M.bovis Genetic Diversity

M.bovis Genetic Diversity Mycoplasmabovis pneumonia is an epidemic worldwide. To understand M.bovis genetic diversity would help develop novel measures to control this disease. Therefore this study was aimed to determine genotype distribution of Chinese strains and the potential global evolution. Firstly three available methods including two M. bovis multilocus sequence typing (MLST) schemes MLST-1 and MLST-2 and pulsed field gel electrophoresis (PFGE) were comparatively used for 44 Chinese strains and M. bovis type strain PG45 originated fromUSA. The results showed a high genetic homogeneity of Chinese isolates. By MLST-1, 43 of 44 (97.7%) Chinese isolate being ST-10, while 1 of 44 ST-34. The MLST-2 scheme clustered 44 Chinese isolates into two sequence types, ST-10 43 of 44 (97.7%) and 1 of 44 ST-32. PFGE clustered 42 of 44 (95.5%) into PT-I. The discrimination index was highest for PFGE (D = 0.160), while both MLST schemes have similar discrimination power (D = 0.110). The agreement rate among three typing methods is 95.4%à ¯Ã‚ ¼Ã‹â€ 95% CIà ¯Ã‚ ¼Ã… ¡84.2%, 99.4%à ¯Ã‚ ¼Ã¢â‚¬ °. The type strain PG45 gave a unique type by all three methods. Additionally, MLST-2 scheme was used to analyze 8 Australia and 8 Israeli isolates. The results showed 8 Israeli strains represent three STs with ST-10 as the most dominant type comprising 50% of the strains, ST-20 (n=2) and ST-28 (n=2). The 8 Australian isolates showed two sequence types ST-10 (n=7) and another sequence type ST-41 (n=1) identified firstly here. The assay of evolutionary relationship by geoBURST Minimum spanning tree (MST) of 60 isolates typed in this study and 207 isolates of 11 countries from the MLST-2 database. It was revealed that similar dominant clone (ST-10 in CC 3) exists in China, Israel, Australia and United States. This may be related to global livestock movements. In conclusion, we firstly demonstrated the remarkable clonality of M. bovis in China and the dominant ST-10 might originate from a common global source. Key words: Mycoplasma bovis; molecular epidemiology; multilocus sequence typing (MLST); pulsed field gel electrophoresis (PFGE); cattle; evolution. Mycoplasma bovis (M.bovis) is the main causative pathogen of bovine mycoplasmosis worldwide such as in North America, Europe (Nicholas and Ayling, 2003), China (Shi et al., 2008), Australia (Morton et al., 2014) and Israel (Lysnyansky et al., 2016). It results in substantial economic losses to producers by causing M.bovispneumonia and mastitis in beef and dairy cattle. M. bovis was first isolated in 1961 in USA from cattle mastitis milk (Hale et al 1962) and has long been considered a player in bovine respiratory diseases (BRD) since 1976 (Thomas et al., 1986). It then appears to have spread via animal movements to, amongst many countries (Nicholas 2002). Today, infection occurs in most European countries and throughout the world. It was estimated that the economic loss caused by M.bovisin United States was up to $108 million per year. In Europe, M.bovis pneumonia constitutes about 30% of calf respiratory diseases (Nicolas and Ayling, 2003; Maunsell et al., 2011). As the prevalence o f M. bovis associated diseases varies widely across the world, there are important trade implications and a pressing need to monitor cattle for M. bovis. However, to date, there are large gaps in our understanding evolutionary relationships of this pathogen isolates between different countries and globally. In China the first M. bovis mastitis was described in 1983 (Chen et al., 1983) and first M. bovis pneumonia in 2008. Since then reports of M. bovis pneumonia and mastitis outbreaks have been frequently described (Shi et al., 2008; Peng et al., 2011). M. bovis pneumonia is characterized by severe respiratory distress, high fever and at postmortem lung lesions including carnification, extensive caseo-necrotic or suppurative foci in the lungs. M.bovis pneumonia caused over 80% morbidity and between 10% to 60% mortality in calves and stockers newly introduced into beef feedlots (Shi et al., 2008). A major contributing factor to this disease is the stress caused by the long distance transportation of calves and stockers between the feedlots and farms (Shi et al., 2008). The