Epitope mapping of Borrelia burgdorferi OspC protein in homodimeric fold

Abstract

In current work, we used recombinant OspC protein derived from B. afzelii strain BRZ31 in the native homodimeric fold for mice immunization and following selection process to produce three mouse monoclonal antibodies able to bind to variable parts of up to five different OspC proteins. Applying the combination of mass spectrometry assisted epitope mapping and affinity based theoretical prediction we have localized regions responsible for antigen‐antibody interactions and approximate epitopes' amino acid composition. Two mAbs (3F4 and 2A9) binds to linear epitopes located in previously described immunogenic regions in the exposed part of OspC protein. The third mAb (2D1) recognises highly conserved discontinuous epitope close to the ligand binding domain 1.

Abbreviations

LD

Lyme disease

MALDI‐TOF

matrix‐assisted laser desorption/ionization time of flight

Introduction

Outer surface protein C (OspC) is one of the most dominant antigens on the surface of LD spirochete, Borrelia burgdorferi sensu lato (Bbsl), and it is considered as an important factor of borrelia infectivity. Although the role of OspC remains somehow elusive, it has been demonstrated that it is essential in the infection establishment, dissemination1, 2, 3 and can promote spirochetes' evasion of macrophages.4 Besides the physiological functions, OspC protein is known for its polymorphic nature and immunogenicity.5 This has been exploited in the development of serologic assays for laboratory confirmation of LD6, 7, 8 and proposed vaccines.9, 10, 11

Regardless the means of immunization and phylogenetic type, the OspC protein can trigger strong immune response (seroconversion) both in human patients and in animals, leading to the production of wide range of specific antibodies in high titres.5, 9, 10 On the other hand, the quality of antibodies and recognized epitopes may vary depending on the conformation state of antigen used.12, 13, 14 The tertiary and quaternary structure of OspC seems to have an impact on production of “more relevant” antibodies possibly due to the steric hindrance, especially when considering discontinuous epitopes.

In the native conformation, OspC consists of two identical monomers (∼22 kDa each) folded in the compact mushroom shape like a four‐helical bundle. The dimer is stabilized by interactions of two α‐1 helices, each belonging to one of the subunits.15 Each subunit has a lipid moiety attached to N‐terminal cysteine anchoring the complex into the outer membrane of B. burgdorferi.16 When displayed on the surface of Bbsl, OspC exposes mainly apical part of the dimer and most probably C‐terminal conserved polyproline II like helix (PVVAESPKKP).12

The tertiary and/or quaternary structure of OspC and related epitopes accessibility seems to play a crucial part in borrelia dissemination and protectiveness of elicited antibodies. It has been proposed earlier, that only non‐denatured protein is able to elicit protective antibodies.10, 17, 18, 19 The importance of preserved tertiary structure also supports the results of experiments with GST‐tag, enhancing the folding, fused to recombinant OspC. GST‐tagged rOspC showed prophylactic capabilities but not the same rOspC after GST tag removal.20 These results were confirmed in the experiments utilising nucleocapsid of hepatitis B virus displaying OspC protein in the native fold as confirmed by cryo‐electron microscopy.11, 13 Nevertheless, the opposite approach based on immunization with synthetic peptides was implemented successfully as well.9, 21, 22, 23

The exact knowledge of epitopes involved in OspC recognition is essential for our understanding of mechanisms lying behind the functional differences in the effect of antibodies targeting OspC exposed on the surface of Bbsl. The antibody screening against the peptide library, the most common method used for the epitope mapping, seems to be insufficient in this case, especially with regards to discontinuous epitopes. In this study, we applied a combination of immunization and affinity assays with fully structured OspC homodimer combined with analysis of surface exposed amino acids, sequence homology and immunoprecipitation of digested antigen coupled with MALDI mass spectrometry analysis.

