Table 1) located upstream and downstream of the hlyC respectively. The wild-type hlyC produces a predicted PCR product of about 1566-bp with these primers (Fig. 5) indicating the absence of tetM insertion at hlyC site. When hlyC is disrupted by inserting tetM along with spiralin gene promoter, the predicted sequence will increase (3072-bp) and could be PCR amplified with hlyC flanking primers (P11, Table 1) (Fig. 5). In our study, we observed a larger PCR product, presumably 3072-bp in size, following insertion of tetM and the spiralin gene promoter into hlyC via homologous recombination (Fig. 6A). Five colonies out of ten (50%) from two repeated experiments were produced with the predicted larger PCR product targeting hlyC when Mini-oriC-HT1 was used. This indicates the success of the specific targeting and inactivation of the hlyC. It also indicates the occurrence of homologous recombination in M. hyorhinis.

Improvement hemolysin targeted inactivation by the inclusion of recA.
To improve the frequency of gene targeting, we inserted recA (from E. coli) controlled by the spiralin gene promoter into Mini-oriC-HT1 to yield Mini-oriC-HT2 targeting plasmid. The previous study indicated the usefulness of recA to enhance homologous recombination in M. mycoides subsp. capri 7 . Similar results were observed with our suicide vectors 12 . Following transformation with Mini-oriC-HT2 plasmid and sub-culture of 5 resistance colonies, we observed the predicted size of PCR product (3072-bp) in all the selected colonies indicating the insertion of tetM at the hemolysin site (Fig. 6B). Similar results were obtained when the experiment repeated. Overall, an about 9/10 (90%) hemolysin targeting efficiency from two repeated experiments was obtained when recA incorporated into the targeting oriC-plasmid (Mini-oriC-HT2). No significant change in the phenotype of the wild-type and mutant colonies was observed (Fig. 6C).
The tetM insertion at the hemolysin site was further investigated with tetM specific PCR (with P9 primers, Table 1), using the 3072-bp PCR product amplified with hlyC flanking primers (P11, Table 1) as a template ( Supplementary Fig. S5). The 3072-bp PCR product amplified with hlyC flanking primers was directly sequenced to analyze the junctions and the presence of tetM along with spiralin gene promoter. Analysis of the DNA sequencing results indicates the presence of sequences of hlyC arms, spiralin gene promoter as well as the tetM (Fig. 7), confirming the insertion of tetM along with the spiralin gene promoter at hlyC site.
Further, we predicted the occurrence of single cross-over between the plasmid and any arm of hlyC leading to the integration of the full Mini-oriC-HT1 plasmid that would result in PCR amplification of tetM thus generating positive PCR results. The phenotype of such transformants should have a hemolysin positive phenotype. To investigate this event, we predicted the integration of the Mini-oriC-HT1 plasmid by a single cross over at the hemolysin right arm site. We designed single cross-over forward and reverse primers (P12, Table 1). The single-cross forward primer (P12-F, Table 1) binds the upstream region of hemolysin right arm (absent if double-cross over had occurred) while the single-cross reverse primer (P12-R, Table 1) binds the tetM present in the Mini-oriC-HT1 plasmid. In our experiment, we failed to obtain a positive PCR product (about 1017-bp) using single-cross primers (Supplementary Fig. S6) indicating that, the insertion of tetM at the hlyC site is due to a double cross-over event. Loss of hemolytic phenotype in the mutants. In M. hyorhinis, disruption of hlyC was expected to diminish its ability to lyse mouse RBCs. The wild-type and the mutant cells grown in KM2 medium and the supernatant were used to assay the hemolytic activity. Mouse RBCs were incubated with the supernatant while the control with PBS. The hemolytic activity was measured at 405 nm. We found that, the supernatant collected from the mutant colonies significantly (P < 0.001) lacks the ability to lyse mouse RBS compared to wild-type bacterial supernatants (Fig. 8). Therefore, this is a first study demonstrated the gene associated with a hemolytic phenotype in mycoplasmas

Discussion
In this study, we have described the construction and the use of a series of M. hyorhinis oriC-plasmids. Although the transformation of M. hyorhinis has been reported earlier with polyethylene glycol (PEG) 5 , the usage of the host-vector systems for genetic studies of M. hyorhinis has not so far been implemented. Therefore, a host-vector system which can be utilized for M. hyorhinis genetic studies is a prerequisite. The electro-transformation conditions optimized in this study are suitable for M. hyorhinis. However, some factors such as the use of minimum inhibitory concentration of tetracycline, high voltage and the excess amount of plasmid DNA should be considered.
