Maximum antigen diversification in a lyme bacterial population and evolutionary strategies to overcome pathogen diversity

Abstract

Natural populations of pathogens and their hosts are engaged in an arms race in which the pathogens diversify to escape host immunity while the hosts evolve novel immunity. This co-evolutionary process poses a fundamental challenge to the development of broadly effective vaccines and diagnostics against a diversifying pathogen. Based on surveys of natural allele frequencies and experimental immunization of mice, we show high antigenic specificities of natural variants of the outer surface protein C (OspC), a dominant antigen of a Lyme Disease-causing bacterium (Borrelia burgdorferi). To overcome the challenge of OspC antigenic diversity to clinical development of preventive measures, we implemented a number of evolution-informed strategies to broaden OspC antigenic reactivity. In particular, the centroid algorithm—a genetic algorithm to generate sequences that minimize amino-acid differences with natural variants—generated synthetic OspC analogs with the greatest promise as diagnostic and vaccine candidates against diverse Lyme pathogen strains co-existing in the Northeast United States. Mechanistically, we propose a model of maximum antigen diversification (MAD) mediated by amino-acid variations distributed across the hypervariable regions on the OspC molecule. Under the MAD hypothesis, evolutionary centroids display broad cross-reactivity by occupying the central void in the antigenic space excavated by diversifying natural variants. In contrast to vaccine designs based on concatenated epitopes, the evolutionary algorithms generate analogs of natural antigens and are automated. The novel centroid algorithm and the evolutionary antigen designs based on consensus and ancestral sequences have broad implications for combating diversifying pathogens driven by pathogen–host co-evolution.

Development of recombinant monoclonal antibodies targeting conserved VlsE epitopes in Lyme disease pathogens

“…The highly divergent vls cassette sequences between phylogenetic sister strains are reminiscent of the rapid amino-acid sequence diversification at the locus encoding the outer surface protein C ( ospC ), another immunodominant antigen of B. burgdorferi (Barbour and Travinsky, 2010). Protein sequences of major ospC alleles diverge in a strain-specific fashion with an average sequence identify of ~75.9% among B. burgdorferi strains in the Northeast US, due to a history of recombination among coexisting strains and diversifying selection driven by host immunity and possibly host-species preferences (Wang et al ., 1999; Brisson and Dykhuizen, 2004; Haven et al ., 2011; Di et al ., 2021). In contrast, the coding sequences of the vls cassettes vary at a significantly higher level between the eight major sequence clusters (~56.3% average sequence identity), while varying at a high level between copies of the genome as well (e.g., 81.0% for B31, 76.5% for N40, and 76%.4 for JD1) (Fig 2).…”

“…The 56 serum samples, consisting of Lyme patient and control sera provided by the US Center for Disease Control and Prevention (CDC, n = 40), Lyme patient sera provided by Dr Maria Gomes-Solecki (University of Tennessee Health Science Center, n = 6), and sera from the reservoir hosts (white-footed mouse, Peromyscus leucopus ) provided also by Dr Maria Gomes-Solecki ( n = 10), have been used and described in previous publications (Ivanova et al ., 2009; Molins et al ., 2014; Di et al ., 2021). Briefly, among the human samples, 25 serum samples were derived from patients with early-stage Lyme disease including those diagnosed as having the skin symptom erythema migran (EM) or as EM convalescence.…”

“…Recombinant VlsE protein from the B31 strain was cloned, over-expressed, and purified using a protocol described preciously (Di et al ., 2021). Briefly, the 585-bp vlsE cassette region (including the direct repeat regions on both ends) of B31-5A3 clone was codon-optimized, synthesized, and cloned into the pET24 plasmid vector which then transfected Escherichia coli BL21 cells.…”

“…We tested sera from naturally infected hosts for reactivity to the IR peptides (Table 1) and recombinant VlsE protein with ELISA using a protocol described preciously (Di et al ., 2021). Briefly, a 96-well MICROLON 600 plate (USA Scientific, Inc., Ocala, FL, USA) was incubated with 10 μg/ml of antigen overnight at 4 °C.…”

“…We tested sera from naturally infected hosts for reactivity to the IR peptides (Table 1) and recombinant VlsE protein with ELISA using a protocol described preciously (Di et al, 2021).…”