Results

Production of recombinant OspC

Recombinant OspCEX_BRZ31, OspCEX_BG, OspCEX_BB, OspCEX_BV and OspCEX_BA were expressed, processed by TEV protease and purified to 99% purity as verified by SDS‐PAGE [Fig. 1(A)]. All five antigens migrated approximately to their calculated molecular weight [Fig. 1(C)]. A monoclonal antibody specific for hexahistidine tag confirmed the identity of recombinant proteins prior the TEV protease cleavage. The amino acid composition was verified by mass spectrometry. The complete folding of OspC after ubiquitin cleavage in phosphate buffer used during immunization, hybridoma cell selection and immunoassays was verified by the 1H nuclear magnetic resonance measurement. The broad dispersion of peaks of amide protons in a region from 6 to 10 ppm confirmed compact folding of the protein (Fig. 2). The ability to form dimers was confirmed by size exclusion chromatography (Supporting Information). To exclude the possibility that the spontaneous disulphide‐dependent dimers or oligomers are formed through the formation of bonds between cysteines C130 and C130'19, 24, 25 the non‐reducing SDS‐PAGE was performed. Only the minor fraction of purified OspC was oxidised, and the dimeric form wasn't abolished [Fig. 1(B)].

Figure 1.

(A) SDS‐PAGE of major fractions obtained during affinity purification (OspCEX), and reverse affinity chromatography after TEV cleavage (OspC cleaved); (B) SDS‐PAGE of OspC cleaved by TEV with a sample prepared under reducing (R) or non‐reducing conditions (Nr) Spontaneously oxidised OspC forming dimers through disulphide bridges (marked by asterisk) were detected only in the minor fraction; (C) SDS‐PAGE of OspC variants after TEV processing used in epitope mapping experiments. M denotes molecular weight marker.

Figure 2.

1H nuclear magnetic resonance measurement of OspC after TEV processing. The broad dispersion of peaks of amide protons in a region from 6 to 10 ppm confirmed compact folding of the protein.

Hybridoma cell selection

For hybridoma cell selection truncated forms of OspC, as well as Bbsl cell lysates, were used both in western blot and ELISA assays. The selection scheme was designed to target those supernatants, which contain antibodies able to recognise epitopes in variable regions of OspC protein. In summary, only the samples showing high affinity to some of the antigens but not to all of them were propagated and purified. Within the group of over 600 tested hybridoma cells only 24 showed the ability to bind to OspC protein. Three of them (3F4, 2D1 and 2A9) exhibited significant differences in the reactivity against proteins included in the test panel and were further propagated for purification.

Affinity assays of purified mAb

Monoclonal antibodies 3F4, 2D1 and 2A9, were purified by affinity chromatography using a column with immobilised protein G to high purity as confirmed by SDS‐PAGE [Fig. 3(A)]. The binding capacity was verified both in ELISA and western blot assays (data not shown). The result confirmed preservation of capacity to bind OspC antigens for all three mAb after purification. The 3F4 mAb was the most sensitive in western blot assays as it was able to bind recombinant as well as native OspC in cell lysates with the exception of recombinant OspC BB, where the signal was less visible. On the other hand, only the recombinant OspC proteins were detected when 2D1 and 2A9 mAb were used [Fig. 3(B)]. This is only in partial agreement with data obtained in ELISA measurements against a panel of antigens utilised in the selection process (Figs. 4 and 5). 2A9 mAb shows low affinity in both immunoassays, which could lead to false negative results in cell lysate samples. On the other hand, the overall affinity of 3F4 and 2D1 mAb in ELISA was similar, which is in contrast to western blot where the 3F4 antibody shows much higher affinity especially visible when cell lysates are regarded.

Figure 3.

(A) SDS‐PAGE of the major fraction obtained during affinity purification of the 3F4 mAb with light chain (L. Ch) and heavy chain (H.Ch) marked (B) WB affinity test of purified mAb to recombinant OspC (100 ng) and whole cell lysate (100 ng, amount of OspC was not detectable by the Coomassie blue staining) from B. garinii (BG), B. afzelii (BA) and B. burgdorferi s. s. (BS). M denotes molecular weight marker.

Figure 4.

The specificity of tested monoclonal antibodies against different variants of OspC proteins as detected by ELISA assay. The antibodies were diluted to final concentration of 0,1 ng/μL (1:10,000).

Figure 5.

The affinity of 3F4, 2D1 and 2A9 mAb to OspC_BRZ31 and OspC_BG. The initial concentration of antibodies prior to dilution was 1 μg/μL.

