Virus-like particles containing multiple antigenic proteins of Toxoplasma gondii induce memory T cell and B cell responses

Although the efforts to develop vaccine against Toxoplasma gondii infection were made for decades, there is currently no licensed vaccine available for humans. Upon discovering a number of T or B cell epitope regions from T. gondii IMC, ROP18 and MIC8, multi-antigen VLPs or combination VLPs were generated. Mice immunized with multi-antigen VLPs or combination VLPs were challenge infected with T. gondii (ME49). T. gondii-specific IgG, IgG isotypes and IgA antibody responses, memory T and B cell responses and protection were evaluated. All the mice survived upon T. gondii challenge infection by multi-antigen VLPs vaccination. Vaccinated mice elicited higher levels of parasite-specific IgG and IgG2a antibody responses in sera, IgA antibody responses in feces, CD4+ and CD8+ T cell responses, and cytokines (IFN-γ, IL-10) responses compared to combination VLPs. In particular, the multi-antigen VLPs vaccination showed significantly higher levels of antibody secreting cell (ASC) responses, CD4+ and CD8+ effector memory T cells, and memory B cells than combination VLPs. Multi-antigen VLPs vaccination showed significant reduction of brain cyst counts and size, and all mice survived. Prediction and analysis of epitopes have indicated that IMC, ROP18 and MIC8 showed partially overlapping epitopes for T and B cells. Our results indicated that antibody responses, memory T and B cells induced by multi-antigen VLPs vaccination might contribute to the complete protection upon T. gondii (ME49) challenge infection.


Introduction
Toxoplasma gondii, a notorious protozoan parasite belonging to the phylum Apicomplexa, is capable of infecting virtually all vertebrate hosts using various transmission routes to cause the zoonotic disease toxoplasmosis [1,2]. Although large variations in the prevalence rates of human toxoplasmosis do exist between countries, it is presumed that approximately third of the world's population have been infected by T. gondii [1,3]. Clinical symptoms associated with the disease in immunocompetent individuals are often asymptomatic or of non-specific origin, which includes myalgia, fever, and other flu-like symptoms [1,2]. However, T. gondii infection can have severe health consequences in pregnant individuals, as these parasites can a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 traverse through the placenta to cause premature abortion and other congenital defects [2,3]. Therapeutic regimen for human toxoplasmosis requires the use of pyrimethamine and sulfadiazine, but side effects and insufficient efficacies against non-tachyzoite stages of the parasite limits their use [4]. Toxovax is currently the only available commercial toxoplasmosis vaccine, albeit being limited to veterinary use with arising safety concerns [5]. These issues, combined with other detriments associated with the treatment, may have created an impetus for the development of a novel vaccine which could effectively block and control the transmission of toxoplasmosis.
The importance of bioinformatics and its growing usage for epitope predictions and vaccine design strategies cannot be overstated. Several T. gondii vaccine studies have already performed in silico epitope analyses of multiple candidate antigens, which may substantially contribute to future vaccine design strategies [6][7][8][9].
Vaccination induced immunological memory responses are critically important in inducing protection against the same pathogen recognized by immune system [10,11]. Naïve CD4 + T cells recognize antigen-MHC complexes and proliferate and differentiate to effector T cells, which provide immediate protection [12]. Although most of the effector T cells subsequently die by apoptosis, a subset of antigen-specific T cells will persist in immune system as memory T cells once pathogens have been eliminated from the host [13]. Multiple memory T cell subpopulations, including but not limited to central memory T cells (T CM ) and effector memory T cells (T EM ), have been identified in humans as of current which can be distinguished based on CD45RO and CD45RA isoform expressions [12]. The T CM shows self-renewal potential and resides in secondary lymphoid organs but lacks effector function, whereas T EM possess immediate effector functions and can rapidly immigrate to peripheral tissues to provide antigen elimination [14]. Increased central memory lymphocyte response induction was observed in cattle vaccinated against the parasite Theileria parva using Tp1 antigen post-challenge [15]. Memory B cells (MB) and plasma cells are the key for maintaining humoral immune response [16]. Microneedle delivery of influenza vaccines have been reported to induce a durable, antigen-specific MB and plasma cell responses in mice [17,18].
Recombinant protein and DNA vaccine studies using potential candidate antigens have been conducted extensively in the past [19][20][21]. Yet, the vaccine efficacies in the aforementioned studies were extremely limited and complete protection was not conferred in mice [22]. Our previous works using virus-like particle vaccines containing single IMC, ROP18, MIC8, ROP4 proteins or multiple proteins have conferred complete protection against T. gondii [22][23][24][25][26]. These studies mainly focused on inducing T. gondii-specific IgG, IgA antibodies, CD4 + T cells, CD8 + T cell responses, and protections in mice [22][23][24][25][26]. Exemplified by earlier works, multi-antigen containing VLPs showed higher number of VLPs particles per μm 2 , higher levels of CD4 + T cell, and germinal center B cell responses, which resulted in better protection [25]. To date, immunological memory T or B cell responses induced by single or multiple proteins containing VLPs, T or B cell epitopes in VLPs and antibody secreting cell response induced upon T. gondii ME49 challenge infection has yet to been reported. As such, in this study, we report the memory T and B responses, T or B cell epitopes, antibody secreting cell (ASC) responses and protections induced by multi-antigen VLPs and combination VLPs upon T. gondii ME49 challenge infections in mice.

