Functional Characterization of 15 Novel Dense Granule Proteins in Toxoplasma gondii Using the CRISPR-Cas9 System

ABSTRACT The analysis of the subcellular localization and function of dense granule proteins (GRAs) is of central importance for the understanding of host-parasite interaction and pathogenesis of Toxoplasma gondii infection. Here, we identified 15 novel GRAs and used C-terminal endogenous gene tagging to determine their localization at the intravacuolar network (IVN), parasitophorous vacuole (PV), or PV membrane (PVM) in the tachyzoites and at the periphery of the bradyzoites-containing cysts. The functions of the 15 gra genes were examined in type I RH strain and 5 of these gra genes were also evaluated in the cyst-forming type II Pru strain. The 15 novel gra genes were successfully disrupted by using CRISPR-Cas9 mediated homologous recombination and the results showed that 13 gra genes were not individually essential for T. gondii replication in vitro or virulence in mice during acute and chronic infection. Intriguingly, deletion of TGME49_266410 and TGME49_315910 in both RH and Pru strains decreased the parasite replication in vitro and attenuated its virulence, and also reduced the cyst-forming ability of the Pru strain in mice during chronic infection. Comparison of the transcriptomic profiles of the 15 gra genes suggests that they may play roles in other life cycle stages and genotypes of T. gondii. Taken together, our findings improve the understanding of T. gondii pathogenesis and demonstrate the involvement of two novel GRAs, TGME49_266410 and TGME49_315910, in the parasite replication and virulence. IMPORTANCE Dense granule proteins (GRAs) play important roles in Toxoplasma gondii pathogenicity. However, the functions of many putative GRAs have not been elucidated. Here, we found that 15 novel GRAs are secreted into intravacuolar network (IVN), parasitophorous vacuole (PV), or PV membrane (PVM) in tachyzoites and are located at the periphery of the bradyzoite-containing cysts. TGME49_266410 and TGME49_315910 were crucial to the growth of RH and Pru strains in vitro. Deletion of TGME49_266410 and TGME49_315910 attenuated the parasite virulence in mice. However, disruption of other 13 gra genes did not have a significant impact on the proliferation and pathogenicity of T. gondii in vitro or in vivo. The marked effects of the two novel GRAs (TGME49_266410 and TGME49_315910) on the in vitro growth and virulence of T. gondii are notable and warrant further elucidation of the temporal and spatial dynamics of translocation of these two novel GRAs and how do they interfere with host cell functions.

IMPORTANCE Dense granule proteins (GRAs) play important roles in Toxoplasma gondii pathogenicity. However, the functions of many putative GRAs have not been elucidated. Here, we found that 15 novel GRAs are secreted into intravacuolar network (IVN), parasitophorous vacuole (PV), or PV membrane (PVM) in tachyzoites and are located at the periphery of the bradyzoite-containing cysts. TGME49_266410 and TGME49_315910 were crucial to the growth of RH and Pru strains in vitro. Deletion of TGME49_266410 and TGME49_315910 attenuated the parasite virulence in mice. However, disruption of other 13 gra genes did not have a significant impact on the proliferation and pathogenicity of T. gondii in vitro or in vivo. The marked effects of the two novel GRAs (TGME49_266410 and TGME49_315910) on the in vitro growth and virulence of T. gondii are notable and warrant further elucidation of the temporal and spatial dynamics of translocation of these two novel GRAs and how do they interfere with host cell functions.
KEYWORDS Toxoplasma gondii, dense granule proteins, subcellular localization, virulence, cysts assay (IFA). Following the successful deletion of 15 gra genes in the RH strain and 5 gra genes in the Pru strain, we investigated the effects of gra deletion on the lytic cycle in vitro and parasite virulence in acute and chronic infection in vivo. Our data indicated that two novel GRAs (TGME49_266410 and TGME49_315910) play crucial roles in the growth and virulence of T. gondii.

