|Home | About | Journals | Submit | Contact Us | Français|
During its life cycle in intermediate hosts, Toxoplasma gondii exists in two interconverting developmental stages: tachyzoites and bradyzoites. This interconversion is essential for the survival and pathogenicity of the parasite, but little is known about the genetic mechanisms that control this process. We have previously generated tachyzoite-to-bradyzoite differentiation (Tbd−) mutants using chemical mutagenesis and a green fluorescent protein-based selection strategy. The genetic loci responsible for the Tbd− phenotype, however, could not be identified. We have now used an insertional mutagenesis strategy to generate two differentiation mutants: TBD-5 and TBD-6 that switch to bradyzoites at 10 and 50% of wild-type levels, respectively. In TBD-6 there is a single insertion of the mutagenesis vector 164 bp upstream of the transcription start site of a gene encoding a zinc finger protein (ZFP1). Disruption of this locus in wild-type parasites reproduces the decreased stage conversion phenotype. ZFP1 is targeted to the parasite nucleolus by CCHC motifs and significantly altered expression levels are toxic to the parasites. This represents the first identification of a gene necessary for efficient conversion of tachyzoites to bradyzoites.
Toxoplasma gondii is an Apicomplexan protozoan parasite that causes congenital disease (hydrocephalus and mental retardation) in children, blindness in immunocompetent individuals, and encephalitis in immunocompromised patients (23, 42, 43). The asexual life cycle of T. gondii occurs in a wide variety of intermediate hosts including humans and is characterized by two interconverting developmental stages: tachyzoite (a rapidly dividing form that is responsible for disease manifestation and is susceptible to drug therapies) and bradyzoite (a slowly growing form that is able to evade the host's immune system and currently available drugs) (7, 39). In response to host immune response and yet unknown cellular signals, tachyzoites differentiate into encysted bradyzoites, which lie dormant in host tissues for months or years. In immunocompromised individuals, clinical disease is due to reactivation of latent bradyzoites that convert to tachyzoites and disseminate throughout the body (23). Thus, this interconversion between the two developmental stages is essential for the survival and pathogenicity of the parasite.
Tachyzoite-to-bradyzoite interconversion has been previously characterized. Several groups have successfully developed conditions that cause tachyzoites to convert into bradyzoites in vitro (31, 33, 41). One of the key signals that controls bradyzoite formation in vivo is the host production of gamma interferon. Some in vitro models of tachyzoite-to-bradyzoite differentiation mimic the “stress” of the host immune response such as treatment with gamma interferon, mitochondrial inhibitors (atovaquone or antimycin), high pH (8.1), and high temperature (43°C) (4, 5, 32, 37, 40). Several bradyzoite-specific antigens (SAG4, BSR4, MAG1, CST1, SAG2C/2D, p21, and BAG1/5) and stage-specific enzymatic paralogs (LDH2 and ENO1) have been identified (3, 14, 21, 32, 39, 45). Control of bradyzoite-specific gene expression has been shown at both transcriptional and translational levels (21, 45). Many developmentally regulated genes and distinct patterns of gene expression have also been identified by using expression profiling with microarrays (12, 24, 30). Despite these developments, little is known about the genes that control the differentiation pathway in T. gondii. Identifying the genetic controls of differentiation will help understand the developmental biology of the parasite and develop measures that would prevent disease propagation or reactivation.
We and others have previously generated tachyzoite to bradyzoite differentiation (Tbd−) mutants (24, 30). Our approach used chemical mutagenesis and selection based on bradyzoite-specific expression of the green fluorescent protein (GFP) (30). In vivo and in vitro analyses of four independently generated Tbd− mutants showed that they are defective in their ability to switch to the bradyzoite form. cDNA microarray analysis of the transcriptional profile of these mutants revealed a hierarchy of developmentally regulated genes, including many bradyzoite-induced genes whose transcripts were reduced in all of the mutants. The work on chemical Tbd− mutants laid a foundation for understanding the developmental pathways of tachyzoite-to-bradyzoite differentiation. However, for various technical reasons, we were unable to identify the genetic loci responsible for the Tbd− phenotype in these mutants. We have now generated tachyzoite-to-bradyzoite differentiation mutants by using insertional mutagenesis. In the present study, we discuss the identification and characterization of two independently generated Tbd− insertional mutants (TBD-5 and TBD-6). We show that TBD-6 is the result of disruption of a locus encoding a nucleolar localized CCHC zinc finger protein.
