Genome sequencing and analysis of BCG vaccine strains

Abstract

Background: Although the Bacillus Calmette-Guérin (BCG) vaccine against tuberculosis (TB) has been available for more than 75 years, one third of the world's population is still infected with Mycobacterium tuberculosis and approximately 2 million people die of TB every year. To reduce this immense TB burden, a clearer understanding of the functional genes underlying the action of BCG and the development of new vaccines are urgently needed.

Methods and findings: Comparative genomic analysis of 19 M. tuberculosis complex strains showed that BCG strains underwent repeated human manipulation, had higher region of deletion rates than those of natural M. tuberculosis strains, and lost several essential components such as T-cell epitopes. A total of 188 BCG strain T-cell epitopes were lost to various degrees. The non-virulent BCG Tokyo strain, which has the largest number of T-cell epitopes (359), lost 124. Here we propose that BCG strain protection variability results from different epitopes. This study is the first to present BCG as a model organism for genetics research. BCG strains have a very well-documented history and now detailed genome information. Genome comparison revealed the selection process of BCG strains under human manipulation (1908-1966).

Conclusions: Our results revealed the cause of BCG vaccine strain protection variability at the genome level and supported the hypothesis that the restoration of lost BCG Tokyo epitopes is a useful future vaccine development strategy. Furthermore, these detailed BCG vaccine genome investigation results will be useful in microbial genetics, microbial engineering and other research fields.

Figure 1. Distribution of single nucleotide proteins (SNPs) and regions of deletion (RDs) in Mycobacterium bovis AF2122/97 and 13 Bacillus Calmette-Guérin (BCG) strains.

Outer circle: coding DNA sequences from the AF2122/97 genome are shown in a pair of concentric rings representing both coding strands; two inner circles: G+C% content and GC deviation (>0% green, <0% purple); other circles, from outer to inner: SNPs (grey) and RDs (red) between AF2122/97 and 13 BCG strains (Mexico, Frappier, Glaxo, Moreau, Phipps, Prague, Sweden, China, Danish, Russia, Tice, Pasteur and Tokyo).

Figure 2. Phylogenetic trees of Mycobacterium bovis Bacullus Calmette-Guérin (BCG), M. bovis and Mycobacterium tuberculosis strains.

(a) Neighbor-joining tree based on 17 housekeeping genes. The tree was rooted using M. bovis AF2122/97. (b) Topological structure tree based on absence genes. Strains of M. bovis BCG are shown in red while strains of M. tuberculosis are shown in blue.

Figure 3. Genealogy of Bacillus Calmete-Guérin (BCG) vaccine strains and Mycobacterium tuberculosis strains.

The genealogy of BCG strains based on Keller et al. ¹⁵, displays a series of genomic alterations including regions of deletion (RDs; squares bordered with a solid line) and single nucleotide proteins (grey squares bordered with a dotted line). The blue squares in the figure represent published RDs, while the white squares represent newly identified RDs. The brown ovals are assumed to be ancestor strains without genome sequences. The solid arrows represent the process of strains living in the lab under human manipulation conditions, while the dotted arrows represent the process of strains living in their natural environments. The BCG and M. tuberculosis strains are divided by dotted lines in the figure.

Figure 4. Distribution of the 483 T-cell epitopes in the 19 Mycobacterium tuberculosis c complex (MTBC) strains.

(a) Total number of the epitopes in each of the 13 Bacillus Calamette-Guérin strains. The table lists the names of the epitopes that are absent from each strain. (b) Four groups of T-cell epitopes in the 19 MTBC strains. Epitopes shown in red are present, while those shown in sky blue are absent.