Complementation of the Mycoplasma synoviae MS-H vaccine strain with wild-type obg influencing its growth characteristics

The temperature-sensitive (ts+) Mycoplasma synoviae vaccine strain MS-H harbors a non-synonymous mutation which results in Glycine to Arginine substitution at position 123 in the highly conserved glycine-rich motif of Obg-fold in the GTP-binding protein Obg. In-silico analysis of the wild-type and mutant Obgs of M. synoviae has indicated that this amino acid substitution affects structure of the protein, potentially leading to abrogation of Obg function in vivo. Present study was conducted to develop the first expression vector for M. synoviae and to investigate the potential effect(s) of complementation of MS-H vaccine with the wild-type obg from 86079/7NS, the parent strain of MS-H. An oriC vector, pKS-VOTL, harboring the 86079/7NS obg gene, downstream of the variable lipoprotein haemagglutinin (vlhA) gene promoter, also cloned from 86079/7NS, was used to transform MS-H. The plasmid was localised at the chromosomal oriC locus of MS-H without any detectable integration at the chromosomal obg locus. Analysis of the MS-H transformants revealed abundant obg transcripts as well as Obg protein, when compared to the MS-H transformed with a similar vector, pMAS-LoriC, lacking obg coding sequence. The MS-H transformants complemented with wild-type Obg maintained their original temperature-sensitivity phenotype (consistent with MS-H vaccine) but, when compared to the MS-H transformed with pMAS-LoriC, had significantly higher (p < 0.05) growth rate and viability at the permissive (33°C) and non-permissive temperature (39.5°C), respectively. Analysis of Obg expression in MS-H and its wild-type parent strain revealed comparatively lower levels of Obg in MS-H. These results indicate that not only the mutation in Obg, but also the level of Obg expression, can confer functional abnormalities in the bacterial host. Furthermore, with the construction of first expression vector for M. synoviae, this study has set foundation for the development of recombinant vaccine(s) based on MS-H.

Temperature-sensitive (ts) mutants of obg have been isolated and studied in different bacteria especially B. subtilis and E. coli [12]. A mutation G79E in the type-II helix d and D84N in the N-terminal Obg domain was shown to be the contributing factor of temperature sensitivity in B. subtilis [6,13]. Corresponding mutations in E.coli ObgE, G80E and D85N, also resulted in no growth at non-permissive temperature of 42˚C [14]. ObgEts mutant of another E. coli strain with point mutation (S314P) in GTPase G5 motif exhibited growth defects at elevated temperature [14,15]. Such observations suggest crucial roles for the N-terminal domain of Obg proteins. The mechanistic insight into how Obg participates in diverse cellular functions is yet to be deciphered but bacterial Obg proteins have been found associated with ribosome(s) and therefore appear to be involved in ribosome biogenesis, maturation and assembly [15][16][17][18][19][20][21]. Consequently, Obg proteins are considered essential for cell viability in various bacteria [3] including Mycoplasma genitalium [22,23].
The Mycoplasma synoviae vaccine MS-H is a ts + strain which is used to control M. synoviae infections in poultry. In our previous study, analysis of single nucleotide polymorphisms (SNPs) in obg gene of MS-H highlighted the putative role of Obg in temperature sensitivity and/or attenuation of MS-H [24]. However molecular basis of the attenuation of MS-H vaccine has not been fully investigated under experimental conditions so far. In fact, failure to understand the mechanism of temperature sensitivity and/or attenuation of MS-H vaccine, and mechanism of M. synoviae pathogenesis in general, is partly due to lack of genetic tools to manipulate this organism at molecular levels. Development of gene transfer methods, as described in our previous study [25], has made it possible to study the role of putative virulence factors in M. synoviae pathogenesis and to underpin the molecular basis of MS-H temperature sensitivity and/or attenuation. This study describes, for the first time, the cloning and expression of a foreign gene (obg) in M. synoviae, and further evaluates the effect of MS-H complementation with wild-type obg on its temperature-sensitivity and/or growth characteristics.

