- Email: [email protected]
Discovery of a recombinant Babesia canis supernatant antigen that protects dogs against virulent challenge infection
Veterinary Parasitology 249 (2018) 21–29
Contents lists available at ScienceDirect
Veterinary Parasitology journal homepage: www.elsevier.com/locate/vetpar
Research paper
Discovery of a recombinant Babesia canis supernatant antigen that protects dogs against virulent challenge infection
MARK
K. Moubria, J. Kleuskensb, J. Van de Crommertb, N. Scholtesb, T. Van Kasterenb, S. Delbecqa, ⁎ B. Carcya, E. Précigouta,1, A. Gorenflota, Th. Schettersa,b,c, a
EA4558 VAP, University of Montpellier, Montpellier, France MSD-AH, Boxmeer, The Netherlands c Department Veterinary Tropical Diseases, University of Pretoria, Onderstepoort, South Africa b
A R T I C L E I N F O
A B S T R A C T
Keywords: Babesia canis recombinant vaccine antigen protective protozoa immunity
Soluble parasite antigens (SPA) in supernatants of in vitro cultures of Babesia canis can be used to vaccinate dogs against virulent B. canis infection. The moment that immunity becomes apparent coincides with the appearance of antibodies against SPA in the serum of the vaccinated animals. This so-called vaccination-challenge serum (VC-serum) was used to precipitate antigens from B. canis culture supernatants in agarose gels. This antigen preparation was then used to analyse the reactivity of sera from vaccinated dogs on western blots. Results: showed that the first appearance of antibody reactivity against a protein that migrated at the 39 kDa position in SDS-PAGE gels was associated with the moment vaccinated dogs started to recover from a virulent challenge infection. In addition, pulse-chase experiments revealed that a 39–40 kDa doublet was released into the supernatant of B. canis cultures starting 15 min after the chase. This doublet was specifically precipitated by VC-serum, thus corroborating that the 39–40 kDa doublet in SPA preparations was of parasite origin. Partial amino acid sequencing allowed the discovery of the gene that encoded the 39–40 kDa doublet (canine Babesia antigen; CBA). The full-length gene was cloned and expressed in E. coli. The recombinant CBA protein (rCBA) was recognized by VC-serum, and antibodies against rCBA precipitated the 39 kDa antigen of SPA preparations and of merozoites of B. canis. In addition, anti-rCBA serum reacted with the surface of B. canis merozoites (but not with B. rossi merozoites) in immunofluorescence. Vaccination of dogs with rCBA induced antibodies against rCBA, which recognized B. canis merozoites. Vaccinated dogs were protected against virulent challenge infection by limiting parasite proliferation. As a result, the development of clinical signs was prevented and the animals self-cured. In contrast, six out of seven non-vaccinated control dogs developed relatively high parasitaemia and serious clinical signs associated with poor tissue perfusion. This antigen can be used to replace the SPA antigen in the conventional B. canis vaccines, which eliminates the need for dog blood and serum for vaccine production.
1. Introduction Babesiosis is a tick-transmitted disease of vertebrates. In dogs the disease is caused by four different large Babesia species (B. canis, B. rossi, B. vogeli and B. presenti) and three different small Babesia species (B. gibsoni, B. conradae and B. vulpes; Uilenberg, 2006; Baneth et al., 2015). In Europe, B. canis is the most prevalent Babesia species of dogs. This species is transmitted by Dermacentor reticulatus ticks. The clinical manifestations vary from anorexia, lethargy and anaemia, to a syndrome in which multiple organs are affected (Matijatko, Torti and Schetters 2012). The infection can be cured chemotherapeutically with imidocarbdipropionate, and to prevent disease two vaccines have been
⁎
1
developed commercially (Schetters, 2005). These vaccines contain antigens from supernatants of in vitro cultures of the parasite (soluble parasite antigens; SPA), adjuvanted with a saponin. The protective effect of these vaccines is on the development of clinical manifestations, not necessarily on the development of the parasite per se (Schetters et al., 1996a). In the same studies it was further shown that vaccinated animals produced antibodies against SPA upon challenge infection as detected by sandwich-ELISA. Such reactivity was not found in serum of control dogs that survived a primary infection with B. canis. Hence, serum from dogs that were vaccinated and subsequently challenged (VC-serum) could be crucial in the discovery of the immunoprotective parasite antigen in SPA preparations. Here we describe the discovery
Corresponding Author at: ProtActivity R&D, Sering 36, 5432 DD CUIJK, The Netherlands. E-mail address: [email protected] (T. Schetters). Deceased
https://doi.org/10.1016/j.vetpar.2017.11.002 Received 14 July 2017; Received in revised form 3 November 2017; Accepted 5 November 2017 0304-4017/ © 2017 Elsevier B.V. All rights reserved.
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
2.2.5. Parasitaemia Percentage of infected red blood cells was determined from blood smears stained with May-Grünwald/Giemsa solutions. The parasitaemia is expressed as the log10 number of red blood cells infected with Babesia parasites per 105 erythrocytes in the peripheral blood. The parasitic load was calculated by adding daily parasitaemia values (cumulative parasitaemia).
and characterization of this antigen in B. canis culture supernatants. In addition, the gene encoding for this antigen was isolated, and the protective effect of a recombinantly expressed protein was evaluated in dogs. 2. Materials and methods 2.1. Parasite
2.2.6. Serum Blood was collected from the vena jugularis and allowed to clot at room temperature for at least 60 min. Cellular contaminants were removed by centrifugation (1500g, 10 min, at room temperature). Serum was stored at −70 °C until use.
The Babesia canis A isolate isolated in 1982 from Dermacentor reticulatus ticks recovered from a dog that had just returned from the Drôme region in France and suffered from babesiosis; (Uilenberg et al., 1989) was used in the experiments. A second B. canis isolate that also derived from France, and which was shown to be immunologically different from B. canis A (named B. canis B; Schetters et al., 1995), and a B. rossi isolate that was originally isolated from a dog from Pietermaritzburg (South Africa; Uilenberg et al., 1989) were additionally used in immunofluorescence experiments. Parasites were stored as stabilates in liquid nitrogen, and passed through splenectomised Beagle dogs to obtain freshly isolated infected blood. Infected blood was taken by venepuncture from the vena jugularis, using heparin (5–10 U/ml, Leo) to prevent clotting. Blood was washed with Babesia medium (Schetters et al., 1992). The challenge inoculum was injected intravenously in experimental animals.
2.2.7. Vaccination-challenge serum (VC-serum) Vaccination-challenge serum was obtained from Beagle dogs that had been vaccinated 2–4 times with SPA of B. canis A with 3-week intervals, and subsequently challenged with homologous B. canis A parasites, as described earlier (Schetters et al., 1996b). Serum was stored at −70 °C until use. 2.3. Soluble parasite antigen from in vitro cultures Babesia canis A was maintained in continuous in vitro culture, following the procedure described previously (Schetters et al., 1997). Briefly, blood was drawn from B. canis A-infected dogs, and washed thrice with culture medium (RPMI 1640, Gibco, containing 25 mM Hepes and 25 mM NaHCO3, pH 7.3, supplemented with 10% (v/v) normal dog serum and 1 g/l reduced glutathione[GSH]). The pellet was suspended with medium to 2% (v/v) packed cell volume (PCV). Of these suspensions 14 ml was cultivated in 25 cm2 tissue culture flasks that were put flat in the incubator so as to obtain 6 mm medium height. The incubator was set at 37C, 5% (v/v) CO2, 90% relative humidity. Caps were loosely tightened to allow gas exchange. Medium was changed daily until the parasitaemia reached 20–30%. At this parasitaemia subcultures were generated at 1% parasitaemia by dilution with a 2% (v/v) normal red blood cell suspension in culture medium. Spent medium from B. canis cultures (taken at 24 and 48 h of a culture) was used as a source of soluble parasite antigen. The supernatant was concentrated 10 times using a dialysis membrane with a 8–10 kDa cut off.
2.2. In vivo phase – vaccination experiment 2.2.1. Animals Beagle dogs of either sex of approximately 6 months of age were used in the experiments. They were obtained from commercial breeders. The animals were clinically healthy, and had no history of babesiosis. The animals were identified by individual transponders that were implanted subcutaneously. Animals were acclimatised for at least one week. The animals received standard dog feed daily. Drinking water was supplied ad libitum. 2.2.2. Vaccination The vaccine contained 50 μg of recombinant CBA protein (see below 2.11) and 250 μg of saponin adjuvant (Supersap®, Desert King, Chile) per dose, in a volume of 1 ml Sörenson buffer (pH 6.0). Two groups of seven dogs each were formed. One group was vaccinated with a single dose of vaccine. These dogs received two booster vaccinations with the same material at three week intervals. A second group of seven dogs did not receive any injections and served as control group. Two weeks after the booster vaccination animals were challenged intravenously with 1 × 106 B. canis A-infected erythrocytes.
2.4. Analytical agarose gel precipitation using VC-serum Analytical agarose gel precipitation was used to determine whether antigens from SPA preparations could be recognized by VC-serum. Microscopic object slides were cleaned with alcohol (70% v/v) and precoated with 1% (v/v) agarose gel solution in Tris-buffered saline solution (10 mM, pH 8.0). Next, 5 ml of the same agarose gel solution was pipetted onto the slide. After cooling to room temperature 3 mm holes were punched into the gel. The centre hole was filled twice with concentrated supernatant of in vitro cultures of B. canis A, and two-fold serial dilutions of vaccination-challenge serum were pipetted in the surrounding holes. Slides were incubated overnight in a humidified chamber at 4C. Precipitation lines were visualised after washing the gels in Tris-buffered saline solution to remove non-precipitated material. The removal of free haemoglobin indicated that washing was complete.
2.2.3. Observations and clinical examinations After challenge infection animals were examined daily for clinical signs of babesiosis for a period of 14 days, when all animals were treated with imidocarbdipropionate (Carbesia®, MSD Animal Health) to cure the infection. Special attention was given to behaviour, spleen size, size of lymph nodes, colour of the mucous membranes of mouth and eye-lid, and the capillary refill time. Clinical observations were scored as a numeric value (Schetters et al., 1994). From these scores the total clinical score value was calculated for each day. Of these, the maximal clinical score value during the observation period was determined. 2.2.4. Haematocrit The haematocrit value is expressed as packed cell volume (PCV) of a sample of venous blood that was taken from the vena jugularis. A haematocrit capillary was filled with blood and centrifuged in a haematocrit centrifuge (Hettich) for 5′ at 10.000 rpm. The packed cell volume was read using a haematocrit reader. For each individual animal the maximal decrease of PCV was calculated, and expressed as a percentage of the value obtained at the day of challenge infection.
2.5. Preparative agarose gel precipitation using VC-serum In order to obtain enough precipitate for further molecular analysis, preparative agarose gels were used. In short, lids of standard polystyrene micro-Elisa titre plates were completely filled with 1% (w/v) agarose gel solution in Tris-buffered saline solution (10 mM, pH 8.0). Three parallel longitudinal slots were cut with a scalpel, at 3–4 mm 22
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
2.9. Immunoprecipitation using rabbit-anti-rCBA antiserum – western blot analysis
distance. Next the central slot was entirely filled with neat SPA preparation, and the two outer slots were filled with neat VC-serum. The gel was incubated overnight in a humidified chamber at 4C to allow the formation of immunoprecipitate. (clearance of haemoglobin from the gel slice was taken as an end point for washing).
