Investigating BB0405 as a novel Borrelia afzelii vaccination candidate in Lyme borreliosis

BB0405 is a surface exposed Borrelia burgdorferi protein and its vaccination protected mice against B. burgdorferi infection. As BB0405 is highly conserved across different B. burgdorferi sensu lato species, we investigated whether vaccination with recombinant BB0405 or through intradermal bb0405 DNA tattoo vaccination could provide protection against different Borrelia species, specifically against Borrelia afzelii, the predominant B. burgdorferi sensu lato genospecies causing Lyme borreliosis across Eurasia. We immunized C3H/HeN mice with recombinant BB0405 or with a codon-optimized bb0405 DNA vaccine using the pVAC plasmid and immunized corresponding control groups mice with only adjuvant or empty vectors. We subsequently subjected these immunized mice to a tick challenge with B. afzelii CB43-infected Ixodes ricinus nymphs. Upon vaccination, recombinant BB0405 induced a high total IgG response, but bb0405 DNA vaccination did not elicit antibody responses. Both vaccine formulations did not provide protection against Borrelia afzelii strain CB43 after tick challenge. In an attempt to understand the lack of protection of the recombinant vaccine, we determined expression of BB0405 and showed that B. afzelii CB43 spirochetes significantly and drastically downregulate the expression of BB0405 protein at 37 °C compared to 33 °C, where as in B. burgdorferi B31 spirochetes expression levels remain unaltered. Vaccination with recombinant BB0405 was previously shown to protect against B. burgdorferi sensu stricto. Here we show that vaccination with either recombinant BB0405 (or non-immunogenic bb0405 DNA), despite being highly conserved among B. burgdorferi sl genospecies, does not provide cross-protection against B. afzelii, mostly likely due to downregulation of this protein in B. afzelii in the mammalian host.

. Multiple studies show that BB0405 is necessary for establishing infection in mice, since bb0405-deletion mutants are unable to be transmitted from ticks and establish infection in mammalian hosts 7,8 . Of importance, vaccination with recombinant BB0405 also protected mice from B. burgdorferi infection by B. burgdorferi-infected ticks 7,8 . Thus, bb0405 is a highly conserved antigen with the potential to form the basis for a vaccine protecting against multiple B. burgdorferi sl genospecies. Most research on new Lyme vaccines focuses on recombinant proteins, but DNA vaccination constitutes an alternative vaccination platform 9 . For instance, a previous study by Wagemakers et al. has shown that DNA vaccination by tattoo with B. afzelii strain PKo Outer surface protein C (OspC) was fully protective against B. afzelii challenge in mice and induced favorable humoral immune responses compared to recombinant protein vaccination 10 . In line with this, we were able to show protection against B. burgdorferi strain N40 in a similar set-up, in which OspC from B. burgdorferi strain N40 was used both as recombinant as well as DNA vaccine (Klouwens et al. manuscript in preparation).
In the current study we aimed to investigate the role of BB0405 in providing protection across B. burgdorferi sl genospecies. To this end, mice were immunized with B. burgdorferi B31-derived recombinant BB0405 or bb0405 DNA vaccine and subsequently challenged with B. afzelii CB43-infected ticks, after which immunogenicity and host protection of the two different vaccination approaches were determined using established methods.

Results
Immunogenicity of bb0405 antigens. As described previously, BB0405 is a highly conserved B. burgdorferi sl surface protein and alignment of the protein BB0405 of B. burgdorferi B31 and B. afzelii CB43 showed an identity of 88% and similarity of 96% at the amino acid sequence level (Fig. 1). To determine whether antibodies against BB0405 would protect across different B. burgdorferi sl genospecies, we performed a vaccination study in mice. Recombinant BB0405 and a DNA vaccine for bb0405, both generated from B. burgdorferi B31, were constructed as well as an empty DNA vaccine, functioning as the negative control. From our previous published and unpublished studies it is known that an empty DNA vaccine, i.e. a pVAX vector without inserted target sequences, does not affect B. burgdorferi sl infection ( 10 and Klouwens et al. manuscript in preparation). Two weeks after the 3rd vaccination, mice were challenged with B. afzelii CB43-infected Ixodes ricinus nymphs. To assess immunogenicity, BB0405 specific total IgG levels were measured by ELISA before and after vaccinations (Fig. 2). As expected, the BB0405 specific total IgG levels were very high in the mice that had received the recombinant BB0405 vaccine. In contrast, this was not the case in the mice that had received the bb0405 DNA vaccine. In fact, the antibody titers in this group were comparable to the control mice that had been vaccinated with the empty DNA vaccine.
