Recombinant Mycobacterium bovis BCG is a promising platform to develop vaccines against Trypansoma cruzi infection

Summary

Chagas disease, caused by the hemoflagelate parasite Trypanosoma cruzi, is one of the most prevalent endemic parasitoses, affecting 7–8 million people. Due to the complexity of the infection, no vaccines are available at present. The extraordinary adjuvant capacity of bacille Calmette–Guérin (BCG) was explored in this work to develop a vaccine candidate to protect against T. cruzi infection using the recombinant BCG (rBCG) vaccine platform. Three antigens of the parasite corresponding to the N and C terminal fragments of the enzyme trans‐sialidase (NT‐TS and CT‐TS, respectively) and a fragment of the cruzipain enzyme (CZf) were cloned into the vectors pUS997 and pUS2000 and transformed into the BCG Pasteur strain. In vaccinated mice, rBCG expressing NT‐TS in pUS2000 plasmid provided the highest protection and the lowest parasitemia after challenging BALB/c mice with a 50% lethal dose of parasites. When mice vaccinated with pUS2000‐NT‐TS were challenged with a 100% lethal dose of parasite, high levels of protection were also obtained, together with a low degree of cardiac lesions 120 days after infection. In immunized mice with pUS2000‐NT‐TS/rBCG clone, the proliferation of CD4⁺ cells from splenocytes stimulated with the TS antigen was significant; this stimulation increased interferon (IFN)‐γ and interleukin (IL)‐17 within CD4⁺ T lymphocytes (LTCD4⁺) cells and IFN‐γ and CD107 expression within LTCD8⁺ cells. Therefore, pUS2000‐NT‐TS/rBCG conferred high levels of protection, which correlated with an immune response orientated towards a T helper type 1 (Th1)/Th17 profile, together with an LTC‐specific response, indicating that rBCG is a promising platform to develop vaccines against T. cruzi.

The use of the recombinant BCG vaccine platform to protect against T. cruzi infection is described. Six constructions of rBCG were obtained and assessed in a mice model of the infection. Mice vaccinated with rBCG expressing N terminal fraction of TS antigen presented the best protection. Also the vaccine protects against the development of cardiac lesions characteristic of the T cruzi infection in the chronic phase of the infection. The use of recombinant BCG vaccine against T cruzi allow a LTC, TH1 and TH17 cellular response that is needed to fight the parasite in accord by previous reports.

Introduction

Chagas disease, caused by human infection of the hemoflagellate parasite Trypanosoma cruzi, affects more than 7 million people worldwide [1]. To date, no vaccine is available to control the human T. cruzi infection, although different preclinical vaccine candidates based on bacterial delivery systems, viral, DNA or protein subunits with adjuvants have been developed and evaluated [2]. Furthermore, many of these vaccines allow the development of an immune response that results in up to 100% of survival in mouse infection models challenged with lethal doses of T cruzi. However, a prophylactic sterilizing vaccine has still not been achieved, which indicates the need to investigate new vaccine strategies.

Interestingly, the use of a recombinant BCG (bacille Calmette‐Guérin) platform to elicit immunity against T. cruzi has still not been described. The use of BCG as a vaccine delivery system has outstanding features, because it has an extraordinary adjuvant capacity and provides immunity against intracellular microorganisms [3]. This effect has been confirmed, for example, in an epidemiological study that determined that BCG directs a T helper type 1 (Th1) response against antigens related or not to the microorganism [4]. This immune response profile is critical for the control of T. cruzi infection [5]. In particular, for this infection a vast amount of complementary information encourages the use of BCG as a vaccine delivery system. For example, in an experimental model, early immunization with soluble extracts of Actinomyces (a genus closely related to mycobacteria) was found to ostensibly modify the course of T. cruzi infection [6]. Moreover, in a study of patients with chronic Chagas disease, our group described that individuals vaccinated with BCG at birth had a lower frequency of evolution towards heart disease than unvaccinated patients [7]. Conversely, the concept of using BCG to improve the immune response against parasites has already been used to design a vaccine against Leishmania braziliensis, a parasite closely related to T. cruzi [8]. In that report, dead parasites formulated with BCG allowed complete remission of the lesions in patients, with the only adverse effect of a local reaction of BCG. Based on these results, we hypothesized that a vaccine formulation with components of BCG could also favorably influence the development of immunoprotection against T. cruzi infection.