All tissue culture with the Tbd− mutants was done with a strain of Toxoplasma engineered to lack hypoxanthine guanine phosphoribosyltransferase (HXGPRT; Pru/hxgprtΔ) and engineered to have bradyzoite-specific expression of GFP (30). Parasites were grown in human foreskin fibroblasts under standard Toxoplasma culture conditions, transfected, and cloned by limiting dilution as described previously (30). In vitro bradyzoite differentiation was induced with high pH and bradyzoite induction assessed by inverted fluorescence microscopy and GFP expression (30). The parasites were scraped and syringe lysed twice each with 27- and 30-gauge needles, and the total number of parasites versus GFP+ parasites counted. Protein localization studies with tagged constructs were performed with the RH strain of Toxoplasma (8).
The vector pHXGPRTxDHFR-TS was used for generation of insertional mutants. This plasmid contains a selection marker, HXGPRT, under the control of dihydrofolate reductase thymidylate synthase (DHFR) promoter (13).
A construct was generated so as to facilitate its integration into the parasite genome by homologous recombination, 164 bp upstream of the transcription start site of ZFP1, identical to the position and orientation of the mutagenesis vector in the original mutant TBD-6. Portions, 7 and 3 kb, respectively, of the 5′ and 3′ flanking genomic DNA (obtained by plasmid rescue of the flanking genomic sequence of TBD-6) were cloned into pBluescript SK(−) at the ApaI/SpeI and NotI sites, respectively. The 5′ flanking sequence obtained from TBD-6 by plasmid rescue was ~12 kb total and included 7 kb of flanking genomic DNA and ~5 kb of the pHXGPRTxDHFR-TS mutagenesis vector initially used for generation and selection of TBD-6.
To knockout ZFP1, 3 kb of the 3′ flanking genomic region was PCR amplified from genomic DNA, TOPO TA cloned (Invitrogen), digested with NotI and SpeI, and cloned in the pmini-HXGPRT at NotI and SpeI sites. The resulting plasmid was digested with XhoI, blunted and digested with ApaI, and ligated to a 5′ flanking region (8-kb PCR fragment from genomic DNA digested with SpeI and blunted and digested with ApaI). The primers used to amplify the fragments were 5′-TTATGGTACCGGTGCGACTGTCCACAGCGACCACTAAGT-3′ and 5′-CCACTCAGAGTTCGTGCCAT-3′ for 3′flanking region and 5′-CGGTTTCTTCAAATCGTCTCGTA-3′ and 5′-CGACAGATCACACGCTAAAGAA-3′ for the 5′ flanking region.
Two yellow fluorescent protein (YFP) fusion constructs, one containing the full-length ZFP1 (amino acids [aa] 1 to 444) and the other containing the amino-terminal region (aa 1 to 141) fused in frame to YFP were generated. PCR products for these two regions were obtained by using a sense primer (5′-AGATCTATGGCACGAACTCTGAGTCCC-3′) and antisense primers 5′-CCTAGGGGCGCTCTGTCCCATCCAGAA-3′ and 5′-CCTAGGTTTCACACCGGCCGACCCTGC-3′, respectively, to amplify products from tachyzoite cDNA, which were TOPO TA cloned, digested with BglII and AvrII (sites underlined in primer sequences), and cloned into identical sites of the ptubFNRYFP construct (kindly provided by David Roos) (34). In these constructs expression of the YFP was driven by a T. gondii tubulin promoter, and chloramphenicol acetyltransferase (CAT) was used as a selection marker.
In order to drive expression of YFP by the endogenous ZFP1 promoter, we replaced the tubulin promoter in the truncated version (1 to 141 aa of ZFP1) of the above-mentioned YFP construct with the endogenous ZFP1 promoter (1,768 bp upstream of transcription start site). The 1,768-bp fragment of the ZFP1 promoter was obtained by PCR amplification from genomic DNA by using the primers 5′-GGTGCGACTGTCCACAGCGACCACTAAGT-3′ and 5′-TATTCCATGGCTTGGTTCCCCGCGGCGAGCTA-3′. This product was TOPO TA cloned, digested with NcoI, blunted and digested with EcoRV, and ligated to the truncated YFP construct (digested with BglII and blunted and digested with PmlI).
