Fig. 1. Targeted disruption of the BAG1 gene. Panel A: parasites were transfected with a linearized targeting construct containing 8.5 kb genomic sequence encompassing the BAG1 locus. The BAG1 gene was disrupted by integration of an HXGPRT minigene into a Hind III-site within intron 3. Arrowheads indicate the positions of primer sequences used for PCR analysis of stably transfected clones (see Materials and methods). The predicted size of PCR products and EcoRI (E) and Pst I (P) restriction fragments expected to hybridize with the BAG1 probe is shown. Panel B: PCR analysis of genomic DNA from parental P(LK) strain parasites (lane 1), and two stably transfected clones (lanes 2 and 3). Presented as negative image of an ethidium bromide-stained gel. Presence of both the wild-type 0.2 kb PCR product and the 2.2 kb recombinant in lane 2 indicates that the targeting construct in this clone integrated into the genome without disrupting the BAG1 locus (probably by nonhomologous recombination at an unrelated site, although pseudodiploid organization at the endogenous locus is also possible [21]). Lane 3 shows only the 2.2 kb PCR product, suggesting that the targeting construct has replaced the wild-type allele by homologous recombination (as diagrammed in panel A). Panel C: hybridization analysis of Pst I- and EcoRI-digested genomic DNA from the putative bag1 knock-out clone (KO), compared with the parental strain (WT). The transgenic clone exhibits restriction patterns consistant with a double crossover at the BAG1 locus (panel A), replacing the endogenous wild-type allele and disrupting gene function.

Fig. 2. Absence of BAG1 expression in the knock-out mutant. Bradyzoite differentiation was induced in bag1 knock-out and parental HXGPRT-deficient P (LK) parasites by alkaline pH-shift [8], and expression of various bradyzoite-specific antigens was observed daily. Panel A: cultures were analyzed 96 h post-infection by double-immunostaining with an anti-BAG1 antiserum (detected with a FITC conjugate; left-hand images), followed by staining with the bradyzoite-specific mAb 4F8 (detected with a rhodamine conjugate; right). No BAG1 protein was detected in the knock-out mutant, but expression of the 4F8 antigen indicates successful induction of bradyzoite differentiation. Panel B: quantitative analysis of bradyzoite antigen expression six days post-induction. Values represent the mean of two samples±SD Antibodies employed: polyclonal mouse anti-BAG1 antiserum [13]; mAb 4F8 [4]; mAb DC11 [25]; anti-P18 mAb T83B1 [26]; anti-P34 mAb T82C2 [26]. All bradyzoite-specific antigens except for BAG1 were induced to the same extent in the bag1 knock-out mutant and the parental strain.

Fig. 3. Electron micrographs of a cell culture infected with the bag1 knock-out mutant 6 days after tissue cyst initiation by alkaline pH. Sections were immuno-stained with mAb CC2 and visualized using Protein A conjugated to 5 nm gold. Panel A: low power micrograph showing a small cyst contain numerous parasites (P) enclosed by a distinct cyst wall (CW). HC, host cell. Scale bar, 1 μm. Panel B, detail of the periphery of a cyst demonstrating intense gold labeling of the cyst wall with mAb CC2 in the bag1 knock-out mutant (arrows). Bar, 100 nm.

Fig. 4. Growth rate of the bag1 knock-out mutant under stress conditions known to induce bradyzoite differentiation in vitro. Parental and mutant parasites were subjected to various treatments known to inhibit replication and induce bradyzoite differentiation [7, 8]: 50 μM Na-nitroprusside (SNP), 2 μM Na-arsenite, or alkaline pH (pH 8.2). Four days post-treatment, cultures were immunostained with bradyzoite-specific mAb 4F8, and 50 positively-staining parasitophorous vacuoles were scored for size (mean±SD; n=2–6). In vitro replication of bradyzoites in the bag1 knock-out mutant was indistinguishable from the parental strain.

Fig. 5. Cyst formation of the bag1 knock-out mutant. Panel A: C57BL/6 mice were infected subcutaneously with 10⁴ bag1 knock-out or parental P(LK) parasites and treated with sulphadiazine in the drinking water from day 14 post-infection. 28 days post-infection, brains were homogenized in 3 ml PBS/brain and aliquots of 30 μl used to determine cyst number. Values represent the mean±SD of seven brains per group. Panel B, genomic DNA isolated from the brains of chronically infected mice was subjected to PCR analysis using BAG1-specific primers, confirming the identity of parental and bag1 knock-out parasites (compare with Fig. 1(B)).

Fig. 6. BAG1 homologs in T. gondii. Clusters of two BAG1 homologs were identified in the T. gondii EST database and aligned with BAG1 (GenBank ID #X82213) and a plant sHSP (Daucus carrota; GenBank™ID #S15525). Only the carboxyl terminal region is shown, as this is the most highly-conserved region of sHSPs [34]; amino acid residues conserved in plants are denoted by *; residues conserved between BAG1 and at least one of the T. gondii ESTs are denoted by ^•. BAG1 homolog 1: dbEST ID #604948; BAG1 homolog 2: dbEST ID #s 466811, 489371, 489445, 574258, 574413, 623333, 1165316.