All animal protocols were reviewed and approved by the Animal Administration and Ethics Committee of Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences. The study was performed in strict compliance with the recommendations set forth in the Animal Ethics Procedures and Guidelines of the People’s Republic of China. All efforts were made to minimize animal suffering and to reduce the numbers of animals used in the experiments.

Toxoplasma gondii-seronegative 8-week-old female and male Kunming mice were purchased from Lanzhou University Laboratory Animal Center, Lanzhou, China. They were housed under pathogen-free conditions at Lanzhou Veterinary Research Institute in controlled room under stable conditions (12-h/12-h dark/light cycle, 50–60% humidity and 22°C temperature). Mice had free access to sterilized food and water ad libitum. Mice were acclimated for 1 week before use. Kunming mice were selected because of their susceptibility to acute and latent infection with T. gondii (14). In all experiments, mice (6–10 mice per group) were observed daily after immunization and challenge.

Tachyzoites of T. gondii type I RH mutant and wt strains and type ToxoDB#9 − PYS and TgC7 − strains were maintained in vitro in human foreskin fibroblast (HFF, ATCC SCRC-1041) monolayers as previously described (25). Freshly egressed tachyzoites were harvested from HFF monolayers by filtration through 3-µm polycarbonate membranes and suspended in sterile phosphate-buffered saline (PBS). T. gondii cysts of type II Prugniuad (Pru) strain were maintained in Kunming mice by oral passage of infectious cysts in mice as described previously (26). The number of parasites (tachyzoites or cysts) of each strain was adjusted in PBS to the final number used in the experiments.

Soluble Toxoplasma antigens were prepared as previously described (27). Briefly, a suspension of T. gondii RH tachyzoites was washed in PBS and subjected to repeated freezing and thawing cycles and then sonicated on ice at 60 W/s. The preparations were centrifuged at 14,000 g for 30 min at 4°C. The supernatants were sterile filtered with 0.22 µm sterile nitrocellulose filters (Sartorius) and the STAg concentration was determined by the Bicinchoninic acid kit (BCA, Sigma, St. Louis, MO, USA), and aliquoted was stored at −80°C until use.

Disruption of GRA17 gene in RH strain has been shown to affect the parasite’s ability to proliferate in cultured cells and its virulence in mice (24). We used CRISPR-Cas9 method to delete GRA17 gene in T. gondii RH strain as described previously (28, 29). Briefly, we generated a single-guide RNA (sgRNA) to disrupt GRA17 gene and modified pSAG1:CAS9-U6:sgUPRT by PCR mutagenesis using specific primers (see Table S1 in Supplementary Material). The DHFR* resistance cassette was amplified as described previously (28, 29). Then, GRA17-specific CRISPR-Cas9 plasmid and the DHFR* resistance cassette amplicons were combined and transfected into T. gondii RH tachyzoites. After selection with 3 µM pyrimethamine, single clones were screened by PCR using specific primers (KO-GRA17-F and KO-GRA17-R) listed in Table S1 in Supplementary Material. Then, we compared the in vitro growth of ΔGRA17 deletion mutant to wt T. gondii tachyzoites. Briefly, confluent monolayers of HFF grown on glass coverslips positioned in the bottom of 6-well plates were infected with freshly purified tachyzoites (2 × 10³/well) of mutant or wild-type (wt) strain. Coverslips were washed to remove extracellular parasites at 3 h after infection, and the parasite’s growth within the PV was assessed at 12 and 24 h by fixation in 4% paraformaldehyde followed by imaging at ×40 using a Leica DM IRB microscope equipped with a high-resolution charge-coupled device camera (Nikon Japan).

Mice were immunized with 5 × 10⁴ ΔGRA17 tachyzoites or mock-immunized in a volume of 200 µl PBS intraperitoneally (i.p.) using a 26-gauge needle (Figure S1 in Supplementary Material). We used i.p. immunization because it can induce both mucosal and systemic immune responses. The type I RH strain was used because immunization with ΔGRA17 RH tachyzoites has been shown to protect mice from lethal infection (24).

We assessed the immune response in ΔGRA17-immunized mice prior to infection in order to determine the pattern of humoral and cellular immune responses of the vaccinated mice compared to non-immune control animals.

In these experiments, the efficacy of immunization with ΔGRA17 mutant strain was tested against acute, latent, and congenital infections in mice (Figure S1 in Supplementary Material).