disease is difficult to control with chemotherapy, and vaccination would be an ideal alternative approach. An insight of the genetic diversity and population structure of M. bovis would assist in the development of novel vaccines, as well as gaining an insight into evolutionary trends. A variety of molecular typing methods have been used for epidemiological characterization of M. bovis strains including random amplified polymorphic DNA (RAPD) analysis (Butler et al., 2001), amplified fragment length polymorphism (AFLP) analysis (Kusiluka et al., 2000; Soehnlen et al., 2012), pulsed field gel electrophoresis (PFGE) (Pinho et al., 2012; Arcangioli et al., 2012), insertion sequence (IS) typing (Miles et al., 2005; Aebi et al., 2012) and multilocus variable number tandem repeats (VNTR) analysis (Pinho et al., 2012; Spergser et al., 2013). In addition, three multi-locus sequence typing (MLST) schemes were recently developed to study population structure, evolution and spread of this pathogen (Manso-Silvan et al., 2012;Register et al., 2015; Rosales et al., 2015). The MLST scheme developed by Manso-Silvan et al. (2012) is based on four housekeeping genes fusA, gyrB, lepAand rpoB and showed a discrimination index of 0.833, while improved MLST scheme have been developed by Rosales et al.2015) here after referred as MLST-1 scheme; and by Register et al. (2015) here after referred as MLST-2 scheme. Both schemes use seven housekeeping genes but they only have one gene in common and therefore theire discrimination power higher than the Manso-Silvan scheme. In the present study, it was aimed to firstly evaluate the three methods MLST-1 and MLST-2 schemes and conventional PFGE by comparing the results in typing 44 Chinese M.bovis isolates, secondly assess the genetic diversity and population structure of M. bovis strains isolated in period of 2007 2014 by using the type strain PG45 as the control., and thirdly explore the evolutionary relationship of Chinese M.bovis isolates with globally diverse isolates. Material and Methods Mycoplasma bovis isolates M. bovis Chinese isolates (n=44) were obtained during 2008 to 2014 from nine Chinese provinces: Hubei (n=25), Anhui (n=1), Fujian (n=2), Hunan (n=1), Jiangxi (n=3), Henan (n=8), Inner Mongolia (n=1), Guangzhou (n= 2) and Shandong (n=1). These M. bovis isolates were mostly from lungs in cases of pneumonia (n=41); together with other sources such as milk with mastitis (n=2); throat swab in case of pneumonia (n=1) and fluid of joint with arthritis (n=1). The M. bovis type strain PG45 was purchased from American Type Culture Collection (ATCC 25523) and also used in this study. DNA samples from 8 Israeli M. bovis isolates were kindly offered by Prof. Dr. Inna Lysnyansky from Kimron Veterinary Institute, Israel, collected during 2013-2014 from pneumonia (n=6), stillbirth (n=1) and arthritis (n =1) in seven regions namely Gilboa (n=1), Beer Tuvia (n=3), Hevel Eilot (n=1), Eshkol (n=1), Jerusalem (n=1), Mateh Yehuda (n=1) and EmekYizrael (n=1). In addition, eight whole genome sequences of Australian M.bovis isolates were retrieved from GenBank representing mastitis, (n=4), lungs (n=1), nose swab (n=1), joint fluid (n=1) and semen culture (n=1) in five regions namely New south Wales (n=2), Queensland (n=1), Tasmania (n=3), South Australia (n=1) and Victoria (n=1) with accession no. SAMN05444185, SAMN05444199, SAMN05444228, SAMN05444239, SAMN05444243, SAMN05444247, SAMN05444250, SAMN05444261) included in this study (Table 1). Growth conditions, species identification and DNA extraction M.bovis isolates were confirmed by species-specific PCR as previously described (Subramaniam et al., 1998). The M.bovis samples were grown in PPLO broth (Difco) supplemented with 0.5% (w/v) sodium pyruvate (Biosharp, China), 0.09% (w/v) yeast extract (BD Biosciences, San Jose, CA, USA)à ¯Ã‚ ¼Ã…’0.004% (w/v) phenol red, 1% (v/v) 10- minimum essential medium (MEM) (Sigma-Aldrich, Saint Louis, MO, USA), 20% (v/v) Hyclone donor horse serum (Invitrogen, Carlsbad, CA, USA) and penicillin G 80,000 IU/100 mL and the final pH was adjusted to 7.6(Khan et al., 2016). DNA from each isolate was extracted using the genomic DNA extraction kit (Tiangen, Beijing, China). Multilocus sequence typing (MLST) MLST-1 scheme