RESULTS
Fifteen novel GRAs are secreted into the IVN, PV, or PVM in the tachyzoites and are localized to the cyst wall or matrix in the bradyzoites. We investigated 15 gra genes (Table 1), designated as dense granule proteins by hyperLOPIT (10). To examine the localization of these 15 GRAs in the tachyzoites and bradyzoites, epitopes tagged with 6Âhemagglutinin (6ÂHA) were introduced into the C-terminal endogenous gene by homologous recombination and confirmed by PCR and DNA sequencing (see Fig. S1A in the supplemental material). Western blotting of the 15 GRAs extracted from tachyzoites verified the insertion of the endogenous epitope tags and showed the expression of the 15 GRAs in T. gondii RH tachyzoites (Fig. S1B). Most of the GRA bands in Western blotting were consistent with the predicted size found in TOXODB (http:// toxodb.org). However, the specific band of TGME49_279350 was smaller than the predicted size, while the bands of TGME49_244530 and TGME49_258462 were bigger than those predicted in TOXODB. TGME49_313440 and TGME49_204320 had the predicted size bands and other small bands. These unpredicted bands in Western blotting may be attributed to posttranslational modifications or an inaccurate prediction of the gene model in TOXODB.
To verify the subcellular localization of GRAs in T. gondii tachyzoites, human foreskin fibroblast (HFFs) were infected with RH::GRAs-6HA for 24 h, and then examined by confocal microscopy using anti-HA antibodies and rabbit anti-GRA12 antibodies as a GRA marker. As shown in Fig. 1, the 15 predicated GRAs were secreted into the IVN, PV, or PVM, and colocalized or partly colocalized with GRA12 located in the IVN, PVM, and PV (36,41). The localizations of the 15 novel GRAs are consistent with that of known GRAs (10). These results indicate that these 15 proteins are GRA proteins.
To further characterize the location of GRAs in the in vitro-induced cyst, tachyzoites were allowed to differentiate into bradyzoites by shifting the pH of medium to 8.2 for 2 days. Cysts were detected by DBA staining, which specially recognizes N-acetylgalactosamine on the bradyzoite-specific cyst wall (42,43), and targeted GRA proteins were detected by anti-HA antibody. By 48 h postdifferentiation (44,45), 15 GRAs were localized to the cyst wall or matrix, which were colocalized or partly colocalized with DBA ( Fig. 2). Combined with previous reports (42) showing that GRAs are secreted into the cyst matrix or wall, our results suggest that these 15 GRAs may play roles in the development of the cysts. Fifteen novel gra genes are successfully disrupted in the type I T. gondii RH strain. To determine the function of the GRAs, each gra locus was disrupted using the CRISPR-Cas9 system (Fig. 3A). A dihydrofolate reductase (DHFR) selectable marker surrounded with the gra gene flanking regions (5UTR-DHFR-3UTR) was used to replace the GRA-coding sequence. The CRISPR-Cas9 plasmid targeting each gra gene and the corresponding homologous drug-selective fragment were cotransfected into freshly egressed tachyzoites. To obtain single clones, transfectants were diluted using a 10-fold gradient dilution method after selection with pyrimethamine. Diagnostic PCRs confirmed the disruption of each gra gene (Fig. 3B). The small fragments (;500 bp) of the gra loci were amplified using diagnostic PCR2 which were not amplified in the knockout strains. The successful replacement of each gra gene was verified by diagnostic PCR1 and PCR3, in which ;1,500 bp fragments were amplified in the mutant strains but were not amplified in the wild-type strain. The results of diagnostic PCRs demonstrated that all 15 gra genes were successfully disrupted in the RH strain by CRISPR-Cas9-mediated homologous recombination (Fig. 3B).
Loss of TGME49_266410 or TGME49_315910 impairs the intracellular replication in the RH strain. Plaque assays were conducted to determine the effect of gra deletion on the lytic cycle of tachyzoites, which represents several processes, including motility, invasion, intracellular growth, and egress (46). HFF cell monolayers grown in 12well tissue culture plates were infected by freshly egressed tachyzoites of the RHDgra and wild-type strains. After 7 days of incubation, cells were stained by crystal violet to visualize the plaques. As shown in Fig. 4A, no significant difference was detected in the number and size of plaques between 13 RHDgra strains (RHD313440, RHD247530, RHD204320, RHD297900, RHD267740, RHD268970, RHD279350, RHD258462, RHD214410, RHD291630, RHD258458, RHD248990, and RHD244530) and the wild-type strain. However, deletion of TGME49_266410 and TGME49_315910 resulted in a significant reduction in the plaque formation ( Fig. 4B and C).
