Genome Sequences of Three Live Attenuated Vaccine Strains of Brucella Species and Implications for Pathogenesis and Differential Diagnosis

Abstract

Live attenuated vaccines play essential roles in the prevention of brucellosis. Here, we report the draft genome sequences of three vaccine strains, Brucella melitensis M5-10, B. suis S2-30, and B. abortus 104M. Primary genome sequence analysis identified mutations, deletions, and insertions which have implications for attenuation and signatures for differential diagnosis.

GENOME ANNOUNCEMENT

Brucella is the etiological pathogen of brucellosis, one of the most important zoonotic infectious diseases worldwide (4, 15). Due to the serious economic loss and public health risk, extensive efforts have been conducted to prevent and control brucellosis through vaccination programs. Live attenuated vaccines are the best vaccine forms and are used worldwide (12, 13). Because most of the present vaccine strains are developed by serial passages on culture media or animals, the definite genetic characteristics and attenuation mechanisms remain largely unknown, which is important for further evaluation of the vaccine strains. Vaccination of the present live attenuated strains evoked several problems, the most important of which is that vaccination induced antibody responses indistinguishable by the current conventional serological tests from those observed in infected animals (3, 8, 11). This limits the extended use of these vaccines in eradication programs based on serological testing and slaughtering of seropositive animals. Therefore, numerous efforts have been made to develop new vaccines by targeted deletion of virulence genes or antigenic ones (1, 5–7, 14). Vaccine strains Brucella melitensis M5-10, B. suis S2-30, and B. abortus 104M are extensively applied for prevention of brucellosis in China. Here, we report the genome sequences of the three strains, which will help to investigate the attenuation mechanism and provide signatures for the differential diagnosis of vaccination from infection.

The genomic DNA of the three strains was sequenced with a HighSeq 2000 sequencer by using a paired-end protocol. The low-quality reads were filtered, and the remaining reads were assembled with CLC bio genomics workbench version 5.5 by using the de novo assembly method. About 1.5 GB of clear data were obtained for each of the strains. After filtering, about 15,000,000 reads meeting the criteria were assembled. A total of 81, 86, and 97 contigs were generated for M5-10, S2-30, and 104M, respectively. The final approximate coverage for these contigs was about 450×.

The genome sequence was then annotated with different tools. Open reading frames (ORFs) were predicted by RAST (2). The rRNA was predicted by using RNAmmer (9), and tRNA was identified with tRNAscan-SE 1.21 (10). The genome sequences of M5-10, S2-30, and 104M were compared with those of virulent strain B. melitensis 16M, B. suis 1330, and B. abortus 9-941. A total of 2,859 (2,287 in the coding region), 7,456 (6,361 in the coding region), and 3,368 (2,881 in the coding region) single nucleotide polymorphisms (SNPs) were observed in M5-10, S2-30, and 104M, respectively. A total of 361, 414, and 24 small deletions and insertions were also identified in M5-10, S2-30, and 104M, respectively. Further confirmation of these genetic differences and their effects on the expression of antigenic proteins will provide signatures for differentiation of vaccination and infection by both molecular and immunological methods. Further detailed analysis and experimental confirmation will be included in our future work.

Nucleotide sequence accession numbers.

The draft genome sequences of B. melitensis M5-10, B. suis S2-30, and B. abortus 104M are available in GenBank under the accession numbers ALOR00000000, ALOS00000000, and ALOQ00000000. The versions described in this paper are the first versions, ALOR01000000, ALOS01000000, and ALOQ01000000.