is based on a partial sequencing of dnaA, metS, recA, tufA, atpA, rpoD and tkt genes (Rosales et al., 2015); For MLST-1 scheme, 44 Chinese isolates and American type strain PG45. The PCR amplification conditions for MLST-1 were used as previously described (Rosales et al., 2015); after amplification, PCR products were further purified and sequenced using PCR Products Extraction Kit (Magnetic Beads) (Enriching Biotechnology, LTD, Wuhan, China) and sequenced. Sequencing reactions were performed by the commercial company (Tianyi Hui Yuan Biological Technology Pvt. Ltd. Wuhan, China).The quality of chromatograms was checked visually and sequence data were assembled and edited using SeqMan software (DNASTAR Inc., Wisconsin, USA). The assembled MLST-1 sequences were compared using non-redundant database (NRDB) comparison tool available in http://pubmlst.org/analysis/ with our previously analyzed 10 strains used as a control to assign allele and Sequence type number (Rosales e t al., 2015). MLST-2 scheme is based on a partial sequencing of adh-1, gltX, gpsA, gyrB, pta-2, tdk and tkt (Register et al., 2015). For MLST-2 scheme, the 44 Chinese strain and PG45 were subjected to PCR, and PCR products were sequenced as above mentioned method. The assembled sequences of all isolates were uploaded to http://pubmlst.org/mbovis/database to identify allele numbers and sequence types (STs). In addition, for the evolutionary assay, 8 Israel strains were typed with the method as described above. Meanwhile, 8 Australian isolates whole genome were annotated using prokka 1.11rapid prokaryotic genome annotation software (Seemann; 2014) at http://www.vicbioinformatics.com. Each locus sequence was extracted from the annotated genome. Pulsed Field Gel Electrophoresis (PFGE) analysis PFGE of 44 Chinese M.bovis field strains and type strain PG45 was performed as previously described (McAuliffe et al., 2004, Arcangioli et al., 2012) with some modifications for agarose block preparation. Briefly, macro-restriction analysis was performed with the restriction enzyme SmaI as follows: Each M.bovis isolate 15 ml culture aliquot was centrifuged at 15000 à ¯Ã¢â‚¬Å¡Ã‚ ´g for 20 min at 40C, the pellet was washed three times with Tris-EDTA buffer and resuspended in 400 à ¯Ã‚ Ã‚ ­l of cold Tris-EDTA buffer (pH 8.0). Agarose plugs were prepared from a 1:1 mixture of the above cell suspension and 2% low-melting-boiling agarose (Bio-Rad). They were then incubated in a lysis buffer containing 10mM Tris-HCl, 1 mM EDTA, 1% lauroyl sarcosine, 1mg of proteinase K per ml for 48 h at 560C. These plugs were washed for 6h with several changes of Tris-EDTA buffer at 40C. The plugs were then cut aseptically into 2 mm sections and equilibrated in 120 à ¯Ã‚ Ã‚ ­l restriction buffer (Prom ega) for 30 min at 40C. Subsequently, plugs were digested with 30U of SmaI (Promega, Shanghai, China) at 240C for 4 h. After digestion loaded in 1% pulsed-field-certified agarose gel (Bio-Rad), and run in a CHEF-DRIII system (Bio-Rad) at 6V/cm, in 0.5à ¯Ã¢â‚¬Å¡Ã‚ ´ TBE buffer at 140C, at 6V/cm with angle of 1200. The initial pulse time was 5s, with a final pulse time of 40s with a running time of 24 h. The lambda DNA ladder PFGE marker (Bio-Rad) was used as a reference. PFGE fragments in the gel were stained with ethidium bromide (EB) (1mg/ml) for 20 min, and destained in distilled water for 1.5 h and visualized under UV transilluminator. Pulsotypes (PT) were assigned numbers consecutively based on differences of more than one band in PFGE patterns upon visual inspection. The banding patterns were analyzed using Dice coefficients with 1% band position tolerance. The clustering of patterns was performed using unweighted pair group matching algorithm (UPGMA) as previously described ( Arcangioliet al., 2012; Timsit et al., 2012). Allelic sequence variance analysis The Sequence Type Analysis and Recombinational Test Version 2 (START2) (Jolley et al., 2001) were used to analyze polymorphic sites, construct UPGM dendrograms and calculate non-synonymous to synonymous ratios (dN/dS). Genetic diversity (H) of each locus and Index of Association (IA) were calculated by using LIAN 3.5 (Haubold and Hudson, 2000) hosted on http://guanine.evolbio.mpg.de/cgi-bin/lian/lian.cgi.pl/query. Global