To further investigate whether the 15 novel GRAs are necessary for parasite replication, HFFs were infected with tachyzoites of the RHDgra and wild-type strains. The parasite replication rate was monitored by counting the number of tachyzoites per PV in at least 200 PVs at 23 h postincubation by using a fluorescence microscope. No (B) Identification of 15 RHDgra strains by diagnostic PCRs. PCR1 and PCR3 were designed to detect the insertion of a homologous fragment into the 59 and 39 of gra genes, respectively. PCR2 assay was used to detect the successful replacement of gra genes by DHFR fragment.

Functional Characterization of Novel Toxoplasma GRAs
Microbiology Spectrum significant changes were observed in the replication rates of 13 RHDgra strains, whereas RHD266410 and RHD315910 exhibited a significant decrease in the parasite intracellular proliferation (Fig. 5A).
To further characterize the effect of gra deletion on the parasite egress efficiency, egress assays were performed. The results showed no significant difference in the egress capacity between the 15 RHDgra strains and the wild-type RH strain (Fig. 5B).
The virulence of TGME49_266410-deficient mutant in the RH strain is attenuated in vivo. To investigate the function of GRAs in vivo, Kunming mice (6 mice per group) were injected intraperitoneally (i.p.) with 100 tachyzoites of each RHDgra strain and the wild-type RH strain. The mice were monitored twice daily postinfection and those reaching their humane endpoint criteria were euthanized. The survival rates of all infected mice are shown in Fig. 5C and D. The knockout of 14 gras (TGME49_313440,  TGME49_247530, TGME49_204320, TGME49_315910, TGME49_297900, TGME49_267740,  TGME49_268970, TGME49_279350, TGME49_258462, TGME49_214410, TGME49_291630,  TGME49_248990, TGME49_244530, and TGME49_258458) had no statistically significant effect on the mouse survival time (Fig. 5C). However, mice infected with tachyzoites of the TGME49_266410 mutant strain had slightly longer survival time compared with that of the wild-type RH strain infected group in acute infection (P , 0.05; Fig. 5D).
TGME49_266410 and TGME49_315910 are critical for the growth of Pru tachyzoites. While 15 gra genes were successfully deleted and their functions were investigated in the type I strain (RH), five of the 15 gra genes were selected to examine the effect of their deletion on the infectivity of the type II strain (Pru). We chose TGME49_266410 and TGME49_315910 genes because their deletion impacts the parasite fitness, and they have low fitness scores assigned through a genome-wide CRISPR/Cas9 knockout study (47). We also chose TGME49_248990 whose expression is high in bradyzoites and thus more likely to play a role in the Pru strain which forms bradyzoite-containing cysts. Given TGME49_258458 The number and size of plaques produced by RHD266410 and RHD315910 strains, both showed a significant reduction compared to the RH strain (***, P , 0.001; ****, P , 0.0001). Each plaque was represented by a symbol.