In the present study, we evaluate BCG as a platform to deliver T. cruzi antigens. Several studies have shown the effectiveness of the enzyme trans‐sialidase (TS) and the protease cruzipain (CZ) as vaccine candidates [9, 10, 11, 12]. Both secretory enzymes are involved in the virulence of the parasite [13]. In a previous work we described that, when a subunit platform vaccine is used, TS presents a much higher level of protection than several candidates tested and that CZ also provides protection, although with less efficacy [14]. Based on that information we report herein the cloning of these antigenic proteins from T. cruzi in a recombinant BCG vaccine platform using two episomal vectors carrying different promoters. Then, the vaccine performance and immune responses of different recombinant BCG vaccines expressing TS and CZ antigens from T. cruzi were assessed and, finally, the cellular response of the clone that afforded the best protection was analyzed.

Materials and methods

Strains and growth conditions

Escherichia coli strain DH5α was grown in Luria–Bertani medium at 37°C with 50 μg/ml kanamycin for transformed strains or without kanamycin for non‐transformed strains. Mycobacterium bovis BCG Pasteur was cultured at 37°C in Middlebrook 7H9 medium (Difco, São Paulo, Brazil) with 10% of oleic acid, albumin, dextrose complex (OADC; Difco), 0·2% glycerol and 0·05% Tween 80 or 7H10 agar (Difco) supplemented with 10% OADC, 0·2% glycerol and kanamycin (25 mg/ml) when necessary.

DNA manipulation reagents

Oligonucleotides were synthesized by Thermo Fisher Scientific (Fremont, CA, USA). Polymerase chain reaction (PCR) reactions were performed using GoTaq Colorless Master Mix (Promega, Madison, WI, USA), as previously described [12]. Restriction enzymes and T4 DNA ligase enzyme were purchased from Promega.

Assembly of recombinant vectors

We selected two TS fragments: CT‐TS, that included the amino acid region from 337 to 627, which we assessed with promising results in a previous work,¹² and NT‐TS fragment from 99 to 349, which was not previously assessed. Furthermore, we selected a CZ fragment that included the region from the amino acid 123 to 338, corresponding to the N‐terminal and catalytic region of the mature enzyme, which was previously described as a promising antigen vaccine candidate [15].

The fragments were amplified by PCR, using primers described in Table 1. The TS (AJ276679) and the CZ (XM_800858.1) sequences were used to design primers with the XbaI and HindIII restriction sites. The fragments, named TS NH2‐terminal end (TS‐NT), TS COO‐terminal end (TS‐CT) and cruzipain fraction (CZf), were cloned into pUS977 and pUS2000 (P_{A N} promoter and 18 kDa gene promoter respectively) [16, 17]. Shuttle vectors able to replicate in E. coli and M. bovis BCG [17].

Table 1.

Nucleotide sequences of primers and probes used in this work

Protein	Primer/probe (5′–3′)	Sequence length	Region included within the gene
NT‐TS	5′‐GGTTCTAGACCACCGTGATTGTGAA‐3	750 bp	296–1046
	5′‐AAGCTTTACAGGACGGAGCTGTAGG‐3
CT‐TS	5′‐CACTCTAGACGGTGATGAAAATTCCGCCTA‐3′	870 bp	1111–1881
	5′‐CTCAAGCTTAAGGTCCTGGCTCAAGAACAA‐3′
CZf	5′‐CACTCTAGACCCCGCGGCAGTGGATTG‐3	645 bp	369–1014
	5′‐CTAAAGCTTACCACCGACCACCGCGG‐3

CZf = cruzipain fraction; bp = base pairs; NT‐TS = NH2‐terminal trans‐sialidase; CT‐TS = COO‐terminal trans‐sialidase.

BCG transformation and analysis of expression

M. bovis BCG Pasteur electrocompetent cells were transformed with the recombinant plasmids using a previously described method [18]. BCG transformants were selected in 7H10 medium with kanamycin and grown in selective 7H9 medium for 5 days. Expression of the recombinant proteins was confirmed by Western blot. Specific proteins were visualized using hyperimmune sera produced in mouse against TS and CZ recombinant proteins (1 : 500) and peroxidase‐conjugated anti‐mouse gamma specific antibody (Sigma‐Aldrich, St Louis, MO, USA) diluted 1 : 2000. Detection was carried out using 3,3′‐diaminobenzidine (DAB; Sigma‐Aldrich). Briefly, for production of hyperimmune sera, TS and CZ were produced in Pichia pastoris and E. coli, respectively; BALB/c mice were then inoculated intraperitoneally on days 0, 14, 21 and 28, and on day 35 sera were collected as previously described [11, 14].