Host-protection via BB0405 immunization. Next, we assessed whether both vaccines were able to provide protection against B. afzelii infection by performing qPCR and culture of several tissues, obtained by sacrificing the mice 21 days after challenge with B. afzelii strain CB43-infected ticks. B. afzelii loads in ear, skin (tick feeding site), heart, bladder and ankle were quantified and normalized to mouse β-actin. In addition, bladder and skin tissues from the feeding site of the ticks were cultured and checked weekly for the growth of spirochetes. Despite the high antibody titers, results of qPCR and culture indicate that vaccination with BB0405, either recombinant or as DNA vaccine, did not protect against heterologous Borrelia challenge in mice ( BB0405 surface expression B. burgdorferi sensu stricto and B. afzelii. To study the location of the protein and also the potential accessibility of the protein to antibodies we performed a proteinase K assay with viable B. burgdorferi ss and B. afzelii spirochetes followed by Western Blot analysis. We show that BB0405 is partly surface exposed as the protein is partly digested upon proteinase K treatment for both 30 and 60 min both in B. burgdorferi ss as well as in B. afzelii. Flagellin B, a periplasmic control protein, was, as expected, not affected by proteinase K treatment (Fig. 3).
Expression of BB0405 or homologous proteins in B. burgdorferi sensu stricto and B. afzelii. Finally, in an attempt to understand why, in contrast to previous observations for B. burgdorferi ss 7,8 , BB0405 did not protect against B. afzelii, we next precisely compared the expression levels of BB0405 between these two different B. burgdorferi sl genospecies under different experimental conditions. We did not find differences in expression of bb0405 at the RNA level between B. burgdorferi ss and B. afzelii grown at 33 °C or 37 °C (data not shown). However, we also assessed BB0405 expression at the protein level. To this end, equal amounts of whole lysates of spirochetes, cultivated in vitro at different temperatures, were run on a Western blot and incubated with BB0405 specific antibodies. As can be appreciated from Fig. 4A, B. afzelii CB43 produces considerably lower amounts of BB0405 at 37 °C, the condition similar to the conditions in the mammalian host. We also quantified the protein bands on the Western Blots (Image J software) and measured the relative density of the bands compared to a corresponding Flagellin B loading control (Fig. 4B). Indeed, the relative density of BB0405 in B. afzelli CB43 cultured at 37 °C was significantly lower compared to BB0405 expression at 33 °C and to B. burgdorferi B31 at 37 °C, offering a possible explanation for our findings.

Discussion
An ideal Lyme vaccine candidate would need to provide protection against the most dominant B. burgdorferi sl genospecies causing Lyme borreliosis. The main B. burgdorferi sl genospecies causing Lyme borreliosis in North-America is B. burgdorferiss, whereas in Europe this is B. afzelii. To explore the potential of BB0405, in the current study we investigated whether vaccination with BB0405-a surface-exposed B. burgdorferi ss protein that is well-conserved among different B. burgdorferi sl genospecies including B. afzelii-protected against heterologous challenge with B. afzelii through tick-bite in an experimental mouse model. Moreover, two different methods of vaccination with BB0405 were used, recombinant protein vaccination and DNA vaccination by tattoo. Finally, to provide further insights into our findings we assessed the differential BB0405 expression of B. burgdorferi ss and B. afzelii under varying conditions. An adequate humoral response is essential in clearing Borrelia spirochetes and specific immunoglobulins play a key role in affording protective host immunity [11][12][13] . Since BB0405 is an outer surface protein, it is surface exposed and therefore likely to be antibody-accessible. Surprisingly, Kung et al. described the interesting phenomenon that no BB0405 specific antibodies are generated during natural infection in mice; however Brooks et al. showed that recombinant BB0405 is immunogenic in rats and antibodies raised against BB0405 could kill B. burgdorferi ss in vitro in the presence of complement 6 . In addition, in mice immunized with recombinant BB0405 a strong and long-lasting antibody response was induced