Functional Characterization of Novel Toxoplasma GRAs
Microbiology Spectrum interaction with GRA44, which interacts with MYR1 for delivering T. gondii effector proteins to the host cell (39), we hypothesized that TGME49_258458 may play a role in the Pru strain. TGME49_214410 was randomly selected as a control. As shown in Fig. 6A, five Pru knockout strains (PruD266410, PruD315910, PruD248990, PruD258458, and PruD214410) were successfully generated and verified by PCR assays. We performed a plaque assay to examine whether these five novel GRAs play any roles in the lytic cycle of the Pru strain. The deletion of TGME49_248990, TGME49_214410, and TGME49_258458 did not affect the propagation of Pru tachyzoites in vitro (Fig. 6B). However, the deletion of TGME49_266410 or TGME49_ 315910 significantly reduced the size and number of plaques, in agreement with the results observed in the RH knockout strains ( Fig. 6B and C). Disruption of TGME49_266410 or TGME49_315910 in the Pru strain reduces brain cyst burden in vivo. To evaluate the role of the selected five GRAs in chronic infection, two doses of PruDgra tachyzoites were inoculated (i.p.) into mice. As shown in Fig. 6D, mice infected with a high dose (5 Â 10 4 tachyzoites) of PruD248990, PruD214410, and PruD258458 took a median of 9 days to reach their humane endpoint, without significant difference with the wild-type strain infected group. However, 83% of mice infected with the same dose of PruD266410 or PruD315910 strain remained alive at 30 days postinfection, which was significantly longer than that observed in the parental Pru strain. In the low dose infection assay (200 tachyzoites), more mice infected with the wild-type Pru tachyzoites reached the

Functional Characterization of Novel Toxoplasma GRAs
Microbiology Spectrum humane endpoint criteria earlier than mice infected with the tachyzoites of PruD266410 and PruD315910 strains (Fig. 6E). The number of brain cysts determined 30 days postinfection showed a marked reduction in the cyst burden in the PruD266410-infected group (median = 56 cysts) and PruD315910-infected group (median = 51 cysts), compared to the wild-type Pruinfected group (median = 460 cysts) (Fig. 6F). These results indicated that TGME49_266410 and TGME49_315910 are important virulence factors in the Pru strain and play a role in chronic infection. We further investigated the ability of each of the five GRA mutants (PruD248990, PruD214410, PruD258458, PruD266410, and PruD315910) to form bradyzoite-containing cysts in vitro. After 2 days differentiation in alkaline pH (pH = 8.2) and ambient air, the expression of the bradyzoite-specific marker DBA was examined to determine the bradyzoite formation (48). Results showed that the in vitro cyst-forming ability of these five GRA mutant strains was similar to that of the wild-type parasites (Fig. 7), suggesting that none of these five GRAs are essential for cyst formation in vitro. Fifteen novel GRAs may play roles in different life cycle stages of T. gondii. Data about the transcription levels of 13 gra genes of different T. gondii lineages, cell cycle phases, life cycle stages, and stage differentiation determined by DNA microarrays was obtained from the TOXODB (https://toxodb.org), except for TGME49_267740 and TGME49_258458 because they had no transcription data in the TOXODB. Among 13 gra genes, the transcriptional levels of TGME49_313440 and TGME49_247530 were significantly different in the three T. gondii genotypes (type I, II and III) (see Fig. S2A in the supplemental material). The expression profiles of 13 gra genes during the parasite cell cycle phase were also analyzed (Fig. S2B). Most gra genes followed a specific cell cycle pattern, with the highest expression level detected at the M and C stages, except six gra genes, including TGME49_315910, TGME49_297900, TGME49_279350, TGME49_258462, TGME49_244530, and TGME49_291630, which had the highest expression level at the G stage. During all cell cycle phases, the expression level of TGME49_244530 was the highest, while TGME49_315910 and TGME49_268970 had the lowest expression. Fig. S2C in supplemental material showed the expression profile of all gra genes during tachyzoite-bradyzoite differentiation, showing an increase in the expression of some gra genes, including TGME49_313440, TGME49_297900, TGME49_248990, and TGME49_214410. However, the expression of 5 gra genes was decreased, including TGME49_247530, TGME49_268970, TGME49_279350, TGME49_258462, and TGME49_2445 30. The remaining four genes (TGME49_204320, TGME49_266410, TGME49_315910, and TGME49_291630) were continuously expressed at the differentiated stages. Across the different developmental stages (Fig. S2D), some GRAs, including TGME49_247530, TGME49_297900, TGME49_279350, TGME49_258462, TGME49_248990, and TGME49_244530, were expressed at different levels. The rest of the gra genes were consistently expressed.