All statistical analyses were performed using SPSS18.0 (SPSS Inc., Chicago, IL, USA). Levels of significance of the differences in the level of antibodies, cytokines, and parasite cyst burdens were compared using one-way ANOVA analysis (for comparing means between ≥three groups) or two-tailed, unpaired Student’s t-test (for comparing means between two groups). The SDs were derived from three independently performed experiments with three replicates per experiment for the in vitro assays. Statistical significance was set at P < 0.05. Mortality was examined by plotting survival curves of the different mouse groups stratified by T. gondii infection and immunization status using Mantel–Cox log-rank test.

We knocked out GRA17 gene to study whether GRA17-deficient RH tachyzoites can be used as a live-attenuated vaccine against acute, chronic, and congenital toxoplasmosis in mice. CRISPR-Cas9 system was used to delete the GRA17 coding sequence (Figure 1A). The sgRNA and DHFR* marker were inserted into the GRA17 coding sequence and stable, single clones were successfully generated and validated by PCR (Figure 1B). We observed a significant defect in the appearance of the PV, along with less intra-vacuolar tachyzoite’s proliferation in the ΔGRA17 mutant compared to the wt strain (Figure 2). Gray-level histogram analysis showed that the PV produced by wild T. gondii RH strain had a significantly larger fraction of darker pixels, which represent the large number of growing intra-vacuolar parasites, whereas a significantly lower number of parasites was observed in the PV produced by the ΔGRA17 mutant strain, as shown by a greater proportion of the lighter pixels (Figure 3).

The prechallenge sera from the immunized mice were monitored for their IgG subclasses and for Th1 and Th2 cytokine profiles. The presence of specific anti-T. gondii IgG antibodies at 28 and 70 days after immunization was determined by quantitative ELISA. At day 28 postimmunization, all immunized mice had seroconverted with a higher level of IgG antibodies compared with naïve mice (Figure 4). Next, we tested whether a Th1 and/or Th2 response was elicited in the immunized mice by evaluating the levels of specific antibodies, IgG1 and IgG2a. At 28 days postimmunization, the levels of IgG2a in immunized mice were significantly higher compared to control naïve mice. In contrast, levels of IgG1 did not increase in immunized mice compared with naïve mice (Figure 4). At 70 days postimmunization, the levels of IgG1 in immunized mice increased, and the levels of IgG and IgG2a remained higher compared to naïve mice (Figure 4). Cytokines released from spleen cells after stimulation with STAg were evaluated by ELISA. As shown in Figure 5, the levels of Th1-type cytokines (IFN-γ, IL-2, and IL-12) in ΔGRA17-immunized mice were significantly higher than that in naïve mice. In the interim, the level of Th2-type cytokine (IL-10) was significantly increased in supernatants from spleen cells of ΔGRA17-immunized mice compared to naïve mice.

Kunming mice, 70 days post ΔGRA17 immunization, were challenged with 1 × 10³ tachyzoites of T. gondii RH, PYS, or TgC7 strain. All immunized mice survived the challenge with the three strains. By contrast, all naïve mice challenged with RH, PYS, or TgC7 strain died within 10 dpi (Figure 6). Assessment of cytokine response after chgallenge with with RH was performed 7 dpi using ELISA. Analysis was performed on serum and peritoneal fluids from ΔGRA17-immunized and infected mice, non-immunized and infected mice, and non-immunized and uninfected mice. The levels of IFN-γ and IL-12 in the sera and peritoneal washes (Figure 7) were significantly higher in non-immunized and infected mice than in the immunized and infected, and non-immunized and uninfected mice.

Seventy days after immunization, both immunized and control mice were orally gavaged with 20 cysts of T. gondii Pru strain. All ΔGRA17-immunized and Pru-infected mice survived compared to only 40% survival of non-immunized and infected mice (Figure 8A). Parasite cyst burden in the brains 5 weeks post-challenge with T. gondii Pru cysts was determined. Infected mice had a larger number of cysts (4,296 ± 687 cysts/brain) compared with immunized mice that had significantly less cyst burden (53 ± 15 cysts/brain) (P < 0.05). Brain cysts were not found in 4 out of 10 (40%) immunized mice by microscopic examination (Figure 8B). Brain tissues of these four immunized mice that did not show any cysts were homogenized and bio-assayed in naïve mice. Anti-T. gondii IgG antibodies were found in two of these four naïve mice suggesting that two brain homogenates derived from immunized and infected mice are infected based on serological evidence and two were completely free of parasites as evidence by negative microscopic and serological results.