For immunoprecipitation of non-radiolabelled antigen preparations, supernatants and B. canis-infected erythrocyte fractions were mixed with each test serum and incubated overnight at 4 °C with constant stirring. The immune complexes that had formed were captured onto protein A-sepharose-CL4 B beads for 1 h at room temperature with constant stirring. After incubation, the beads were centrifuged at 3000g for 15 min and supernatant was discarded. Immune complexes were recovered from the beads by acid elution according to the manufacturer’s instructions. The eluate was collected, filtered through a 0.22 μm pore size membrane and frozen at − 80 °C until use for SDSPAGE and western blotting.
2.6. SDS-PAGE and western blotting Protein characteristics and antigenic properties were analysed by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDSPAGE; Laemmli, 1970) and Western blot analysis (Towbin et al., 1979). Briefly, protein preparations were separated on 12% (w/v) reducing SDS-PAGE and transferred to nitrocellulose membranes. Blots were blocked with 5% (w/v) skimmed milk in Tris-buffered saline solutions (TBS; 20 mM Tris-HCl, pH 7.2, 150 mM NaCl).
2.10. Immunofluorescence with antisera against rCBA
2.7. Pulse-chase experiments
Babesia canis A was cultured in vitro to a parasitaemia of 7–10%. A culture sample (500 μl) was collected and centrifuged at 1200g for 5 min and washed five times in RPMI 1640. The erythrocytes were coated onto wells of Teflon-coated microscopic slides, air-dried and fixed for 10 min in ice-cold acetone/methanol solution (4:1). After three washes with PBS, slides were incubated for 30 min with twofold serial dilutions of test sera in PBS in a moist chamber, washed and then incubated 30 min with FITC-conjugated second antibody (Sigma). Slides were mounted with Citifluor™ (EMS Acquisition Corporation) and observed with immersion oil (1000X) using a fluorescence microscope (Axio Scope, Zeiss).
Babesia canis A was maintained in culture in vitro as described above (T-75 flasks). When parasitemia reached 10%, the culture supernatant was removed and the erythrocytes were washed in RPMI 1640 and then resuspended in methionine-free RPMI medium supplemented with 10% (v/v) NDS and 1 g/L GSH. After 4 h of incubation, L35 S-methionine (Amersham) was added (65 μCi/ml) for 15 min. The erythrocytes were then pelleted by centrifugation and washed twice in RPMI 1640 medium and then dispensed into a new flask (T-25) with complete culture medium and incubated for various chase-times (15 min, 30 min, 1 h or 4 h). At each chase-time, the erythrocytes were pelleted and the culture supernatant was collected. Each sample was centrifuged for 20 min at 10 000g at 4 °C, the supernatant was collected and filtered through a 0.22 μm pore size membrane and frozen at − 80 °C until use for SDS-PAGE and autoradiography.
2.11. Cloning of CBA gene and preparation of recombinant CBA protein The full length CBA gene, containing both the C- and N- terminal hydrophobic sequences, was amplified from Babesia canis genomic DNA using the following primer set: FW(5′accATGATGCTGCTCTTCGCCTTG3′) and REV (5′GCGAAAAA CATGAGTGGGACC3′) and cloned with a C-terminal HIS tag, behind the T7 promotor into a pET101 expression vector. After expression of the protein in E. coli BL21 (DE3), cells were lysed and antigen was dissolved in 6 M urea, 300 mM NaCl, 50Mm NaH2PO4 Ph 8.0. The protein was bound to a FF HisTrap™ column (GE Healthcare) and refolded on the column by exchanging the denaturing buffer for a non-denaturing buffer, containing a redox couple (0.4 g/L oxidized glutathione; 1.8 g/L reduced glutathione; 6 g/L CHAPS; 1 mM EDTA; 0.1 M NaCl; 50 mM Tris pH 8.0). After refolding, protein was eluted in a 0–500 mM imidazole gradient and stored at −70C until use. 2.12. Detection of antibodies against recombinant CBA by ELISA For the detection of antibodies against recombinant CBA by ELISA, a hydrophilic form of the protein was produced. The coding sequence from E19 to A301 was cloned in pIVEX 2.3MCS vector (Roche) by assembly PCR using 5′-CGCTTAATTAAACATATGACCGAAAACAC TATACTTTTATCCAATGTAGAATTCC as forward primer and 5′TTAGTTAGTTACCGGATCCCTTAAGCGGTAGTAAGTTTAGGAGAAGC as reverse primer. Amplification and assembly was done using Phusion DNA polymerase (New England Biolabs), and resulting plasmids were transformed in E. coli DH5alpha. A recombinant plasmid was selected and verified by sequencing, and then transformed in E. coli BL21(DE3) for protein expression. The recombinant protein, without any hydrophobic sequence but with a HIS-tag appended at the C-terminal position, was then produced in a 3-l bioreactor (Infors) at 30 °C. After harvesting, cells were lysed in HIS-tag lysis buffer (NaH2PO4 50 mM, NaCl 500 mM, imidazole 10 mM, pH = 8.0) using a high pressure homogenizer (Avestin) operating at 1200 bar. Non-soluble material was removed by centrifugation and Triton X-100 (1% [v/v]) was added to the clarified lysate, which was subsequently loaded on a 5 ml HisTrap™ column. The column was washed with NaH2PO4 50 mM, NaCl 1 M, imidazole 10 mM, pH = 8.0 to remove non-bound material. The
2.8. Immunoprecipitation using VC-serum − autoradiography For conventional immunoprecipitation the protocol of Hines et al. (1989) was used. Parasites were metabolically labelled with L- 35Smethionine to allow detection of parasite-encoded proteins. When parasitaemia reached 10%, overnight incubation was performed in methionine-deficient culture medium (Gibco) supplemented with 50 μCi of L- 35S-methionine (Amersham) per ml. After incubation, the labelled cells and the corresponding supernatants were collected and processed for immunoprecipitations assays. For the analysis of antigens from culture supernatants, culture supernatants were centrifuged for 15 min at 15 000g, filtered through a membrane of 0.22 μm and depleted of serum immunoglobulins by incubation with protein A-sepharose-CL4 B for 1 h at room temperature with constant stirring. The beads were centrifuged at 3000g for 15 min and supernatant was collected. For the analysis of cell-associated antigens, radiolabelled infected red blood cells were washed twice by centrifugation at 700g for 5 min in culture medium without serum and solubilized with 10 vols of lysis buffer 50 mM Tris, 5 mM EDTA, 100 mM NaCl, 1% (v/v) nonidetP40, 1 mM phenylmethylsulfonylfluoride (PMSF), 0.1 mM N-tosyl-Llysine-chloromethylketone (TLCK). Insoluble material was removed by centrifugation at 10000g at 4 °C. For immunoprecipitation, the radiolabelled lysates (106 cpm) and the different radiolabelled supernatants (200 μl) were mixed with each test serum and incubated overnight at 4 °C with constant stirring. The immune complexes that had formed were captured onto protein A-sepharose-CL4 B beads for 1 h at room temperature with constant stirring. After incubation, the beads were centrifuged at 3000g for 15 min and supernatant was discarded. Immune complexes were recovered from the beads by acid elution according to the manufacturer’s instructions. The eluate was collected, filtered through a 0.22 μm pore size membrane and frozen at − 80 °C until use for SDS-PAGE and autoradiography. 23
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
vaccinated dogs that were challenged with B. canis A. As control, serum from adjuvant control dogs was used (serum was derived from group C and A respectively, described in Schetters et al., 1992). At all strips the canine immunoglobulin heavy chain was detected at the 50 kDa position. Results showed that serum from vaccinated dogs taken at day 1–12 after infection (except for the serum sample of day 2 p.i.) reacted with a 100 kDa antigen. Serum from adjuvant control dogs did not recognize this antigen throughout the observation period (Fig. 2). From day 5 after infection onwards the serum samples of the vaccinated dogs recognized a second antigen of Mr 39 kDa specifically. Serum from adjuvant control dogs did not recognize this antigen. Serum from dogs of both groups recognized a 70 kDa antigen from day 7–8 after infection onwards.
Fig. 1. Analytical agarose gel precipitation. Immunoprecipitates formed by interaction of SPA and different dilutions of VC-serum. Note two distinct precipitation lines. Figure is photographically inverted for clarity.
3.2. Pulse-chase radioactively labelling of B. canis parasites In order to exclude the possibility that these antigens were of erythrocyte rather than parasite origin, metabolic labelling studies were performed. Since canine erythrocytes do not produce protein de novo, the only proteins that can become labelled in in vitro cultures of B. canis are parasite derived. Babesia canis A was cultivated in vitro and pulsed for 15 min with radioactively labelled methionine. Results showed that from 15 min after the pulse onwards a doublet of 40 kDa became radioactively labelled, and appeared the most prominent (Fig. 3). It appeared that, whereas at 120 min after the pulse the upper band of the doublet was more intense than the lower band, another two hours later the lower band of the doublet was more intense. Additional proteins of Mr 80, 48, 34, and 25 kDa were found but these were less pronounced (Fig. 3). A protein of 70 kDa could not be detected. Together with the western blot results, the data suggested that the major B. canis antigen in SPA preparations that was recognized by VC-serum was a 39 kDa antigen.
recombinant protein was then eluted using Tris 50 mM, imidazole 400 mM pH = 8.0. The protein was stored at −80 °C until use in a standard ELISA test for the detection of antibodies against the recombinant CBA protein in the serum of vaccinated dogs. 3. Results 3.1. Analysis of immune complexes from preparative gel diffusion Vaccinated dogs that were protected against B. canis infection produced antibodies against SPA from day 5 after infection onwards as detected by sandwich-ELISA (Schetters et al., 1996a). The antigen in a sandwich-Elisa is captured in its native form i.e. not denatured and stretched as in western blotting techniques. In addition, for a sandwichElisa to become positive, the relevant antigens must complex with at least two antibodies; the capture antibody and the reporter (=second) antibody. Hence, in order to determine which antigens were recognized in the sandwich-Elisa, immunoprecipitation assays that required complex formation between different antibodies and antigens were used. It was decided to use agarose gel immunoprecipitation. First, analytical agarose gel immunodiffusion was used to determine whether SPA preparations contained antigens that could be complexed with VCserum. Results showed that there were two distinct precipitation lines (Fig. 1). In order to further characterize the precipitated antigens, preparative gel diffusion was performed using B. canis A-SPA preparations and VC-serum in thick agarose gels. A slice of gel containing the precipitate was cut from the gel and washed in PBS to remove free proteins. Next the gel slice was treated for SDS-PAGE and western blotting. Blot strips were incubated with daily serum samples from a group of
3.3. Immunoprecipitation of radioactively labelled antigens from B. canis a cultures To verify that indeed a 39–40 kDa antigen from radioactively labelled B. canis parasites could be recognized by VC-serum, radioactively labelled antigens from in vitro cultures of B. canis A were prepared. Three different fractions were analysed; antigens from supernatants (SUP), and two fractions from B. canis A-infected erythrocytes, the purified merozoite fraction (MERO) and the cytosolic fraction of the lysed infected erythrocytes (CYT). Each of these fractions was used for immunoprecipitation with VC-serum. The precipitated antigens were separated on SDS-PAGE and the gel was subsequently developed by autoradiography. Results showed that amongst other proteins, a doublet of Mr of 39–40 kDa was precipitated from each of the antigen preparations Fig. 2. Western blot analysis of reactivity of sera from SPA-vaccinated dogs that were challenged with B. canis A parasites (V). As control sera from adjuvant control dogs were used (N). Antigen on the blots was derived by preparative agarose gel immunoprecipitation. Numbers below the strips indicate the day post infection at which serum was collected. C = no serum incubation. Image is manipulated to create pairs of blot strips per day (slight colour overlay).