Two hemagglutinin (HA) epitope tag constructs were generated by digestion of the YFP fusion constructs (the full-length and short versions) with AvrII and blunting and digestion with KpnI. The resulting fragment was ligated to a pNotI-HA-Gra3′ plasmid (kindly provided by Peter Bradley) which had been digested with NotI, blunted, and digested with KpnI. In this construct the tubulin promoter was utilized for expression in T. gondii.
Transfections by electroporation were carried out as described previously (2, 20). Insertional mutagenesis was carried out by transfecting Pru/hxgprtΔ parasites with 20 μg of the pHXGPRTxDHFR-TS vector linearized at the EcoRV or HindIII site; nonhomologous integration was enhanced by restriction enzyme-mediated integration with DpnII (2, 20). Transformants were selected for stable integration with 50 μg of mycophenolic acid/ml and 50 μg of xanthine/ml under tachyzoite conditions, and parasite clones were isolated by limiting dilution in 96-well plates as previously described (21). Fluorescence-activated cell sorting analysis of bradyzoites was performed as described earlier (30). Briefly, insertionally mutagenized parasite populations were subjected to sequential rounds of switching to bradyzoites, sorted for GFP− or GFP-faint parasites, and expanded under tachyzoite conditions. Two clones, Tbd− mutants TBD-5 and TBD-6, were identified by their reduced ability to express GFP under bradyzoite conditions. Identification of the mutagenesis site and recovery of flanking genomic DNA sequence was performed by plasmid rescue as described previously (1). Briefly, genomic DNA was prepared from each of the mutants, digested with SpeI, ApaI, BamHI, or BglI, ligated, and transformed into Escherichia coli, and ampicillin-resistant clones were selected. Plasmid clones were sequenced with primers specific to the mutagenesis vector to obtain the flanking genomic sequences (1).
For Southern blot analysis, parasite genomic DNA was extracted by using standard procedures (30). A total of 10 μg of genomic DNA from each clone was digested with BamHI, loaded onto a 1% agarose gel, transferred to a Hybond N+ membrane, cross-linked at 1,200 J using a UV cross-linker (Stratagene), and hybridized with a 1.6-kb probe of ZFP1 using Express Hyb (Clontech). The probe was generated from wild-type (WT) parasite tachyzoite cDNA by amplification with the primers 5′-CTGAGCGATGAGTTTTCTCGTT-3′ and 5′-CGTTCACTTCCCTCCAGATAA-3′. For Northern blot analysis, total RNA was isolated from tachyzoites and bradyzoites (72 h after switch) by using TRIzol reagent (Gibco-BRL). Then, 10 to 20 μg of total parasite RNA was loaded onto a formaldehyde gel, transferred onto a Hybond-N membrane, cross-linked, and hybridized with the 1.6-kb ZFP1 probe described above (12). Northern blot analyses were also performed by using probes for TgTwinscan proteins 1508 and 1510. Primer pair 5′-GCATTAACGAGGCCCTATCA-3′ and 5′-TTCTGAAGACCCGGAATCAC-3′ and primer pair 5′-ATGCCGGGTATGGACGACAAG-3′ and 5′-CTTGAAGGCGGTCTTCGCTGT-3′ were used to make probes for Twinscan protein 1508 and Twinscan protein 1510, respectively. Primers 5′-GCTGGAGAGAATCAATGTGT-3′ and 5′-GAAGCAAATGTCGTATAGGG-3′ were used to amplify 500 bp of the T. gondii β-tubulin gene, and primers 5′-GTTCGGCCGCAAGAAGAA-3′ and 5′-ATCAGGGATCTGTCGTACGCA-3′ were used to amplify 400 bp of the T. gondii 40S ribosomal gene, both of which were used as a loading controls. Quantitation was performed using densitometer and ImageQuant software. Values were adjusted for background levels and normalized to the loading control.