OriC-plasmids have been successfully used in a number of Mollicutes 15,16,18,19 . Here, the origin of replication region of M. hyorhinis genome has been identified and was used to construct oriC-plasmids for this bacterium for the first time. The prediction of large oriC (LoriC) region of M. hyorhinis chromosome has identified two dnaA boxes upstream and downstream the dnaA, and this LoriC was proved to be functional in the pGEMT-LoriC plasmid. In the M. hyorhinis transformants, the pGEMT-LoriC plasmid was found to replicate as a free extra-chromosomally before integrating into the host chromosome during passaging, similarly to the pBOT1 plasmid in Spiroplasma citri 26 .
The length of the oriC sequences has been shown to influence the recombination efficiency. Furthermore, the presence of a complete dnaA is not essential to driving the replication in oriC-plasmids as in Spiroplasma citri 13,19 . Therefore, a plasmid containing a minimum oriC fragment was constructed to decrease the frequency of plasmid integration into the oriC-region of M. hyorhinis 15 . Indeed, free extra-chromosomal oriC-plasmids were obtained when the pGEMT-MoriC plasmid containing a minimum oriC-fragment (876-bp) used, similar to previous observations 14,19 . Tetracycline was chosen as the selection marker as it has been used in multiple mycoplasma species and that M. hyorhinis strain HUB-1 stored in our laboratory is tetracycline sensitive 27 . The oriC-based plasmids described here have several useful features that will provide flexibility in future genetic studies. The multiple cloning sites located upstream and downstream of tetM and the spiralin gene promoter make these oriC-plasmids compatible for the subsequent cloning of foreign DNA (such as GFP or other antigens) for many applications in M. hyorhinis. Indeed, this may encourage in vivo imaging following M. hyorhinis infection or even development of recombinantly engineered vaccines when foreign antigens are expressed.
The hemolytic activity has been previously described in some mycoplasmas, however, there has been no previous study showing that a particular gene has hemolytic activity in mycoplasmas. In this study, we investigated the function of hlyC by using our oriC-plasmids, and the data show that, the hlyC is responsible for the hemolytic activity observed in M. hyorhinis.  Fig. S7). The conservancy of hlyC sequence in these strains may indicate its potential in the pathogenesis of these strains although of the fact that the strains of M. hyorhinis express differences in virulence 28 . However, the association between the hlyC and virulence of M. hyorhinis remained to be evaluated. Due to the available genome sequences of M. hyorhinis and the importance  Table 1): forward integration primer (P10-F, Table 1) binds specifically to M. hyorhinis genome upstream the LoriC, while reverse integration primer (P10-R, Table 1) binds specifically to tetM of the pGEMT-LoriC plasmid (B) to yield a product of predicted size of 4037 bp. The gel image was cropped and full-length gel is included in the Supplementary Fig. S9. of this bacterium in the development of arthritis as well as cancers in vitro, the development of oriC-plasmids for M. hyorhinis, could allow identification of virulence factors, and understand its pathogenesis. In conclusion, our study has developed useful oriC-plasmids for M. hyorhinis and also identified the association between hlyC and the hemolysis phenotype for the first time.  Wan, Hong Kong) to the liquid KM2 medium and was incubated at 37 °C until the growth of the visible colonies. For the growth of mutants, tetracycline hydrochloride (Sigma-Aldrich) was added at 5.0 μg/ml and 2.5 µg/ml to the solid and liquid media respectively.

Construction of oriC-plasmids.