24
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
3.4. Discovery of the gene encoding the 39 kDa B. canis protein So far, results indicated that the major B. canis antigen from SPA preparations that is specifically recognized in the sandwich-Elisa is the 39 kDa antigen of a 39–40 kDa doublet. The antigen was named canine Babesia antigen (CBA). In order to obtain a sufficient amount of CBA for amino acid sequence analysis, preparative agarose gel precipitation was performed using B. canis SPA preparations and VC-serum in thick agarose gels. The immunoprecipitate was separated on SDS-PAGE and the entire gel blotted onto nitrocellulose paper. Two strips taken from the left and right side of the blot were developed with VC-serum to locate the immunoreactive band at the 39 kDa position. Next, a strip of approximately 2 mm at the 39–40 kDa position was cut from the entire nitrocellulose blot. The molecules recovered from this strip were analysed by trypsin digestion and subsequent mass spectrometry (Proteomics Centre, Nijmegen University, the Netherlands). Aside from trypsin and haemoglobin fragments, six unique peptides, (not present in public databases) were discovered (depicted in bold Fig. 5). Next, genomic DNA of B. canis A was sequenced (BaseClear, Leiden, the Netherlands) and blasted with the nucleotide sequences deduced from the partial amino acid sequences that had been discovered. Results showed that all six peptides clustered in a single gene that encodes for a 33 kDa protein (Fig. 5). The gene contains a putative signal sequence (1–17). At the C-terminal side the gene encodes for a transmembrane sequence (301–321) that is required for GPI-anchoring. It is predicted that threonine (pos 300) is the amino acid to which the GPI-anchor is attached. The core amino acid sequence contains two cysteines (enlarged), one at pos 78 and one at pos 264.
Fig. 3. Autoradiograph of supernatants of in vitro cultures of B. canis A that were pulsed with L- 35S-methionine for 15 min and separated on SDS-Page. Supernatants were taken at different time points after the pulse (indicated in minutes at the top of each lane).
3.5. Recombinant CBA production and characterization The full length CBA gene, containing both the C- and N-terminal hydrophobic sequences, was amplified from B. canis A genomic DNA and cloned in E. coli BL21. Thirty grams of wet cell paste were lysed through a French press cell. The lysate was centrifuged and the soluble fraction was pipetted off. There was a small amount of recombinant protein detectable in the soluble fraction (Fig. 6). The pelleted material containing the inclusion bodies of the bacteria was dissolved in urea buffer. This fraction contained the recombinant CBA (rCBA) antigen. In addition, a doublet around 26 kDa was found. Finally, the expressed recombinant protein was refolded on a His-Trap column. Upon analysis on SDS-PAGE, the protein presented as a doublet of 37–40 kDa. Minor protein bands were found at 29 and 26 kDa position. Thirty grams of wet cell paste yielded 5 mg of recombinant protein with a purity of approximately 85% as estimated from a Coomassie Brilliant Bluestained gel (Fig. 6). To determine whether the purified recombinant protein was recognised by VC-serum, the protein was dot-blotted onto nitrocellulose paper and incubated with VC-serum. As a control, SPA of culture supernatant of B. canis A cultures was used. Results showed that VC-serum indeed recognized the recombinant protein. Normal dog serum did not recognize the recombinant protein (Fig. 7). The SPA of culture supernatant was positive with both sera due to the presence of normal dog serum in the culture supernatants. The recombinant protein could not be detected by sandwich-Elisa (data not shown), indicating that the protein did not express the two antigenic epitopes that are required to
Fig. 4. Autoradiograph of radioactively labelled antigens from in vitro cultures of B. canis A that were immunoprecipitated with VC-serum or normal dog serum and separated on SDS-PAGE. MERO = merozoite antigens, CYT = antigens from the cytosol of B. canis Ainfected erythrocytes, SUP = antigens from supernatants of in vitro cultures of B. canis A.
(Fig. 4). The 39 kDa protein from the doublet appeared to be more pronounced in the cytosol of infected erythrocytes, whereas in the culture supernatant the 40 kDa protein of the doublet was more pronounced. Serum from control dogs did not precipitate these molecules.
Fig. 5. Deduced amino acid sequence of the CBA gene of B. canis A. Six peptides that were detected in the native 39 kDa protein of B. canis A are presented between brackets in bold. Predicted transmembrane sequences are underlined.
25
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
was detected in supernatant (IP-SUP; very weak), lysate of infected erythrocytes (IP-LYS) and merozoites (IP-MERO). Such reactivity was not found in immunoprecipitates obtained with normal rabbit serum. Additional proteins were detected at the 35 kDa and 25 kDa (very weak) position in immunoprecipitates of lysate of infected erythrocytes (IP-LYS) and merozoites (IP-MERO). 3.6. Evaluation of the vaccine potential of rCBA in dogs Having established that the recombinant CBA antigen of B. canis A was specifically recognized by serum of immune dogs, and induced antibodies against the 40 kDa doublet of B. canis A parasites, the protective activity of the recombinant protein was evaluated in a vaccination-challenge experiment in dogs. Dogs were vaccinated at day 0 and received two booster vaccinations at days 21 and 42. Upon vaccination, dogs produced antibodies against rCBA as measured by Elisa. This response was boosted after the second vaccination at day 21. The additional booster vaccination at day 42 did not increase antibody titres further (Fig. 10). The serum recognized B. canis A merozoites in immunofluorescence (Fig. 11). Animals were challenged 14 days after the second booster vaccination (day 56). In non-vaccinated control dogs the infection became patent two days after challenge. In vaccinated dogs the first parasites were detected two days later, at day 4 post infection (p.i.; Fig. 12). Parasitaemia gradually increased in both groups of dogs, and peaked around day 8–9 p.i. From that day of infection onwards, the parasitaemia in vaccinated dogs progressively decreased, whereas in nonvaccinated dogs parasitaemia remained relatively high until day 12 p.i. and declined thereafter. The cumulative daily parasitaemia (=parasitic load) was statistically significant lower in the vaccinated group as compared to that of the non-vaccinated control group (P < 0.04, Student’s t-test, two-sided, unequal variances; Table 1). In non-vaccinated control dogs, the packed cell volume started to decrease immediately the day after challenge infection. In contrast, in vaccinated dogs this decrease started one day later (Fig. 13). In general, the decrease in PCV was on average 10% ( ± 3.0) less in the vaccinated dogs as compared to controls over the entire period. This was also reflected in the average maximal PCV decrease of the dogs in each group (Table 1). This difference of 10.5% was statistically significant (p = 0.05, Student’s t-test, two sided). During the post infection period the clinical signs of dogs of both groups were relatively mild until day 11 p.i. when six out of seven nonvaccinated dogs showed pale mucosae and had increased capillary refill times. This was reflected in the clinical score value that remained high until the end of the observation period when all dogs were treated (Fig. 14). The clinical score value of the vaccinated dogs was statistically significant lower than that of non-vaccinated dogs at day 3 and day 5 p.i., and from day 9 p.i. onwards when all vaccinated dogs recovered from the infection. The average maximal clinical score value of the vaccinated group was statistically significant lower than that of the control group (p < 0.01, Student’s t-test, two sided, equal variances; Table 1).
Fig. 6. Coomassie stained SDS-PAGE gel on which samples at different stages of recombinant CBA preparation were separated. M = markers; Soluble fraction = supernatant of lysed E.coli that expressed the CBA gene; Purified CBA = inclusion bodies dissolved in urea buffer; Refolded CBA = protein eluted from His-Trap column after refolding.
Fig. 7. Reactivity of VC-serum with recombinant CBA protein (rCBA). Recombinant protein was spotted onto nitrocellulose membranes (top rows)). As control SPA from supernatants from in vitro.
render a positive signal. Antiserum raised against rCBA was raised in rabbits and used to control whether the recombinant antigen induced antibodies against B. canis. Three different Babesia strains were used to assess specificity of the antiserum. Immunofluorescence showed that the antiserum recognized merozoites of B. canis A and B. canis B on blood smears of in vitro cultures of these strains. No fluorescence was found when blood smears of B. rossi cultures were used (Fig. 8). To determine the molecular specificity of the rabbit anti-rCBA serum, immunoprecipitation studies were performed. Different fractions of B. canis A in vitro cultures were prepared for immunoprecipitation: Supernatant (SUP), lysate of B. canis A-infected erythrocytes (LYS), merozoites isolated from the lysate (MERO), and the cytosol remaining after removal of merozoites (CYT). As a control, a lysate of uninfected red blood cells was used (RBC). Immunoprecipitation with rabbit anti-rCBA antiserum was performed, followed by western blotting and incubation with goat-anti-rabbit-immunoglobulin conjugate to detect the precipitated antigen. As a control antigen, B. canis A merozoites were used (no immunoprecipitation). Control immunoprecipitations were done with serum from the rabbit before immunization with rCBA (normal rabbit serum). Results are presented in Fig. 9. Immunoprecipitates showed a very high signal at the 50 kDa position where the rabbit immunoglobulin heavy chain migrated. No additional proteins were precipitated from lysates of uninfected red blood cells (IP-RBC). Reactivity against a 40 kDa protein
4. Discussion In a variety of Babesia-host models it has been shown that animals can be protected against clinical manifestations after experimental infection by vaccination using soluble parasite antigens from in vitro Babesia cultures (SPA; Schetters and Montenegro-James 1995). Based on this principle, commercial vaccines for the prevention of canine babesiosis have been developed and marketed in Europe (Schetters, 2005). Until to date the identity of the immunoprotective antigens has not been revealed. This is mainly due to the presence of large amounts of serum (10–40%) in the culture supernatants, which hampers the use of standard purification techniques. Using a combination of techniques we discovered a 39 kDa B. canis antigen that is present in culture 26
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
Fig. 8. Immunofluorescence on blood smears of in vitro cultures of different Babesia strains. Cells were incubated with rabbit serum raised against rCBA of B. canis A. Nuclei were stained with DAPI.
Fig. 9. Western blot analysis of immunoprecipitates (IP-xxx) obtained with rabbit antiserum raised against rCBA (left panel) and normal rabbit serum (right panel). Merozoite antigen of B. canis A was used as control (MERO). SUP = antigens from supernatants of in vitro cultures of B. canis A, LYS = lysed B. canis A-infected red blood cells; CYT = antigens from the cytosol of B. canis A-infected erythrocytes, RBC = lysate of uninfected red blood cells. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
1 gene described recently (Zhou et al., 2016). These authors have shown that monospecific mouse antiserum against the recombinant BcMSA1 protein stained B. canis merozoites, and reacted with a 39–40 kDa doublet in lysates of B. canis-infected erythrocytes (Zhou et al., 2016). This further corroborates the hypothesis that the 39 kDa and 40 kDa proteins are encoded for by a single gene. The amino acid sequence has a predicted GPI anchoring site, which lends support to the hypothesis that it is a merozoite membrane protein. It is hypothesized that the 39 kDa protein could be the GPI-anchored 40 kDa merozoite membrane protein from which the GPI-anchor is cut off. There are many similarities between CBA of B. canis and Bd37 of B. divergens. In the B. divergens-gerbil model we have found earlier that supernatant of in vitro cultures of B. divergens contains a 37 kDa merozoite surface antigen (Bd37; Carcy et al., 1995, Delbecq et al., 2002). The protein is GPI-anchored to the merozoite surface, and is released into the environment after the GPI-anchor is cut. The recombinant Bd37
supernatants, and antibodies against which are associated with protective immunity. The 39 kDa antigen appears to be a doublet that can be radioactively labelled with L- 35S-methionine (this work). A doublet of similar size that could be labelled with 14C-glucosamine has been described in the supernatants of B. canis before (Azzar et al., 1990). It is likely that this is the same doublet as described here. In pulse-chase experiments the 39 kDa protein signal increases between 120 and 240 min after the pulse, simultaneously with a relative decrease in the intensity of the 40 kDa protein. This suggests that the 39 kDa molecule derives from the 40 kDa protein. With the resolution of part of the amino acid sequence of the 39 kDa antigen, the gene encoding for the protein (canine Babesia antigen; CBA) could be discovered in the B. canis genome. The CBA-gene was expressed and the resulting rCBA protein was recognized by VC-serum, and monospecific rabbit antiserum against rCBA stained B. canis merozoites and reacted with the native 40 kDa antigen of B. canis. The CBA gene is similar to the BcMSA27
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
Fig. 12. Dynamics of the parasitaemia in dogs after experimental infection with B. canis A infected erythrocytes. One group of dogs was vaccinated with rCBA of B. canis A (blue circles), non-vaccinated dogs served as control (red diamonds). Data represent the group average of log-transformed number of infected erythrocytes per 105 red blood cells (n = 7). Error bars indicate the standard deviation.