The transcription start site of ZFP1 was identified by using 5′ rapid amplification of cDNA ends (5′RACE) as previously described (12). A total of 1 μg of total RNA was treated with calf intestinal phosphatase and tobacco acid pyrophosphatase and ligated to the 5′RACE adapter (Ambion). This was converted to first-strand cDNA by using the Thermoscript RT-PCR System (Invitrogen) and a random decamer primer. The cDNA was used as a template for the 5′RACE nested PCR by using the RLM-RACE kit (Ambion). The primers used for 5′RACE were 5′-CCATCCTTTCTCTCACTTGTT-3′ and 5′-GCTGATGGCGATGAATGAACACTG-3′ (5′outer RACE) and 5′-CCACTCAGAGTTCGTGCCAT-3′ and 5′-CGCGGATCCGAACACTGCGTTTGCTGGCTTTGATG-3′ (5′inner RACE). For 3′RACE, 1 μg of total RNA was converted into first stand cDNA by using the Thermoscript RT-PCR System and the 3′RACE adapter (Ambion). This cDNA was then used as a template for the nested PCR by using the RLM-RACE kit (Ambion) with the primers 5′-CCAGTTGAGCGAGGCGAAAT-3′ and 5′-GCGAGCACAGAATTAATACGACT-3′ (3′outer RACE) and the primers 5′-CGACAGATCACACGCTAAAGAA-3′ and 5′-CGCGGATCCGAATTAATACGACTCACTATAGG-3′ (3′inner RACE).
Two peptides from ZFP1, DSARHFPFATGNGN (aa 10 to 23) and QSEIDTPSDPKSTD (aa 29 to 42), were synthesized, coupled to a keyhole limpet hemocyanin carrier protein (Imject Maleimide Activated mcKLH; Pierce), and coupled with an equal volume of RIBI adjuvant (RIBI Immunochem Research, Inc.). Two constructs for the production of recombinant proteins were generated in a T7 expression vector pET28a(+) (8). For this, cDNA corresponding to aa 2 to 159 of the ZFP1 was amplified with the primers 5′-AAAACATATGGCACGAACTCTGAGTGGG-3′ and 5′-CCCCGGATCCTCTCACTTGTTCTATAGTCCA-3′, and cDNA corresponding to aa 176 to 444 was amplified with the primers 5′-AAAACATATGACCAAGTCTAGGCAGAAGAGA-3′ and 5′-AAAAGGATCCGGCGCTCTGTCCCATCCA-3′. Using the NdeI and BamHI sites (underlined in the primers), the PCR products were cloned into identical sites of the expression vector upstream of a His6 tag, resulting in a His tag fusion protein. The constructs were sequenced in their entirety and transformed into BL21(DE3) strain of E. coli (8). Bacterial expression was induced with 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside) for 4 to 5 h at 37°C and recombinant protein purified from bacterial lysates by using a nickel-agarose resin (QIAGEN). BALB/c strain mice were injected intraperitoneally on days 0, 21, and 42 with 50 μg of recombinant protein or keyhole limpet hemocyanin-coupled peptide resuspended in 100 μl of phosphate-buffered saline (PBS) and emulsified with an equal volume of RIBI adjuvant. Blood was collected from the tail vein prior to initial immunization and after each boost, and the serum fraction was assayed for specific antibody content.
Total parasite lysates were prepared from tachyzoites and bradyzoites (72 h after pH switch) in sodium dodecyl sulfate-polyacrylamide gel electrophoresis sample lysis buffer containing dithiothreitol. Western blot analyses were performed by standard protocols (8). Briefly, lysates were separated on 8% sodium dodecyl sulfate-polyacrylamide gels, transferred overnight onto nitrocellulose membrane, blocked in 5% dry milk-0.1% Tween 20-PBS, and incubated with antisera diluted in blocking solution. Bound antibodies were detected with horseradish peroxidase-conjugated anti-mouse or anti-rabbit antibodies (1:3,000 dilution) and developed by using enhanced chemiluminescence (Amersham).