Ethics approval and consent to participate
All experimental procedures involving animals were reviewed and approved by Kyung Hee University IACUC (permit number: KHUASP (SE) -18-050). Animals were housed in pathogen-free animal facility with easy access to food and water. All of the researchers involved in the study were trained for proper animal handling.

Prediction of T cell and B cell epitopes
Amino acid sequence of T. gondii IMC (accession number: ADV15617), ROP18 (accession number: CAJ27113), and MIC8 (accession number: AAK19757) were retrieved from NCBI GenBank sequence database. Prediction of T cell and B cell epitopes of IMC, ROP18 and MIC8 was performed using IEDB web server (http://tools.iedb.org). MHC binding affinity, rescale binding affinity, C-terminal cleavage affinity, and transport associated with antigen processing (TAP) transport efficiency propensity scores were calculated for T cell epitope prediction. B cell epitopes were predicted based on the propensity scale method, which evaluates biophysical parameters of the amino acids such as linear epitope, beta-turns, surface exposure, flexibility, antigenicity, and hydrophilicity.

Immunization and challenge infection in mice
Seven weeks old BALB/c mice were divided into groups (n = 20 per group) for immunization with either multi-antigen VLPs or combination VLPs. On week 0 and 4, mice were intranasally (IN) immunized with 90 μg of multi-antigen VLPs whereas three VLPs expressing IMC, ROP18, or MIC8 were combined at equal ratio (30 μg from each VLP, total 90 μg protein) for combination VLP immunization. Mice were challenge-infected with 500 bradyzoites of ME49 strains via oral route 4 weeks after receiving second immunization. Ten mice were sacrificed from each group at week 4 post-challenge for brain and spleen collection as previously described [22] and brain tissues were prepared for brain cyst counting as previously described [23]. Remaining mice were monitored daily to observe changes in body weight and survival until death. A total of 70 mice were used in the present study, of which 55 were sedated and humanely euthanized by CO 2 inhalation upon loss of 20% or more body weight post-infection. The entire experiment lasted 3 months, which involved 2 months of immunization phase followed by 1 month of challenge infection period. All efforts have been made to minimize animal suffering and distress, in compliance with the animal welfare policy.

T. gondii-specific antibody responses in sera and feces
Sera and feces of mice were collected from all groups 4 weeks post-challenge infection. Fecal antibodies were collected by immersing feces with 500 μl 0.1 M PBS and incubating it at 37˚C for 1h, which were subsequently centrifuged at 2,000 rpm for 10 min. Harvested supernatants were stored at -20˚C until use. T. gondii-specific IgG, IgG1, IgG2a and IgA antibody responses were determined by enzyme-linked immunosorbent assay (ELISA) as described previously [22]. Briefly, 96-well flat-bottom immunoplates were coated with 4 μg/mL T. gondii antigen in 100 μL of 0.05 M carbonate bicarbonate buffer (pH 9.6) and incubated overnight at 4˚C. The next day, each wells were incubated with 100 μL of serum samples (diluted 1:100 in PBST) for 2 h at 37˚C. HRP-conjugated IgG, IgG1, IgG2a, and IgA secondary antibodies were diluted 1:2000 in PBST and 100 μL of these were added to each well to determine T. gondii-specific antibody responses.