DISCUSSION
Despite the significant advances in the knowledge of T. gondii virulence factors, the spectrum and functions of the effector proteins necessary to enable T. gondii infection remain unclear (1,4,(7)(8)(9). GRAs are the most studied and important effector proteins in T. gondii due to their pivotal roles in mediating host-parasite interaction and parasite adaption to the intracellular environment. Most GRAs are localized in the PV, PVM, IVN, and several to the host cytoplasm or nucleus. GRAs localized in the PVM or the host cell affect numerous signaling pathways in the host cell (29) and alter host cell gene expression (24,28) and immunity (26,27). The PVM-or IVN-localized GRAs are involved in host cytoskeleton restructuring, nutrient acquisition, and protein translocation. However, the functions of the newly identified putative GRAs are still unknown. Here, we investigated the intracellular localization and functions of 15 novel GRAs in T. gondii type I (RH) and type II (Pru) strains in vitro and in vivo.
The majority of GRAs are synthesized in the rough ER and then exported from the ER, trafficking through the Golgi to the dense granules, then finally secreted from the parasite to a particular location to perform its function (23). In this study, subcellular localizations of 15 novel GRAs were investigated in the tachyzoites and bradyzoites of the T. gondii RH strain. We found that all 15 novel GRAs were secreted into the IVN, PV, and PVM, and colocalized or partly colocalized with GRA12 in the tachyzoites. Similar findings are reported for other GRAs (11,38,49). GRA proteins are also associated with the wall of the bradyzoite-containing cysts to maintain the parasite viability and ability to evade the host immune response (42,50,51). In the bradyzoite stage, the 15 novel GRAs were localized in the cyst wall or cyst matrix, which is consistent with the localizations of other GRAs (42,43,48).
The posttranslational modification plays crucial roles in the function of some GRAs (52)(53)(54). In the present study, the unpredicted bands shown in the Western blotting may be attributed to posttranslational modification, such as procession of aspartyl protease V (ASP5) or phosphorylation by WNG1 (parasite-secreted kinase) (52)(53)(54)(55). Except the major bands which matched the predicted size, the other bands of TGME49_313440 and TGME49_204320 detected by Western blotting may be attributed to protein degradation or the processing of TGME49_ 204320 by ASP5, considering that this putative protein harbors an arginine-arginine-leucine (RRL) motif termed TEXEL (T. gondii export element) (52,53,55). The phosphorylation of TGME49_244530 by WNG1 (parasite-secreted kinase), reported previously (54), may cause the bigger bands observed in the Western blotting. The unpredicted bigger band of TGME49_ 258462 may be caused by posttranslational modifications or an imprecise gene prediction in TOXODB (40). Apart from the major bands, there were other bands in the Western blotting of TGME49_258462. Given that two RRLs were predicted in TGME49_258462, the other bands that are detected in the Western blotting may be caused by the effect of the cleaving of ASP5 (52,53,55). Further experiments are needed to verify whether posttranslational modifications are the underpinning mechanisms for these observations.
Like most GRAs that are not essential for the parasite growth in vitro (56), CRISPR-Cas9-mediated knockout of 13 gra genes had no significant impact on the parasite replication ability, as indicated by the limited differences in the size and number of plaques formed in HFFs by the knockout and wild-type strains, and the insignificant difference in intracellular replication between the 13 gra knockout strains and the wild-type strains. This result suggests that none of these 13 novel GRA proteins was individually essential for the parasite survival in vitro, which was consistent with the high CRISPR fitness score (47). However, these proteins may play a role in the parasite fitness in other intracellular environment, such as interferon gamma-activated host cells (36,38).
On the other hand, some GRAs play important roles in T. gondii. Disruption of these GRAs impair the parasite growth, including GRA17 (22,57), GRA39 (11), GRA41 (58), GRA44 (39), and PPM3C (59). GRA17 along with GRA23 influence PVM permeability and transportation of small molecules (22). GRA41 is critical for regulating calcium homeostasis and egress (58). GRA44 and PPM3C mediate the effector export (39,59). In the present study, deletion of TGME49_266410 and TGME49_315910 in both RH and Pru strains resulted in a marked growth defect of tachyzoites, suggesting that TGME49_266410 and TGME49_315910 play roles in T. gondii propagation. Whether these effects are due to the involvement of TGME49_266410 and TGME49_315910 in the uptake or trafficking of nutrients like GRA17 and GRA23, or protein export like GRA44 and PPM3C, remains to be investigated.