For the immunofluorescence assays, human foreskin fibroblasts on glass coverslips in 24-well plates were infected with parasites for 24 to 36 h under tachyzoite and for 72 h under bradyzoite conditions. The samples were fixed with 3% formaldehyde (20 min at room temperature), quenched with 100 mM glycine (2 min at room temperature), permeabilized with 0.2% Triton X-100 in PBS (20 min at room temperature), blocked with 3% bovine serum albumin in PBS (1 h at room temperature or overnight at 4°C), and incubated with primary and secondary antibodies for 1 h each at room temperature (30). For localization experiments with YFP fusion or HA-tagged constructs, RH strain parasites were transfected with 50 to 75 μg of nonlinearized DNA by restriction enzyme-mediated integration. After 24 h of transfection, the parasites were placed under drug selection with 20 μM chloramphenicol to select for stable transfectants. Antibodies were used in the following dilutions: 1:200, recombinant ZFP1; 1:20, mouse anti-HA rhodamine-conjugated antibody (Roche Diagnostics); 1:3,000, T. gondii tubulin (26); 1:200, T. gondii SRS9 (kindly provided by Seon-Kyeong Kim); and 1:1,000, Dolichos biflorus agglutinin. All secondary antibodies were used at 1:1,000. Hoechst or DAPI (4′,6′-diamidino-2-phenylindole) stains, each at 1:10,000 dilution, were used for nuclear staining. All samples were visualized on an Olympus BX60 microscope, and images were collected by using a Hamamatsu Orca digital charge-coupled device camera with Image Pro Plus 4.0 software (MediaCybernetics). Quantitation was performed by using densitometer and ImageQuant software. Values were adjusted for background levels and normalized to the loading control.
Parasites were transfected (upstream region disruption or knockout constructs), and stable parasite populations were selected with mycophenolic acid-xanthine. Genomic DNA was obtained and used as a template for PCR to identify parasites with disruption of the upstream or coding regions of ZFP1. The primers 5′-CGACAGCAGACAACTTTCCTTCTA-3′ and 5′-GGATTTGCAGACGAGAAAGAGGAA-3′ (for upstream region disruption screen) and the primers 5′-CGTCCTGAGCGACGCCGCT-3′ and 5′-GTGGCTTTCAGGTGATTGGGT-3′(for coding region knockout screen) were designed so as to yield a product only if homologous recombination occurred. Parasites populations with a positive PCR result were cloned by limiting dilution, retested by PCR, positives recloned, tested again by PCR, and confirmed by Southern blot analysis.
To identify genes involved in regulating tachyzoite to bradyzoite conversion, we insertionally mutagenized a Pru/hxgprtΔ strain with bradyzoite-specific expression of GFP and selected for parasite clones that were GFP− or GFP faint under bradyzoite inducing conditions as described previously (30). Two tachyzoite-to-bradyzoite differentiation mutants (TBD-5 and TBD-6) generated by independent rounds of insertional mutagenesis were identified and characterized. Switch efficiency was measured by analyzing the percentage of vacuoles that expressed GFP. Compared to the WT strain (which routinely had >90% switch efficiency under the conditions tested), TBD-5 and TBD-6 had 10 and 50% switch efficiencies, respectively, in vitro when exposed to high-pH bradyzoite induction for 3 days. Other bradyzoite-specific markers (Dolichos biflorus lectin staining of cyst wall; bradyzoite-specific surface antigen SRS9) were decreased compared to the WT strain, at levels comparable to the GFP reduction in each mutant. TBD-5 did not switch to bradyzoites even after prolonged exposure (6 to 7 days) to high-pH conditions. In contrast, TBD-6 switched with similar efficiency to WT when exposed to harsher conditions (pH 8.5) or for longer time periods (6 to 7 days). Both TBD-5 and TBD-6 had similar growth rates compared to the WT strain under tachyzoite and bradyzoite conditions. Both the leaky phenotype of the Tbd− mutants and the WT growth rates under tachyzoite conditions are similar to the characteristics of the Tbd− mutants generated previously by chemical mutagenesis (24, 30). The ability to switch at WT levels with harsher or longer bradyzoite stress conditions, however, was a unique property of TBD-6.
