Antigenic Variation in the Lyme Spirochete: Insights into Recombinational Switching with a Suggested Role for Error-Prone Repair

Abstract

The Lyme disease spirochete, Borrelia burgdorferi, uses antigenic variation as a strategy to evade the host's acquired immune response. New variants of surface-localized VlsE are generated efficiently by unidirectional recombination from 15 unexpressed vls cassettes into the vlsE locus. Using algorithms to analyze switching from vlsE sequencing data, we characterize a population of over 45,000 inferred recombination events generated during mouse infection. We present evidence for clustering of these recombination events within the population and along the vlsE gene, a role for the direct repeats flanking the variable region in vlsE, and the importance of sequence homology in determining the location of recombination, despite RecA's dispensability. Finally, we report that non-templated sequence variation is strongly associated with recombinational switching and occurs predominantly at the 5' end of conversion tracts. This likely results from an error-prone repair mechanism operational during recombinational switching that elevates the mutation rate > 5,000-fold in switched regions.

A Borrelia burgdorferi mini‐vls system that undergoes antigenic switching in mice: investigation of the role of plasmid topology and the long inverted repeat

“…Although the switching frequency that we observed was low, we observed use of the full complement of available SNPs (Fig. ) as expected from our previous switching studies on the wild‐type vls locus (Verhey et al, ). The reason for the highest level of switching being in the center of vlsE remains open to speculation at this time.…”

“…We found that three of the four mini‐ vls constructs underwent switching at vlsE following mouse infection. However, the switching rate that we observed for these constructs was approximately 7x10 ‐4 switches per amplicon per week compared with our previously reported switching rate of 7x10 ‐1 switches per amplicon per week for the wild‐type vls locus on lp28‐1 (Verhey et al, ). The reason for this 10 3 ‐fold decrease in activity is likely multi‐factorial.…”

“…Spirochetes lacking the vlsE locus are infectious, but they are cleared after two to three weeks and cannot mount a persistent infection in immunocompetent mice; however, they can cause long‐term infections in SCID mice (Bankhead and Chaconas, ; Rogovskyy and Bankhead, , Rogovskyy et al, ; Magunda and Bankhead, ). Although recombinational switching seems to be a process that depends upon sequence homology (Verhey et al, ), it does not require the key protein for homologous recombination, RecA (Liveris et al, ; Dresser et al, ) or a wide variety of other DNA replication/recombination/repair proteins (Dresser et al, ; Lin et al, ) that are involved in antigenic variation at pilE in Neisseria gonorrhoeae (Vink et al, ). The only required protein identified to date (also involved in recombination at pilE) is the branch migrase RuvAB (Dresser et al, ; Lin et al, ) indicating that there is at least the formation of Holliday junctions or other branched structures that can be migrated (Tsaneva et al, ; Lloyd and Sharples, ).…”

Antigenic variation in the Lyme spirochete: detailed functional assessment of recombinational switching at vlsE in the JD1 strain of Borrelia burgdorferi

“…However, all of the other proteins involved in pilE antigenic variation are absent in the type strain B. burgdorferi B31 or are not required for variation at vlsE ; this includes RecA, which plays a central role in bacterial homologous recombination (Liveris et al , ; Dresser et al , ). Although RecA is not required in B31, recombination boundaries are associated with regions of higher sequence similarity, and donor sequences near both ends of each silent cassette are less frequently utilized, pointing to a requirement for sequence identity both to begin and end the strand invasion process (Verhey et al , ).…”

“…Using unbiased alignments and an inference algorithm to identify recombination events, we quantified their length, the rate of recombination and the rate of mutations believed to arise from error‐prone repair processes. In addition, the range of recombination donor sequences and their properties, and the effect of immune selection at vlsE in B. burgdorferi B31 were also analyzed (Verhey et al , ). We have also applied this sequencing and analysis technology to detect low levels of switching behavior from genetically manipulated vls systems that would have appeared to be inactive using other assays (Castellanos et al , ).…”