High antibody titres induced by protein subunit vaccines using Mycobacterium ulcerans antigens Hsp18 and MUL_3720 with a TLR-2 agonist fail to protect against Buruli ulcer in mice

Strains and culture conditions

Escherichia coli Rosetta2 containing plasmid pET30b-Hsp18 (strain TPS681) or pDest17-MUL_3720 (strain TPS682) was grown at 37 °C in Luria-Bertani (LB) broth (Difco, Becton Dickinson, MD, USA) supplemented with 100 µg/ml ampicillin (Sigma-Aldrich, USA) or 50 µg/ml kanamycin to express 6xHIS-tagged Hsp18 or MUL_3720 recombinant protein. Mycobacterium ulcerans (strain Mu_1G897 from French Guiana (De Gentile et al., 1992)) was grown at 30 °C in 7H9 broth or 7H10 agar (Middlebrook, Becton Dickinson, MD, USA) supplemented with oleic acid, albumin, dextrose and catalase growth supplement (OADC) (Middlebrook, Becton Dickinson, MD, USA), and 0.5% glycerol (v/v). M. bovis BCG (strain Sanofi Pasteur) used for vaccinations was grown at 37 °C in 7H9 broth or 7H10 agar supplemented with OADC. Mycobacterial colony counts from cultures or tissue specimens were performed using spot plating as previously described (Wallace et al., 2017).

Recombinant protein expression

Overnight cultures of strains TPS681and TPS682 were diluted to OD₆₀₀ = 0.05 in LB broth. Each culture was incubated at 37 °C with shaking at 200 rpm until OD₆₀₀ = 0.6–0.7, then 1 mM IPTG (Isopropyl b-D-1-thiogalactopyr-anoside) was added to induce protein expression. The cells were incubated for a further four hours to express the protein. To harvest the protein, cells were resuspended in wash buffer (8 M urea, 150 mM sodium chloride, 10% glycerol) and sonicated at amplitude 60 (QSonica Ultrasonic Liquid Processor S-4000, Misonix) until the solution turned clear. The lysate was filtered with a 0.22 µM filter (Millipore) to remove cellular debris and the protein was column-purified using anti-histidine resin (ClonTech). The resin was washed ten times with 10x column volumes of wash buffer mixed with an increasing proportion of tris buffer (20 mM Tris–HCl, 150 mM sodium chloride, 10% glycerol) until the column was washed with only tris buffer. The resin was washed a further two times with tris buffer containing 20 mM imidazole. Protein was eluted in tris buffer containing 200 mM imidazole and dialysed in phosphate buffered saline (PBS) before concentration using a 3K MWCO PES concentration column (Pierce). Protein was endotoxin purified using Triton X-114 phase separation until less than 0.1 endotoxin unit/ml (detectable limit), measured by PierceTM LAL choromogenic endotoxin quantitation kit (ThermoFisher) as per manufacturer’s instructions.

Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE)

Samples were denatured in an equal volume of 2 x sample loading buffer (40% (v/v) 0.5M Tris-HCL pH 6.8, 10% glycerol, 1.7% (w/v) SDS, 10% 2-β-mercaptoethanol, 0.13% (w/v) bromophenol blue in distilled water) at 100 °C for 5 min. Ten microlitres of each sample and SeeBlue^® Plus2 pre-stained protein standard (Invitrogen) were loaded into a 0.5 mm 12% polyacrylamide gel under reducing conditions, as previously described (Laemmli, 1970). The gel was run in running buffer (0.3% (w/v) Tris, 1.44% (w/v) glycine and 0.1% (w/v) SDS in distilled water) for 1 h at 150 volts (Mini-protean vertical electrophoresis cell, Bio-Rad). The gels were stained in Coomassie stain (45% methanol, 10% acetic acid 0.25% (w/v) Coomassie brilliant blue in distilled water) for 1 h and destained in Coomassie destain (33% Methanol, 10% acetic acid, 60% distilled water) until the protein bands could be identified.