Ethics statement
Fourteen-day old specified pathogen free chickens (n = 3) were used in this study. Throughout the experiment, chickens were housed in high-efficiency particulate air-filtered negative pressure isolators within animal house facility of Asia Pacific Centre for Animal Health, Faculty of Veterinary and Agricultural Sciences, The University of Melbourne, Werribee, Victoria, Australia. Chickens were provided irradiated food and acidified water ad libitum. Purified recombinant Obg-MBP protein, emulsified with Montanide ISA 71 VG (Seppic Inc., Puteaux, France) adjuvant, was used to immunise chickens by intramuscular injection. Similarly, two booster injections were also administered and blood samples of approximately 1 ml were taken from jugular vein after 14 days following each injection. At termination of the experiment, birds were bled (maximum 5% of the body weight) from jugular vein with a 10 ml syringe and 21-gauge needle and then euthanized using intravenous injection of phenobarbitone. The University of Melbourne Animal Ethics Committee specifically approved this whole study (approval number 0911553.1).

Bacterial strains and culture conditions
M. synoviae strains/transformants, listed in Table 1, were grown to late exponential phase (p H 6.8) at 37˚C, unless otherwise stated, in mycoplasma broth (MB) as described previously [26], except MB and mycoplasma agar (MA) contained 3 μg mL -1 tetracycline (Sigma, St. Louis, Missouri, USA) for the growth of tetracycline resistant M. synoviae strains transformed with pMAS-LoriC and pKS-VOTL plasmids. Escherichia coli α-Select (DH5 α derivative) competent cells (Bioline, Alexandria, New South Wales, Australia) were used as host for gene cloning and propagation of different plasmids following manufacturer's instructions.

Construction of the vlhA-obg-oriC plasmid (pKS-VOTL)
In B. subtilis, obg is part of an operon consisting of an upstream gene spo0B and two downstream genes pheB and pheA (Kok et al., 1994). In M. synoviae, obg may also be part of a large operon consisting of upstream secD and downstream trpS and thrS genes. As obg gene regulatory elements are not known in M. synoviae, vlhA gene promoter region was selected to drive obg gene expression ( Fig 1A) and vlhA gene promoter region was joined to obg CDS using splicing by overlap extension (SOE) PCR as described before [30] with modifications (S1 File). Amplification of vlhA promoter region was conducted using vlhA-extF and vlhA-intR primers in PCR#1 and obg CDS was amplified using obg-intF and obg-extR primers in PCR#2 (Fig 1B;  S1 Table). Overlapping nucleotide sequences from these two PCRs were extended in the overlap extension PCR (PCR#3) using the oligonuclotides vlhA-extF and obg-extR (Fig 1B; S1  Table). Amplification conditions for the above 3 PCRs are described in S1 File. Amplicons of expected size were confirmed by agarose gel electrophoresis ( Fig 1C) and vlhA-obg amplicon  Table) to confirm inserts (vlhA-obg, LoriC-tetM) cloned into pKS-VOTL ( Fig 1D). PureYield Plasmid Midiprep (Promega) of pKS-VOTL (~500 ng μL -1 ) was used to transform M. synoviae strain MS-H as described previously for pMAS-LoriC [25]. MS-H vaccine parent strain, 86079/7NS, was also transformed with pMA-S-LoriC and pKS-VOTL, and resulting transformants, at in vitro passage number 8, were used as control only in Obg expression (Western immunoblotting) and growth comparison studies.

Southern blot hybridisation
Genomic DNA from untransformed and transformed MS-H was extracted using Chen and Kuo [31] method as modified and described by Shahid et al. [25]. Genomic DNA from pKS-VOTL MS-H transformants, untransformed MS-H and pKS-VOTL plasmid was digested with restriction enzyme BglII (New England Biolabs, Wilbury Way, Hitchin, England) according to manufacturer's recommendations. Digoxigenin (DIG)-labelled probes for tetM, oriC and vlhA-obg were synthesised using the Roche PCR-based method [32] by incorporating 0.5 μL of 1 mM DIG-11-dUTP (Roche Diagnostics, Mannheim, Baden-Württemberg, Germany) per PCR reaction. Blotting of the restriction fragments, separated on 1% agarose gel, onto positively charged Nylon membrane, hybridisation to DIG-labelled probes for tetM, oriC and vlhA-obg, and detection of chemiluminescent signals, was achieved as described previously [25].