Preliminary Southern blot analyses revealed that in TBD-5 the vector apparently integrated tandemly in multiple copies. TBD-5 was not pursued because of the multiple tandem insertions and the resulting complicated pattern on Southern blot analyses; as a result, we could not exclude the possibility of insertions at multiple loci. Hence, we focused on characterizing TBD-6 in which the bradyzoite conversion efficiency of 50% of WT levels was a highly reproducible phenomenon and in which only one vector appeared to have integrated.
Using genomic sequences flanking the mutagenesis vector, 5′RACE, 3′RACE, and PCR from gDNA and cDNA, we reconstructed the genomic locus disrupted in TBD-6 (Fig. (Fig.1).1). The analysis revealed that the mutagenesis vector inserted 164 bp 5′ of the transcription start site for a predicted open reading frame (ORF). The genomic locus was 3,411 bp, with three introns with typical GU/AG consensus-splice sites, and a 1,332-bp predicted ORF. The 5′ and 3′ untranslated regions were 269 and 283 bp, respectively. By searching the expressed sequence tag (EST) database, we identified five tachyzoite ESTs and one predicted protein (TgTwinscan 1509) corresponding to this gene. BLASTX analysis revealed high homology to zinc finger proteins, and motif finder algorithms (http://us.expasy.org/prosite/) identified three CCHC zinc finger motifs. We will therefore refer to this gene as ZFP1. Nuclear localization signal (NLS) prediction programs (http://cubic.bioc.columbia.edu/predictNLS/) predicted NLSs, but no RNA recognition motifs were identified. Using ORF predictions from the T. gondii genome sequencing effort (http://toxoDB.org/), we analyzed the genomic sequences adjacent to the vector insertion site. The closest predicted ORF was ~2.6 kb downstream of the vector insertion site (Twinscan protein 1510) with the closest EST (TgEST zya73d03.y1) at 5.5 kb. Upstream there was a predicted ORF (Twinscan protein 1508) at 15 kb and the closest EST (TgEST 95050737) at 21 kb.
Insertional mutagenesis and transfection methods can result in multiple insertions (as seen in TBD-5) or point mutations in other genes. To confirm that the TBD-6 switch phenotype was due to insertion of the mutagenesis vector at the ZFP1 locus, we disrupted the identical region in WT parasites by homologous recombination. A parasite clone in which there was a disruption of the ZFP1 upstream region was identified by Southern blot analysis (Fig. (Fig.2A).2A). This mutant (subsequently referred to as “ZFP1 upstream disruption mutant”) was identical to the TBD-6 mutant in (i) the location of the vector insertion (164 bp 5′ of the transcription start site of ZFP1), (ii) the orientation of the mutagenesis vector, and (iii) the lack of other genomic insertions.
To study the phenotype of ZFP1 upstream disruption mutant, parasites were subjected to bradyzoite inducing conditions for 72 h, scraped, and syringe lysed, and GFP-positive parasites counted relative to the total number of parasites (Fig. (Fig.2B).2B). A parasite clone with heterologous integration of the HXGPRT vector was used as a control. TBD-6 and ZFP1 upstream disruption parasites both had similar bradyzoite switch efficiencies (~49% ± 3% and ~59% ± 4%, respectively) compared to WT (~93% ± 2%) and heterologously inserted controls (~82% ± 5%). This indicated that integration into the identical genomic locus as TBD-6 and disruption of the ZFP1 upstream region decreased bradyzoite conversion efficiency.
To attempt to disrupt the ZFP1 coding region, a knockout construct was designed that included HXGPRT plus extensive flanking regions from the ZFP1 locus but missing 942 bp of the coding region (see Materials and Methods). However, despite significant efforts (12 independent transfections), we were not able to get a knockout of ZFP1 (data not shown). The fact that this locus was readily targeted for homologous recombination upstream of the transcription start site but not disruption of the actual coding region argues against a technical problem and instead strongly suggests that some level of expression of this gene is essential.