We used pGEM ® -T vector to construct the oriC-plasmids. Images of PCR ethidium bromide-stained agarose gels were acquired and with Gel Doc ™ XR + System with Image Lab ™ Software (Bio-Rad, USA). The tetracycline gene, tetM (ID: AGI19285.1) was PCR amplified (using P1 primers, Table 1) from pSE-1 vector provided by Professor Xiao Shaobo (Huazhong Agricultural University, Wuhan, China) and cloned into pGEM ® -T vector at SpeI/PstI restriction enzymes sites by infusion cloning methods to generate pGEMT-tetM plasmid. Since common bacterial cloning vectors such as pGEM ® -T vector or genes from E. coli cannot express in the M. hyorhinis from their own promoter, the spiralin gene promoter (from Spiroplasma citri, GI: 2384684, amplified from pSE-1 vector using P2 primers, Table 1) was cloned upstream of tetM at SpeI restriction enzyme site to generate pGEMT-Sp/tetM plasmid. The large fragment of the origin of replication (LoriC) of M. hyorinis was identified (Fig. 1), PCR amplified from of M. hyorinis genomic DNA using P3 primers, Table 1 and cloned at ApaI restriction enzyme site to generate pGEMT-LoriC plasmid.
To amplify a minimum oriC fragment (MoriC), the flanking regions of dnaA that contains the AT-rich, were PCR amplified with P4A primers, Table 1 as fragment A (upstream dnaA, 563-bp) and with P4B primers, Table 1 as fragment B (downstream dnaA, 313-bp). These two fragments were joined by overlapping PCR (using P4A-F and P4B-R primers, Table 1) to form a single fragment of 876-bp as minimum oriC (Fig. 1). This minimum oriC PCR product was cloned into pGEMT-Sp/tetM at ApaI restriction enzyme site to generate the pGEMT-MoriC plasmid as a minimum oriC-plasmid. The design of the fragments A and B were shown in the diagram presented in (Fig. 1), while the primers in (Table 1).
We chose to disrupt hlyC (NCBI Sequence: WP_o14582557.1) because we hypothesized that if hlyC confers hemolytic activity, it may be easy to measure the hemolytic phenotype. To construct plasmid targets hlyC, both hemolysin left and right arms were at least 450-bp to provide adequate homologous sequences to facilitate recombination at the target gene 10,11,22 . These arms were PCR amplified and cloned to flank tetM as well as the spiralin gene promoter since we aimed to knock-in a functional tetM at the hemolysin site. The left arm (LA) of hlyC (amplified from M. hyorhinis DNA using P5 primers, Table 1) was cloned into pGEMT-MoriC plasmid upstream spiralin gene promoter at SacII restriction enzyme site while right arm (RA) of hlyC (amplified from M. hyorhinis DNA using P6 primers, Table 1) cloned downstream of tetM at SalI restriction enzyme site. The resultant plasmid named mini-oriC targeting plasmid-1 (Mini-oriC-HT1) and was used to inactivate hlyC by homologous recombination.
The recombinase gene (recA) of E. coli was found to enhance the homologues recombination in M. mycoides subsp. capri 7 . Therefore, the recA (GI:446885887) was PCR amplified from E. coli BL-21 DNA using P7 primers, Table 1. At the same time, the spiralin gene promoter was PCR amplified with P8 primers, Table 1 and joined by overlapping PCR with recA (using P8-F and P7-R primers, Table 1) to form a single fragment which was cloned into Mini-oriC-T1 plasmid at NcoI restriction enzyme site. The resulted plasmid named mini-oriC targeting plasmid-2 (Mini-oriC-HT2) and was also evaluated for its functionality to inactivate hlyC by homologous recombination. All cloning steps were verified by colony PCR and analysis of DNA sequencing. The diagram of the vectors constructions was shown in (Fig. 3), while the primers used to amplify or join the PCR fragments into each, were listed in (Table 1).

Figure 8.
Hemolytic activity of the wild-type and hlyC mutants of M. hyorhinis. The supernatant of wildtype and hlyC mutants of M. hyorhinis were collected and incubated with mouse RBCs. PBS was used as a negative control. The released hemoglobin was determined by measuring the OD at 405 nm. All reactions were performed in triplicate. Data are presented as Means ± SD (n = 3). P < 0.001 versus control.