Genetic Polymorphism Of Brucella Isolates In China

Bacteria of the genus Brucella are the etiological agents of brucellosis, which is azoonotic disease endemic in many areas of the world. This disease remains endemicin some countries, especially in developing ones and it is still a substantial economicburden for many areas of the world. The genus Brucella is divided into six classicalspecies: B.abortus (bovine), B.melitensis (caprine and ovine), B. ovis (ovine), B.canis(canine), B. suis (porcine), and B.neotomae (only seen in the desert wood rat) on thebasis of host specificity, antigenic differences and biochemical characteristics. Fourout of six species are pathogenic for humans, i.e. B. abortus, B. melitensis, B. suis andrarely B. canis. B. melitensis and B. abortus are highly pathogenic and is a frequentcause of human brucellosis. The conservation of this genus makes it difficult toestablish the true relationships between some classical Brucella species and biovars.Brucella has several host and widely distributed foci. Different location ofBrucella strains lead differences in biochemical features and pathogenicity. Alsodifferent species of Brucella have diversity in host range and virulence. Gene obtainedand deletion which relevant with adaption to the environment, host specificity andvirulence are an important mechanism of bacterial evolution. Understanding theevolution and virulence characteristics of Brucella depends on the knowlege ofgenetic differences of Brucella. In present, the typing methods of Brucella arephenotyping and genotyping. A variety of methods exist in the genotyping, and typingmethods have advantages and disadvantages, especially the typing results aredifference but have some contacts. In particular, multi-locus sequence typing (MLST)and multi locus VNTR analysis (MLVA) method possess repeatability and highresolution. However, due to the variation of Brucella, some atypical strains cannot betyped in the actual work. A better typing methods, or associated with a variety of methods may be better to solve the typing problem in Brucella.The comparative analysis of ten Brucella whole genome and Brucellawhole-genome microarray hybridization results identified a total of53DFR, and26in chromosome I, moreover the number of gene differences were102.27DFR werein chromosome II, including a total of194genes, and the DFR density inchromosome II was higher than I. The cluster analysis of53DFR in19standardstrains suggested that all abortus and melitensis, suis biovar4/5were clustered intoone branch, and suis biovar1/2/3, canis, neotomae and ovis were another.129Brucella isolates were also analyzed and it consisted two branches, one was abortusand melitensis isolates, another for suis and canis. DFR may be the acquisition ormissing in evolutionary process and there is no relationship with strains hereditary.18abortus isolates in the1970s and14canis isolates in the1980s were Clusteringanalyzed. It indicates that the propagation of abortus Brucella isolates is only inwestern area in the1970s,basically among several adjacent several provinces.However, in the1980s, the propagation of canis Brucella isolates limites in China'seastern region. The first canis Brucella strain may originate from Xinjiang provinceand then spread in Jiangshu and Zhejiang Provinces.The product length amplified by traditional MLST(CMLST) primers aregenerally400to500bp, and the valid sequences are between350to450bp afterremoving low quality sequence on both ends. Therefore, the available sequenceinformation by the conventional MLST is greatly limited. With the development ofsequencing instrument and technology, the length per sequencing reaction was greatlyimproved. In this study, all sequences sequenced by the3730sequencer werestatistical analyzed and it showed that at least800bp reading length with high qualitycould be generated. To improve the resolution of MLST for Brucella, and also extendthe application of MLST in other pathogenic bacteria, in the present study, we testedthe feasibility of improving resolution of MLST by increase of the sequencing length,defined as extended MLST (EMLST). To compare with the CMLST method,the nucleic acid length of all loci for EMLST were increased from128bp to301bpwith the proportion increasing about26.12%to71.33%.129Chinese Brucella isolates were analyzed with CMST and EMLST,23STswere defined by CMLST method, and35STs by EMLST. Except for aroA,gyrB andInt-hyp,alleles identified of six loci were at least increased one allele, such as thatcobQ increased five alleles and omp25increased2alleles. Alleles identified of eachloci dependent on polymorphic sites. polymorphic sites of six loci of nine genes wereincreased by increasing amplified product length.All data of CMLST and EMLSTmethod were analyzed by splittree, eBURST and BioNumerics software. It indicatesthat some of the strains that could be differentiated by CMLST were clearlydifferentiated by EMLST and the resolution of the EMLST is greatly improved.EMLST method can differentiate some canis strains from suis strains.EMLST has ahigher genotyping resolution than CMLST.27STs were defined in foreign Brucella strains by MLST analysis, and15STswere firstly defined in129Brucella of China which had23STs defined.8STs firstlydefined for abortus Brucella strains were ST8,ST28-33and ST42. ST33,ST34andST35were firstly defined for melitensis isolations. There were4STs and3STs firstlydefined respectively for suis and canis Brucella strains, that is,ST36-37and ST40-41for suis isolationsï¼› ST38,ST39and ST40for canis strains.PCR amplification and electrophoresis were used to distinguish in traditionalMLVA method, but an instability and time-consuming drawback. However,15highlypolymorphic MLVA sites in120isolates were analyzed by this method in our study,and obtained87types. There were13Brucella strains defined the same type81andtype6was consist of5Brucella isolations.The cluster analysis showed that abortusgenotypes were clustered into one branch, suis and canis for one, melitensis were theother one. We analyzed the evolutionary relationships or phylogenetic relationships of all genotypes for all strains by using maximum parsimony method. There is ahypothesis that canis strains may be evolved by the suis strains and abortus strainsmay be due to the melitensis strains.The CMLST, EMLST and MLVA data of120isolates showed that EMLSTcompared CMLST has more branches and a higher resolution. MLVA possessed thehighest resolution in the three methods, and it can nearly type each strains. However,all three methods can not differentiated biovars of Brucella strains.