Fig. 10. Dynamics of the antibody response of dogs upon immunization with rCBA of B. canis A (blue curve). Control dogs received no vaccinations (red curve). Arrows indicate the time point of immunization. Data represent the group average antibody titre as measured in Elisa using rCBA without the hydrophobic terminal sequences (n = 7). Error bars indicate standard deviation.
Table 1 Essential disease parameters of Babesia infection. Effect of vaccination on the essential parameters of Babesia infection in dogs after experimental infection with B. canis A infected erythrocytes. One group of dogs was vaccinated with rCBA of B. canis A, nonvaccinated dogs served as control. Data represent the group average and standard deviation. Asterisk indicates that these values are statistically different from those of controls.
protein induces protective immunity against virulent challenge infection (Delbecq et al., 2006). Because of the similarities between Bd37 of B. divergens and CBA of B. canis (including the presence of two cysteine residues in the core protein that could be involved in 3D structuration of the protein), it is tempting to speculate that they are homologues proteins. However, there are no significant sequence homologies between the two molecules or any other known sequences in public databases (e.g. MSA1 of B. bovis; reviewed in Carcy et al., 2006). Because of this it is felt premature to designate the gene encoding the 39–40 kDa protein BcMSA1 as suggested by Zhou and collaborators (Zhou et al., 2016). Upon analysis of a series of recombinantly expressed Bd37 proteins, it appeared that presence of at least one hydrophobic terminal sequence to the hydrophilic core sequence was indispensable for the induction of immunity against heterologous B. divergens strains (Delbecq et al., 2006). Because of this, we cloned and expressed the entire CBA gene (including the two hydrophobic terminal sequences) for vaccination purposes. The rCBA protein induced protection against B. canis infection in dogs. Six out of 7 control dogs had to be chemotherapeutically treated to prevent succumbing to infection, whereas all vaccinated dogs controlled the challenge infection before the end of the observation period. The nature of the protection was different from that obtained with SPA from in vitro culture supernatants in that a striking antiparasite effect was found (Schetters, 2005). A major difference between vaccination with rCBA and SPA is that immunization with SPA does not induce antibodies against the 39–40 kDa antigen, but apparently primes dogs to develop a rapid anamnestic response against the 39 kDa protein from day 6 after challenge onwards (this work and Schetters et al., 1996a,b). In contrast, dogs that were vaccinated with rCBA had antibodies already at the time of challenge infection, which could actually
Parasite Load Max PCV decrease Max Clinical Score
Vaccinated
Control
3,5 ± 1,9* 46,6 ± 7,3* 1,3 ± 0,5*
7,2 ± 3,5 57,1 ± 10,5 2,7 ± 0.7
neutralize part of the parasites of the challenge inoculum. This would be reflected in a prolonged prepatent period and delayed onset of haematocrit decreases, as if a lower challenge dose was used (Schetters et al., 2009). This is corroborated by the fact that the rate at which the PCV values decrease in both groups is similar when the curve of the vaccinated group is shifted two days. The anti-parasite effect is not restricted to a direct effect on the challenge inoculum, however, as evidenced by the fact that the parasitaemia during the post infection period is much lower in the group of vaccinated dogs than in the controls. In addition, assuming a delayed onset of disease due to an immediate effect on the challenge inoculum, it can be deduced that vaccinated dogs control the infection 7 days after the PCV started to decrease. From that day onwards the parasitaemia in the vaccinated dogs progressively decreased. In the control group, the infection developed more chronically, which led to severe clinical symptoms 11 days after the PCV started to decrease. The most important clinical symptoms were pale mucosae and a prolonged capillary refill time. It has been suggested that pale mucosae
Fig. 11. Immunofluorescence of B. canis A-infected erythrocyte. Cells were incubated with serum of dogs that were vaccinated rCBA of B. canis A. Nuclei were stained with DAPI.
28
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
employees of Merck Animal Health. Research done by KM, SD, BC, EP and AG was in part funded by Merck AH. Acknowledgements This publication is dedicated to Eric Précigout who passed away during the course of the project. The authors thank all participating laboratory staff and students for their contribution and assistance in these studies during the course of this project since 1994. Special thanks to Silvie Randazzo for continuous support. References Azzar, G., Ravisson, J., Got, R., 1990. Characterization and purification of culture-derived soluble glycoproteins of Babesia canis. Parasitol. Res. 76, 578–580. Baneth, G., Florin-Christensen, M., Cardoso, L., Schnittger, L., 2015. Reclassification of Theileria annae as Babesia vulpes sp. nov. Parasites Vectors 8, 207. Carcy, B., Precigout, E., Valentin, A., Gorenflot, A., Schrevel, J., 1995. A 37-kilodalton glycoprotein of Babesia divergens is a major component of a protective fraction containing low molecular-mass culture-derived exoantigens. Infect. Immun. 63, 811–817. Carcy, B., Precigout, E., Schetters Th Gorenflot, A., 2006. Genetic basis for GPI-anchor merozoite surface antigen polymorphism from Babesia and resulting antigenic diversity. Vet. Parasitol. 138, 33–49. Delbecq, S., Précigout, E., Vallet, A., Schetters, Th., Gorenflot, A., 2002. Babesia divergens: cloning and biochemical characterisation of Bd37. Parasitology 125 305–212. Delbecq, S., Hadj-Kaddour, K., Randazzo, S., Kleuskens, J., Schetters, T., Gorenflot, A., Precigout, E., 2006. Hydrophobic moieties in recombinant proteins are crucial to generate efficient saponin-based vaccine against Apicomplexan Babesia divergens. Vaccine 24, 613–621. Hines, S.A., McElwain, T.F., Buening, G.M., Palmer, G.H., 1989. Molecular characterization of Babesia bovis merozoite surface proteins bearing epitopes immunodominant in protected cattle. Mol. Biochem. Parasitol. 37, 1–9. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature 227, 680–685. Matijatko, V., Torti, M., Schetters Th, P., 2012. Canine babesiosis in Europe: how many diseases? Trends Parasitol. 28, 99–105. Schetters, Th., Montenegro-James, S., 1995. Vaccines against babesiosis using soluble parasite antigens. Parasitol. Today 11, 456–462. Schetters, Th.P.M., Kleuskens, J.A.G.M., Scholtes, N.C., Bos, H.J., 1992. Vaccination of dogs against Babesia canis infection using parasite antigens derived from in vitro culture. Parasite Immunol. 14, 295–305. Schetters, Th.P.M., Kleuskens, J.A.G.M., Scholtes, N.C., Pasman, J.W., Bos, H.J., 1994. Vaccination of dogs against Babesia canis infection using antigen from culture supernatant with an emphasis on clinical babesiosis. Vet. Parasitol. 52, 219–233. Schetters, Th., Kleuskens, J., Scholtes, N., Bos, H.J., 1995. Strain variation limits protective activity of vaccines based on soluble Babesia canis antigens. Parasite Immunol. 17, 215–218. Schetters, Th.P.M., Scholtes, N.C., Kleuskens, J.A.G.M., Bos, H.J., 1996a. Not peripheral parasitaemia but the level of soluble parasite antigen in plasma correlates with vaccine efficacy against Babesia canis. Parasite Immunol. 18, 1–6. Schetters, Th., Kleuskens, J., Scholtes, N., Pasman, J., Bos, H.J., 1996b. Vaccination of dogs with soluble Babesia canis antigens derived from in vitro culture is strain specific. New Dimensions in Parasitology: Keynote Papers from the VIII International Congress of Parasitology; Acta Parasitologica Turcica. pp. 543–550. Schetters, Th.P.M., Kleuskens, J.A.G.M., Scholtes, N.C., Goovaerts, D., Pasman, J., Bos, H.J., 1997. Vaccination of dogs against Babesia canis. Vet. Parasitol. 73, 35–41. Schetters, Th.P.M., Kleuskens, J., Van De Crommert, J., De Leeuw, P., Finizio, A.-L., Gorenflot, A., 2009. Systemic inflammatory responses in dogs experimentally infected with Babesia canis: a haematological study. Vet. Parasitol. 162, 7–15. Schetters, Th.P.M., 2005. Vaccination against canine babesiosis. Trends Parasitol. 21, 179–184. Towbin, H., Staehelin, T., Gordon, J., 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. U. S. A. 76, 4350–4354. Uilenberg, G.G., Franssen, F.F.J., Perié, N.M., Spanjer, A.A.M., 1989. Three groups of Babesia canis distinguished and a proposal for nomenclature. Vet. Quarterly 11, 33–40. Uilenberg, G., 2006. Babesia – a historical overview. Vet. Parasitol. 138, 3–10. Zhou, M., Cao, S., Luo, Y., Liu, M., Wang, G., Moumouni, P.F.A., Jirapattharasate, C., Iguchi, A., Vudriko, P., Terkawi, M.A., Löwenstein, M., Kern, A., Nishikawa, Y., Suzuki, H., Igarashi, I., Xuan, X., 2016. Molecular identification and antigenic characterization of a merozoite surface antigen and a secreted antigen of Babesia canis (BcMSA1 and BcSA1). Parasit. Vectors 9, 257.
Fig. 13. Dynamics of the haematocrit in dogs after experimental infection with B. canis A infected erythrocytes. One group of dogs was vaccinated with rCBA of B. canis A (diamonds), non-vaccinated dogs served as control (circles). Data represent the group average of packed cell volume (PCV) expressed as percentage relative to the value at the day of challenge infection (n = 7). Error bars indicate the standard deviation.