that provided protective immunity against tick-transmitted infection with B. burgdorferi 8 . In the current study we also show a robust humoral immune response in mice vaccinated with recombinant BB0405 compared to control mice, corroborating the observation that recombinant BB0405 is immunogenic. In contrast, we did not observe antibody responses against BB0405 in the mice vaccinated with the bb0405 DNA vaccine. Interestingly, the same DNA vaccination approach has been successful in the past for other Borrelia outer surface proteins 10 . In our study this could indicate that the mice cells were not able to transcribe, translate or translocate bb0405 and that the utility of DNA vaccination against B. burgdorferi sl is highly dependent on the target. However, as murine infection with B. burgdorferi ss, either by syringe or ticks, does not result in BB0405-specific antibody responses, it could also be that bb0405 needs the adjuvants used in recombinant protein vaccinations and is not immunogenic by itself 8 . www.nature.com/scientificreports/ In the study described here, we have challenged the BB0405 (B. burgdorferi B31 derived)-immunized mice with B. afzelii infected ticks. We did not observe any protection of BB0405, neither as recombinant vaccine nor as DNA tattoo. Based on our observations, BB0405 does not seem to be a suitable vaccine candidate for the European situation, as it does not provide cross-protection between B. burgdorferi sl species-in this case B. burgdorferi ss and B. afzelii-despite high antibody titers after vaccination with recombinant protein. Although we have performed a heterologous challenge, the in silico analysis showed high sequence homology between these proteins. Indeed in Figs. 3 and 4 we show that the raised antibodies recognize and bind to BB0405 expressed by B. afzelii. It seems therefore unlikely that antigen recognition can explain the observed lack of vaccine efficacy. Future studies could assess whether using different species-specific BB0405 homologues from different B. burgdorferi sl genospecies could protect against heterologous challenge.  B31 and Borrelia afzelii strain CB43) grown at different temperatures. A temperature of 33 °C, resembling (feeding) tick conditions, and 37 °C, resembling conditions in the mammalian host, was used to culture the spirochetes. Lysates of the in vitro cultured spirochetes were obtained, as described in the materials and method section, and subjected to Western blot. Three independent experiments were performed. The first experiment is displayed on the left, the second experiment in the middle and the third experiment is displayed on the right. The far left lane is the protein weight marker. Blots were loaded with 2.5 ug/well whole lysates of spirochetes and were cut in half just between the bands of Flagellin B (41 kDa) and BB0405 (22 kDa) to enable separate incubations. The blots displayed at the top are incubated with anti-flagellin rabbit IgG 1:1000 (as a loading control) and subsequently with secondary antibody anti-rabbit IgG-HRP 1:1000. Blots displayed at the bottom are incubated with pooled serum from mice vaccinated with recombinant BB0405 1:500 and subsequently with secondary antibody anti-mouse IgG-HRP 1:2000, respectively. Blot images were cropped (Image acquisition tools Microsoft Powerpoint). Imaging was performed using ImageQuant LAS 4000 and quantification using Image J (Wayne Rasband, National Institutes of Health, USA, Java 1.8.0_77(32-bit), http:// image j.nih.gov/ij). Full length blots are presented in Supplementary Fig. S1-S6. (B) Protein expression of BB0405 as determined by Western blot in panel A was quantified and normalized against the relative density of loading control Flagellin B. Relative density was determined using ImageJ software (National Institute of Health). A quantitative comparison between samples on the same blot and within the same experiment were made. Statistical differences between groups were calculated using an unpaired parametric t-test. Error bars represent mean ± s.e.m. n.s.: P > 0.05. www.nature.com/scientificreports/ As stated above, antigen specificity of the generated antibodies is unlikely to explain the observed lack of efficacy. Therefore we firstly wanted to determine whether BB0405 was indeed surface expressed in B. afzelii strain CB43, as has been shown B. burgdorferi ss strain B31 8 . We have shown