Assessment of the protection triggered by different constructions

The different constructions were assessed in a mouse model of T. cruzi infection that consisted of BALB/cCmedc female mice infected with the Tulahuén strain of T. cruzi. The groups assessed and the analyzed parameters are described in Table 2. In all groups of mice, antibodies and DTH were assessed 30 days after the last immunization in non‐challenged animals. Then, we evaluated the efficacy of all rBCG constructions using a sublethal challenge and we selected the more promising construction to be assessed by a lethal challenge. In animals immunized with rBCG carrying NH2‐terminal trans‐sialidase in pUS2000 (pUS2000‐NT‐TS), the trypanolytic activity of sera, histology and cellular response was also performed. The Tulahuén strain of T. cruzi employed in these studies was maintained by serial passages in C57BL/6 suckling mice, as previously described [19]. The experimental procedures for immunization and infection as well as the protocols for immunological and histological analysis were reproduced three times to corroborate the results.

Table 2.

rBCG constructions assessed as vaccine candidates against T. cruzi infection in this work and parameters analyzed for each formulation

Group	Immunized formulation	Experimental parameters evaluated
pUS977‐NT‐TS	rBCG carrying pUS977‐NT‐TS	Non‐lethal challenge with 500 parasites, antibodies, DTH, survival, parasitemia
pUS2000‐NT‐TS	rBCG carrying pUS2000‐NT‐TS	Non‐lethal challenge with 500 parasites and lethal challenge with 1000 parasites, antibodies, DTH, trypanolysis, survival, parasitemia, histology, cellular response
pUS977‐CT‐TS	rBCG carrying pUS977‐CT‐TS	Non‐lethal challenge with 500 parasites, antibodies, DTH, survival, parasitemia
pUS2000‐CT‐TS	rBCG carrying pUS2000‐CT‐TS	Non‐lethal challenge with 500 parasites, antibodies, DTH, survival, parasitemia
pUS977‐CZf	rBCG carrying pUS977‐CZf	Non‐lethal challenge with 500 parasites, antibodies, DTH, survival, parasitemia
pUS2000‐CZf	rBCG carrying pUS2000‐CZf	Non‐lethal challenge with 500 parasites, antibodies, DTH, survival, parasitemia
BCG Pasteur	BCG Pasteur strain	Non‐lethal challenge with 500 parasites and lethal challenge with 1000 parasites, antibodies, DTH, trypanolysis, survival, parasitemia, histology, cellular response
PBS	PBS	Non‐lethal challenge with 500 parasites and lethal challenge with 1000 parasites antibodies, DTH, trypanolysis, survival, parasitemia, histology, cellular response

T. cruzi = Trypanosoma cruzi; NT‐TS = NH2‐terminal trans‐sialidase; CT‐TS = COO‐terminal trans‐sialidase; rBCG = recombinant bacille Calmette–Guérin; DTH = delayed‐type hypersensitivity; CZf = cruzipain fraction; PBS = phosphate‐buffered saline.

Immunization schedules and infection protocol

To evaluate the efficacy of different rBCG constructions BALB/cCmedc female mice (6 weeks old) were vaccinated twice at 30‐day intervals and challenged 30 days after the last immunization with a dose of 500 bloodstream trypomastigotes of Tulahuén strain, equivalent to the 50% lethal dose (LD₅₀). Mice were randomly allocated to eight groups of five to six animals each. The groups are described in Table 2. Parasitemia was monitored at days 14, 21 and 28 post‐infection by examining 5 μl of blood under microscope, as previously described [11]. Survival was recorded daily until day 180 post‐immunization. A second round of experiments to validate the efficacy of the construction pUS2000‐NT‐TS was performed with a dose of 1000 parasites equivalent to the 100% lethal dose (LD₁₀₀). These challenges were performed in pUS2000‐NT‐TS, BCG Pasteur and PBS groups of mice, and parasitemia and survival were recorded as indicated above. The animals used in all experiment procedures were obtained from the Litoral Institute of Veterinary Sciences, National University of Litoral (ICIVET‐CONICET UNL), Argentina. The Animal Care and Use Committee, according to the institutional guidelines, approved all protocols for animal studies (Resolution no. 7/15). ICIVET is certified according to the ISO 9001 standards and declared by the Argentine Accreditation Organization (OAA) in accordance with the requirements of Good Laboratory Practices (GLP) to conduct non‐clinical studies, with specific scope for pharmaceutical products (http://www.cmc.unl.edu.ar/docs/broshure_english_web.pdf).

Humoral immune response determination

To evaluate serum antibody levels, samples of blood were obtained from mice of all groups before immunization, 30 days after the first immunization and 30 days after the second immunization. The antibody response was evaluated by indirect enzyme‐linked immunosorbent assay (ELISA) using purified recombinant TS and CZ proteins, as previously described [14]. Each protein was used separately in a concentration of 500 ng per well, diluted in carbonate–bicarbonate buffer, pH 9.6. The ELISA plates were blocked with 5% phosphate‐buffered saline (PBS) bovine serum albumin (BSA), and mouse sera were added at a 1 : 50 dilution in 1% PBS–BSA for 1 h at 37°C. Anti‐mouse immunoglobulin (Ig)G antibody conjugated to peroxidase (Abcam, Cambridge, MA, USA) was diluted (1 : 6000) and incubated for 1 h at 37°C. ELISA wells were washed with PBS‐Tween 20 0·05% was performed three times between all steps. Reaction samples were read at 450 nm in an ELISA reader (Bio‐Tek Instruments, Winooski, VT, USA) after incubation with 50 μl ready‐to‐use trimethylbenzidine (Invitrogen, Carlsbad, CA, USA) and stopped by adding 25 μl of 2 N H₂SO₄.