Epitope mapping

The prediction of regions responsible for interactions with each of purified mAb (3F4, 2D1 and 2A9) with OspC antigens was based on the evaluation of the cross‐reactivity measurements with five different OspC variants (OspC_BRZ31, OspC_BG, OspC_BB, OspC_BV and OspC_BA) in ELISA assay. 3F4 mAb were able to detect OspC_BRZ31 and OspC_BG. 2D1 mAb was able to bind to all antigens with the similar readouts except for the OspC_BV where the absorbance was significantly lower. 2A9 antibody was active only against OspC_BG, OspC_BRZ31 and OspC_BV but very low absorbances were obtained (Figs. 4 and 5). The data are summarised in Table 1.

Table 1.

Summary of Cross‐Reactivity Data

mAb/Ag	BRZ31	BG	BB	BV	BA
3F4	+++	+++	+	−	−
2D1	+++	+++	+++	+	++
2A9	++	++	++	−	−

+++ ‐ designate strong interaction, + ‐ designate weak interaction, − ‐ designate no interaction

The cross‐reactivity data were extrapolated to multiple sequence alignment of tested antigens to theoretically predict the possible regions responsible for antibody‐antigen interactions. The 3F4 mAb‐binding pattern correlates the most with the amino acid composition of C‐terminal part of helix 2 and N‐terminal part of helix 3 interconnected by coil 3 of OspC_BRZ31 and OspC_BG (⁸⁵TKKLEXLIKNXGEL¹⁰³). The highest level of similarity with affinity data for 2A9 mAb was determined in the coil 4 and the C‐terminal part of helix 6 (¹³²ATDAXAK¹³⁸). For 2D1 highly conserved part of the helix 1 (⁴⁴SSIDELA⁵⁰) was in agreement with ELISA assays (Fig. 6).

Figure 6.

Prediction of interaction areas. The multiple sequence alignment of OspC antigens used in ELISA assays. Antibodies able to bind each of antigens are given on the right side of the alignment. The positions of amino acids corresponding to cross‐reactivity data in recognised variants of OspC for each antibody are highlighted under the alignment. Buried amino acids are labeled by symbol “b”. The completely buried regions are labeled by “B” and highlighted in grey colour.

To confirm the theoretical predictions, the trypsin or pepsin digest of OspCEX_BRZ31 used for immunization was incubated with each mAb in the pull‐down assay. The fragments were subsequently eluted from the G‐protein sepharose beads with immobilised mAb and identified using MALDI mass spectrometry analysis (Fig. 7). When the results were compared to negative control mAb (anti‐AHP protein antibody), there are apparent significant peaks in each case, which were identified based on the known amino acid composition. For 3F4 mAb two peaks with m/z 2239.2 (⁸⁸KLEELIKNPGELKAEISEAK¹⁰⁷) and m/z 2367.2 (⁸⁸KLEELIKNPGELKAEISEAKK¹⁰⁸) respectively were detected. In the case of 2A9 two peaks of m/z 1687.7 (¹¹⁷LKDSNAQLGVQNGAATDAR¹³⁵) and m/z 1929.0 (¹¹⁹DSNAQLGVQNGAATDAR¹³⁵) were observed. Both results were in agreement with previous theoretical prediction (Fig. 6). In the case of 2D1 mAb the MS analysis after trypsin digest showed that the mixture was enriched by fragment 38EVEALLSSIDELAAQAIGQKI⁵⁸ (m/z 2085.1 and m/z 2107.0) as predicted. Moreover, in the mixture, after pepsin digest treated with 2D1 mAb the ¹¹³FTKKLKDSNAQLGVQNGAATDARAKAAILK¹⁴³ (m/z 3000.5 and m/z 3071.5) fragment was detected, although in prediction alignment this position was evaluated by low expectation. When both 3000.5 and 2085.1 peptides were mapped into the published tertiary structure, there was apparent co‐localization with shared surface area (Fig. 8). The close proximity suggests that the 2D1 binds to both peptides forming a discontinuous epitope. The major role in specificity seems to be played by moiety ⁴⁴SSIDELA⁵⁰ as predicted and the interaction is stabilized, at least partially, by completely conserved moiety ¹⁴⁰AILK¹⁴³ in the C‐terminal region of helix 5 and the following coil. The conformation nature of recognised epitope is also supported by the results of ELISA and western blot. Although very similar absorbance was measured for 2D1 and 3F4 mAb under native conditions in ELISA, there is a significant difference in ability to bind tested antigens under denaturating conditions in western blot.