Northern blot hybridisation
Cell pellets from approximately 50 mL M. synoviae cultures were obtained by centrifugation at12,000×g for 40 min at 4˚C and then immediately processed for total RNA extraction using TriPure reagent (Roche) as per manufacturer's recommendations. Concentration of RNA as determined by spectrophotometry (NanoDrop 1000; Thermo Fisher Scientific, Wilmington, Delaware, USA) at A260nm was adjusted to~1μg μL -1 and then stored at -70˚C until further use. RNA was mixed (1:1) with RNA loading buffer and then separated on 1% agarose gel, made in MOPS (4-morpholine propanesulfonic acid) buffer containing 1.85% formaldehyde (Sigma), using 1× MOPS (20 mM MOPS, 5 mM Na-Ac, 1 mM EDTA, pH 7.0) as running buffer. RNA gel was soaked in 20× SSC buffer to remove formaldehyde. DIG labelling of the probes (obg and vlhA) was performed as described above. Blotting of RNA onto positively charged Nylon membrane, hybridisation of DIG-labelled vlhA-obg, obg and vlhA probes (S1 Table) and detection of hybridised probes was performed as described for Southern blot hybridisation [25]. Northern blotting was performed in two separate experiments using approximately 25 and 50 μg total RNA per lane. Each blot was probed with three different probes as described above and a qualitative analysis was performed to determine if obg is transcribed under vlhA promoter or not.

Production of chicken and rabbit polyclonal antibodies to M. synoviae Obg
Recombinant M. synoviae Obg was expressed in, and purified from, E. coli as Obg-MBP and used to inoculate chickens for polyclonal anti-Obg-MBP chicken serum (S2 File). Using Gen-Script OptimumAntigen design tool, several highly hydrophilic/antigenic epitopes were predicted in M. synoviae Obg. Based on criteria for surface probability, homology to mouse or rabbit proteins, flexibility, secondary structures, signal peptide and solubility, four peptides (Obg_1 TSHLQIKIEDDFET, Obg_2 GKGGRGNNKFKTSK, Obg_3 NTAPRIAENGMPGE and Obg_7 KMNRNHFKVTGKKI) were selected and synthesized (GenScript, Piscataway, New Jersey, USA). New Zealand rabbits were immunized with each of these peptides and affinity-purified polyclonal antibodies were produced (GenScript) and used in immunoblotting as described below.

SDS-PAGE and Western immunoblotting
Whole cell proteins from M. synoviae cultures were subjected to SDS-PAGE and gels stained with Coomassie brilliant blue R-250 (Bio-Rad, Gladesville, New South Wales, Australia). For Western immunoblotting, unstained gels were used for protein transfer onto PVDF membrane using Trans-Blot Turbo RTA Mini PVDF Transfer Kit (Bio-Rad) using Trans-Blot Turbo Transfer System (Bio-Rad). Blotting membrane was blocked with 10% BSA overnight at 4˚C, and then incubated with either polyclonal chicken anti-Obg-MBP serum or rabbit polyclonal anti-Obg peptide primary antibodies for 1 h at room temperature. After washing three times in PBS-T (PBS containing 0.1% Tween-20), the blot was incubated for 1 h with either rabbit anti-chicken or goat anti-rabbit IgG HRP conjugate (Millipore, Temecula, California, USA) (1:1000 dilution). The blot was developed with a solution of SIGMAFAST 3-3' Diaminobenzidine and Urea Hydrogen Peroxide tablets (Sigma) dissolved in distilled water. For comparative analysis of Obg expression, whole cell proteins from M. synoviae strains/clones MS-H, MS-H-C28, MS-H-T75, 86079/7NS, 7NS-C38 and 7NS-T30 were quantified using 2-D quant kit (GE Healthcare, Piscataway, NJ, USA) as per manufacturer's recommendations. Loading volumes for each sample were adjusted, using 2-D quant kit estimated whole cell protein concentrations, in order to load approximately equal amounts of whole cell proteins onto the SDS-PAGE gels and subsequently subjected to Western immunoblotting. Immunoblots were scanned using CanoScan 8800F (Canon, Sydney, NSW, Australia). Intensity of Obg specific bands was quantified with ImageJ 1.50i (Wayne Rasband, NIH, USA), using volume tool (rectangle) and local background subtraction, and relative concentration (fold change) estimated from at least three independent immunoblotting experiments.