To assess the effect of disruption of the upstream region of ZFP1, we performed Northern blot analysis. In WT parasites under tachyzoite conditions, the ZFP1 gene was expressed at low, but detectable levels; in TBD-6, the major band was equally apparent but, in addition, a smear extending upward was observed (Fig. (Fig.3).3). Quantitation of the ZFP1 mRNA signal relative to the loading control revealed that the signal was 198% ± 47% (average ± standard error) in TBD-6 parasites compared to the WT. Northern blot analysis of ZFP1 mRNA in the upstream disruption mutant showed results identical to TBD-6, and these results were seen in multiple independent experiments (data not shown). Despite loading 10 to 20 μg of total parasite RNA, we could not detect the ZFP1 transcript in bradyzoite RNA, indicating that message abundance was likely lower under bradyzoite conditions. We characterized the transcription start site, the polyadenylation site, and checked for alternate splicing in TBD-6 relative to the WT strain. The main PCR products for all experiments (5′RACE, 3′RACE, and full-length gene PCR) were indistinguishable in both WT and TBD-6, as would be expected since the majority of message in TBD-6 was similar to the WT strain. Since PCR is biased toward smaller products, it is feasible that predicted PCR products from TBD-6 of longer length were not adequately amplified and visualized.
To verify that insertion of the mutagenesis vector has not affected other neighboring genes, we analyzed the expression levels of two annotated genes closest to the mutagenesis vector (TgTwinscan predicted protein 1510 [~2.6 kb downstream of ZFP1] and TgTwinscan predicted protein 1508 [15 kb upstream of ZFP1]). Northern blot analysis of total tachyzoite RNA from WT, TBD-6, and upstream disruption mutants revealed that TgTwinscan 1508 had detectable signal and similar levels in WT and mutant parasites (data not shown). In contrast, we could not see detectable signal on Northern blot analysis for TgTwinscan 1510 in WT or mutant parasite strains, even after prolonged exposure (data not shown). Analysis of the EST database for TgTwinscan 1510 revealed a single EST (TgEST zya73d03.y1) from a partially sporulated oocyst cDNA library. There were no ESTs from either tachyzoite or bradyzoite stages, a finding correlating with our inability to detect message for this gene. Although we cannot definitively exclude that there were no polar effects whatsoever, these results indicate that insertion of the mutagenesis vector affected tachyzoite expression of ZFP1 but not other adjacent genes.
To determine the localization of the ZFP1, we generated four polyclonal antisera (anti-peptide antibodies for regions A and B and antibodies to bacterially expressed recombinant proteins for regions C and D) (Fig. (Fig.1B).1B). None of these stained parasites in immunofluorescence assays (under tachyzoite or bradyzoite conditions using various methods of parasite fixation), although the antisera to regions C and D worked in Western blot analysis (see below). We thus generated YFP- and HA-tagged fusion constructs; one in which the full-length gene (1 to 444 aa) and the other in which aa 1 to 141 were fused to YFP or HA tags (Fig. (Fig.1C).1C). The constructs were transfected into RH/hxgprtΔ parasites and expressed protein visualized by fluorescence microscopy (YFP) or by immunofluorescence assays (HA). Constructs with the full-length ZFP1 tagged with either YFP or HA localized to the nucleolus (Fig. (Fig.4).4). Constructs in which aa 1 to 141 were fused to YFP or HA lacked the CCHC zinc finger domains and localized to the parasite nucleus (Fig. (Fig.4).4). This suggests that the CCHC domains may function as nucleolar targeting signals, as has been demonstrated in other systems (44). We were not able to get stable transformants with any of the four constructs in which ZFP1 was expressed from a strong tubulin promoter, indicating that overexpression may be toxic to the parasites.
We confirmed our results by expression of ZFP1-YFP (1- to 141-aa fusion) under the control of the endogenous ZFP1 promoter. The overall expression with this construct was extremely low; however, we could identify parasites in which, as expected, nuclear localization was seen (Fig. (Fig.4,4, row 4). This strongly indicates that the localization data obtained by expression under a strong tubulin promoter are correct and not a result of mistargeting of the protein due to overexpression.