Figure 7.

Digested peptides enriched by immunoextraction were subjected to MS analysis to confirm the presence of specific peptides corresponding to predicted regions recognised by mAb.

Figure 8.

Determined localization of epitopes mapped into the OspC dimer structure (pdb entry 1ggq).

Discussion

In the current work, we identified epitopes of 3 mAbs (3F4, 2A9, 2D1) elicited against OspC protein variable regions. Two independent approaches (affinity data based theoretical prediction and MALDI‐TOF assisted epitope mapping) were employed for determining the location as well as the amino acid composition of recognised epitopes.

The theoretical prediction was based on affinity data obtained from five different OspC proteins. Predicted epitopes were subsequently compared with results from MS assisted epitope mapping of OspC_BRZ31 used for immunization. The comparison of determined epitopes showed that both methods produce comparable results for 2A9 and 3F4 mAbs and only a partial agreement was found for 2D1 mAb. The ⁸⁵TKKLEXLIKNXGEL¹⁰³ located at C‐termianl part of helix 2 and subsequent coil which was predicted for 3F4 antibody covers a substantial part of peptide ⁸⁸ KLEELIKNPGELKAEISEAK¹⁰⁷ detected in solution enriched by immunoprecipitation. The localization of the recognised epitope for 2A9 mAb was predicted mainly in the coil connecting helix‐3 and helix‐4 with eight amino acids (¹³²ATDAXXK¹³⁸), which correspond to MS detected peptide ¹¹⁹DSNAQLGVQNGAATDAR ¹³⁵. Minor differences in predicted and measured peptides were probably caused by the outcome of enzymatic digest taking place prior to pull‐down assays. The exception, when the epitope was predicted only partially, was the one for 2D1 mAb. Although the affinity data from ELISA assays and western blots suggested that the 2D1 mAb might recognise discontinuous epitope (Figs. 3 and 4), such pattern was not identified in the predictive alignment (⁴⁴SSIDELA⁵⁰). However, the MS analysis confirmed the presence of two different peptides 38EVEALLSSIDELAAQAIGQKI⁵⁸ and ¹¹³FTKKLKDSNAQLGVQNGAATDARAKAAILK¹⁴³ with ¹⁴⁰AILK¹⁴³ being the second part of epitope, thus forming discontinuous epitope consisting of central part of helix‐1 and C‐terminal of helix‐4 close to ligand binding domain.26 This inaccuracy is inherent in all prediction procedures based on sequence similarities as it is impossible to know which non‐contiguous residues along a sequence make up discontinuous epitope, especially when the particular part of the epitope have no effect on specificity. Moreover, the typical discontinuous epitope consists of 15–22 amino acids, but only 3–5 of them are significantly involved (functional epitope) in interaction with paratope.

The identified linear epitopes obtained for 3F4 and 2A9 mAbs colocalize with previously identified immunogenic parts of OspC protein. 2A9 mAb recognised region previously described as immunodominant and later included in the experimental OspC polyvalent vaccine.9 The topology of epitope corresponding to 3F4 mAb is in close proximity to the immunogenic region in OspC protein described by Yang.14 Both published epitopes, as well as epitopes determined in our experiments, are surface exposed and may be accessible in solution. The part of discontinuous epitope recognised by 2D1 mAbs has been previously observed in the study published by Paul Arnaboldi where the screening of peptide library was employed. The AILK moiety was described in his publication as epitope with diagnostic potential. Similar result confirming the immunogenicity of AILK containing region was obtained by Pulzova et al.22 More over the VETLL moiety was identified in the same study as well, which is located next to the SIDELA region. On the other hand, neither SIDELA moiety nor VETLL moiety did show any significant affinity to sera of LD patients in peptide library screening experiments.6 The difference may be explained either by disruption of 2D1 MAb epitope conformation10 or by differences in antibody reactivity in human patients and rodent hosts.27