evolution and minimum spanning tree (MST) analysis The evolutionary relationship between isolates and M.bovis population structure was determined using PHYLOViZ (Fransciso et al., 2012) and evaluated by minimum spanning tree (MST) created using eBURST (geoBURST) algorithm (Francisco et al., 2009). MST for MLST-2 was performed for 257 isolates from 11 countries including 60 strains (44 China, 8 Israeli and 8 Australia isolates)   typed in this study and 207 isolates   retrieved January, 2017 (Supplementary Table 3) from the M.bovis MLST-2 database www.pubmlst.org/mbovis. Statistical analysis The discriminatory ability of both MLST methods and PFGE was calculated using Simpsons index of diversity as previously described (Hunter and Gaston, 1998). Congruence between both typing techniques was measured using the adjusted Rand Coefficient and Wallace Coefficient (Severiano et al., 2011). All statistical analyses were performed using the freely available online tool (http://darwin.phyloviz.net/ComparingPartitions/) Results The comparison of M.bovis typing with three methods MLST-1 analysis A total of 44, out of 10 were previously typed (Rosales et al., 2015) were also used for control and typed by MLST-1. The mean GC contents of seven gene fragments ranged from 29.15% (dnaA) to 37.23% (tufA) while it was 37.4 % in the whole M. bovis HB0801 genome (Qi et al., 2012). For each of seven loci, allelic variation was analyzed including polymorphic sites, guanine-cytosine(GC) content, synonymous and non-synonymousratios (dN/dS)(Table 2).The number of polymorphic sites per locus ranged from 4 (6.2%) in recA to 19 (29.6 %) in dnaA, and a total of 64 polymorphic sites for all seven genes were identified. The number of alleles observed ranged from 2 (metS, recA, tufA, atpA, and tkt) to 3 (dnaA and rpoD). The genetic diversity (H) for each locus was 0.0879 for dnaA and 0.0444 for metS, recA, tufa atpA and tkt. The dN and dS substitutions ranged from 0.0000 to 0.0623. In summary, all 44 Chinese M.bovis isolates typed by MLST-1 were divided into two STs namely ST-10 and ST-34 (Table 1).The ST-10 (with allelic profile of 2,6,2,2,2,5,3) was most numerically dominant, comprising 97.7%à ¯Ã‚ ¼Ã‹â€ 43/44à ¯Ã‚ ¼Ã¢â‚¬ °of Chinese M.bovis isolates including the Chinese strain HB0801 (Fig.1). In addition, ST-34 (allelic profile of 11,6,2,2,2,5,3) contains only one strain SZ; while ST-1(allelic profile of 1,1,1,1,1,1,1) represented by strain PG45 was identified (Table 1). Genetic relatedness amongst the 44 Chinese M.bovis strains showed two clades A and B. Clade A contained the majority (97.7%) of isolates (43/44) including the Chinese strain HB0801, while clade B contained one Chinese strain SZ (ST-34). M.bovis PG45 type strain was an outlier of these two clades (Fig.1). The geoBURST and MST analysis clustered 44 Chinese in the clonal complex CC2, whereas reference strain PG45 (ST-1) in CC1 (Table 1) as previously described (Rosales et al., 2015) MLST-2 analysis All 44 M.bovis isolates were examined by MLST-2. The mean GC contents of seven gene fragments ranged from 28.76% (tdk) to 35.61% (gyrB).The number of polymorphic sites per locus ranged from 8 in gyrB (8.66%) to 22(23.91%) in gpsA and a total of 92 polymorphic sites were identified (Table 2). The numbers of alleles identified were 2 for adh-1, gpsA, gyrB, pta2 and tkt and, 3 for gltX. The genetic diversity obtained 0.328 for adh-1 to 0.962 for gpsA (Table 2).   The Chinese strains were distributed into two different sequence types. ST-10 with allelic profile 4,3,3,3,5,3,4 was the most numerically dominant type, comprising 97.7% (43/44) of Chinese isolates; and ST-32 had only one isolate, SZ respectively. All M.bovis isolates tested in this study were clustered into two major clades A and B based on genetic relatedness by UPGMA. Clade A was comprised of 97.7% (43/44) of Chinese isolates including the Chinese strain HB0801. Whereas Clade-B contains one Chinese isolate. Same as above, M.bovis PG45 type strain was an outlier of these two clades (Fig. 2) PFGE typing The 44 Chinese M.bovis strains, and type strain PG45 were subjected to PFGE following the use of restriction enzyme SmaI. All isolates were typeable and the banding profile of the isolates ranged from 6 to 10 bands (from