Western Blotting

Proteins were separated on a 12% polyacrylamide gel as per the method for SDS-PAGE. After separation proteins were transferred to a nitrocellulose membrane in tris-glycine transfer buffer (1.5 mM Tris, 12mM glycine, 15% methanol (v/v) in distilled water) for 1 h at 100 volts (Mini Trans-Blot Cell, Bio-Rad). The nitrocellulose membrane was blocked in blocking buffer (5% (w/v) skim milk powder and 0.1% Tween-20 in PBS) overnight at 4 °C. The membrane was incubated in blocking buffer containing anti-6xHIS-HRP antibody (Roche Applied Science) at 1:500 dilution. The membrane was washed in PBS containing 0.1% Tween-20 and then exposed to developing solution (Western Lighting Chemiluminescence kit, Perkin Elmer) according to manufacturer’s guidelines. Chemiluminescence was detected using an MF ChemiBIS gel imaging system (DNR Bio-Imaging Systems).

Analysis of electrostatic interaction between protein antigen and lipopeptide formulations

The association between each protein and R₄Pam₂Cys was measured by mixing 25 µg of protein with increasing amounts of lipopeptide in 50 µl PBS in a 96-well plate (Nunc, Thermo Scientific). The formation of protein-lipopeptide complexes through electrostatic interaction was measured by an increase in light absorbance. Plates were read at dual wavelengths of 505 and 595 nm on plate reader (LabSystems Multiskan Multisoft microplate reader).

Lipopeptide vaccine preparation

Each vaccine dose contained 25 µg protein added to R₄Pam₂Cys at a ratio of 1:5 mole of protein to lipopeptide. PBS was added to a final volume of 100 µl and the combination sonicated in a water bath for 30 s. Control vaccine preparations were made containing 25 µg protein alone or R₄Pam₂Cys lipopeptide alone and sonicated before administration.

Ethics statement for animal experiments

All animal experiments were performed in full compliance with national guidelines (articles R214-87 to R214-90 from French “rural code”) and European guidelines (directive 2010/63/EU of the European Parliament and of the council of September 22, 2010 on the protection of animals used for scientific purposes). All protocols were approved by the Ethics Committee of region Pays de la Loire under protocol nos. CEEA 2009.14 and CEEA 2012.145. Animals were maintained under specific pathogen-free conditions in the animal house facility of the Centre Hospitalier Universitaire, Angers, France (agreement A 49 007 002). Six-week old female C57BL/6 and BALB/c mice were obtained from Charles River Laboratories (Saint-Germain-Nuelles, France) and housed at CHU Angers. Food and water were given ad libitum. Animals were euthanized by inhalation of CO₂ gas, delivered using a gradual fill method in a chamber containing the animal. All personnel using this technique were appropriately trained to operate the equipment, evaluate animal vital signs and confirm death.

Vaccination of animals

The synthesis and purification of the branched cationic lipopeptide, R₄Pam₂Cys, was performed as previously described (Chua et al., 2011; Sekiya et al. 2017; Wijayadikusumah et al., 2017). Each vaccine dose contained 25 µg protein formulated in PBS with R₄Pam₂Cys at a 1:5 molar ratio of protein to lipopeptide in a final volume of 100 µl. The protein alone control formulation contained 25 µg protein per dose diluted in PBS. The R₄Pam₂Cys alone formulations contained the same amount of lipopeptide used in each of the protein + adjuvant formulations, calculated by the 1:5 molecular ratio (with the omission of the protein from the solution). The R₄Pam₂Cys alone formulations were diluted to the correct concentration in PBS. Live-attenuated M. bovis BCG strain ‘Sanofi Pasteur’ was grown to log phase and stored at −80 °C in 20% glycerol until use. Bacteria were washed with PBS and resuspended in 200ul, before administration at 4.7 × 10⁵ bacteria per dose. All vaccines and control formulations were sonicated for 5 min in a waterbath sonicator before being administered.

For vaccination using R₄Pam₂Cys, animals were inoculated subcutaneously at the base of tail (100 µl per dose at 50 µl per flank) and boosted 21 days later with the same formulations. Mice vaccinated with approximately 1 × 10³ CFU M. bovis BCG resuspended in PBS at the base of tail (100 µl per dose at 50 µl per flank).