Trypanolytic activity assay

To assess trypanolytic activity obtained with pUS2000‐NT‐TS vaccine, antibodies obtained 30 days after the second immunization were evaluated and compared with antibodies obtained at the same time in mice from the BCG Pasteur and PBS formulations. Bloodstream trypomastigotes (5 × 10⁴) were incubated for 6 h at 37°C with each serum sample (diluted 1 : 20) and a fresh human serum as a source of complement in a final volume of 50 µl RPMI‐1640 (gibco, Carlsbad, CA, USA). The number of live trypomastigotes was determined using a Neubauer chamber. Lysis (0%) was determined based on the number of parasites incubated only with RPMI and complement.

Delayed‐type hypersensitivity (DTH)

A DTH test was performed 30 days after the last immunization in mice of all groups by intradermal challenge with 5 μg of TS or CZ in the right hind footpad 28 days after completion of the immunization schedule. Hind footpad thickness was measured before antigen injection and after 48 h with a Vernier caliper (Stronger). Results were expressed as the increment in millimeters of footpad thickness induced by inoculation.

Spleen cell culture and cytokine determination

To evaluate the cellular immune response with pUS2000‐NT‐TS, BCG or PBS, mice were vaccinated twice at 30‐day intervals, then euthanized 14 days after the last immunization dose to analyze ex‐vivo cytokine production of splenocytes stimulated with TS. Briefly, spleens were aseptically harvested and homogenized. Red blood cells were eliminated and splenocytes were resuspended in RPMI‐1640 medium (gibco) supplemented with 10% fetal bovine serum, 2% penicillin (100 μg/ml), streptomycin (100 U/ml) and 0·4 mM 2‐mercaptoethanol. Cells (1 × 10⁶/ml/well) were cultured in 48‐well plates (Nunc/Thermo Fisher Scientific) in supplemented RPMI. Splenocytes were stimulated with TS (10 µg/ml) obtained as indicated in item 2.5 or RPMI as control. Concanavalin A (2·5 μg/ml) was used as the positive stimulation control. After 48 h at 37°C and 5% CO₂, cells were incubated with 75 ng/ml of 12‐myristate 13‐acetate (PMA; Sigma‐Aldrich) and monensin (BD Pharmingen, San Jose, CA, USA), according to the manufacturer’s instructions. Four hours later, cells were washed twice with PBS, incubated with anti‐Fc III/II receptor antibody for 30 min and stained with fluorescein isothiocyanate (FITC) anti‐CD4, peridinin chlorophyll cyanin 5·5 (PerCP Cy5·5) anti‐CD8 and phycoerythrin (PE) anti‐CD107α for 30 min. Then, cells were washed and stained for allophycocyanin (APC) anti‐interferon (IFN)‐γ, PerCP Cy5·5 anti‐interleukin (IL)‐17, PE anti‐IL‐10 using the IFN‐γ kit for intracellular staining from BD Pharmingen. To assess the proliferative capacity, cells marked with CD4 were also intracellularly stained with PerCP Cy5·5 anti‐Ki67, using the forkhead box protein 3 (FoxP3) kit from Miltenyi Biotec (Bergisch Gladbach, Germany), according to the manufacturer’s instructions. Samples were analysed on an Attune NxT cytometer (Invitrogen) and analyses were performed using FlowJo software (TreeStar, Inc., Ashland, OR, USA).

Histopathology

Hearts from immunized and T. cruzi‐infected mice were removed on day 180 post‐immunization. Mouse hearts were fixed in buffered formalin and embedded in paraffin. Sections (5 µm) were then stained with hematoxylin and eosin to evaluate tissue parasitism and cardiac lesions, as previously reported [11]. Briefly, cardiac tissue was examined in blinded experimental groups and scored as follows: normal tissue (score 0); mild foci = slight infiltration with damage of one or two myocardial fibers (score 1); moderate‐sized foci = aggregated infiltrates compromising three to five muscle fibers (score 2); and intense foci = heavy accumulation of mononuclear cells with destruction of more than five muscle fibers (score 3). In addition, mouse hearts were also stained with picrosirius red staining to evaluate fibrosis, and were classified as follows: normal tissue (score 0); slight myocardial fibers (score 1); moderate myocardial fibers (score 2); and intense myocardial fibers (score 3). A total cardiac tissue damage score by combining these measures was determined as follows: score of hematoxylin and eosin staining × score of picrosirius red staining.