When we compare our results with OspC sequences published previously28, 29, 30, 31, 32 we can estimate that linear epitopes recognised by 3F4 and 2A9 mAbs can be found only in a small fraction of OspC proteins. The fragment ⁸⁵TKKLEXLIKNXGEL¹⁰³corresponding to 3F4 mAb epitope was found in 8 instances (6.4%) out of 125 unique OspC amino acid sequences, ¹³²ATDAXXK¹³⁸ recognised by 2A9 was observed in 16 cases (12.8%). Discontinuous epitope ⁴⁴SSIDELA⁵⁰∼¹⁴⁰AILK¹⁴³ is on the other hand highly conserved with 74 (59.2%) instances found among compared OspC sequences. Moreover, the most variable position in ⁴⁴SSIDELA⁵⁰∼¹⁴⁰AILK¹⁴³ moieties is ⁵⁰serine at the N‐terminal of the first part of the epitope with other amino acids nearly fully conserved across all OspC sequences taken into account (Supporting Information 2–6).

Conclusions

In conclusion, we can summarise that the approach used during our study enable us to determine the epitope location for all three mAbs (3F4, 2A9 and 2D1) and to approximate amino acid composition. Such an approach seems to be well applicable in other variable proteins as well. These findings confirmed feasibility of chosen combination of methods using experimental data from immunoassays, analysis of multiple sequence alignments and MS assisted epitope mapping for rapid epitope determination. Moreover, we were able to recognised highly conserved discontinuous epitope ⁴⁴SSIDELA⁵⁰∼¹⁴⁰AILK¹⁴³, which may have great implications in further development of diagnostic as well as prophylactic tools concerning LD.

Material and Methods

General

All common chemicals were purchased from Sigma‐Aldrich (USA) in at least analytical grade. Enzymes used in DNA cloning were obtained from New England BioLabs, USA) and all filtration devices were purchased from Merck‐Millipore (Germany).

OspC protein production

Two types of OspC antigens were used in this study here designated OspCEX (immunizations) and OspC (hybridoma cells selection and affinity studies). Recombinant OspC was derived from the gene for ospC, but the signal sequence and the first 10 amino acids were omitted.33 The amplified coding sequence was fused to ubiquitin with hexahistidine tag by ligation into the pETM60 plasmid34 giving the OspCEX protein. The Ubiquitin with hexahistidine tag was subsequently cleaved by TEV protease and the OspC with additional four N‐terminal amino acids (GAME) was obtained.

The DNA coding OspC was obtained by PCR amplification of genomic DNA isolated from five strains of Borrelia burgdorferi sensu lato: B. afzelii BRZ31, access. Nr. JN828672 35, 36 (BRZ31); B garinii, access. Nr. CAH56465 (BG); B. burgdorferi, access. Nr. L42871 (BB); B. valaisiana, access. Nr. WP_012664746 (BV) and B. afzelii, access. Nr. ABA42057 (BA).

Two sets of primers (ospCEX_F: ATACCATGGAGGCATCTACTAATCCTGATG and OspCEX_R: CATGGATCCTTAAGGTTTTTTTGGACTTTCTG) were employed for amplification and introduction of NcoI/BamHI cloning sites. Emerald Master Mix (Takara Bio, USA) was used for PCR according to manufacturer instructions with annealing temperature 53°C for 20 sec. The PCR product was treated with corresponding endonucleases for 2 h at 37°C (NcoI/BamHI), separated by electrophoresis, excised and isolated (NucleoSpin Gel and PCR Clean‐up kit, Macherey‐Nagel, Germany) prior the ligation into the pETM60_Ub3 plasmid containing N‐terminal ubiquitin and hexahistidine tag to promote solubility and purification.34

All ligation steps have been performed according to the manufacturer's instructions in the reaction mixture containing 1U T4 DNA ligase, 2 μl of T4 buffer, 50 ng of linearized vector and corresponding insert (vector: insert molar ratio 1:9). The total volume of the reaction mixture was 20 μL. The ligation mixtures were incubated for 16h at 16°C, 5 μL was used subsequently for competent cells transformation. For general cloning and plasmid maintenance the E. coli DH10B strain was used, while E. coli BL21(DE3) RILP have served as the host strain for protein expression (Agilent Technologies, USA).