Temperature-sensitive phenotyping, growth curves and loss of viability study of MS-H and 86079/7NS transformants
Temperature-sensitive phenotype of MS-H, MS-H-C28, MS-H-T75, 86079/7NS, 7NS-C38 and 7NS-T30 was determined by vlhA Q-PCR and conventional microtitration as described previously [29]. For growth rate comparisons, transformant cultures, at an initial titre of~10 7 CCU mL -1 , were diluted (1:1000) in MB containing 3 μg mL -1 tetracycline, and grown in triplicates (40 mL each) at 33˚C for 192 h. Samples were taken in 24 h intervals and subjected to titration using titration plates incubated at 33˚C for two weeks. Colour-changing units per ml (CCU mL -1 ) were calculated as described before [33]. Generation times, during the exponential phase of cultures, were calculated from the log 10 -transformed CCU mL -1 values plotted against time using the formula provided by Microbiology Laboratories (http://inst.bact.wisc. edu/inst/index.php; dated 26.2.2015): where k is the growth rate (generations per hour), Xt is higher CCU mL -1 , X0 is lower CCU mL -1 and t is the time interval between Xt and X0 (in hours). Generation time is reciprocal of k i.e. 1/k (hours per generation). Temperature shift experiment was conducted to compare loss of viability of transformant cultures as described previously [29], with some modifications adopted from Taschner et al. [34] and Addinall et al. [35]. Briefly, MS-H-C28, MS-H-T75, 7NS-C38 and 7NS-T30 cultures, all at an initial titre of~10 7 CCU mL -1 , were diluted (1:10) in MB and grown in triplicates in 40 mL aliquots at 39.5˚C for 72 h. A 0.5 mL aliquot was taken from each culture in 12 h intervals and subjected to titration in a microtitre plate incubated at 33˚C for two weeks.

Statistical analysis
For quantitative analysis of Western blots, intensities of Obg specific bands were averaged from at least three independent experiments and relative fold change was estimated for Obg protein.
Growth curve and loss of viability data is expressed as mean ± standard deviation from three independent experiments. Using GraphPad Prism (version 5.01), log 10 -transformed CCU mL -1 values at each time interval, from three independent experiments, for each transformant were compared using Student's t-test. Differences with p values < 0.05 were considered statistically significant and indicated by " Ã ".

Integration of pKS-VOTL plasmid in oriC region of MS-H vaccine strain
Following BglII digestion, genomic DNA of three MS-H transformant clones (MS-H-T75, MS-H-T78 and MS-H-T90) was subjected to Southern blot hybridisation to locate pKS-VOTL plasmid in MS-H cells. BglII digestion of unintegrated pKS-VOTL was expected to produce a single band of~12.2 kb which was detected by DIG-labeled tetM probe (Fig 2A and 2B). In case of integration at the chromosomal oriC locus, two fragments of~6.7 and 10.2 kb were expected to be produced, and only one band of~10.2 kb hybridised to tetM probe as expected. Absence of~12.2 kb band in all pKS-VOTL transformants, at both 4 th and 8 th passage level, indicated no extrachromosomal form of the pKS-VOTL existed in MS-H (Fig 2A and 2B). DIG-labeled oriC probe was expected to bind to a single BglII fragment of~12.2 kb in case of free pKS-VOTL plasmid (Fig 2A and 2C). A single DNA fragment of 4.6 kb from untransformed MS-H, digested by BglII, hybridised to oriC probe as expected (Fig 2C). In case of pKS-VOTL transformants, following integration at oriC locus, two fragments of 6.7 and 10.2 kb hybridised to oriC probe as expected (Fig 2C). A DIG-labeled probe, vlhA-obg, spanning the joining site of vlhA promoter region with obg CDS, was also used to localise integration site of the plasmid. Only a~10.2 kb band was detected by vlhA-obg probe, indicating absence of free pKS-VOTL plasmid in transformants. DIG-labeled vlhA-obg probe hybridised to bands of expected sizes (~2.1 and 3.3 kb for obg and vlhA loci, respectively), following BglII digestion of the genomic DNA from untransformed and transformed MS-H, indicating absence of homologous recombination event occurring in transformants at any other locus except chromosomal oriC (Fig 2A and 2D).  Fig 3A). Slight fluctuation of rRNA bands in the left three samples was observed which might have been caused by different concentration of salts in the specimens loaded on each lane (Fig 3A). Northern blot analyses of approximately equal quantities (~50 μg per lane) were conducted. Hybridisation of the blot with vlhA-obg probe resulted in detectable transcripts of~1.4 and/or 1.2 kb in all clones examined; however, intensity of both bands from three different MS-H pKS-VOTL transformants was noticeably higher than all other clones. A further transcript of~3.9 kb was detected only in the three pKS-VOTL transformants of MS-H (Fig 3B). Hybridisation of the blot with the obg probe generated two bands of~1.4 and~1.2 kb in transformants  (Fig 3C), and almost undetectable in northern blots with lower quantity (~25 μg per lane) of total RNA (S1 Fig). The vlhA probe (specific for the conserved 5 0 end of the highly expressible vlhA gene) detected a transcript of~2.6 kb with variable intensity in all clones (Fig 3D). The vlhA probe was used only to differentiate vlhA transcripts from obg transcripts. These results indicated that the transcript of~3.9 kb was transcribed from integrated plasmid harboring wild-type obg under vlhA promoter and transcripts of~1.4 and/or 1.2 kb were most likely transcribed from endogenous obg.