Some GRAs contribute to parasite virulence in mice (11,22,38,57,59,60). This effect was also observed in TGME49_266410 in the present study. Disruption of TGME49_266410 significantly attenuated the virulence of both RH and Pru strains. The reduction in virulence of the knockout strains is likely a consequence of a slow growth of the mutant strain in mice. In contrast, disruption of TGME49_315910 did not attenuate the virulence of the RH strain; however, it attenuated that of the Pru strain, which is likely to be related to the high virulence of the RH strain (1,2). The activities of some GRAs have been shown to be straindependent (32). Thus, the function of TGME49_315910 might be genotype/strain-specific, given that there is a difference in six amino acids in TGME49_315910 between type I (GT1 strain) and type II (ME49 strain) (http://toxodb.org).
In chronic infection, PruD266410 and PruD315910 had significantly formed fewer brain cysts in mice. Considering the in vitro growth kinetics showing significant growth defect in PruD266410 and PruD315910, the reduced cyst-forming ability of PruD266410 and PruD315910 might be caused by elimination of most tachyzoites inoculated (i.p.) into the mice, although some of the tachyzoites managed to arrive to the brain and form cysts. The exact mechanism by which TGME49_266410 and TGME49_315910 affects the parasite propagation and virulence remains to be investigated. Although PruD266410 and PruD315910 showed attenuated virulence, they can form brain cysts in mice and are thus not promising vaccine candidates against toxoplasmosis. However, TGME49_266410 and TGME49_315910 could still be great candidate genes to generate double or triple gene knockout mutants as live-attenuated vaccines.
Transcriptome data available in the TOXODB showed that the expression patterns of gra genes vary by different T. gondii genotypes, cell cycle phases, life cycle forms, and bradyzoite differentiation. The different expression of TGME49_313440 and TGME49_247530 in different T. gondii genotypes indicate that they may have strain-specific roles like GRA15 (32). Several GRAs are upregulated in the bradyzoites and play an important role in the establishment or maintenance of cysts in the mouse brain, such as GRA55 (12,48,50,51). Among the 15 characterized GRAs, TGME49_248990 had the highest expression in bradyzoites. However, disruption of this gra gene did not change the bradyzoite differentiation rate in vitro. Although 13 GRA proteins were not involved in the replication and infectivity of T. gondii, they may play roles in the other developmental stages of this parasite.
In conclusion, our data revealed the roles of two novel GRAs TGME49_266410 and TGME49_315910 in T. gondii virulence. Further investigations are needed to unravel the molecular mechanisms and kinetics of translocation, and their molecular interaction with host cell organelles, all are important elements in understanding T. gondii manipulation of host cell machinery. Our study, along with others, show that GRAs are key virulence factors utilized by T. gondii to facilitate infection and colonization of the host cells, and provide possible targets for the development of novel therapeutics for T. gondii.

MATERIALS AND METHODS
Host cell and parasite culture. Human foreskin fibroblast (HFF) cells (American Type Culture Collection; ATCC SCRC-1041) were cultured in Dulbecco's Modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 10 mM HEPES (pH 7.2), 100 U/mL penicillin, and 100 mg/mL streptomycin, as described previously (36). Cultured HFFs were maintained in a humidified atmosphere containing 5% CO 2 at 37°C. The tachyzoites of T. gondii strains, including RHDku80, PruDku80, and gra knockout strains, were maintained in confluent HFF monolayers under the same conditions, except that the concentration of FBS was reduced to 2%. HFF cells heavily infected by tachyzoites were scraped off and passed through 27-gauge needles. The released tachyzoites were purified using a 5 mm Millipore filter, counted using a hemocytometer, and used for the assays described below.