Protein expression and cell disruption

Large‐scale expressions were performed in 3 L flasks using 500 mL of LB Broth Low Salt media (Duchefa Biochemie, Netherlands), with corresponding antibiotic at 37°C/220 rpm. When the OD₆₀₀ reached ∼0.6, the culture was induced with 0.6 mM IPTG and incubated at 22°C for 16 h. Cells were cooled down on ice, harvested by centrifugation (8000 g for 20 min at 4°C) and resuspended in 12 mL of loading buffer (100 mM Tri HCl pH 7,2; 300 mM NaCl; 0,01% Tween 20; 10 mM imidazole; 20% glycerol). Cell suspension was sonicated on ice (12.7 mm probe, 1 s pulse with amplitude 30, 4 s pause, working time 6 min.) using Q700 sonicator (Qsonica, USA)

Protein purification

The cleared supernatant (centrifugation 20.000g/45min./4°C) was filtered through 0,22 μm syringe filter and applied onto the 5 mL HisTrap column (GE Healthcare, USA) equilibrated with loading buffer. After thorough washing (10 CV) the non‐specifically bound proteins were removed by 59 mM imidazole in loading buffer followed by peaks of HiSalt (100 mM Tri HCl pH 7,2; 500 mM NaCl; 0,01% Tween 20; 10 mM imidazole; 20% glycerol) and LowSalt (100 mM Tri HCl pH 7,2; 0,01% Tween 20; 10 mM imidazole; 20% glycerol) buffers switching after 5 mL (repeated six‐times) and 5 CV of 108 mM imidazole in loading buffer. The elution was achieved by 3 CV of elution buffer (100 mM Tri HCl pH 7.2; 300 mM NaCl; 0,01% Tween 20; 500 mM imidazole; 20% glycerol). All fractions were examined by SDS‐PAGE for purity and stored either at −80°C in the same buffer (long term storage) or at +4°C up to one month.

Ubiquitin removal

Purified OspCEX was dialysed overnight at 4°C (4 kDa MWCO) against the TEV buffer (50 mM Tri HCl pH 7,2; 300 mM NaCl; 0,01% Tween 20; 20% glycerol; 1 mM ethylenediaminetetraacetic acid) with constant stirring. Subsequently, the TEV protease was added at a mass ratio of 70 : 1 (i.e., for 1 mg of purified protein, 14 ng of TEV protease was added). After 16 h of incubation at 4°C under mild agitation, the protein sample was loaded onto nickel column. The column was washed with the loading buffer free of imidazole and flow‐through fractions were collected to obtained protein without ubiquitin was stored.

Production of mAbs

MAbs were raised against OspCEX_BRZ31 (conjugated with Ub3) to promote the production of specific antibodies (Veterinary Research Institute, Brno, Czech Republic). In summary, white females of BALB/c mice were injected three times with 100 μg of purified OspCEX intraperitoneally in the complete Freund's adjuvans on the day 0 and on the day 14 and 28 in incomplete Freund's adjuvans. Spleen cells from a single mouse were used for the hybridoma fusion with Sp2/0 myeloma cells in ratio 1:10. For the cultivation, a foetal bovine serum depleted of bovine IgG (Low IgG FBS) was used.

The selection of positive hybridoma cells against three different OspC proteins and corresponding Bbsl lysates (BRZ_31, BG, BB) was done with indirect ELISA and western blot assays. The selection process was configured to detect antibodies capable of binding to variable parts of OspC protein. Only those hybridoma cells producing mAbs with the high affinity to some of OspC proteins used (but not binding to all of them) were selected for propagation.

Selected mAbs were purified using the standard procedure with HiTrap Protein G HP column (GE Healthcare, USA) according to manufacturer instructions. Pure fractions were dialysed against PBS buffer with 20% glycerol, diluted to final concentration 1 mg/mL and stored at −80°C.