Overexpression of wild-type Obg in pKS-VOTL transformed MS-H and 86079/7NS
Expression of wild-type Obg under vlhA promoter from MS-H and 86079/7NS transformants with pKS-VOTL was demonstrated by SDS-PAGE of whole cell proteins (Fig 4A and 4C). Polyclonal chicken serum raised against Obg-MBP could not detect a band of~50 kDa in M. synoviae strains (S2 Fig). Polyclonal rabbit sera against M. synoviae Obg peptides Obg_1, Obg_2, Obg_3 and Obg_7 detected a band of~50 kDa in pKS-VOTL transformants; however, polyclonal rabbit serum against Obg_1 and Obg_3 mainly reacted to~50 kDa band (Fig 4B  and 4D) whilst Obg_2 and Obg_7 reacted to several bands including one of~50 kDa (results not shown). Specificity of the immunostaining for M. synoviae Obg was further confirmed by staining of a band of~77 kDa corresponding to recombinant Obg-MBP (Fig 4B). Using Obg_1 polyclonal rabbit serum, the pKS-VOTL transformed 86079/7NS clones (7NS-T30, 7NS-T20, 7NS-T8 and 7NS-T4) exhibited strong reaction with~50 kDa band whilst pMA-S-LoriC transformants (7NS-C38 and 7NS-C12) and untransformed 86079/7NS exhibited weaker, but detectable, reaction with Obg_1 polyclonal rabbit serum (Fig 4D).  (Fig 4E), the MS-H and MS-H-C28 showed no detectable reaction with the Obg_1 polyclonal rabbit serum at two different concentrations examined, whilst 86079/7NS and 7NS-C38 showed detectable reaction which increased on doubling the protein concentration (Fig 4F and S3 Fig). The reaction of 7NS-T30 with rabbit polyclonal anti-Obg peptide primary antibodies was also higher,~2.7 ± 0.5-fold with Obg_1 (n = 5), than that of MS-H-T75 (Fig 4F and S3 Fig).   (Fig 5A). Examination of the cultures using conventional microtitration also showed that all cultures, except 86079/7NS, exhibited no growth (change of MB color) at the non-permissive temperature, and hence, were ts + (results not shown).  (Fig 5B, panel b). , were grown at 39.5˚C for 72 h, aliquots collected at 12 h intervals, and then titrated by incubation at 33˚C for two weeks. All cultures were grown in MB containing tetracycline (3 μg mL -1 ). Bars indicate standard deviations of log 10 -transformed CCU mL -1 values from three independent cultures and " Ã " indicates significant differences (p < 0.05) using Student's t-test.

Growth promoting effect of Obg was only evident in pKS-VOTL transformed MS-H
https://doi.org/10.1371/journal.pone.0194528.g005 Complementation of MS-H with wild-type obg Using temperature shift experiment, it was shown that both transformants MS-H-T75 and MS-H-C28 couldn't grow at the non-permissive temperature. However, MS-H-T75 gradually lost its viability at 39.5˚C and remained viable at least up to 72 h of incubation at 39.5˚C. MS-H-C28 lost viability relatively quickly and couldn't survive after 36 h of incubation at 39.5˚C (Fig 5B, panel c). Similarly, 7NS-T30 and 7NS-C38 couldn't grow at 39.5˚C. Contrary to MS-H-T75, 7NS-T30 completely lost viability at 72 h of incubation at 39.5˚C while 7NS-C38, like MS-H-C28, lost viability after 36 h of incubation (Fig 5B, panel d).