Generation of GRA knockout parasite strains. Selected gra genes were disrupted by the CRISPR-Cas9 mediated homologous recombination as described previously (36,61). For construction of a knockout plasmid of each gra gene, we used a template plasmid, pSAG1:CAS9-U6-SgUPRT, expressing CAS9 and a single guide RNA targeting the UPRT in T. gondii, to replace the UPRT gRNA with the specific gRNA targeting each gra gene. A drug-selective plasmid was constructed by fusing the 59 and 39 Functional Characterization of Novel Toxoplasma GRAs Microbiology Spectrum homologous arms of each gra gene amplified from T. gondii genomic DNA, the DHFR fragment amplified from pUPRT-DHFR-D plasmid, and the pUC19 fragment amplified from plasmid pUC19 using the multifragment cloning method by a CloneExpress II one-step cloning kit (Vazyme). This drug-selective plasmid, validated by sequencing, was used as a template to amplify the fragment of 5UTR-DHFR-3UTR which was then extracted using a gel extraction kit (Omega). The knockout plasmids, validated by sequencing, were collected by Endo-free plasmid DNA minikit (Omega). The specific knockout plasmid (;35 mg) and corresponding homologous drug-selective fragment (5UTR-DHFR-3UTR; ;20 mg) were cotransfected into freshly egressed tachyzoites by electroporation (62). The gra-knockout transfectants were obtained by selection with 3 mM pyrimethamine, and the single clones were obtained by using 96well tissue culture plates and modified limiting dilution. The confirmation of gra-knockout strains was carried out by PCR assays (Fig. 3A). All the primers used to construct the RHDgras are listed in Table S1 in supplemental file 1.
Endogenous C-terminal epitope tagging. For C-terminal endogenous tagging of gra genes, a specific CRISPR plasmid targeting the locus near the STOP codon of each gra gene was obtained, and the homologous fragment containing 6 Â hemagglutinin (6ÂHA) and DHFR was amplified using p6ÂHA-LIC-DHFR as a template and a pair of specific primers. One of the primers was designed with 42 bp of the 39 region of gra gene without a STOP codon, and the other primer was designed with 42 bp of the gra gene just after the corresponding SgRNA. The successfully sequenced plasmid (;35 mg) and the corresponding purified fragment (;20 mg) were cotransfected into the tachyzoites of RHDku80 strain. After drug selection in 96-well tissue culture plates, independent clones were confirmed by sequencing, PCRs, IFA, and Western blotting. Primers used for the generation of epitope-tagged strains are listed in Table S2 in the supplemental material.
Induction of bradyzoite differentiation. The tachyzoites of T. gondii were differentiated into bradyzoites in vitro, as previously described (44,45). The tachyzoites of strains with C-terminal HA-tagged GRA, the PruDgra or wild-type Pru strains were used to infect confluent HFF cells cultured on coverslips at the bottom of 12-well culture plates. The infected HFF cells were washed 2 h after infection using the differentiation medium with pH 8.2 and incubated at 37°C in ambient air. To maintain the alkalinity of the medium, the differentiation medium was replaced daily. The samples were analyzed by IFA after 48 h postdifferentiation (44,45). The Dolichos biflorus agglutinin (DBA) positive vacuoles were assigned as bradyzoite-containing cysts and the percentage of cyst differentiation was calculated based on the results obtained from three independent experiments.
Detection of GRA proteins by Western blotting and immunofluorescence. For detection of GRAs and verification of the success of C-terminal tagging, extracellular tachyzoites were collected, centrifuged, and washed with cold phosphate-buffered saline (PBS) twice (10 min, 1,000 Â g). The purified tachyzoites were lysed by using RIPA lysis buffer on ice for 1 h. The supernatant was collected after centrifugation of the lysates and was blended with 4 Â sample loading buffer and boiled for 15 min at 100°C. These samples were analyzed by SDS-PAGE and subsequently transferred to polyvinylidene fluoride (PVDF) membrane by wet electroblotting, as described previously (36). For Western blotting, the primary antibodies were rabbit antialdolase (at 1:500), and rabbit anti-HA (at 1:1,000); the secondary antibody was goat antirabbit (at 1:5,000). Antibodies were obtained from Cell Signaling.
