Proximity biotinylation reveals novel secreted dense granule proteins of Toxoplasma gondii bradyzoites

Abstract

Toxoplasma gondii is an obligate intracellular parasite which is capable of establishing life-long chronic infection in any mammalian host. During the intracellular life cycle, the parasite secretes an array of proteins into the parasitophorous vacuole (PV) where it resides. Specialized organelles called the dense granules secrete GRA proteins that are known to participate in nutrient acquisition, immune evasion, and host cell-cycle manipulation. Although many GRAs have been discovered which are expressed during the acute infection mediated by tachyzoites, little is known about those that participate in the chronic infection mediated by the bradyzoite form of the parasite. In this study, we sought to uncover novel bradyzoite-upregulated GRA proteins using proximity biotinylation, which we previously used to examine the secreted proteome of the tachyzoites. Using a fusion of the bradyzoite upregulated protein MAG1 to BirA* as bait and a strain with improved switch efficiency, we identified a number of novel GRA proteins which are expressed in bradyzoites. After using the CRISPR/Cas9 system to characterize these proteins by gene knockout, we focused on one of these GRAs (GRA55) and found it was important for the establishment or maintenance of cysts in the mouse brain. These findings highlight new components of the GRA proteome of the tissue-cyst life stage of T. gondii and identify potential targets that are important for maintenance of parasite persistence in vivo.

Fig 1. MAG1-BirA* localizes to the PV and biotinylates proteins in the cyst vacuole.

(A) Diagram of the construct encoding the promoter and full genomic sequence of MAG1 fused to BirA* along with a 3xHA C-terminal epitope tag. (B) IFA of MAG1-BirA* showing that the fusion protein appropriately traffics to the tachyzoite PV and colocalizes with the known protein GRA14. Scale bar: 5 μm (applicable to all panels). (C) IFA of MAG1-BirA*-expressing parasites, showing the bradyzoite PV is labeled in a biotin-dependent manner (yellow arrowheads, +Biotin row). Endogenously biotinylated apicoplasts are observed with and without biotin (white arrows, -Biotin row). Scale bar: 5 μm (applicable to all panels in C). (D) Western blot of whole-cell lysates of parental (PrugniaudΔku80Δhxgprt) and MAG1-BirA*-expressing parasites -/+ biotin. Lysates were probed with streptavidin-HRP, revealing an increase in biotinylated proteins in MAG1-BirA*-expressing parasites upon addition of biotin. The MAG1-BirA* fusion protein is predicted to be ~105 kDa (arrowhead). (E) Western blot showing that MAG1-BirA* fusion migrates to ~105 kDa as detected by antibody against HA epitope tag.

Fig 2. Identification of novel dense granule proteins, GRA55-GRA59 by MAG1-BioID.

(A) IFA of tachyzoites with rabbit anti-HA antibodies shows strong staining of the parasitophorous vacuole for each novel GRA that colocalizes with GRA14, demonstrating that these are novel dense granule proteins. Gene numbers and novel designations are as follows: GRA55 (TgME49_309760), GRA56 (TgME49_309930), GRA57 (TgME49_217680), GRA58 (TgME49_268790), and GRA59 (TgME49_313440). Scale bar: 5 μm. (B) IFA of bradyzoites with mouse anti-HA antibodies shows staining of each GRA at the cyst periphery and vacuolar space. Transgenic parasites expressed GFP, driven by LDH2 promoter, upon successful bradyzoite conversion. Images were taken three days after growth under alkaline-stressed conditions. Scale bar: 5 μm.

Fig 3. Deletion of GRAs 55–59 using CRISPR/Cas9.

(A) IFA shows the absence of HA signal in the PV. GRA14 is a marker of the PV and is shown as a control. Scale bar: 5 μm. (B) PCR verification of the deletion of GRAs 55–59. Attempted amplification of genes in knockout parasite strains show an absence of the coding sequence in the knockout strain but present in the wild-type control (gene check). The insertion of HXGPRT into the correct locus is verified by PCR from upstream of the gene of interest into the selectable marker (KO check) which is present only in the knockouts (PruΔku80 genomic DNA used as positive control).

Fig 4. GRA55 is a novel GRA which is not important for in vitro growth or short-term virulence.

(A) Complementation strategy of GRA55 into the UPRT locus. The complemented gene contains a C-terminal 3xHA epitope tag and is driven by the native promoter. (B) IFA of GRA55-HA, Δgra55, and GRA55c parasites. Rabbit anti-HA shows strong staining in the vacuole of GRA55-HA and GRA55c parasites, colocalizing with GRA14. HA signal is absent in Δgra55 parasites. Scale bar: 5 μm. (C) Western blot demonstrating GRA55-HA migrating at the predicted size of 35.2 kDa (the size of GRA55-HA without the predicted signal peptide) and showing similar expression levels to the endogenously tagged strain. As expected, there is no HA signal in the Δgra55 strain. RON5C is used as a loading control [50]. (D) Quantitation of plaques formed by GRA55-HA, Δgra55, and GRA55c parasites. Results are shown as box-whisker plot with the middle line representing the median, the bottom and top of box representing the 25^th and 75^th percentile, and whiskers corresponding to smallest and largest plaques. (E, F, G) C57BL/6 female mice were infected with indicated doses of GRA55HA, Δgra55, and GRA55c parasites and Kaplan-Meier survival curves were generated. There was no statistical difference between the virulence of Δgra55 (F) vs. GRA55-HA (E) or GRA55c (G) parasites. (H) CBA/J mice were infected with 250 parasites each of the indicated strain. Mice were euthanized and mouse brains were examined by IFA after 30 days infection to enumerate T. gondii cyst burden. Δgra55 caused 10-fold lower burden of cysts when compared to GRA55-HA infection and GRA55c infection (p < 0.05 by two-tailed T-test). (I) Numerical counts of cysts per mouse brain (* = mouse dead prior to end of experiment).