Discussion
Effect of Obg on growth characteristics has been documented in several bacterial species [6,16,[36][37][38]; however, this is the first study examining the role of Obg in any mycoplasma species under experimental conditions so far. A previous study in our group suggested a link between mutation in Obg and ts phenotype of M. synoviae [24]. In the current study, the effect of MS-H obg gene SNP (G367A) on temperature sensitivity and growth characteristics of M. synoviae was further investigated under experimental conditions. The obg gene was amplified from vaccine parent strain 86079/7NS, cloned into pKS-VOTL, and successfully transformed into MS-H. Plasmid was shown to be integrated completely into the chromosomal oriC and no incidence of homologous recombination at obg or any other locus in the MS-H genome was observed. The~3.9 kb vlhA-obg transcript (transcribed from integrated plasmid), detected by Northern blot hybridisation, may have comprised the downstream region of the plasmid pKS-VOTL (obg-containing plasmid) as there was no transcription termination signal cloned for M. synoviae obg. Two bands of~1.4 and 1.2 kb detected by vlhA-obg and obg probes in total RNA of MS-H pKS-VOTL transformants, indicated abundance of chromosomal obg transcript (s). It is likely that the expression of wild-type obg may have enhanced the transcription of chromosomal obg in the transformed cells by unknown mechanism(s) which might involve either the upregulation or enhanced activity of transcription factor complexes, or the inability of transcription regulators, e.g., miRNA-like RNAs [39] to cope with this overexpression.
The results presented here suggest that the expression of endogenous Obg was less abundant in untransformed MS-H and MS-H transformed with pMAS-LoriC (non-obg plasmid) as compared to untransformed 86079/7NS and its pMAS-LoriC transformant. The abundance of endogenous Obg protein in 86079/7NS as compared to MS-H was more evident in their transformants with wild-type obg-containing plasmid (pKS-VOTL). Interestingly, MS-H5, one of the ts − MS-H reisolates with back mutation at position 123 of Obg [24,29], exhibited detectable levels of endogenous Obg beyond that of MS-H. Previous studies have reported that missense mutations might accelerate the degradation of the mutant protein by proteases and hence reduce the concentration of the mutant gene product [40][41][42]. It has been shown that ts + E. coli mutants possess proteins with a rapid turnover, e.g. a ts + RNA polymerase was found hydrolysed rapidly, especially at the nonpermissive temperature [43]. Therefore, it may be possible that reduced level of MS-H Obg is due to Gly123Arg substitution, resulting in mutant protein instability and rapid degradation. It may be speculated that mutant protein instability and lower concentration may not hinder its biological functions at a permissive temperature but structural defect and concentration lower than a certain critical level may block its biological functions at higher temperature.
In this study, effect of obg complementation on growth characteristics of MS-H was investigated. Subcultures of MS-H pKS-VOTL transformant colonies, grown in MB containing tetracycline, exhibited faster growth rates (observed by rapid acidification of the broth medium) as compared to MS-H pMAS-LoriC transformant colonies at permissive temperatures of 33 and 37˚C. Growth curve analysis of MS-H and 86079/7NS transformants (with both plasmids) showed clear growth promoting effect of Obg only in MS-H, suggesting that MS-H Obg with Gly123Arg substitution is functionally impaired. When cultures of MS-H and 86079/7NS transformants were titrated, and shifted to 33˚C following incubation at nonpermissive temperature, pKS-VOTL transformants of both MS-H and 86079/7NS maintained viability for longer time than pMAS-LoriC transformants. The ability to maintain viability, despite being growth arrested at the nonpermissive temperature (39.5˚C), suggests at least partial biological function of the wild-type Obg expressed from pKS-VOTL.
Obg has binding site for ppGpp (guanosine 5 0 -diphosphate 3 0 -diphosphate) which is abundantly synthesized under stress conditions [13] and is responsible for the readjustment of gene expression according to the endured conditions [44]. Obg might hydrolyze pppGpp into ppGpp and regulate pppGpp/ppGpp ratio which has been linked to replication fork stability and cell survival [7,12,[44][45][46]. Recently, Obg has been proposed as a checkpoint protein in the assembly of 50S ribosome subunit, sensing the energy status of the cell via (p)ppGpp levels and linking ribosome assembly to other global growth control pathways [47]. In present study, MS-H complementation with wild-type obg could not restore ts phenotype, identical to that of the wildtype strain (86079/7NS), suggesting that contribution of Obg mutation towards MS-H temperature sensitivity phenotype may involve a complex mechanism. Furthermore, complete genome sequencing of MS-H and 86079/7NS has revealed that MS-H harbours several other mutations, some of which are mis-sense mutations in other essential genes (unpublished data). However, it is not yet clear if those mutation(s), either independently or in conjunction with obg gene SNP, could influence the temperature sensitivity and/or attenuation of MS-H.