Statistical analyses

Data were analyzed using non‐parametric tests (Kruskall–Wallis test for k samples followed by Mann–Whitney U‐test for comparisons between two samples). The Mantel–Cox log‐rank test was used to evaluate survival curves. All analyses were performed using GraphPad Instat version 4.0 software (GraphPad, San Diego, CA, USA). The between‐group significant differences are indicated as *P < 0·05 and **P < 0·01.

Results

TS and CZ proteins fragments of T. cruzi were successfully expressed in rBCG strains

Cloning of TS and CZ fragments into pUS977 and pUS2000 resulted in the construction of six recombinant plasmids. Expression of the three proteins in rBCG using pUS977 and pUS2000 vectors was determined by Western blot (Fig. 1) in the sizes expected (pUS977‐NT‐TS = 27 kDal; pUS2000‐NT‐TS = 27 kDal; pUS977‐CT‐TS = 32 kDal; pUS2000‐CT‐TS = 32 kDal; pUS977‐CZf = 24 kDal; pUS2000‐CZf = 24 kDal).

Fig. 1.

Western blot demonstrating the expression of trans‐sialidase (TS) and cruzipain (CZ) antigens by Mycobacterium bovis bacille Calmette–Guérin (BCG). Characterization of the fractions COO‐terminal trans‐sialidase (CT‐TS) and NH2‐terminal trans‐sialidase (NT‐TS) (a) and the cruzipain fraction (CZf) (b) expressed in rBCG under control of two different mycobacterial promoters. (a) Lane 1, BCG Pasteur wild‐type; lane 2, rBCG transformed with COO‐terminal trans‐sialidase in pUS2000 (pUS2000‐CT‐TS); lane 3, rBCG transformed with COO‐terminal trans‐sialidase in pUS977 (pUS977‐CT‐TS); lane 4, rBCG transformed with pUS2000‐NT‐TS: NH2‐terminal trans‐sialidase in pUS2000 (pUS2000‐NT‐TS); lane 5, rBCG transformed with pUS977‐CT‐TS; lane 6, molecular weight marker. (b) Lane 1, rBCG transformed with cruzipain fraction in pUS2000 (pUS2000‐CZf); lane 2, rBCG transformed with cruzipain fraction in pUS977 (pUS977‐CZf); lane 3: molecular weight marker.

Mice immunized with rBCG strains showed humoral and cellular immunological parameters

The humoral response in all immunized groups presented a low level of antibody production (Fig. 2). However, on day 60, specific total IgG production rates showed statistical significance between the groups vaccinated with rBCG (P < 0·001) and the control groups (Fig. 2a,b). To analyze the specific T cell responses of the rBCG strains in vivo, a footpad assay for DTH was performed. As shown in Fig. 2c,d, all mice immunized with rBCG strains and challenged with each specific respective antigen showed a marked increase in footpad thickness compared with control groups (P < 0·01).

Fig. 2.

Immunological parameters of the immune response elicited by immunization with recombinant bacille Calmette–Guérin (BCG) strains. (a,b) Total immunoglobulin (Ig)G antibody levels in mice immunized with different vaccine formulations. The results are presented as mean ± standard deviation (bars) of the optical densities of the indirect enzyme‐linked immunosorbent assay (ELISA) for each experimental group. (a) Total IgG antibody from mice immunized with rBCG strains that express N‐ and C‐terminal trans‐sialidase (TS) fractions. (b) Total IgG antibody from rBCG strains of cruzipain (CZ). Blood was collected and evaluated on days 0 (preimmunization), 30 and 60 post‐immunizations. (c,d) Delayed‐type hypersensitivity test in all immunized mice. (c) Footpad thickness was measured before and 48 h after inoculation of 5 µg of trans‐sialidase (TS). (d) Footpad thickness was measured before and 48 h after inoculation of 5 µg of CZ. Results are expressed as ‘delta H’: the difference between the values obtained after and before inoculation. Data [mean ± standard deviation (s.d.)] represent one of three independent experiments (n = 5/group, *P < 0·05, **P < 0·01, ***P < 0·001).

pUS2000‐NT‐TS induced the best protection in vaccinated mice against an infection challenge with parasites

The immunoprotective potential of rBCG strains against T. cruzi was evaluated in terms of survival for 120 days after challenge. All immunized mice were infected with a dose of parasites able to kill 40–60% of non‐immunized mice. The results of parasitemia and survival rates are shown in Table 3. pUS2000‐NT‐TS in the rBCG strain achieved the highest protection percentages (80%) with the lowest parasitemia. Then, a 100% lethal dose of parasites was administered to mice immunized with pUS2000‐NT‐TS and significant survival rates were observed after challenge (P < 0·01) (Fig. 3a). In addition, the sera ability of immunized mice to lyse the trypomastigote form of the parasites was found to be significantly increased (Fig. 3b). Based on these results, subsequent experiments were performed to more clearly characterize pUS2000‐NT‐TS vaccine, as described below.