Standard in vitro plaque assay. A plaque assay was performed as previously described (36,63). Briefly, 500 tachyzoites were added to confluent monolayers of HFF cells grown on the surface of 12-well tissue culture plates. After 7 days (for the RHDgra and wild-type RH strains) or 12 days (for the PruDgra and wild-type Pru strains) of undisturbed incubation, cells were washed twice with warm PBS, and fixed with 4% PFA for 20 min. Subsequently, cells were stained with 2% crystal violet for 20 min and washed twice with PBS. The number of plaques formed by the replicating tachyzoites was quantified by ImageJ software.
Intracellular replication and egress. Confluent HFF cells growing on 6-well tissue culture plates were infected by 10 5 tachyzoites of T. gondii per well. Infected cells were washed three times with DMEM after 1 h invasion to remove any unbounded tachyzoites. For intracellular replication assay, infected HFF cells were fixed with 4% PFA after 23 h of incubation (63). The cells were stained with mouse anti-SAG1 followed by a secondary goat antimouse IgG conjugated with Alexa Fluor 488. The number of tachyzoites per PV was counted, including 200 randomly selected PVs per sample (36). For the parasite egress experiment, cell culture plates were maintained at 37°C for another 32 to 36 h after removing tachyzoites that remained extracellular. Then, the wells were washed with warm PBS and treated with 3 mM calcium ionophore A23187 (Sigma) diluted in DMEM. Once the egress started, the cultured plates were immediately fixed, and the percentage of egressed or nonegressed PVs was determined, as previously described (63). Three independent experiments were performed for each assay.
Virulence assessment during acute and chronic infection. Female Kunming mice (6 to 7 weeks old) were purchased from the Center of Laboratory Animals, Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Science. All experimental procedures involving the use of mice were reviewed and approved by the Animal Ethics Committee of Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences (approval no. 2021-008). Every effort was made to reduce the suffering of the animals. Prior to the start of experiment, all mice were habituated for 1 week before allocation to the experimental groups. In acute infection, 100 freshly egressed tachyzoites of the RHDgra mutant strains and the wild-type RH strain suspended in 200 mL PBS were injected intraperitoneally (i.p.) into mice (6 mice/strain) (64,65). The viability of the parasites used to infect mice was determined by a plaque assay. For survival rates, the mice were monitored twice daily and the mice that had reached their humane endpoint were immediately euthanized.
We evaluated the roles of five GRAs during chronic infection. Briefly, mice were inoculated by high dose (5 Â 10 4 tachyzoites) or low dose (200 tachyzoites) of five mutant strains (PruD248990, PruD214410, PruD258458, PruD266410, and PruD315910) and the wild-type Pru strain by i.p. route. Mice were monitored twice daily for up to one mouth unless they exhibited humane endpoint criteria sooner. We also investigated the cyst-forming ability of the two strains, PruD261410 and PruD315910, which exhibited the highest survival rates compared to the wild-type Pru strain. In brief, the brains of mice infected by 200 tachyzoites of PruD261410, PruD315910, or the Pru strain were collected and homogenized in 1 mL PBS at 30 days postinfection and the number of cysts was counted, as previously described (36).
Bioinformatics analysis of gra genes. Bioinformatic information on the gra genes were obtained from T. gondii genome database (http://toxodb.org). The transcriptional patterns of the main archetypal lineages (genotypes I, II, and III), cell cycle expression profiles (66), different developmental stages (oocysts, tachyzoites, and bradyzoites) (67), and during bradyzoite differentiation (68) were analyzed using the Robust Multiarray Average (RMA) algorithm of the Partek Genomics Suite package (Partek, Inc., St. Louis, MO, USA). The genomic features obtained included the phenotype value, the number of exons, predicted signal peptide, transmembrane domains (TMHMM), and molecular weight.
Statistical analysis. Statistical analyses were performed using GraphPad Prism (version 9.0). All data were based on three independent experiments. The results shown are the means 6 standard deviations (SD). The significant difference between 2 groups or $ 3 groups were determined by two-tailed, unpaired Student's t test, and one-way analysis of variance (ANOVA), respectively. The difference between groups was considered statistically significant when the P values were , 0.05.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 1.3 MB.