M. ulcerans challenge

Mice were challenged on day 35 by subcutaneous injection on the tail with 1 × 10⁴ CFU M. ulcerans (Mu_1G897) resuspended in 50 µl PBS. Mice were allowed to recover and monitored for up to 40 days after infection and euthanised when tail ulceration was observed wherein sera were obtained for immunological analysis.

Serum antibody titre measurements

Serum was prepared from blood obtained from mice at day 0, day 18, day 33 and day 63. Antibody titres were measured using enzyme linked immunosorbent assay (ELISA) as per methods described in Chua et al. (2011). Briefly, ELISA plates (Nunc, Thermo Scientific) were coated overnight with 5 µg/ml protein diluted in PBSN₃ and blocked with BSA₁₀PBS for 2 h at room temperature. Plates were washed with PBS containing 0.05% Tween-20 (PBST). Neat sera were sequentially diluted in BSA₅PBST and incubated at room temperature for 6 h. Bound antibody was detected by adding horse radish peroxidase conjugated rabbit anti-mouse IgG (Dako, Glostrup, Denmark) or rat anti-mouse IgM, IgG1, IgG2a, IgG2b or IgG3 antibodies (Southern Biotech, USA) at a concentration of 1:400 in BSA₅PBST for 2 h. Plates were developed with developing solution (hydrogen peroxide, citric acid and ABTS) and incubated for 10-15 min with gentle agitation to observe a colour change. The reaction was stopped with 50 mM sodium fluoride. Plates were read at dual wavelengths of 505 and 595 nm on plate reader (LabSystems Multiskan Multisoft microplate reader). The titers of antibody are expressed as the reciprocal of the highest dilution of serum required to achieve an OD₆₀₀ of 0.2.

Statistical analysis

Graphpad Prism software (GraphPad Software v7, CA, USA) was used to perform statistical analyses on the antibody titre. Antibody titres were analysed using two-way ANOVA with Tukey’s correction for multiple comparisons. The time-to-ulceration data were displayed as a Kaplan–Meier plot and statistical significance was determined using a Log-Rank (Mantel-Cox) test. For all tests *p <0.05, **p <0.01 and ***p <0.001 and **** p <0.0001 were considered statistically significant.

Recombinant MUL_3720 and Hsp18 both bound to R₄Pam₂Cys

Recombinant MUL_3720 and Hsp18, expressed from inducible E. coli expression vectors, were prepared for use as antigens in the vaccine formulations (Table S1). Purification of the recombinant proteins was confirmed by SDS-PAGE and Western blot analyses of the eluate (Fig. 1). DLS analysis was then performed to identify whether recombinant MUL_3720 or Hsp18 would electrostatically bind to either the positively charged lipopeptide adjuvant R₄Pam₂Cys, or its negatively charged counterpart, E₈Pam₂Cys. The optical density of solutions containing these constituents at a wavelength of 450 nm (OD₄₅₀) is related to the particle size of molecules in solution, reflecting the strength of the ionic interaction between protein and lipopeptide (Chua et al., 2011). MUL_3720 preferentially bound to R₄Pam₂Cys compared to E₈Pam₂Cys (Fig. 2A, Table S2). This is shown as a gradual increase in optical density following the addition of increasing amounts of R₄Pam₂Cys to a constant amount of MUL_3720. At a 5-fold molar excess of protein to lipopeptide the OD ₄₅₀plateaued, suggesting MUL_3720 bound most strongly to R₄Pam₂Cys at a 1:5 protein to lipopeptide ratio. Conversely, when E₈Pam₂Cys was added to MUL_3720 the optical density remained static and did not increase with increasing lipopeptide concentrations, indicating a lack of binding. Hsp18 also appeared to bind preferentially to R₄Pam₂Cys and also at a 1:5 ratio of Hsp18 to R₄Pam₂Cys (Fig. 2B). Therefore, two protein-adjuvant formulations were prepared using MUL_3720 with R₄Pam₂Cys and Hsp18 with R₄Pam₂Cys, both at a 1:5 protein to lipopeptide molar ratio.