Table 3.

Parasitemia and survival in immunized BALB/c mice

	Parasitemia			Survival 120 dpi (%)
	14 dpi	21 dpi	28 dpi
pUS977‐NT‐TS	49 (12‐80)	74 (29‐120)	50 (0‐100)	3/5 (60)
pUS2000‐NT‐TS	16 (8‐30)	34 (17‐52)	14 (6‐19)	4/5 (80)
pUS977‐CT‐TS	76 (42‐140)	120	150	1/5 (20)
pUS2000‐CT‐TS	49 (14‐100)	93 (23‐150)	59 (2‐120)	3/5 (60)
pUS977‐CZf	99 (77‐125)	‐	‐	0/5 (0)
pUS2000‐CZf	47 (11‐100)	43 (14‐22)	35 (0‐70)	2/5 (40)
BCG Pasteur	48 (24‐102)	12 (13‐20)	7 (6‐8)	2/5 (40)
PBS	32 (9‐54)	96 (24‐140)	65 (200‐0)	2/5 (40)

Data correspond to one representative immunization and challenge experiment. Bacille Calmette–Guérin (BCG) or phosphate‐buffered saline (PBS) were used as negative controls. Two weeks after the last injection, mice were challenged with 500 bloodstream trypomastigotes of Trypanosoma cruzi (n = 5/group). Parasitemias were studied by direct microscopic observation under standard conditions at 14, 21 and 28 days post‐infection. Results were expressed as median (rank) of parasites/50 microscopic fields (×400). The survival after a final‐point 120 dpi is indicated, showing the rate of survival during acute and chronic infection. Data represent one of three independent experiments. NT‐TS = NH2‐terminal trans‐sialidase; CT‐TS = COO‐terminal trans‐sialidase; CZf = cruzipain fraction.

Fig. 3.

(a) Survival rates of immunized mice after a high‐dose Trypanosoma cruzi challenge. Mice immunized with NH2‐terminal trans‐sialidase in pUS2000 (pUS2000‐NT‐TS), bacille Calmette–Guérin (BCG) Pasteur and phosphate‐buffered saline (PBS) (n = 5/group) were challenged with 1000 bloodstream trypomastigotes of T. cruzi 14 days after the last immunization. (b) Trypomastigotes lysis assay was performed using serum from pUS2000‐NT‐TS, BCG Pasteur and PBS mice. **P < 0·01. The results represent one of three independent experiments.

pUS2000‐NT‐TS immunization induced protection against tissue damage

Histological analysis during the chronic phase of pUS2000‐NT‐TS‐vaccinated mice revealed a decrease of heart inflammatory infiltrates compared to the PBS group of mice, which presented diffused inflammatory foci in the heart (histological score = 2–4, Fig. 4a–d). Picrosirus red staining of tissues revealed significantly reduced fibrosis in vaccinated/chronically infected mice, with a reduction in the fibrotic area in the heart tissue compared to the BCG‐ and PBS‐vaccinated chronically infected control mice (Fig. 4e–h, P < 0·05, Fig. 4i). The score obtained for pUS2000‐NT‐TS‐vaccinated mice was significantly reduced (Fig. 4i), indicating that vaccination was effective in controlling the immunopathology and fibrosis in chronic chagasic mice.

Fig. 4.

Analysis of the cardiac lesions in immunized and control mice. (a–d) Hematoxylin and eosin (H&E) staining (blue: nuclear staining, pink staining: muscle cytoplasm) of heart tissue sections harvested at 120 days post‐infection from immunized and control mice. (e–h) Picrosirus red staining of hearts after 120 days post‐infection (green: muscle/cytoplasm, red: collagen fibers). (i) Total cardiac tissue damage score combining both stainings at 120 days post‐infection *P < 0·05.

pUS2000‐NT‐TS immunized mice showed immunological parameters that included components of a Th1 response