Fig. 14. Dynamics of the clinical score of dogs after experimental infection with B. canis A infected erythrocytes. One group of dogs was vaccinated with rCBA of B. canis A (diamonds), non-vaccinated dogs served as control (crosses). Data represent the group average of the clinical score (n = 7). Error bars indicate the standard deviation.
as seen in babesiosis reflects anaemia. The fact that at day 14 p.i. the average PCV values of the non-vaccinated dogs (of which 6 out of 7 had pale mucosae), was not significantly different from the vaccinated dogs that had normal mucosae (29.3 ± 6.1 and 28.7 ± 3.8 respectively), does not support that suggestion. It is more likely that, due to (compensated) hypotension, which has been shown to occur in this experimental model (Schetters et al., 2009), tissue perfusion was impaired. Vaccination of dogs with rCBA limited the development of this pathological condition because of an effect on parasite proliferation. In conclusion, an immunoprotective B. canis protein was isolated from the supernatants of in vitro cultures of B. canis A parasites. The antigen that appears as a doublet on western blots, is encoded for by a single gene, and named Canine Babesiosis Antigen (CBA). A vaccine based on the recombinantly produced full-length CBA protein (including the hydrophobic terminal sequences) induced protective immunity against homologous challenge infection in dogs. Whether the antigen also protects against heterologous challenge infection remains to be determined. Declaration of interest Authors NS and TvK are current, and TS, JK and JvdC are former
29
Contents lists available at ScienceDirect
Veterinary Parasitology journal homepage: www.elsevier.com/locate/vetpar
Research paper
Discovery of a recombinant Babesia canis supernatant antigen that protects dogs against virulent challenge infection
MARK
K. Moubria, J. Kleuskensb, J. Van de Crommertb, N. Scholtesb, T. Van Kasterenb, S. Delbecqa, ⁎ B. Carcya, E. Précigouta,1, A. Gorenflota, Th. Schettersa,b,c, a
EA4558 VAP, University of Montpellier, Montpellier, France MSD-AH, Boxmeer, The Netherlands c Department Veterinary Tropical Diseases, University of Pretoria, Onderstepoort, South Africa b
A R T I C L E I N F O
A B S T R A C T
Keywords: Babesia canis recombinant vaccine antigen protective protozoa immunity
Soluble parasite antigens (SPA) in supernatants of in vitro cultures of Babesia canis can be used to vaccinate dogs against virulent B. canis infection. The moment that immunity becomes apparent coincides with the appearance of antibodies against SPA in the serum of the vaccinated animals. This so-called vaccination-challenge serum (VC-serum) was used to precipitate antigens from B. canis culture supernatants in agarose gels. This antigen preparation was then used to analyse the reactivity of sera from vaccinated dogs on western blots. Results: showed that the first appearance of antibody reactivity against a protein that migrated at the 39 kDa position in SDS-PAGE gels was associated with the moment vaccinated dogs started to recover from a virulent challenge infection. In addition, pulse-chase experiments revealed that a 39–40 kDa doublet was released into the supernatant of B. canis cultures starting 15 min after the chase. This doublet was specifically precipitated by VC-serum, thus corroborating that the 39–40 kDa doublet in SPA preparations was of parasite origin. Partial amino acid sequencing allowed the discovery of the gene that encoded the 39–40 kDa doublet (canine Babesia antigen; CBA). The full-length gene was cloned and expressed in E. coli. The recombinant CBA protein (rCBA) was recognized by VC-serum, and antibodies against rCBA precipitated the 39 kDa antigen of SPA preparations and of merozoites of B. canis. In addition, anti-rCBA serum reacted with the surface of B. canis merozoites (but not with B. rossi merozoites) in immunofluorescence. Vaccination of dogs with rCBA induced antibodies against rCBA, which recognized B. canis merozoites. Vaccinated dogs were protected against virulent challenge infection by limiting parasite proliferation. As a result, the development of clinical signs was prevented and the animals self-cured. In contrast, six out of seven non-vaccinated control dogs developed relatively high parasitaemia and serious clinical signs associated with poor tissue perfusion. This antigen can be used to replace the SPA antigen in the conventional B. canis vaccines, which eliminates the need for dog blood and serum for vaccine production.
1. Introduction Babesiosis is a tick-transmitted disease of vertebrates. In dogs the disease is caused by four different large Babesia species (B. canis, B. rossi, B. vogeli and B. presenti) and three different small Babesia species (B. gibsoni, B. conradae and B. vulpes; Uilenberg, 2006; Baneth et al., 2015). In Europe, B. canis is the most prevalent Babesia species of dogs. This species is transmitted by Dermacentor reticulatus ticks. The clinical manifestations vary from anorexia, lethargy and anaemia, to a syndrome in which multiple organs are affected (Matijatko, Torti and Schetters 2012). The infection can be cured chemotherapeutically with imidocarbdipropionate, and to prevent disease two vaccines have been
⁎
1
developed commercially (Schetters, 2005). These vaccines contain antigens from supernatants of in vitro cultures of the parasite (soluble parasite antigens; SPA), adjuvanted with a saponin. The protective effect of these vaccines is on the development of clinical manifestations, not necessarily on the development of the parasite per se (Schetters et al., 1996a). In the same studies it was further shown that vaccinated animals produced antibodies against SPA upon challenge infection as detected by sandwich-ELISA. Such reactivity was not found in serum of control dogs that survived a primary infection with B. canis. Hence, serum from dogs that were vaccinated and subsequently challenged (VC-serum) could be crucial in the discovery of the immunoprotective parasite antigen in SPA preparations. Here we describe the discovery
Corresponding Author at: ProtActivity R&D, Sering 36, 5432 DD CUIJK, The Netherlands. E-mail address: [email protected] (T. Schetters). Deceased
https://doi.org/10.1016/j.vetpar.2017.11.002 Received 14 July 2017; Received in revised form 3 November 2017; Accepted 5 November 2017 0304-4017/ © 2017 Elsevier B.V. All rights reserved.
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
2.2.5. Parasitaemia Percentage of infected red blood cells was determined from blood smears stained with May-Grünwald/Giemsa solutions. The parasitaemia is expressed as the log10 number of red blood cells infected with Babesia parasites per 105 erythrocytes in the peripheral blood. The parasitic load was calculated by adding daily parasitaemia values (cumulative parasitaemia).
and characterization of this antigen in B. canis culture supernatants. In addition, the gene encoding for this antigen was isolated, and the protective effect of a recombinantly expressed protein was evaluated in dogs. 2. Materials and methods 2.1. Parasite
2.2.6. Serum Blood was collected from the vena jugularis and allowed to clot at room temperature for at least 60 min. Cellular contaminants were removed by centrifugation (1500g, 10 min, at room temperature). Serum was stored at −70 °C until use.
The Babesia canis A isolate isolated in 1982 from Dermacentor reticulatus ticks recovered from a dog that had just returned from the Drôme region in France and suffered from babesiosis; (Uilenberg et al., 1989) was used in the experiments. A second B. canis isolate that also derived from France, and which was shown to be immunologically different from B. canis A (named B. canis B; Schetters et al., 1995), and a B. rossi isolate that was originally isolated from a dog from Pietermaritzburg (South Africa; Uilenberg et al., 1989) were additionally used in immunofluorescence experiments. Parasites were stored as stabilates in liquid nitrogen, and passed through splenectomised Beagle dogs to obtain freshly isolated infected blood. Infected blood was taken by venepuncture from the vena jugularis, using heparin (5–10 U/ml, Leo) to prevent clotting. Blood was washed with Babesia medium (Schetters et al., 1992). The challenge inoculum was injected intravenously in experimental animals.
2.2.7. Vaccination-challenge serum (VC-serum) Vaccination-challenge serum was obtained from Beagle dogs that had been vaccinated 2–4 times with SPA of B. canis A with 3-week intervals, and subsequently challenged with homologous B. canis A parasites, as described earlier (Schetters et al., 1996b). Serum was stored at −70 °C until use. 2.3. Soluble parasite antigen from in vitro cultures Babesia canis A was maintained in continuous in vitro culture, following the procedure described previously (Schetters et al., 1997). Briefly, blood was drawn from B. canis A-infected dogs, and washed thrice with culture medium (RPMI 1640, Gibco, containing 25 mM Hepes and 25 mM NaHCO3, pH 7.3, supplemented with 10% (v/v) normal dog serum and 1 g/l reduced glutathione[GSH]). The pellet was suspended with medium to 2% (v/v) packed cell volume (PCV). Of these suspensions 14 ml was cultivated in 25 cm2 tissue culture flasks that were put flat in the incubator so as to obtain 6 mm medium height. The incubator was set at 37C, 5% (v/v) CO2, 90% relative humidity. Caps were loosely tightened to allow gas exchange. Medium was changed daily until the parasitaemia reached 20–30%. At this parasitaemia subcultures were generated at 1% parasitaemia by dilution with a 2% (v/v) normal red blood cell suspension in culture medium. Spent medium from B. canis cultures (taken at 24 and 48 h of a culture) was used as a source of soluble parasite antigen. The supernatant was concentrated 10 times using a dialysis membrane with a 8–10 kDa cut off.
2.2. In vivo phase – vaccination experiment 2.2.1. Animals Beagle dogs of either sex of approximately 6 months of age were used in the experiments. They were obtained from commercial breeders. The animals were clinically healthy, and had no history of babesiosis. The animals were identified by individual transponders that were implanted subcutaneously. Animals were acclimatised for at least one week. The animals received standard dog feed daily. Drinking water was supplied ad libitum. 2.2.2. Vaccination The vaccine contained 50 μg of recombinant CBA protein (see below 2.11) and 250 μg of saponin adjuvant (Supersap®, Desert King, Chile) per dose, in a volume of 1 ml Sörenson buffer (pH 6.0). Two groups of seven dogs each were formed. One group was vaccinated with a single dose of vaccine. These dogs received two booster vaccinations with the same material at three week intervals. A second group of seven dogs did not receive any injections and served as control group. Two weeks after the booster vaccination animals were challenged intravenously with 1 × 106 B. canis A-infected erythrocytes.
2.4. Analytical agarose gel precipitation using VC-serum Analytical agarose gel precipitation was used to determine whether antigens from SPA preparations could be recognized by VC-serum. Microscopic object slides were cleaned with alcohol (70% v/v) and precoated with 1% (v/v) agarose gel solution in Tris-buffered saline solution (10 mM, pH 8.0). Next, 5 ml of the same agarose gel solution was pipetted onto the slide. After cooling to room temperature 3 mm holes were punched into the gel. The centre hole was filled twice with concentrated supernatant of in vitro cultures of B. canis A, and two-fold serial dilutions of vaccination-challenge serum were pipetted in the surrounding holes. Slides were incubated overnight in a humidified chamber at 4C. Precipitation lines were visualised after washing the gels in Tris-buffered saline solution to remove non-precipitated material. The removal of free haemoglobin indicated that washing was complete.
2.2.3. Observations and clinical examinations After challenge infection animals were examined daily for clinical signs of babesiosis for a period of 14 days, when all animals were treated with imidocarbdipropionate (Carbesia®, MSD Animal Health) to cure the infection. Special attention was given to behaviour, spleen size, size of lymph nodes, colour of the mucous membranes of mouth and eye-lid, and the capillary refill time. Clinical observations were scored as a numeric value (Schetters et al., 1994). From these scores the total clinical score value was calculated for each day. Of these, the maximal clinical score value during the observation period was determined. 2.2.4. Haematocrit The haematocrit value is expressed as packed cell volume (PCV) of a sample of venous blood that was taken from the vena jugularis. A haematocrit capillary was filled with blood and centrifuged in a haematocrit centrifuge (Hettich) for 5′ at 10.000 rpm. The packed cell volume was read using a haematocrit reader. For each individual animal the maximal decrease of PCV was calculated, and expressed as a percentage of the value obtained at the day of challenge infection.
2.5. Preparative agarose gel precipitation using VC-serum In order to obtain enough precipitate for further molecular analysis, preparative agarose gels were used. In short, lids of standard polystyrene micro-Elisa titre plates were completely filled with 1% (w/v) agarose gel solution in Tris-buffered saline solution (10 mM, pH 8.0). Three parallel longitudinal slots were cut with a scalpel, at 3–4 mm 22
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
2.9. Immunoprecipitation using rabbit-anti-rCBA antiserum – western blot analysis
distance. Next the central slot was entirely filled with neat SPA preparation, and the two outer slots were filled with neat VC-serum. The gel was incubated overnight in a humidified chamber at 4C to allow the formation of immunoprecipitate. (clearance of haemoglobin from the gel slice was taken as an end point for washing).
For immunoprecipitation of non-radiolabelled antigen preparations, supernatants and B. canis-infected erythrocyte fractions were mixed with each test serum and incubated overnight at 4 °C with constant stirring. The immune complexes that had formed were captured onto protein A-sepharose-CL4 B beads for 1 h at room temperature with constant stirring. After incubation, the beads were centrifuged at 3000g for 15 min and supernatant was discarded. Immune complexes were recovered from the beads by acid elution according to the manufacturer’s instructions. The eluate was collected, filtered through a 0.22 μm pore size membrane and frozen at − 80 °C until use for SDSPAGE and western blotting.