Vaccination induced strong protein-specific antibody responses

Prior to challenge with M. ulcerans, the ability of the vaccine candidates to generate murine immune responses was assessed. ELISAs were utilized to measure the antibody (IgG) titres in sera obtained from two strains of mice (BALB/c and C57BL/6) immunized with either MUL_3720 + R₄Pam₂Cys or Hsp18 + R₄Pam₂Cys after the primary vaccination dose (day 18) and a secondary dose (day 33) (Fig. 3, Table S3).

Vaccination with MUL_3720 recombinant protein alone or MUL_3720 + R₄Pam₂Cys were capable of inducing MUL_3720-specific antibody titres in both BALB/c and C57BL/6 strains of mice (Figs. 3A, 3B). Primary vaccination with MUL_3720 protein alone induced MUL_3720-specific antibody responses that significantly increased following a vaccine boost (p < 0.0001 and p = 0.0005 for BALB/c and C57BL/6, respectively). Additionally, MUL_3720 + R₄Pam₂Cys generated MUL_3720 specific antibody responses after primary vaccination (p  <  0.0001 in BALB/c and C57BL/6), which were increased after the secondary boost (p <0.0001 in BALB/c and p = 0.0035 in C57BL/6). The titres after the boost in particular were greater than MUL_3720 alone vaccination (p  < 0.0001 in BALB/c and p = 0.0075 in C57BL/6). Mice that were not vaccinated with recombinant MUL_3720 (R₄Pam₂Cys alone and BCG) did not have an increase in MUL_3720-specific antibodies compared to naïve mice.

Vaccination with Hsp18 recombinant protein alone or Hsp18 + R₄Pam₂Cys induced Hsp18-specific antibody titres in both strains of mice (Figs. 3C, 3D). Vaccine boost with Hsp18 recombinant protein alone induced significantly higher Hsp18-specific antibody responses in BALB/c mice compared to a single vaccination with Hsp18 protein (p  <  0.0001). Boosting with protein alone in C57BL/6 did not significantly increase antibody titres. Hsp18 + R₄Pam₂Cys induced Hsp18-specific antibody responses in both mouse strains after primary vaccination (p  <  0.0001 in BALB/c and C57BL/6) and the Hsp18-specific antibody titre significantly increased after booster vaccination (p  <  0.0001 in BALB/c and p = 0.0006 in C57BL/6). In all strains, the antibody titres induced by Hsp18 + R₄Pam₂Cys were significantly higher than vaccination with Hsp18 protein alone (p  <  0.0001 in BALB/c and C57BL/6) (Figs. 3C, 3D) with negligible levels of antibodies seen in mice vaccinated with only R₄Pam₂Cys, or BCG.

Measurement of IgG antibody subtypes following MUL_3720 + R₄Pam₂Cys and Hsp18 + R₄Pam₂Cys vaccination

Quantifying levels of IgG antibody shows that the predominant isotypes produced by MUL_3720 were IgG1 and IgG2_b (Fig. 3E) with no significant difference between these isotype titres. The antibody titres for both isotypes were highest prior to infection with M. ulcerans (day 33) and decreased after infection by day 63. This vaccine was capable of inducing IgG2a antibodies, which was detected also on day 33, however in smaller amounts than IgG₁ (p = 0.0300; Fig. 3E).

Similar to vaccination with MUL_3720, Hsp18 was also capable of inducing strong IgG antibody titres. The predominant isotype was IgG1 which Hsp18 + R₄Pam₂Cys elicited more than any other isotype including IgG2_a (p = 0.0317) and IgG2_b (although, not significant) (Fig. 3F). Again, these titres were highest at day 33 and decreased significantly after infection on day 63. This trend was also observed after vaccination with Hsp18 alone (p < 0.0001 vs IgG2_a and p = 0.0001 vs IgG2_b, respectively at day 33).

