Abstract
VMP-like sequence (vls) antigenic variation systems are present in every Lyme disease Borrelia strain with complete genome sequences. The linear plasmid-encoded vls system consists of a single expression site (vlsE) and contiguous array(s) of silent cassettes that have ~90% identity with the central cassette region of the cognate vlsE gene; antigenic variation occurs through random, segmental, and unidirectional recombination of vls silent cassette sequences into the vlsE expression site. Automated annotation programs do not accurately recognize vls silent cassette sequences, so these regions are not correctly annotated in most genomic sequences. In this study, the vls sequences were re-analyzed in the genomic sequences of 31 available Lyme disease Borrelia and one relapsing fever Borrelia organisms, and this information was utilized to systematically compare the vls systems in different species and strains. In general, the results confirm the conservation of the overall architecture of the vls system, such as the head-to-head arrangement of vlsE and a contiguous series of vlsS silent cassette sequences and presence of inverted repeat sequences between the two regions. However, the data also provide evidence for the divergence of the vls silent cassette arrays through point mutations, short indels, duplication events, and rearrangements. The probable occurrence of convergent evolution toward a vls system-like locus is exemplified by Borrelia turcica, a variable large protein (Vlp) expressing organism that is a member of the relapsing fever Borrelia group.