2.6. SDS-PAGE and western blotting Protein characteristics and antigenic properties were analysed by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDSPAGE; Laemmli, 1970) and Western blot analysis (Towbin et al., 1979). Briefly, protein preparations were separated on 12% (w/v) reducing SDS-PAGE and transferred to nitrocellulose membranes. Blots were blocked with 5% (w/v) skimmed milk in Tris-buffered saline solutions (TBS; 20 mM Tris-HCl, pH 7.2, 150 mM NaCl).
2.10. Immunofluorescence with antisera against rCBA
2.7. Pulse-chase experiments
Babesia canis A was cultured in vitro to a parasitaemia of 7–10%. A culture sample (500 μl) was collected and centrifuged at 1200g for 5 min and washed five times in RPMI 1640. The erythrocytes were coated onto wells of Teflon-coated microscopic slides, air-dried and fixed for 10 min in ice-cold acetone/methanol solution (4:1). After three washes with PBS, slides were incubated for 30 min with twofold serial dilutions of test sera in PBS in a moist chamber, washed and then incubated 30 min with FITC-conjugated second antibody (Sigma). Slides were mounted with Citifluor™ (EMS Acquisition Corporation) and observed with immersion oil (1000X) using a fluorescence microscope (Axio Scope, Zeiss).
Babesia canis A was maintained in culture in vitro as described above (T-75 flasks). When parasitemia reached 10%, the culture supernatant was removed and the erythrocytes were washed in RPMI 1640 and then resuspended in methionine-free RPMI medium supplemented with 10% (v/v) NDS and 1 g/L GSH. After 4 h of incubation, L35 S-methionine (Amersham) was added (65 μCi/ml) for 15 min. The erythrocytes were then pelleted by centrifugation and washed twice in RPMI 1640 medium and then dispensed into a new flask (T-25) with complete culture medium and incubated for various chase-times (15 min, 30 min, 1 h or 4 h). At each chase-time, the erythrocytes were pelleted and the culture supernatant was collected. Each sample was centrifuged for 20 min at 10 000g at 4 °C, the supernatant was collected and filtered through a 0.22 μm pore size membrane and frozen at − 80 °C until use for SDS-PAGE and autoradiography.
2.11. Cloning of CBA gene and preparation of recombinant CBA protein The full length CBA gene, containing both the C- and N- terminal hydrophobic sequences, was amplified from Babesia canis genomic DNA using the following primer set: FW(5′accATGATGCTGCTCTTCGCCTTG3′) and REV (5′GCGAAAAA CATGAGTGGGACC3′) and cloned with a C-terminal HIS tag, behind the T7 promotor into a pET101 expression vector. After expression of the protein in E. coli BL21 (DE3), cells were lysed and antigen was dissolved in 6 M urea, 300 mM NaCl, 50Mm NaH2PO4 Ph 8.0. The protein was bound to a FF HisTrap™ column (GE Healthcare) and refolded on the column by exchanging the denaturing buffer for a non-denaturing buffer, containing a redox couple (0.4 g/L oxidized glutathione; 1.8 g/L reduced glutathione; 6 g/L CHAPS; 1 mM EDTA; 0.1 M NaCl; 50 mM Tris pH 8.0). After refolding, protein was eluted in a 0–500 mM imidazole gradient and stored at −70C until use. 2.12. Detection of antibodies against recombinant CBA by ELISA For the detection of antibodies against recombinant CBA by ELISA, a hydrophilic form of the protein was produced. The coding sequence from E19 to A301 was cloned in pIVEX 2.3MCS vector (Roche) by assembly PCR using 5′-CGCTTAATTAAACATATGACCGAAAACAC TATACTTTTATCCAATGTAGAATTCC as forward primer and 5′TTAGTTAGTTACCGGATCCCTTAAGCGGTAGTAAGTTTAGGAGAAGC as reverse primer. Amplification and assembly was done using Phusion DNA polymerase (New England Biolabs), and resulting plasmids were transformed in E. coli DH5alpha. A recombinant plasmid was selected and verified by sequencing, and then transformed in E. coli BL21(DE3) for protein expression. The recombinant protein, without any hydrophobic sequence but with a HIS-tag appended at the C-terminal position, was then produced in a 3-l bioreactor (Infors) at 30 °C. After harvesting, cells were lysed in HIS-tag lysis buffer (NaH2PO4 50 mM, NaCl 500 mM, imidazole 10 mM, pH = 8.0) using a high pressure homogenizer (Avestin) operating at 1200 bar. Non-soluble material was removed by centrifugation and Triton X-100 (1% [v/v]) was added to the clarified lysate, which was subsequently loaded on a 5 ml HisTrap™ column. The column was washed with NaH2PO4 50 mM, NaCl 1 M, imidazole 10 mM, pH = 8.0 to remove non-bound material. The
2.8. Immunoprecipitation using VC-serum − autoradiography For conventional immunoprecipitation the protocol of Hines et al. (1989) was used. Parasites were metabolically labelled with L- 35Smethionine to allow detection of parasite-encoded proteins. When parasitaemia reached 10%, overnight incubation was performed in methionine-deficient culture medium (Gibco) supplemented with 50 μCi of L- 35S-methionine (Amersham) per ml. After incubation, the labelled cells and the corresponding supernatants were collected and processed for immunoprecipitations assays. For the analysis of antigens from culture supernatants, culture supernatants were centrifuged for 15 min at 15 000g, filtered through a membrane of 0.22 μm and depleted of serum immunoglobulins by incubation with protein A-sepharose-CL4 B for 1 h at room temperature with constant stirring. The beads were centrifuged at 3000g for 15 min and supernatant was collected. For the analysis of cell-associated antigens, radiolabelled infected red blood cells were washed twice by centrifugation at 700g for 5 min in culture medium without serum and solubilized with 10 vols of lysis buffer 50 mM Tris, 5 mM EDTA, 100 mM NaCl, 1% (v/v) nonidetP40, 1 mM phenylmethylsulfonylfluoride (PMSF), 0.1 mM N-tosyl-Llysine-chloromethylketone (TLCK). Insoluble material was removed by centrifugation at 10000g at 4 °C. For immunoprecipitation, the radiolabelled lysates (106 cpm) and the different radiolabelled supernatants (200 μl) were mixed with each test serum and incubated overnight at 4 °C with constant stirring. The immune complexes that had formed were captured onto protein A-sepharose-CL4 B beads for 1 h at room temperature with constant stirring. After incubation, the beads were centrifuged at 3000g for 15 min and supernatant was discarded. Immune complexes were recovered from the beads by acid elution according to the manufacturer’s instructions. The eluate was collected, filtered through a 0.22 μm pore size membrane and frozen at − 80 °C until use for SDS-PAGE and autoradiography. 23
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
vaccinated dogs that were challenged with B. canis A. As control, serum from adjuvant control dogs was used (serum was derived from group C and A respectively, described in Schetters et al., 1992). At all strips the canine immunoglobulin heavy chain was detected at the 50 kDa position. Results showed that serum from vaccinated dogs taken at day 1–12 after infection (except for the serum sample of day 2 p.i.) reacted with a 100 kDa antigen. Serum from adjuvant control dogs did not recognize this antigen throughout the observation period (Fig. 2). From day 5 after infection onwards the serum samples of the vaccinated dogs recognized a second antigen of Mr 39 kDa specifically. Serum from adjuvant control dogs did not recognize this antigen. Serum from dogs of both groups recognized a 70 kDa antigen from day 7–8 after infection onwards.
Fig. 1. Analytical agarose gel precipitation. Immunoprecipitates formed by interaction of SPA and different dilutions of VC-serum. Note two distinct precipitation lines. Figure is photographically inverted for clarity.
3.2. Pulse-chase radioactively labelling of B. canis parasites In order to exclude the possibility that these antigens were of erythrocyte rather than parasite origin, metabolic labelling studies were performed. Since canine erythrocytes do not produce protein de novo, the only proteins that can become labelled in in vitro cultures of B. canis are parasite derived. Babesia canis A was cultivated in vitro and pulsed for 15 min with radioactively labelled methionine. Results showed that from 15 min after the pulse onwards a doublet of 40 kDa became radioactively labelled, and appeared the most prominent (Fig. 3). It appeared that, whereas at 120 min after the pulse the upper band of the doublet was more intense than the lower band, another two hours later the lower band of the doublet was more intense. Additional proteins of Mr 80, 48, 34, and 25 kDa were found but these were less pronounced (Fig. 3). A protein of 70 kDa could not be detected. Together with the western blot results, the data suggested that the major B. canis antigen in SPA preparations that was recognized by VC-serum was a 39 kDa antigen.
recombinant protein was then eluted using Tris 50 mM, imidazole 400 mM pH = 8.0. The protein was stored at −80 °C until use in a standard ELISA test for the detection of antibodies against the recombinant CBA protein in the serum of vaccinated dogs. 3. Results 3.1. Analysis of immune complexes from preparative gel diffusion Vaccinated dogs that were protected against B. canis infection produced antibodies against SPA from day 5 after infection onwards as detected by sandwich-ELISA (Schetters et al., 1996a). The antigen in a sandwich-Elisa is captured in its native form i.e. not denatured and stretched as in western blotting techniques. In addition, for a sandwichElisa to become positive, the relevant antigens must complex with at least two antibodies; the capture antibody and the reporter (=second) antibody. Hence, in order to determine which antigens were recognized in the sandwich-Elisa, immunoprecipitation assays that required complex formation between different antibodies and antigens were used. It was decided to use agarose gel immunoprecipitation. First, analytical agarose gel immunodiffusion was used to determine whether SPA preparations contained antigens that could be complexed with VCserum. Results showed that there were two distinct precipitation lines (Fig. 1). In order to further characterize the precipitated antigens, preparative gel diffusion was performed using B. canis A-SPA preparations and VC-serum in thick agarose gels. A slice of gel containing the precipitate was cut from the gel and washed in PBS to remove free proteins. Next the gel slice was treated for SDS-PAGE and western blotting. Blot strips were incubated with daily serum samples from a group of
3.3. Immunoprecipitation of radioactively labelled antigens from B. canis a cultures To verify that indeed a 39–40 kDa antigen from radioactively labelled B. canis parasites could be recognized by VC-serum, radioactively labelled antigens from in vitro cultures of B. canis A were prepared. Three different fractions were analysed; antigens from supernatants (SUP), and two fractions from B. canis A-infected erythrocytes, the purified merozoite fraction (MERO) and the cytosolic fraction of the lysed infected erythrocytes (CYT). Each of these fractions was used for immunoprecipitation with VC-serum. The precipitated antigens were separated on SDS-PAGE and the gel was subsequently developed by autoradiography. Results showed that amongst other proteins, a doublet of Mr of 39–40 kDa was precipitated from each of the antigen preparations Fig. 2. Western blot analysis of reactivity of sera from SPA-vaccinated dogs that were challenged with B. canis A parasites (V). As control sera from adjuvant control dogs were used (N). Antigen on the blots was derived by preparative agarose gel immunoprecipitation. Numbers below the strips indicate the day post infection at which serum was collected. C = no serum incubation. Image is manipulated to create pairs of blot strips per day (slight colour overlay).