Antibody secreting cell (ASC) and cytokine analysis
For ASC assays, 96-well culture plates were coated with T. gondii antigen at 4˚C overnight, and splenocytes (10 6 cells/well) in RPMI-1640 culture media were added to coated plates after blocking. Secreted antibody levels were determined 3 days after incubation at 37˚C. To detect cytokines, splenocytes (10 6 cells/well) were cultured in 96-well culture plates with T. gondii antigen (5 μg/mL) or medium alone which was used as control at 37˚C with 5% CO 2 for 3 days. The culture supernatants were collected at 72 h for determination of IFN-γ and IL-10. The levels of IFN-γ and IL-10 were examined using BD OptEIA Set according to the manufacturer's instructions (BD Biosciences).

Statistics
Statistical analyses were performed using PC-SAS 9.3 software (SAS Institute, Cary, NC, USA). Data sets were expressed as mean +/-SEM and compared using One-way ANOVA with Duncan's post hoc analysis. A P value < 0.01 was considered statistically significant.

VLPs vaccination induces T. gondii-specific IgA, IgG, IgG1 and IgG2a antibody responses
To evaluate the level of T. gondii-specific antibody induced by multi-antigen VLPs or combination VLPs vaccines, we measured the IgG, IgG1 and IgG2a antibody responses in serum and measured the IgA antibody responses in feces upon challenge (Fig 2). Multi-antigen VLPs or combination VLPs immunized group (Multi-antigen Cha or Combination Cha) showed higher levels of T. gondii-specific IgG and IgG2a antibody responses compared to naïve challenged group (Naïve Cha) (Fig 2A and 2C, � P < 0.01). Multi-antigen Cha also showed significantly higher levels of T. gondii-specific IgA antibody responses in feces compared to Combination Cha or Naïve Cha ( Fig 2D, � P < 0.01).

VLPs vaccination induces CD4 + T cells and CD8 + T cells responses
To determine T cell responses in immunized mice upon challenge, flow cytometer analysis was performed using mouse spleen cells. As shown in Fig 3, at 4 weeks post-challenge, higher populations of CD4 + T cells and CD8 + T cells were found in multi-antigen VLPs (Multi-antigen Cha, � P < 0.01) or combination VLPs (Combination Cha, � P < 0.01) immunized mice compared to naïve challenge (Naïve Cha) control mice. Significantly higher levels of CD4 + and CD8 + T cell responses were found in Multi-antigen Cha compared to those in Combination Cha (Fig 3B and 3C, � P < 0.01).

T and B cell epitopes prediction and VLPs vaccines-induced memory cell responses
As shown in Table 1, many T cell and B cell epitopes from T. gondii IMC, ROP18 and MIC8 in VLPs were predicted by IEDB tools. The top score (more than 0.7) T cell epitopes (9-mer) were identified and the overlapped B cell epitopes (more than 5 parameters) were selected (Table 1).

VLPs vaccination induces cytokine responses
The splenocytes from VLPs immunized mice were cultured to determine the production of Th1 and Th2-like cytokines. Multi-antigen VLPs (Multi-antigen Cha) vaccination induced significantly higher levels of IFN-γ and IL-10 compared to combination VLPs (Combination Cha) or naïve challenged control mice (Naïve Cha) (Fig 5D and 5E, � P < 0.01). Higher level of IFNγ was detected than that of IL-10 (Fig 6D and 6E).

VLPs vaccinations significantly inhibit T. gondii replication and protect mice from T. gondii challenge infection
Mice immunized with VLPs were challenge infected with T. gondii (ME49) and the cysts of T. gondii in the brain were observed. As shown in Fig 7A, the lower number of cysts was found in multi-antigen VLPs immunized mice (Multi-antigen Cha) compared to combination VLPs immunized mice (Combination Cha) or naïve challenged control mice (Naïve Cha) (Fig 7A, � P < 0.01). Cysts size was found to be significantly reduced in Multi-antigen Cha and Combination Cha compared to Naïve Cha (Fig 7B and 7C, � P < 0.01). The mice were monitored daily to observe the survival rates and body weight changes upon challenge infection. As shown in Fig 8A and 8B, all mice immunized with multi-antigen VLPs showed 100% of survival whereas combination VLPs showed 50% and naïve mice showed 0% of survival. Multiantigen VLPs showed lower levels of body weight loss compared to combination VLPs at day 27 and 30 post-challenge infection.