Immunization with ΔGRA17 was well tolerated during pregnancy as evidenced by the absence of abortion in immunized mice, number of pups per litter, survival rate, and BW of pups born to immunized dams. Also, it is important to note that an immunopathologic response was not induced by immunization with ΔGRA17 strain.

A major goal of vaccine development is the induction of adequate immune responses against antigens delivered by the mucosal route. ΔGRA17-immunized mice developed a high level of anti-T. gondii IgG at 28 and 70 days after immunization. This level of antibodies plays important role in protection against subsequent infection with T. gondii tachyzoites and in controlling T. gondii during the chronic phase of infection by preventing cysts’ reactivation (34). We next analyzed the antibody types elicited against immunization with ΔGRA17 mutant by studying the antibody subclasses (IgG1 and IgG2a). Our results showed that immunization with live-attenuated ΔGRA17 strain promoted both cell- (Th1) and humoral- (Th2) mediated immune responses in Kunming mice. Immunized mice exhibited IgG2a as the predominant antibody’s class during early infection followed by upregulation of IgG1 in the chronic phase, probably as a result of Th1 to Th2 switching. A high ratio of IgG2a to IgG1 antibody titers observed at 28 days after immunization indicated that Th1 immune response was induced. However, at 70 days after immunization the level of IgG1 antibody increased, suggesting the induction of a mixed Th1/Th2 immune response. These results demonstrated a sequential upregulation of antiparasite IgG2a and IgGl antibodies. This isotypes’ pattern is supported by the patterns of cytokines detected in the supernatants from spleen cell cultures where the levels of Th1-type cytokines (IFN-γ, IL-2, and IL-12) and Th2-type cytokine (IL-10) in ΔGRA17-immunized mice were significantly higher compared to naïve mice. The fact that IgG2a is known to be elicited by IFN-γ in mice (35) is consistent with our finding where a bias to Th1 response (i.e., higher IgG2a/IgG1) was correlated with early IFN-γ production, pointing to the important role of IFN-γ in eliciting high IgG2a responses against avirulent T. gondii, therefore, influencing the immunogenicity of this vaccine.

High levels of Th1-type cytokines including IFN-γ, IL-2, and IL-12 developed in immunized mice might have provided immune-protection to mice against challenge with a lethal dose of wt strains (type I RH or two ToxoDB#9 strains). Deaths occurred by days 7–10, with 100% mortality in non-immunized and infected mice, whereas all immunized mice survived. Seven days following a subsequent challenge with T. gondii RH strain a significantly higher (P < 0.001) level of IFN-γ and IL-12 was detected in the peritoneal washes and sera of non-immunized and infected mice compared to ΔGRA17-immunized and infected mice. This finding suggests that the response of the naïve and infected mice was of an inflammatory nature and that ΔGRA17-immunized and infected mice developed a balanced pro-(Th1) and anti-inflammatory (Th2) cytokine response. The remarkably high levels of IFN-γ and IL-12 and the associated over-inflammation are probably the basis for adverse health complications developed in naïve and infected mice as previously reported (36, 37). T. gondii infection is known to induce a Th1-biased inflammatory response by stimulating T cells, macrophages, natural killer cells, dendritic cells, and neutrophils to produce IFN-γ and IL-12 (38–41). These two cytokines are central to the development of immunity against T. gondii infection. Suppression of IL-2 by effector T cells triggered by T. gondii infection can cause Treg cells to lose their capacity to balance Th1 immune responses, eventually leads to immunopathogenesis and the death of mice during acute T. gondii infection (42). On the other hand, the Th2 cytokine IL-10 is essential for counterbalancing the exacerbated inflammatory Th1 response that induces immunopathologies (36). Therefore, the balanced level of Th1/Th2 cytokines observed in ΔGRA17-immunized and infected mice was adequate to control the dissemination of tachyzoites, but not excessive enough to elicit significant inflammation. The immune response triggered by ΔGRA17 immunization also fully protected mice against oral infection with type II Pru cysts. Our results showed that only 40% non-immunized mice and orally infected 20 Pru cysts survived, whereas all immunized mice (100%) survived. It is worth mentioning that although the number of cysts in immunized mice was significantly reduced, immunization did not prevent the establishment of chronic infection and cyst formation in 80% of the immunized mice.