24
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
3.4. Discovery of the gene encoding the 39 kDa B. canis protein So far, results indicated that the major B. canis antigen from SPA preparations that is specifically recognized in the sandwich-Elisa is the 39 kDa antigen of a 39–40 kDa doublet. The antigen was named canine Babesia antigen (CBA). In order to obtain a sufficient amount of CBA for amino acid sequence analysis, preparative agarose gel precipitation was performed using B. canis SPA preparations and VC-serum in thick agarose gels. The immunoprecipitate was separated on SDS-PAGE and the entire gel blotted onto nitrocellulose paper. Two strips taken from the left and right side of the blot were developed with VC-serum to locate the immunoreactive band at the 39 kDa position. Next, a strip of approximately 2 mm at the 39–40 kDa position was cut from the entire nitrocellulose blot. The molecules recovered from this strip were analysed by trypsin digestion and subsequent mass spectrometry (Proteomics Centre, Nijmegen University, the Netherlands). Aside from trypsin and haemoglobin fragments, six unique peptides, (not present in public databases) were discovered (depicted in bold Fig. 5). Next, genomic DNA of B. canis A was sequenced (BaseClear, Leiden, the Netherlands) and blasted with the nucleotide sequences deduced from the partial amino acid sequences that had been discovered. Results showed that all six peptides clustered in a single gene that encodes for a 33 kDa protein (Fig. 5). The gene contains a putative signal sequence (1–17). At the C-terminal side the gene encodes for a transmembrane sequence (301–321) that is required for GPI-anchoring. It is predicted that threonine (pos 300) is the amino acid to which the GPI-anchor is attached. The core amino acid sequence contains two cysteines (enlarged), one at pos 78 and one at pos 264.
Fig. 3. Autoradiograph of supernatants of in vitro cultures of B. canis A that were pulsed with L- 35S-methionine for 15 min and separated on SDS-Page. Supernatants were taken at different time points after the pulse (indicated in minutes at the top of each lane).
3.5. Recombinant CBA production and characterization The full length CBA gene, containing both the C- and N-terminal hydrophobic sequences, was amplified from B. canis A genomic DNA and cloned in E. coli BL21. Thirty grams of wet cell paste were lysed through a French press cell. The lysate was centrifuged and the soluble fraction was pipetted off. There was a small amount of recombinant protein detectable in the soluble fraction (Fig. 6). The pelleted material containing the inclusion bodies of the bacteria was dissolved in urea buffer. This fraction contained the recombinant CBA (rCBA) antigen. In addition, a doublet around 26 kDa was found. Finally, the expressed recombinant protein was refolded on a His-Trap column. Upon analysis on SDS-PAGE, the protein presented as a doublet of 37–40 kDa. Minor protein bands were found at 29 and 26 kDa position. Thirty grams of wet cell paste yielded 5 mg of recombinant protein with a purity of approximately 85% as estimated from a Coomassie Brilliant Bluestained gel (Fig. 6). To determine whether the purified recombinant protein was recognised by VC-serum, the protein was dot-blotted onto nitrocellulose paper and incubated with VC-serum. As a control, SPA of culture supernatant of B. canis A cultures was used. Results showed that VC-serum indeed recognized the recombinant protein. Normal dog serum did not recognize the recombinant protein (Fig. 7). The SPA of culture supernatant was positive with both sera due to the presence of normal dog serum in the culture supernatants. The recombinant protein could not be detected by sandwich-Elisa (data not shown), indicating that the protein did not express the two antigenic epitopes that are required to
Fig. 4. Autoradiograph of radioactively labelled antigens from in vitro cultures of B. canis A that were immunoprecipitated with VC-serum or normal dog serum and separated on SDS-PAGE. MERO = merozoite antigens, CYT = antigens from the cytosol of B. canis Ainfected erythrocytes, SUP = antigens from supernatants of in vitro cultures of B. canis A.
(Fig. 4). The 39 kDa protein from the doublet appeared to be more pronounced in the cytosol of infected erythrocytes, whereas in the culture supernatant the 40 kDa protein of the doublet was more pronounced. Serum from control dogs did not precipitate these molecules.
Fig. 5. Deduced amino acid sequence of the CBA gene of B. canis A. Six peptides that were detected in the native 39 kDa protein of B. canis A are presented between brackets in bold. Predicted transmembrane sequences are underlined.
25
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
was detected in supernatant (IP-SUP; very weak), lysate of infected erythrocytes (IP-LYS) and merozoites (IP-MERO). Such reactivity was not found in immunoprecipitates obtained with normal rabbit serum. Additional proteins were detected at the 35 kDa and 25 kDa (very weak) position in immunoprecipitates of lysate of infected erythrocytes (IP-LYS) and merozoites (IP-MERO). 3.6. Evaluation of the vaccine potential of rCBA in dogs Having established that the recombinant CBA antigen of B. canis A was specifically recognized by serum of immune dogs, and induced antibodies against the 40 kDa doublet of B. canis A parasites, the protective activity of the recombinant protein was evaluated in a vaccination-challenge experiment in dogs. Dogs were vaccinated at day 0 and received two booster vaccinations at days 21 and 42. Upon vaccination, dogs produced antibodies against rCBA as measured by Elisa. This response was boosted after the second vaccination at day 21. The additional booster vaccination at day 42 did not increase antibody titres further (Fig. 10). The serum recognized B. canis A merozoites in immunofluorescence (Fig. 11). Animals were challenged 14 days after the second booster vaccination (day 56). In non-vaccinated control dogs the infection became patent two days after challenge. In vaccinated dogs the first parasites were detected two days later, at day 4 post infection (p.i.; Fig. 12). Parasitaemia gradually increased in both groups of dogs, and peaked around day 8–9 p.i. From that day of infection onwards, the parasitaemia in vaccinated dogs progressively decreased, whereas in nonvaccinated dogs parasitaemia remained relatively high until day 12 p.i. and declined thereafter. The cumulative daily parasitaemia (=parasitic load) was statistically significant lower in the vaccinated group as compared to that of the non-vaccinated control group (P < 0.04, Student’s t-test, two-sided, unequal variances; Table 1). In non-vaccinated control dogs, the packed cell volume started to decrease immediately the day after challenge infection. In contrast, in vaccinated dogs this decrease started one day later (Fig. 13). In general, the decrease in PCV was on average 10% ( ± 3.0) less in the vaccinated dogs as compared to controls over the entire period. This was also reflected in the average maximal PCV decrease of the dogs in each group (Table 1). This difference of 10.5% was statistically significant (p = 0.05, Student’s t-test, two sided). During the post infection period the clinical signs of dogs of both groups were relatively mild until day 11 p.i. when six out of seven nonvaccinated dogs showed pale mucosae and had increased capillary refill times. This was reflected in the clinical score value that remained high until the end of the observation period when all dogs were treated (Fig. 14). The clinical score value of the vaccinated dogs was statistically significant lower than that of non-vaccinated dogs at day 3 and day 5 p.i., and from day 9 p.i. onwards when all vaccinated dogs recovered from the infection. The average maximal clinical score value of the vaccinated group was statistically significant lower than that of the control group (p < 0.01, Student’s t-test, two sided, equal variances; Table 1).
Fig. 6. Coomassie stained SDS-PAGE gel on which samples at different stages of recombinant CBA preparation were separated. M = markers; Soluble fraction = supernatant of lysed E.coli that expressed the CBA gene; Purified CBA = inclusion bodies dissolved in urea buffer; Refolded CBA = protein eluted from His-Trap column after refolding.
Fig. 7. Reactivity of VC-serum with recombinant CBA protein (rCBA). Recombinant protein was spotted onto nitrocellulose membranes (top rows)). As control SPA from supernatants from in vitro.
render a positive signal. Antiserum raised against rCBA was raised in rabbits and used to control whether the recombinant antigen induced antibodies against B. canis. Three different Babesia strains were used to assess specificity of the antiserum. Immunofluorescence showed that the antiserum recognized merozoites of B. canis A and B. canis B on blood smears of in vitro cultures of these strains. No fluorescence was found when blood smears of B. rossi cultures were used (Fig. 8). To determine the molecular specificity of the rabbit anti-rCBA serum, immunoprecipitation studies were performed. Different fractions of B. canis A in vitro cultures were prepared for immunoprecipitation: Supernatant (SUP), lysate of B. canis A-infected erythrocytes (LYS), merozoites isolated from the lysate (MERO), and the cytosol remaining after removal of merozoites (CYT). As a control, a lysate of uninfected red blood cells was used (RBC). Immunoprecipitation with rabbit anti-rCBA antiserum was performed, followed by western blotting and incubation with goat-anti-rabbit-immunoglobulin conjugate to detect the precipitated antigen. As a control antigen, B. canis A merozoites were used (no immunoprecipitation). Control immunoprecipitations were done with serum from the rabbit before immunization with rCBA (normal rabbit serum). Results are presented in Fig. 9. Immunoprecipitates showed a very high signal at the 50 kDa position where the rabbit immunoglobulin heavy chain migrated. No additional proteins were precipitated from lysates of uninfected red blood cells (IP-RBC). Reactivity against a 40 kDa protein
4. Discussion In a variety of Babesia-host models it has been shown that animals can be protected against clinical manifestations after experimental infection by vaccination using soluble parasite antigens from in vitro Babesia cultures (SPA; Schetters and Montenegro-James 1995). Based on this principle, commercial vaccines for the prevention of canine babesiosis have been developed and marketed in Europe (Schetters, 2005). Until to date the identity of the immunoprotective antigens has not been revealed. This is mainly due to the presence of large amounts of serum (10–40%) in the culture supernatants, which hampers the use of standard purification techniques. Using a combination of techniques we discovered a 39 kDa B. canis antigen that is present in culture 26
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
Fig. 8. Immunofluorescence on blood smears of in vitro cultures of different Babesia strains. Cells were incubated with rabbit serum raised against rCBA of B. canis A. Nuclei were stained with DAPI.
Fig. 9. Western blot analysis of immunoprecipitates (IP-xxx) obtained with rabbit antiserum raised against rCBA (left panel) and normal rabbit serum (right panel). Merozoite antigen of B. canis A was used as control (MERO). SUP = antigens from supernatants of in vitro cultures of B. canis A, LYS = lysed B. canis A-infected red blood cells; CYT = antigens from the cytosol of B. canis A-infected erythrocytes, RBC = lysate of uninfected red blood cells. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
1 gene described recently (Zhou et al., 2016). These authors have shown that monospecific mouse antiserum against the recombinant BcMSA1 protein stained B. canis merozoites, and reacted with a 39–40 kDa doublet in lysates of B. canis-infected erythrocytes (Zhou et al., 2016). This further corroborates the hypothesis that the 39 kDa and 40 kDa proteins are encoded for by a single gene. The amino acid sequence has a predicted GPI anchoring site, which lends support to the hypothesis that it is a merozoite membrane protein. It is hypothesized that the 39 kDa protein could be the GPI-anchored 40 kDa merozoite membrane protein from which the GPI-anchor is cut off. There are many similarities between CBA of B. canis and Bd37 of B. divergens. In the B. divergens-gerbil model we have found earlier that supernatant of in vitro cultures of B. divergens contains a 37 kDa merozoite surface antigen (Bd37; Carcy et al., 1995, Delbecq et al., 2002). The protein is GPI-anchored to the merozoite surface, and is released into the environment after the GPI-anchor is cut. The recombinant Bd37
supernatants, and antibodies against which are associated with protective immunity. The 39 kDa antigen appears to be a doublet that can be radioactively labelled with L- 35S-methionine (this work). A doublet of similar size that could be labelled with 14C-glucosamine has been described in the supernatants of B. canis before (Azzar et al., 1990). It is likely that this is the same doublet as described here. In pulse-chase experiments the 39 kDa protein signal increases between 120 and 240 min after the pulse, simultaneously with a relative decrease in the intensity of the 40 kDa protein. This suggests that the 39 kDa molecule derives from the 40 kDa protein. With the resolution of part of the amino acid sequence of the 39 kDa antigen, the gene encoding for the protein (canine Babesia antigen; CBA) could be discovered in the B. canis genome. The CBA-gene was expressed and the resulting rCBA protein was recognized by VC-serum, and monospecific rabbit antiserum against rCBA stained B. canis merozoites and reacted with the native 40 kDa antigen of B. canis. The CBA gene is similar to the BcMSA27
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
Fig. 12. Dynamics of the parasitaemia in dogs after experimental infection with B. canis A infected erythrocytes. One group of dogs was vaccinated with rCBA of B. canis A (blue circles), non-vaccinated dogs served as control (red diamonds). Data represent the group average of log-transformed number of infected erythrocytes per 105 red blood cells (n = 7). Error bars indicate the standard deviation.