ELISA assays

Indirect ELISA assays were performed for determination of binding activity of purified mAbs against five OspC proteins (BRZ31, BG, BB, BV and BA) under native conditions. Up to 500 ng of each antigen in PBS buffer (137 mM NaCl, 2.7 mM KCl, 8 mM Na₂HPO₄, 2 mM KH₂PO₄, pH 7.2) was applied to Maxisorp 96 well plates (Nunc, Denmark) and coated overnight at 4°C. Wells were blocked with 1% (w/w) bovine serum albumin in PBS‐T (PBS containing 0.05% Tween‐20) for 1h at RT, and appropriate concentrations of mAbs were added and incubated under agitation for 1 hour at RT. Bound recombinant antibodies were detected with 1:10,000 secondary anti‐mouse IgG HRP conjugated antibody (Sigma‐Aldrich, USA) and TMB substrate (TestLine, Czech Republic). The absorbances were measured by PowerWave 340 ELISA reader instrument (BioTek, USA) at 450 nm.

Western blot assay

Recombinant OspC proteins (BRZ_31, BG, BB) and corresponding Bbsl lysates (100 ng) were separated by SDS‐PAGE and electroblotted onto Immobilon‐P Transfer membrane (Merck‐Millipore, Germany) under custom set conditions (1A/30V/35 min.) using TransblotTurbo (Bio‐Rad, USA). After blocking (1% BSA in PBS for 1 hour), the membrane was incubated with the corresponding mAb overnight at 4°C, washed three times with PBS‐T and detected with secondary anti‐mouse IgG alkaline phosphatase conjugated antibody (Sigma). The membrane was equilibrated for 15 minutes in detection buffer (100 mM Tris pH 9.5; 100 mM NaCl; 5 mM MgCl₂) preceding the membrane development.

Prediction of mAbs binding sites

For theoretical prediction of the possible binding sites the overlaps of identical regions in antigens used were evaluated. The multiple sequence alignment of BRZ_31, BG, BB, BV and BA was prepared using Clustal Omega algorithm37, 38 in the default setup and minor manual adjustments based on the published OspC coordinates were done.39 Subsequently, positions of buried amino acids where the low probability of involvement in the antigen‐antibody interface can be suspected were determined using NetSurfP ver. 1.140 (SAS – solvent accessible surface) and OspC protein structural data.15 The evaluation of conserved amino acid positions corresponding to the results of ELISA assays (binding pattern) was done manually. Each amino acid position in the alignment was scored based on the similarity to cross‐reactivity data formulated in “binding pattern”. Positions identical to the binding patterns were scored by three points. When the identity was supported by adjacent amino acid (either corresponding to the binding pattern or complete identity) the score was increased by one for each of the surrounding amino acids with the maximal possible score 5. When the surrounding amino acid did not support the binding pattern or it was predicted as buried the score was decreased by one point for each amino acid.

MALDI‐TOF assisted epitope mapping

Immunoextraction coupled with MALDI‐TOF (matrix‐assisted laser desorption/ionization time of flight) mass spectrometry was used for analysis of epitope‐containing peptides. The method is based on immunoaffinity extraction of specific peptides on immobilised mAb followed by MS.41

30 μg of OspC_BRZ31 was digested with trypsin or pepsin (1:50), and the reaction was terminated after 30 minutes. The digest was diluted in PBS, and 100 μg mAb was added to final volume 500 μL and incubated for 1 hour under gentle agitation at RT. 100 μL of Protein‐G Sepharose (GE Healthcare, USA) was added subsequently and the suspension was incubated for next 12 hours under mild agitation at 4°C. The final mixture was centrifuged for 2 min (2000 g/4°C), the supernatant has been discarded and replaced by 500 μL of immunoprecipitation buffer (25 mM Tris‐HCl pH 7.2; 150 mM NaCl). The wash step was repeated four times before the elution was achieved by adding 20 μL of elution buffer (100 mM glycine – HCl pH 2.7). In the last step, the Sepharose beads were pelleted by centrifugation (2500 g for 5 min.) and the supernatant was carefully collected and analyzed by MALDI‐TOF MS on a fee for service bases (Proteomics Core Facility, CEITEC, Czech Republic).

Supporting information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information

Supporting Information