that BB0405 is indeed partly expressed at the cell surface by performing a Western Blot analysis after treating viable B afzelii B43 spirochetes with proteinase K. We used B. burgdorferi B31 as a control. The reason that there is only partial surface expression in both B. burgdorferi sl strains can be explained by BB0405 being a transmembrane protein and thus partly intracellular or membrane bound, explaining the lower protein products on the Western Blots of Fig. 3. Secondly, there are over 20 different species within the B. burgdorferi sl complex and these are capable of molecular adaptations in order to ensure efficient transmission by their vector and to survive in different environments of their hosts [14][15][16][17][18] . Temperature is a key environmental factor known to affect B. burgdorferi sl gene expression. More specifically, it has been described by Ojaimi et al. that BB0405 is upregulated by temperature suggesting upregulation in conditions that mimic the situation in the mammalian host 17 . We therefore wondered whether a difference in protein expression of BB0405 between American and European B. burgdorferi sl genospecies could explain the lack of protection in our vaccination experiment. At the protein level we observed a distinct and significantly lower expression of BB0405 in B. afzelii CB43 spirochetes grown at 37 °C, as compared to the expression in the same spirochetes grown at 33 °C. More importantly, expression of BB0405 in B. afzelii CB43 spirochetes grown at 37 °C was also significantly lower compared to B. burgdorferi B31 spirochetes grown at the same temperatures. This difference was not apparent at the RNA level. However, it is well-known that RNA expression levels do not necessarily correspond to the protein expression levels 19 . Thus, the BB0405 protein seems to be evidently less expressed by B. afzelii CB43 in conditions reflecting the mammalian host and this might explain the observed lack of protection against B. afzelii strain CB43, despite the presence of specific antibodies upon vaccination with recombinant BB0405. An alternative explanation for the lack of protection could be an indirect effect of failure of antibody recognition. B. burgdorferi sl is able to alter surface protein in vivo via differential gene expression and via VlsE recombination which might impact spirochete recognition by host-generated antibodies [20][21][22][23] . There are multiple examples of surface-exposed Borrelial antigens that are masked by neighboring (abundant) proteins 15,24 . Although BB0405 is conserved across B. burgdorferi sl genospecies, it could be possible that expression of other surface proteins is different in B. burgdorferi ss and B. afzelii, which may account for different surface topology and outcome of antibody-mediated protections.

Scientific Reports
In conclusion, being a surface-exposed, immunogenic and well-conserved B. burgdorferi sl protein, BB0405 was shown to be an interesting vaccine candidate to protect against Lyme borreliosis caused by B. burgdorferi ss. However, we here show that vaccination with B. burgdorferi ss-derived BB0405 does not protect against heterologous challenge with B. afzelii through tick-bite, and our data suggest that this could be due to lower expression of the BB0405 homologue at the protein level in B. afzelii in the mammalian host. Nevertheless, experiences with the OspA vaccine have shown that a vaccine does not have to protect against all B. burgdorferi sl genospecies to be commercially viable. Future experiments should investigate whether multivalent BB0405 vaccines are able to protect against multiple B. burgdorferi sl genospecies.

Methods
Ethics statement. All experiments were reviewed and approved by the Animal Research Ethics Committee of the Academic Medical Center, Amsterdam, The Netherlands (protocol 208AK-1 and 271AA-1). Experiments have been conducted according to European and national guidelines http://eur-lex.europ a.eu/legal -conte nt/EN/ TXT/?uri=celex :32010 L0063 ) and in compliance with the ARRIVE guidelines (http://www.nc3rs .org.uk/page. asp?id=1357).

Recombinant BB0405 protein.
The bb0405 gene was cloned without the N-terminal leader sequence into pET28a (Invitrogen), produced in E. coli, and purified using Ni-NTA resin as detailed elsewhere 8 . Sequence alignment between BB0405 from different strains was performed using Clustal Omega 25 .