MUL_3720 + R₄Pam₂Cys and Hsp18 + R₄Pam₂Cys do not protect against the onset of BU

As both vaccines were capable of inducing protein-specific antibody responses, they were tested in a murine challenge model to measure their protective efficacy. Efficacy was measured by time delay to the onset of ulceration in a mouse tail infection model. There is a progression of clinical symptoms for Buruli ulcer in this model (Fig. 4). Once ulceration has been reached the disease would likely continue until the tail became necrotic. Therefore, the experimental endpoint was deemed to be the point of ulceration.

After the scheduled vaccinations, mice were challenged via subcutaneous tail inoculation with 1 × 10⁴ CFU of M. ulcerans and observed for up to 40 days. In BALB/c and C57BL/6 mice there was no significant difference between the time to ulceration between control mice (mice not vaccinated with recombinant protein, such as R₄Pam₂Cys alone and BCG) and mice vaccinated with either MUL_3720 + R₄Pam₂Cys or Hsp18 + R₄Pam₂Cys (Figs. 5A, 5B). There was also no significant difference in the time to ulceration between mice that were vaccinated with MUL_3720 + R₄Pam₂Cys or Hsp18 + R₄Pam₂Cys and BCG, the benchmark for mycobacterial vaccine efficacy. Signs of infection in mice were visible by day 63 (Tables S4 and S5) and all mice reached ulceration by day 75, 40 days post-M. ulcerans challenge (Figs. 5A, 5B).

Antibody titres do not correlate with protection against M. ulcerans

High antibody titres were observed in all mice vaccinated with either recombinant MUL_3720 or Hsp18, particularly in the secondary response after booster vaccination (Figs. 3A–3D) prior to M. ulcerans challenge. However, mice vaccinated with protein alone or protein plus lipopeptide adjuvant all succumbed to infection by day 75. The sera from mice at the day 63 was used to quantify antibody titres during infection. At day 63 all mice still had detectable protein-specific antibodies against the recombinant protein with which they were vaccinated (Figs. 3A–3D). In BALB/c mice (Figs. 3A, 3C) the antibody titres at day 63 were lower than after the secondary response prior to challenge (p < 0.0001 Hsp18 + R₄Pam₂Cys and not significant for MUL_3720 + R₄Pam₂Cys) but remained significantly higher than at day 0 (p < 0.0001 for both Hsp18 + R₄Pam₂Cys and MUL_3720 + R₄Pam₂Cys). In C57BL/6 mice (Figs. 3B and 3D), antibody titres against MUL_3720 or Hsp18 from mice vaccinated with either protein alone or protein plus lipopeptide adjuvant were also significantly decreased at day 63 compared to the secondary response at day 35 (p < 0.0001 for MUL_3720 + R₄Pam₂Cys and Hsp18 + R₄Pam₂Cys, respectively). Similar to BALB/c mice, the day 63 respective protein-specific antibodies for MUL_3720 + R₄Pam₂Cys and Hsp18 + R₄Pam₂Cys were significantly higher than at day 0 (p < 0.0001 for MUL_3720 + R₄Pam₂Cys and Hsp18 + R₄Pam₂Cys).

Challenge with M. ulcerans did not induce protein-specific antibody levels comparable to vaccination with MUL_3720 or Hsp18

MUL_3720 and Hsp18 recombinant proteins are immunogenic and capable of inducing protein-specific antibody responses after vaccination. However, only minor detectable antibody responses against either recombinant MUL_3720 or Hsp18 at day 63 (Figs. 3A–3D) were found in mice vaccinated with R₄Pam₂Cys alone or BCG then challenged with M. ulcerans. These responses are much lower than the protein-specific antibody responses generated from MUL_3720 or Hsp18 vaccinated mice, particularly in C57/BL6 mice (p < 0.0001) (Figs. 3A–3D). Animals from both mouse strains that were vaccinated with R₄Pam₂Cys alone or BCG showed no increase in protein-specific antibody responses against either recombinant MUL_3720 and Hsp18 on day 63 post-M. ulcerans challenge (Figs. 3A–3D), even though these two proteins are both expressed in M. ulcerans.