Fig. 10. Dynamics of the antibody response of dogs upon immunization with rCBA of B. canis A (blue curve). Control dogs received no vaccinations (red curve). Arrows indicate the time point of immunization. Data represent the group average antibody titre as measured in Elisa using rCBA without the hydrophobic terminal sequences (n = 7). Error bars indicate standard deviation.
Table 1 Essential disease parameters of Babesia infection. Effect of vaccination on the essential parameters of Babesia infection in dogs after experimental infection with B. canis A infected erythrocytes. One group of dogs was vaccinated with rCBA of B. canis A, nonvaccinated dogs served as control. Data represent the group average and standard deviation. Asterisk indicates that these values are statistically different from those of controls.
protein induces protective immunity against virulent challenge infection (Delbecq et al., 2006). Because of the similarities between Bd37 of B. divergens and CBA of B. canis (including the presence of two cysteine residues in the core protein that could be involved in 3D structuration of the protein), it is tempting to speculate that they are homologues proteins. However, there are no significant sequence homologies between the two molecules or any other known sequences in public databases (e.g. MSA1 of B. bovis; reviewed in Carcy et al., 2006). Because of this it is felt premature to designate the gene encoding the 39–40 kDa protein BcMSA1 as suggested by Zhou and collaborators (Zhou et al., 2016). Upon analysis of a series of recombinantly expressed Bd37 proteins, it appeared that presence of at least one hydrophobic terminal sequence to the hydrophilic core sequence was indispensable for the induction of immunity against heterologous B. divergens strains (Delbecq et al., 2006). Because of this, we cloned and expressed the entire CBA gene (including the two hydrophobic terminal sequences) for vaccination purposes. The rCBA protein induced protection against B. canis infection in dogs. Six out of 7 control dogs had to be chemotherapeutically treated to prevent succumbing to infection, whereas all vaccinated dogs controlled the challenge infection before the end of the observation period. The nature of the protection was different from that obtained with SPA from in vitro culture supernatants in that a striking antiparasite effect was found (Schetters, 2005). A major difference between vaccination with rCBA and SPA is that immunization with SPA does not induce antibodies against the 39–40 kDa antigen, but apparently primes dogs to develop a rapid anamnestic response against the 39 kDa protein from day 6 after challenge onwards (this work and Schetters et al., 1996a,b). In contrast, dogs that were vaccinated with rCBA had antibodies already at the time of challenge infection, which could actually
Parasite Load Max PCV decrease Max Clinical Score
Vaccinated
Control
3,5 ± 1,9* 46,6 ± 7,3* 1,3 ± 0,5*
7,2 ± 3,5 57,1 ± 10,5 2,7 ± 0.7
neutralize part of the parasites of the challenge inoculum. This would be reflected in a prolonged prepatent period and delayed onset of haematocrit decreases, as if a lower challenge dose was used (Schetters et al., 2009). This is corroborated by the fact that the rate at which the PCV values decrease in both groups is similar when the curve of the vaccinated group is shifted two days. The anti-parasite effect is not restricted to a direct effect on the challenge inoculum, however, as evidenced by the fact that the parasitaemia during the post infection period is much lower in the group of vaccinated dogs than in the controls. In addition, assuming a delayed onset of disease due to an immediate effect on the challenge inoculum, it can be deduced that vaccinated dogs control the infection 7 days after the PCV started to decrease. From that day onwards the parasitaemia in the vaccinated dogs progressively decreased. In the control group, the infection developed more chronically, which led to severe clinical symptoms 11 days after the PCV started to decrease. The most important clinical symptoms were pale mucosae and a prolonged capillary refill time. It has been suggested that pale mucosae
Fig. 11. Immunofluorescence of B. canis A-infected erythrocyte. Cells were incubated with serum of dogs that were vaccinated rCBA of B. canis A. Nuclei were stained with DAPI.
28
Veterinary Parasitology 249 (2018) 21–29
K. Moubri et al.
employees of Merck Animal Health. Research done by KM, SD, BC, EP and AG was in part funded by Merck AH. Acknowledgements This publication is dedicated to Eric Précigout who passed away during the course of the project. The authors thank all participating laboratory staff and students for their contribution and assistance in these studies during the course of this project since 1994. Special thanks to Silvie Randazzo for continuous support. References Azzar, G., Ravisson, J., Got, R., 1990. Characterization and purification of culture-derived soluble glycoproteins of Babesia canis. Parasitol. Res. 76, 578–580. Baneth, G., Florin-Christensen, M., Cardoso, L., Schnittger, L., 2015. Reclassification of Theileria annae as Babesia vulpes sp. nov. Parasites Vectors 8, 207. Carcy, B., Precigout, E., Valentin, A., Gorenflot, A., Schrevel, J., 1995. A 37-kilodalton glycoprotein of Babesia divergens is a major component of a protective fraction containing low molecular-mass culture-derived exoantigens. Infect. Immun. 63, 811–817. Carcy, B., Precigout, E., Schetters Th Gorenflot, A., 2006. Genetic basis for GPI-anchor merozoite surface antigen polymorphism from Babesia and resulting antigenic diversity. Vet. Parasitol. 138, 33–49. Delbecq, S., Précigout, E., Vallet, A., Schetters, Th., Gorenflot, A., 2002. Babesia divergens: cloning and biochemical characterisation of Bd37. Parasitology 125 305–212. Delbecq, S., Hadj-Kaddour, K., Randazzo, S., Kleuskens, J., Schetters, T., Gorenflot, A., Precigout, E., 2006. Hydrophobic moieties in recombinant proteins are crucial to generate efficient saponin-based vaccine against Apicomplexan Babesia divergens. Vaccine 24, 613–621. Hines, S.A., McElwain, T.F., Buening, G.M., Palmer, G.H., 1989. Molecular characterization of Babesia bovis merozoite surface proteins bearing epitopes immunodominant in protected cattle. Mol. Biochem. Parasitol. 37, 1–9. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature 227, 680–685. Matijatko, V., Torti, M., Schetters Th, P., 2012. Canine babesiosis in Europe: how many diseases? Trends Parasitol. 28, 99–105. Schetters, Th., Montenegro-James, S., 1995. Vaccines against babesiosis using soluble parasite antigens. Parasitol. Today 11, 456–462. Schetters, Th.P.M., Kleuskens, J.A.G.M., Scholtes, N.C., Bos, H.J., 1992. Vaccination of dogs against Babesia canis infection using parasite antigens derived from in vitro culture. Parasite Immunol. 14, 295–305. Schetters, Th.P.M., Kleuskens, J.A.G.M., Scholtes, N.C., Pasman, J.W., Bos, H.J., 1994. Vaccination of dogs against Babesia canis infection using antigen from culture supernatant with an emphasis on clinical babesiosis. Vet. Parasitol. 52, 219–233. Schetters, Th., Kleuskens, J., Scholtes, N., Bos, H.J., 1995. Strain variation limits protective activity of vaccines based on soluble Babesia canis antigens. Parasite Immunol. 17, 215–218. Schetters, Th.P.M., Scholtes, N.C., Kleuskens, J.A.G.M., Bos, H.J., 1996a. Not peripheral parasitaemia but the level of soluble parasite antigen in plasma correlates with vaccine efficacy against Babesia canis. Parasite Immunol. 18, 1–6. Schetters, Th., Kleuskens, J., Scholtes, N., Pasman, J., Bos, H.J., 1996b. Vaccination of dogs with soluble Babesia canis antigens derived from in vitro culture is strain specific. New Dimensions in Parasitology: Keynote Papers from the VIII International Congress of Parasitology; Acta Parasitologica Turcica. pp. 543–550. Schetters, Th.P.M., Kleuskens, J.A.G.M., Scholtes, N.C., Goovaerts, D., Pasman, J., Bos, H.J., 1997. Vaccination of dogs against Babesia canis. Vet. Parasitol. 73, 35–41. Schetters, Th.P.M., Kleuskens, J., Van De Crommert, J., De Leeuw, P., Finizio, A.-L., Gorenflot, A., 2009. Systemic inflammatory responses in dogs experimentally infected with Babesia canis: a haematological study. Vet. Parasitol. 162, 7–15. Schetters, Th.P.M., 2005. Vaccination against canine babesiosis. Trends Parasitol. 21, 179–184. Towbin, H., Staehelin, T., Gordon, J., 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. U. S. A. 76, 4350–4354. Uilenberg, G.G., Franssen, F.F.J., Perié, N.M., Spanjer, A.A.M., 1989. Three groups of Babesia canis distinguished and a proposal for nomenclature. Vet. Quarterly 11, 33–40. Uilenberg, G., 2006. Babesia – a historical overview. Vet. Parasitol. 138, 3–10. Zhou, M., Cao, S., Luo, Y., Liu, M., Wang, G., Moumouni, P.F.A., Jirapattharasate, C., Iguchi, A., Vudriko, P., Terkawi, M.A., Löwenstein, M., Kern, A., Nishikawa, Y., Suzuki, H., Igarashi, I., Xuan, X., 2016. Molecular identification and antigenic characterization of a merozoite surface antigen and a secreted antigen of Babesia canis (BcMSA1 and BcSA1). Parasit. Vectors 9, 257.
Fig. 13. Dynamics of the haematocrit in dogs after experimental infection with B. canis A infected erythrocytes. One group of dogs was vaccinated with rCBA of B. canis A (diamonds), non-vaccinated dogs served as control (circles). Data represent the group average of packed cell volume (PCV) expressed as percentage relative to the value at the day of challenge infection (n = 7). Error bars indicate the standard deviation.
Fig. 14. Dynamics of the clinical score of dogs after experimental infection with B. canis A infected erythrocytes. One group of dogs was vaccinated with rCBA of B. canis A (diamonds), non-vaccinated dogs served as control (crosses). Data represent the group average of the clinical score (n = 7). Error bars indicate the standard deviation.
as seen in babesiosis reflects anaemia. The fact that at day 14 p.i. the average PCV values of the non-vaccinated dogs (of which 6 out of 7 had pale mucosae), was not significantly different from the vaccinated dogs that had normal mucosae (29.3 ± 6.1 and 28.7 ± 3.8 respectively), does not support that suggestion. It is more likely that, due to (compensated) hypotension, which has been shown to occur in this experimental model (Schetters et al., 2009), tissue perfusion was impaired. Vaccination of dogs with rCBA limited the development of this pathological condition because of an effect on parasite proliferation. In conclusion, an immunoprotective B. canis protein was isolated from the supernatants of in vitro cultures of B. canis A parasites. The antigen that appears as a doublet on western blots, is encoded for by a single gene, and named Canine Babesiosis Antigen (CBA). A vaccine based on the recombinantly produced full-length CBA protein (including the hydrophobic terminal sequences) induced protective immunity against homologous challenge infection in dogs. Whether the antigen also protects against heterologous challenge infection remains to be determined. Declaration of interest Authors NS and TvK are current, and TS, JK and JvdC are former
29