We assessed the protection conferred by immunization against transplacental transmission and results showed that immunized mice infected with 200 tachyzoites of T. gondii RH strain on day 18 of gestation did not show any significant differences in the litter size, survival rate, or mean BW of pups at day 5 after birth compared to pups from non-immunized and uninfected mice. Also, immunization protected completely the dams form developing clinical signs of toxoplasmosis. In contrast, non-immunized and infected dams developed clinical disease and aborted. Furthermore, the litter size, survival rate, and mean BW of their pups were significantly reduced. The small litter size and low survival rate of pups born to non-immunized dams can be attributed to T. gondii infection because parasite DNA was detected in the aborted fetuses. The reduced BWs of pups may be caused by the limited care and postnatal malnutrition as a consequence of infection of the dams.

Likewise, in the immunized pregnant mice orally gavaged with Pru cysts on day 12 of gestation the litter size, survival rate and mean BW of the pups at day 35 after birth were not different from that of non-immunized and uninfected mice. In contrast, the litter size, survival rate, and mean BW of pups from non-immunized and Pru-infected mice were significantly reduced (P < 0.05). It can be argued that vaccination in pregnant mice with a live-attenuated parasite strain may be fraught with the possibility of transmission of infection to the fetus. However, although the immune responses triggered by ΔGRA17 immunization did not completely block the vertical transmission, a significantly less tissue cyst burden was observed in the brain of pups from immunized and Pru-infected mice. As expected, all pups of mock-immunized and Pru-infected dams had more brain cysts. With regard to the brain tissue cyst load in infected dams, there was a significantly larger number of brain cysts in non-immunized than immunized dams. These results indicate that immunization with ΔGRA17 can induce a partial protective immunity against subsequent i.p. infection with T. gondii tachyzoites RH strain during the third trimester of gestation and oral infection with Pru cysts during the second trimester of gestation.

Further analysis of the cytokines secreted by maternal spleen cells of immunized dams revealed higher levels of Th1-type cytokines (IFN-γ, IL-12, and IL-2) and Th2 cytokine (IL-10) in immunized and Pru-infected compared to non-immunized and infected mice, and non-immunized and uninfected mice. This cytokines’ pattern in immunized and infected mice is similar to that observed at 70 days after immunization in non-pregnant mice. But, T. gondii is known to stimulate a strong Th1 response characterized by the early classical activation of macrophages and the production of pro-inflammatory mediators, such as IL-12, IFN-γ, and nitric oxide. Th1 response can suppress the development of alternative activation of macrophages and, thus, Th2 responses. Thus, while high level of anti-T. gondii Th1 immune response is required to control the materno–fetal transmission of T. gondii (43) and prevent congenital transmission (44, 45), it can adversely affect the pregnancy (37).

The increased IL-10 level in our study is interesting because this cytokine plays a critical role in downregulating IFN-γ responses in C57BL/6 mice following T. gondii infection. IL-10-deficient C57BL/6 mice displayed elevated IL-12 levels and consequently increased IFN-γ and TNF-α responses, which resulted in hepatic inflammation and necrosis (36). Previous work showed a higher rate of materno–fetal transmission of T. gondii in IFN-γ-deficient C57BL/6 mice compared to wt C57BL/6 mice (43). On the other hand, the pregnancy outcome of C57BL/6 mice infected with T. gondii improved by treatment with recombinant IL-10 and deteriorated in IL-10-deficient mice compared to T. gondii-infected naïve mice (46). Increased level of the Th2-type cytokine, IL-10, restrains Th1 immune response and Th1-mediated immunopathology, while maintaining a Th2 immune response (47, 48).

During pregnancy, the maternal immune response is shifted toward a Th2 cytokine profile in order to accommodate the growing fetus and to ensure maintenance of the pregnancy. Therefore, while the high level of Th1-type cytokines in the immunized mother plays an important role in countering congenital transmission (44, 45), Th2-type cytokine IL-10 is needed to balance Th1 immune response in order to maintain the pregnancy. In agreemnt with these findings, we observed that pregnancy was not successful in mice mated 1 month after immunization where Th1 response was dominating. The role of IL-10 in preventing immunopathology during toxoplasmosis is a topic of great interest. CD4⁺ T lymphocyte cells derived from mice infected with T. gondii-irradiated tachyzoites were found to co-produce IFN-γ and IL-10 (49). Also, IL-10 production was found to require a “reactivation” step involving IL-12 production and subsequent IFN-γ production by CD4 T cells (50). Therefore, simultaneous production of antagonistic IL-10 and IFN-γ cytokines by CD4 Th1 cells in response to T. gondii suggests that conditions that maintain anti-T. gondii Th1 immunity also promote a negative feedback mechanism for limiting Th1 hypersensitivity and inflammatory tissue damage.