Discussion
Vaccination is the most effective method in protecting the host against infectious diseases [30].
Although the efforts to develop vaccine against T. gondii infection were made for decades, efficacy issues have hindered its development [22,31]. T. gondii vaccine-induced memory T and B cell responses and related humoral and cellular immunity study are largely lacking. In the present study, multiple antigenic proteins containing influenza VLPs vaccine were found to be completely protective and correlated with memory T and B cell responses. The VLPs we generated targeted T. gondii multiple antigenic proteins, which possess T and B cell epitopes. A highly effective vaccine must be capable of eliciting strong antigen-specific antibody responses as well as ensuring development of persisting memory B and T cells [30,32]. Antibody production post-vaccination is of utmost importance, as their contribution to immunity and protection against pathogens are crucial for controlling disease progression [33]. In our current study, we found that multi-antigen VLPs elicited significantly higher levels of T. gondii-specific total IgG and IgG2a isotype antibody secreting cells cultured in vitro compared to combination VLPs, correlating with higher levels of IgG and IgG2a antibody responses in sera. The dominant Toxoplasma gondii vaccine targeting multiple antigens IgG2a isotype antibody response is known to be more effective than other isotypes in pathogen clearance [34][35][36], and their role in complement activation, antibody-dependent cellular cytotoxicity induction, as well as virus removal via macrophages have already been documented [37]. To investigate the vaccine efficacy, mice brain was collected at week 4 after challenge infection. T. gondii (ME49) were separated individually from mice brain and counted under the microscope. The number of brain cysts was significantly reduced in multi-antigen or combination VLPs immunized mice compared to naïve mice upon challenge infection (A, � P < 0.01). The cysts size was significantly reduced in Multi-antigen Cha (18 μm) or combination Cha (24.8 μm) compared to Naïve Cha (32 μm) (B and C, � P < 0.01).
https://doi.org/10.1371/journal.pone.0220865.g007 Although memory B cells may not confer direct protection against pathogenic diseases, they appear to be invoked upon depletion of pre-existing antibody levels which are required to prevent infection [30]. Antibody responses detected in sera upon VLPs vaccination were low [25] and increased upon T. gondii ME49 challenge infection in our current study. Memory B cells were found to be increased upon challenge infection, indicating memory B cell response was initiated because of low level of antibody response which were not able to provide protection against T. gondii infection as a critical first line of defense.
Protection against intracellular pathogens requires T cell-mediated responses [38]. T cell differentiation and memory and effector T cells play a significant role in immunity against pathogenic agents [39]. Upon exposure to pathogen, memory T cells from vaccination undergo rapid clonal expansion and induce effective immune response compared to response from the primary immune response [39]. Memory T cell populations have been further divided into "central memory" (T CM ) or "effector memory" (T EM ) subsets that differ by phenotype and function [40,41]. The importance of T CM or T EM subsets with regards to vaccinemediated protection against viral challenge infection remains elusive [30]. In our current study, we found that multi-antigen VLPs showed higher levels of both CD4 + and CD8 + effector memory T cells compared to central memory T cells, indicating that effector memory T cell responses play an important role in contributing the protection against T. gondii infection. Combination VLPs showed no significant increase in CD4 + and CD8 + effector memory T cells and central memory T cell responses which might contribute to the low levels of protection induced by combination VLPs. These results may be attributed to higher density or purity of T. gondii virus-like particles, which were observed in multi-antigen VLPs compared to combination VLPs [25].
Vaccines containing T or B cell epitopes can elicit both cellular and humoral immune responses, the T and B cell epitopes prediction is important in designing epitope-based vaccines [42][43][44][45]. In the current study, we found a number of T or B cell epitope regions from antigenic proteins of IMC, ROP18 and MIC8 from T. gondii by epitope prediction tools. Epitope prediction results outlined in Table 1 have revealed that MIC8 contained more T and B cell epitope regions than other antigens, indicating its expression in the VLPs might have contributed more to immunological protection than others. Since MIC8 VLPs vaccination has shown great protection against highly virulent T. gondii RH challenge infection [22], T and B cell epitope portions from MIC8 could be used for epitope-based vaccine designs.

Conclusions
Multi-antigen VLPs vaccination induced significantly higher levels of CD4 + CD8 + effector memory T cell and memory B cells compared to combination VLPs. T and B cell epitopes from IMC, ROP18 and MIC8 in multi-antigen VLPs might contribute to induce memory T and B cells responses, which result in effective CD4 + , CD8 + T cell responses, antibody secreting cell responses, IgG, IgG2a, IgA antibody responses in sera or feces and complete protection against T. gondii ME49 challenge infection.