To examine the pUS2000‐NT‐TS T cell profile, splenocytes were gated for CD4⁺ T and CD8⁺ T cells before and after in‐vitro stimulation with recombinant TS. After antigen priming, CD4⁺ T cells were analyzed for proliferative capacity (Ki67⁺) and intracellular cytokine profile (IFN‐γ, IL‐17 and IL‐10). Before antigen stimulation of the splenocytes, percentages of CD4⁺ T (range = 12·2–16·7%) and CD8⁺ T (range = 5·55–7·85%) cells were comparable in vaccinated and control mice (Fig. 5a). After antigen stimulation, CD4⁺ T cells (Ki67⁺) from pUS2000‐NT‐TS‐immunized mice exhibited a significant increase (Fig. 5b, P < 0·01–0·001). Many of the mice immunized with the pUS2000‐NT‐TS and BCG Pasteur exhibited an increase in intracellular IL‐17 cytokine within CD4⁺ T cells (Fig. 5c, P < 0·01, 0·05). Conversely, no differences in the IL‐10 within CD4⁺ T cells were observed between pUS2000‐NT‐TS immunized mice and PBS‐inoculated mice (Fig. 5d). Regarding IFN‐γ production, a higher percentage of intracellular IFN‐γ was detected in CD4⁺ and CD8⁺ cells from pUS2000‐NT‐TS immunized mice than in PBS‐inoculated mice. (Fig. 5e,f). Finally, CD107a⁺ was significantly increased in CD8 T cells from vaccinated mice, indicating the potential cytotoxic activity from these cells.

Fig. 5.

Splenic T cells cytokine profile elicited by NH2‐terminal trans‐sialidase in pUS2000 (pUS2000‐NT‐TS) immunization. Mice were immunized and splenocytes were harvested 2 weeks after last immunization. (a) Splenocytes were incubated with fluorescein isothiocyanate (FITC)‐conjugated anti‐CD4 and peridinin cyanin 5·5 (PerCPCy5·5)‐conjugated anti‐CD8 antibodies, and T cell subsets were monitored by flow cytometry. (b–g) Splenocytes were stimulated in vitro for 48 h with recombinant antigens. The mean percentage of FITC⁺CD4⁺T cells that were Ki67⁺ (PerCPCy5·5, b), IL‐17⁺ (PerCPCy5·5, c), interleukin (IL)‐10⁺ [phycoerythrin (PE), d] and interferon (IFN)‐γ⁺ [allophycocyanin (APC), e] is shown. The mean percentage of PerCPCy5·5⁺CD8⁺T cells that was IFN‐γ (APC, f) and the mean percentage of antigen‐specific PerCPCy5·5⁺CD8‐PE⁺CD107a⁺ T cells that was IFN‐γ⁺ (PE, g) are shown.

Discussion

Several reports have described the ability of rBCG as a delivery platform for subunit protein vaccines and the immune protection conferred against various pathogens [3, 20, 21]. This system induces the T cell immune response necessary to fight against intracellular pathogens [3]. In the case of T. cruzi infection, this profile of immune response is desirable to control the infection [5]. In this work, different rBCG vaccines expressing fragments of the TS and CZ antigens were constructed and evaluated. In the case of TS, we expressed a C‐terminal fraction of the protein, not including the shed acute‐phase antigen (SAPA) region, which has a decoy effect for the immune response [22]. In a previous work we described that this fragment, expressed in E. coli and administered with cage‐like particles, allowed immune protection [12]. Conversely, the N‐terminal fragment of the TS protein has not been previously evaluated alone in any vaccine; therefore, this work is the first study, to our knowledge, of its antigenic performance. In the case of CZ, the N‐terminal fraction was used, as the best performance of this CZ region was previously described when compared with the C‐terminal region [15]. When the clones were inoculated in mice, the antibody response was low but significant for all clones with respect to the controls. The humoral response obtained is compatible with several previous studies performed for rBCG expressing antigens from other pathogens, in which the IgG response was reported to be low [21, 23]. In contrast, a strong DTH response was obtained, indicating a cellular activation compatible with the type of BCG response. These results confirm the in‐vivo expression of the proteins for each construction. In a previous work, using whole TS protein (excluding SAPA) or whole CZ antigens in a subunit vaccine, the best protection was obtained for the TS antigen [14]. In the present study, although the immune response was similar for TS or CZ antigens, both fragments of TS antigens had better rates of survival than the CZ antigen when challenges were performed in accord with our previous study.

When mice were challenged with a 50% lethal dose of the parasite, pUS2000 tended to be more protective than pUS977. This result agrees with the report by Seixas et al. [17], who found that both vectors triggered specific immune responses and protection against Leptospira, although pUS2000 gave better results. This difference could be due to the fact that the promoter of pUS2000 allows greater expression in vivo. This vector contains the M. leprae 18‐kDa antigen gene promoter and reaches low levels of expression in standard cell cultures but higher levels in macrophage cultures [16]. Indeed, a weakness of the rBCG platform is that the in‐vivo expression of the recombinant protein cannot be determined; however, the in‐vitro expression may indicate an order of magnitude of the amount of protein that different constructions can express [24]. Dennehy and Williamson [24] reviewed different reports in which the expression levels at a dose of 10⁶ colony‐forming units (CFU) produce 1–20 ng of proteins, and up to 100 ng for the highest reported expressions. Then, in comparison with our previous works in which 10 µg of protein were used as subunit formulations in vaccine candidates [14], the high level of response obtained with the rBCG platform is noteworthy. Based on this result, we hypothesize that the use of alternative expression systems to improve the level of expression in vivo would be able to generate total protection against T. cruzi infection.