Generation of the BB0405 DNA vaccine. The pVAXhTPA-BB0405 DNA vaccine was designed as described before in Wagemakers et al. 10 . In the BB0405 gene sequence of B. burgdorferi B31 (NCBI Reference Sequence: NC_001318.1) the 18aa signal sequence (predicted by SignalP 4.0 web-based software, CBS, Lyngby, Denmark) was replaced with the human tissue plasminogen activator (hTPA) signal sequence (genbank AAA61213.1) 26 . The resulting sequence was codon-optimized to mouse tRNA usage with Java Codon Adaptation tool (Braunschweig, Germany) 27 . At the 5′ end a BamH1 and a Kozak sequence were added, and at the 3′ end a sequence encoding a double stop codon and a Xho1 were added. The insert was synthesized (BaseClear, Leiden, The Netherlands) and ligated into a BamH1/Xho1 restricted empty pVAX vector (Invitrogen, Carlsbad, CA, USA). The plasmid was amplified using a Nucleobond Xtra EF kit (Macherey-Nagel, Düren, Germany) and resuspended in DNase free water.
Vaccination experiments. Six to eight weeks old female C3H/HeN mice were purchased from Charles River and the experiment was designed with 3 groups of eight mice each. The vaccination experiment was carried out as described in a previous publication of our group 10 . The first group was vaccinated with a recombinant BB0405 vaccine (recBB0405), the second group with a BB0405 DNA vaccine, and the third group consisted of an empty vector DNA vaccination group as a negative control. Mice were vaccinated at t = 0, t = 14 and t = 28 days and sera were collected at each time point for use in ELISA experiments. For the recombinant BB0405 vaccine 10 μg protein was emulsified with complete Freund's adjuvant at t = 0 and 5 μg in incomplete Freund's adjuvant at t = 14 and t = 28 days (Fig. 5). All vaccinations were administered subcutaneously. For the pVAX-hTPA-BB0405 DNA vaccine and the negative control hair was removed from the mice abdomens using hair removal  Borrelia afzelii detection and quantification. Murine bladder and skin samples were cultured in modified Kelly Pettenkofer (MKP) medium with rifampicin, 50 µg/ml and phosphomycin, 100 µg/ml) at 33 °C. The cultures were checked weekly (for a total of 8 weeks) for the presence of motile spirochetes with dark field microscopy as described before 28 . For all samples DNA was extracted using Qiagen Blood and Tissue kit (Qiagen, Venlo, The Netherlands). Quantitative (q)PCR was used to quantify B. afzelii DNA in mouse tissues and was performed according to previously described protocol 10 . OspA primers were used for quantification; forward 5′-AAA AAT ATT TAT TGG GAA TAG GTC T-3′ and reverse 5′-CAC CAG GCA AAT CTA CTG AA-3′, mouse Beta-actin forward 5′-AGC GGG AAA TCG TGC GTG -3′ and reverse primer 5′-CAG GGT ACA TGG TGG TGC C-3′ were used for normalization. The qPCRs were performed on the LightCycler480 (Roche, Nutley, NJ, USA) using SYBR green dye (Roche) using the following PCR protocol: 95 °C 6 min, and 60 cycles of 95 °C 10 s, 60 °C 20 s and 72 °C 20 s. Reactions were performed in triplicate. Results were analyzed using LinRegPCR software (Amsterdam, The Netherlands) 29 . Negative and positive controls were included in each qPCR run. A positive Borrelia load was determined by at least one melting curve exactly matching the positive control (skin tissue known to be infected with B. afzelii). In case of a negative value for OspA in a sample, the OspA value was replaced by the value of the OspA detection limit. The detection limit is defined as the highest dilution of the control sample in which OspA is still detectable + 3xSD.
RNA and cDNA synthesis. B. burgdorferi B31 and B. afzelii CB43 were cultured in MKP medium at 33 °C or 37 °C to 1 × 10e6 spirochetes/ ml as assessed by a Petroff-Hausser counting chamber and dark-field microscopy. The cultures were centrifuged for 10 min at 10.000 rpm and the pellets were dissolved in 750 µl RNA Later (Qiagen) stored at − 80 °C until further use. All samples were thawed simultaneously and centrifuged for 5 min at 12.000 rpm. The pellets were then subsequently used for isolation of RNA using the Nucleospin RNA isolation kit (Macherey-Nagel), according to the manufacturer's instructions. Subsequently, RNA samples were digested by DNase for a second time using the Qiagen RNase-Free DNase Set (#79254) and then cleaned up using the RNeasy MinElute Cleanup Kit (#74104, Qiagen). A total of 10 µl of each RNA sample was then used to generate cDNA and heated for 5 min at 85 °C and then cooled to 23   License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.