Rita Melo , Luciana Richer , Daniel L. Johnson , Maria Gomes-Solecki
Published: March 18, 2016DOI: 10.1371/journal.pone.0151850
http://journals.plos.org/plosone/articl ... ne.0151850
Some interesting things noted from the Discussion section of the full text:Abstract
Oral vaccination strategies are of interest to prevent transmission of Lyme disease as they can be used to deliver vaccines to humans, pets, and to natural wildlife reservoir hosts of Borrelia burgdorferi. We developed a number of oral vaccines based in E. coli expressing recombinant OspC type K, OspB, BBK32 from B. burgdorferi, and Salp25, Salp15 from Ixodes scapularis. Of the five immunogenic candidates only OspC induced significant levels of antigen-specific IgG and IgA when administered to mice via the oral route. Antibodies to OspC did not prevent dissemination of B. burgdorferi as determined by the presence of spirochetes in ear, heart and bladder tissues four weeks after challenge. Next generation sequencing of genomic DNA from ticks identified multiple phyletic types of B. burgdorferi OspC (A, D, E, F, I, J, K, M, Q, T, X) in nymphs that engorged on vaccinated mice. PCR amplification of OspC types A and K from flat and engorged nymphal ticks, and from heart and bladder tissues collected after challenge confirmed sequencing analysis. Quantification of spirochete growth in a borreliacidal assay shows that both types of spirochetes (A and K) survived in the presence of OspC-K specific serum whereas the spirochetes were killed by OspA specific serum. We show that oral vaccination of C3H-HeN mice with OspC-K induced significant levels of antigen-specific IgG. However, these serologic antibodies did not protect mice from infection with B. burgdorferi expressing homologous or heterologous types of OspC after tick challenge.
This raises a number of questions related to human host antibody response to OspC and the effects on serology testing.The ticks used for challenge were produced from a culture of B. burgdorferi isolated from heart of P. leucopus infected with field caught ticks from an endemic area of Lyme disease. Given that there are at least 25 phyletic types of OspC , ,  and that seven (A, B, C, D, I, K, N) cause disseminated infection in humans, ,  we investigated the types of OspC present in ticks used for challenge (Fig 4B). Next generation sequencing analysis of genomic DNA purified from engorged nymphal ticks identified 11 OspC phyletic types: A, D, E, F, I, J, K, M, Q, T, X, four of which were invasive types A, D, I, K. We confirmed by direct PCR that these ticks contained B. burgdorferi type K in addition to type A, which we used as an example of another invasive OspC type (Fig 5). Furthermore, we show the presence of B. burgdorferi containing the same pair of OspC phyletic types in heart and bladder from mice orally immunized with OspC-K showing that oral immunization did not prevent dissemination of B. burgdorferi carrying the homologous type of OspC (type K) nor the heterologous type A. Others have shown that OspC protection is type specific after challenge with homologous strains via needle challenge ,  and also via tick challenge , but not after challenge with heterologous strains via needle challenge .
Seven weeks after the last immunization and four weeks after tick challenge (day 105), antibody production to OspC decreased to levels below OD450 <0.5, (Fig 6A) which is the minimum necessary to ensure some level of protection against OspA . Levels of antibodies remained low until day 170, three months after challenge. These data suggest that production of antibody to OspC by the host depends on the presence of OspC pressure via immunization as we previously observed for OspA . In a study to investigate the anamnestic immune response to B. burgdorferi, Gilmore et all found that immunization with OspC type A waned quickly and did not protect from challenge with naturally infected ticks . Alternatively, production of anti-OspC antibody may have been down regulated by the tick inoculation of B. burgdorferi during feeding. Furthermore, B. burgdorferi is able to negatively regulate OspC expression in presence of anti-OspC antibodies as a mechanism of immune evasion , , . If OspC is not presented on the surface of the spirochete then antibodies against OspC will not find the protein for recognition and additional immune response processing and killing. The fact that synthesis of OspC by B. burgdorferi is involved exclusively in transmission from tick to mammal and not from mammal to tick  lends further support to our hypothesis.
In this study we also show that oral vaccination of mice with OspC induced significant levels of antigen-specific IgG equally distributed between IgG1 and IgG2a before we infected mice via tick challenge. After tick challenge, total IgG was reduced to about half and we observed that the difference was due to the loss of IgG1, an immunoglobulin subclass associated with TH2 responses, but not IgG2a an immunoglobulin associated with TH1 responses (Fig 6B). Interestingly a study by Caine et al shows that elimination of OspC decreases bloodstream burdens of B. burgdorferi in TH1 biased C3H-HeN but not in TH2 biased Balb/c mice  suggesting that OspC expression is necessary for B. burgdorferi dissemination in the TH1 C3H-HeN background but not in TH2 Balb/c. In our study we observed that in the absence of B. burgdorferi expressing OspC, the mouse (C3H-HeN) produced equal levels of IgG1 and IgG2a to the OspC immunogen indicative of a TH1/TH2 response before challenge. After the host was exposed to B. burgdorferi expressing OspC upon challenge, dissemination of B. burgdorferi proceeded efficiently in the presence of anti-OspC antibody and the immune response seems to have switched from the initial TH1/TH2 response raised against the OspC-K immunogen, to a TH1 biased response to the pathogen.
Finally, our borrelicidal assay data suggested that challenge with B. burgdorferi carrying an OspC distinct from (type A) or homologous to (type K) the vaccinogen (type K) may have lead to immune evasion given that in either case we were able to quantify increasing amounts of live type A and type K spirochetes after treatment with serum from OspC-K immunized mice which was independent of the route of immunization (Fig 7). Our data seems to corroborate previous findings in that pathogenic B. burgdorferi impairs the complement pathway of the host and thus survives in serum . The results we observed for anti-OspC borrelicidal activity support Bockenstedt’s findings in that neutralization activity was weak even against autologous strains . Although we observed a degree of type-specific killing with anti-OspC type K mouse serum (loss of 1log of spirochetes after treatment) it was nevertheless insufficient. Others have shown that anti-OspC borrelicidal activity is dependent on complement and that serum produced against a cocktail of OspC types (α-ABKD) killed each of the 4 types of spirochetes included in the vaccine cocktail . In that study, the authors did not investigate killing of each type of spirochete by its respective type-specific serum. Thus, it is possible that the OspC strain cross-protection observed may be related to the effect of one type of OspC in the cocktail rather than all four, and it certainly does not mean that serum containing all types of OspC antibodies will kill all types of B. burgdorferi.
Our results suggest that oral vaccines based in OspC type K are ineffective. Vaccine failure in mice could be due to factors related to the biology of the spirochete (B. burgdorferi early shut down of OspC expression on its surface), due to factors related to the immunological response to the vaccine (ie, the quick waning of the antibody response to OspC after immunization or due to tick induced shutdown of B cell production of IgG antibodies), due to the genetic background of the mice (OspC expression is needed for B. burgdorferi dissemination in C3H-HeN but not in Balb/c mice), and to partial neutralization of B. burgdorferi by homologous serum and inefficient production of cross-protective neutralizing borreliacidal antibodies to OspC by the mammalian host as shown here and in , .