Regarding the vector–antigen combination, NT‐TS in pUS2000 provided the best protection. This result may be explained by findings of Hoft et al. [10], who showed that the presence of the immunodominant INOVSQUY epitope of TS is critical for the immune protection of BALB/c mice against T. cruzi infection [25]. This epitope is present in the NT‐TS but not in CT‐TS fragments. The high protection elicited by NT‐TS‐pUS2000 clone was confirmed by the low mortality obtained when mice were challenged with a lethal dose of parasites. This high protection correlated with the increased level of trypanolysis that the sera of vaccinated mice were able to promote. This result is in accord with another report, showing that the quality of the antibody response is far more important than the titers of antibodies in a T cruzi vaccine candidate [26]. This trypanolysis is as likely to be meditated by complement as in the absence of complement; we did not obtain a significant lysis (data not shown). As well as increasing the survival of infected mice, the vaccine reduced the cardiac lesions that are produced during the chronic phase of the infection. This effect has also been described using other platforms of vaccines to control T. cruzi infection [11, 12, 27, 28].

To determine the involvement of the cellular immune response in the protection conferred by the vaccine formulation, we analyzed the specific T cells triggered by the immunized mice with NT‐TS‐pUS2000. Both CD4⁺ and CD8⁺ T subpopulations are known to play a pivotal role in controlling T. cruzi infection [13]. When cellular proliferation of CD4⁺ cells stimulated with the specific antigen was analyzed by staining with anti‐Ki67, a significant proliferation of cells was determined. IFN‐γ and IL‐17 levels were also increased in CD4⁺ T cells stimulated with the specific antigen. Together, these results indicate the activation of CD4⁺ T cells with both a Th1 and Th17 profile. The production of IFN‐γ by CD4⁺ T cells is important for the control of the parasite by means of activated macrophages. Moreover, IFN‐γ production by TS‐experienced CD4⁺ T cells is required for optimal activation and IFN‐γ production by CD8⁺ T effector cells [29, 30, 31, 32]. Accordingly, in pUS2000‐NT‐TS mice the frequency of IFN‐γ produced by CD8⁺CD107⁺ T cells was increased after TS stimulation. This IFN‐γ increase suggests that this vaccine candidate favors the priming and activation of CD8 lymphocytes in a polyfunctional way, which is critical to fight T. cruzi infection [27]. Regarding the Th17 profile that was obtained together with the Th1 profile in the present work, it should be noted that mixed profiles of Th1 and Th17 triggered by the BCG vaccine were reported in mice [33] and humans [34]. Interestingly, a Th17 response profile was recently found to be critical for protection against T. cruzi [25, 35]. Regarding IL‐10, CD4⁺ T cells stimulated with the specific antigen increased their expression of the cytokine, although the levels were not significant in relation to the control mice. Then, NT‐TS‐stimulated splenocytes from infected pUS2000‐NT‐TS mice increased the levels of IFN‐γ, IL‐17 and, to a lower extent, those of IL‐10 in vaccinated mice with respect to control, suggesting a highly inflammatory but controlled response. This type of response is currently considered a good predictor of success for a T. cruzi vaccine [36, 37, 38]. This profile of the immune response is compatible with the reduction of lesions observed in histological analysis of the heart in the vaccinated mice. Consequently, it can be hypothesized that a similar response pattern would be desirable and achievable in human vaccines.

We conclude that, despite a low level of expression of the antigen in the rBCG platform, the protection was extremely significant. We hypothesize that new plasmids that might improve the expression in rBCG would achieve a very high protection against T. cruzi infection. There are several alternatives to improve expression, such as the use of episomal multicopy plasmids, inducible promoters, BCG codon optimization, strong promoters or fusion proteins [20, 21]. As demonstrated for HIV, there is still a long way to go to achieve a satisfactory heterologous expression in rBCG [20]. Evaluation of all the variants in combination is an almost impossible task for a single working group. However, the acquisition of information and the description of better expression systems will allow us to improve the response obtained with this vaccine. An improvement in the expression of recombinant proteins will allow us to extend the study of the T. cruzi vaccine candidate to other models of T. cruzi strain/mouse strains.

Disclosures

None of the authors have any conflicts of interest in relation to the content of the present work.