It is possible that Obg overexpression, as observed in this study, exert an inhibitory effect on some unknown factor(s), required only under stress conditions (e.g., high temperature), and that the overexpression of Obg may have abolished in vivo function of unknown factor(s). In S. coelicolor and S. griseus, bacteria undergoing morphological development, overexpression of Obg had no effect on normal vegetative growth but inhibited aerial mycelium and thus spore formation [38,48]. Aerial mycelium followed by spore development is initiated under adverse environmental conditions and the level of Obg decreased dramatically just after the onset of aerial mycelium formation, and essentially no Obg was detected in cells during spore formation. Obg seems to mediate transition from vegetative to non-vegetative growth by sensing the intracellular GTP/GDP levels [38,48]. Obg-GTP form normally promotes vegetative growth while Obg-GDP stimulates aerial mycelium and spore formation. It is known that under stress conditions the level of GTP expression is decreased and therefore it is speculated that stress may induce a direct signal for Obg down regulation. It has been proposed that Obg overexpression suppress aerial mycelium and spore formation, and this suppression results from the presence of the excess Obg bound to GTP under conditions when Obg would normally bound to GDP [12,38].
Depletion or overexpression of Obg lead to growth defects in E. coli [14,49,50], B. subtilis [6,51], S. coelicolor [38] and C. crescentus [16]. Changing the level of Obg expression has significant effects on cell morphology, nucleoid appearance, chromosome segregation and regulation of chromosomal functions [12]. Consistent with these observations, overexpression of wild-type Obg in MS-H (transformant MS-H-T75), and 86079/7NS (transformant 7NS-T30), had no deleterious effect on growth at permissive temperature (33˚C) but might have growth inhibitory effect at nonpermissive temperature (39.5˚C). In M. synoviae, Obg concentration might also be regulated in response to transition from favorable to unfavorable growth conditions/temperatures; however, further studies using obg native promoter or other inducible promoters are required to prove this hypothesis.
It was puzzling that transformation of the wild-type parent strain 86079/7NS with the wildtype obg-bearing construct and no-obg bearing vector resulted in temperature sensitivity phenotype (clones 7NS-T30 and 7NS-C38 respectively) in this study. Temperature sensitivity of 7NS-T30 might be explained by the overexpression of Obg (as discussed above for MS-H-T75) while temperature sensitivity of 7NS-C38 was unexpected. The real cause of this result could not be found in the current study. It is perhaps worthwhile examining if spontaneous mutations in plasmid-encoded proteins (e.g., antibiotic resistance proteins, replication initiation proteins) may have occurred causing replication abnormalities in the host only at the high or nonpermissive temperature. Mutations in replication associated proteins encoded by plasmids can eliminate, reduce or even stimulate temperature sensitivity to host bacterial strains [52][53][54]. A number of mutations in TrfA, the only plasmid-encoded protein required for initiation of replication of the broad-host-range mini-RK2 plasmid, have been shown to induce temperature sensitivity in E. coli and Pseudomonas aeruginosa [54] while intragenic secondsite copy up (cop) mutations of trfA eliminated or strongly reduced temperature sensitivity in P. aeruginosa [53]. Future studies should also include transforming 86079/7NS or any other ts − wild-type M. synoviae field strain with a construct containing mutant (MS-H type) obg gene, or generate such a mutant using reverse genetic techniques, and to examine its temperature sensitivity phenotype.

Conclusion
In conclusion, complementation of ts + MS-H vaccine with wild-type obg could not restore ts − phenotype; however, wild-type obg expression had growth promoting effect at the permissive temperature whilst it improved the ability to survive at the nonpermissive temperature. Furthermore, with the development of first expression system in M. synoviae, development of the recombinant MS-H vaccine may also be feasible. To establish the role of Obg mutations in temperature sensitivity and/or attenuation of MS-H vaccine, further work is essential, either by expressing obg under its natural or a regulable (inducible) promoter allowing for the examination of Obg on M. synoviae growth at expression level lower than that observed in this study.