We analyzed ZFP1 levels in WT and TBD-6 parasites. Lysates from tachyzoites and bradyzoites (72 h after bradyzoite induction) were analyzed by using a mouse polyclonal antisera (to region D) (Fig. (Fig.1B).1B). This revealed a parasite-specific band of 65 kD in both WT and TBD-6 parasites (Fig. (Fig.5A).5A). Using anti-tubulin antibodies to normalize for parasite loading, anti-SRS9 antibodies to confirm bradyzoite conversion, and densitometric analysis of the bands, we identified that in WT and TBD-6 parasites the ZFP1 was significantly lower in bradyzoites than in tachyzoites (Fig. (Fig.5B).5B). Using three independently generated tachyzoite and bradyzoite preparations, the ZFP1 bradyzoite abundance was 39% ± 7% compared to tachyzoites. These results confirm the Northern blot results in which we were unable to detect ZFP1 mRNA under bradyzoite conditions. The upstream disruption mutant had similar results to TBD-6 for both ZFP1 and SRS9 levels and antisera to regions C and D both gave similar results (data not shown).
Tachyzoite-to-bradyzoite interconversion is an essential component of the asexual life cycle of T. gondii and responsible for the majority of disease in adults (23, 42). We and others have previously generated tachyzoite-to-bradyzoite differentiation mutants and used these to characterize bradyzoite biology (24, 30). However, the genetic loci responsible for the Tbd− phenotype were not identified in these mutants. We have now generated two Tbd− mutants by insertional mutagenesis and shown the feasibility of this approach for generating and characterizing the genetic loci of interest. Importantly, and similar to what has previously been demonstrated, both Tbd− mutants have leaky phenotypes in that a subset of parasites still differentiate under bradyzoite conditions. However, we confirmed that insertional mutagenesis can generate mutants with phenotypes as robust as those generated with chemical mutagenesis (24). This indicates that disruption of a single genomic locus is sufficient to confer a robust Tbd− phenotype.
In TBD-6, the mutagenesis vector disrupted the promoter of a T. gondii zinc finger protein containing CCHC motifs. Zinc finger motifs are very common and reported to occur in up to 1% of the mammalian genome (17). Eukaryotic proteins with these motifs are thought to be involved in RNA binding or single-stranded DNA binding (25, 38). CCHC zinc finger genes can also have diverse roles in transcriptional modulation, translational control, protein degradation pathways, and growth regulation (6, 27-29, 35). In T. gondii, transcriptional and translational control mechanisms play important roles in stage differentiation. A large number of stage-specific genes are transcriptionally regulated, although the mechanisms of transcriptional control for most of these have not been elucidated (12, 19). Translational control mechanisms are also important in regulating differentiation in T. gondii. A number of posttranscriptionally controlled, stage-specific genes have been identified as well as the recent identification of a parasite-specific eukaryotic initiation factor-2 (eIF2) kinase required for stress-induced translation control (21, 36, 45). In addition, mRNA stability is an important independent factor in regulating tachyzoite- and bradyzoite-specific transcript abundance in T. gondii (11). Tachyzoites and bradyzoites have vastly different growth rates and reduced replication of tachyzoites appears to be a prerequisite for bradyzoite development (5). Interestingly, some of the Tbd− mutants identified previously grew faster than WT parasites under bradyzoite-inducing conditions (24).
The ZFP1 localizes to the nucleolus with targeting dependent on a region that includes the CCHC motifs. In other systems, CCHC zinc finger proteins without NLS or RNA recognition motifs but with nucleolar localization have been identified (44, 46). In these situations, consistent with our observations with ZFP1, nucleolar targeting is apparently mediated by the zinc finger domains. In addition, although no NLS motifs are identified in ZFP1 in aa 1 to 141 (http://cubic.bioc.columbia.edu/predictNLS/), this region did target the nucleus. Another algorithm (http://psort.nibb.ac.jp/) does identify a potential NLS motif in this region that may thus be responsible for the nuclear targeting. In TBD-6, disruption of the upstream region of ZFP1 altered the RNA size distribution. However, all attempts to incur significant changes in expression levels of ZFP1 (either by knockout or overexpression) failed, suggesting that such changes are not tolerated by the parasites. Multiple instances where overexpression of zinc finger proteins is toxic and/or induces cellular apoptosis have been reported (10, 18, 22). Given the diverse and essential roles of zinc finger proteins, it is not surprising that altered levels are toxic to cells.
ZFP1 may be needed to initiate differentiation but is apparently less important for bradyzoite maintenance. In order to regulate the initiation of differentiation, ZFP1 is present under tachyzoite conditions. The presence of this protein may therefore provide a response mechanism for tachyzoites when they encounter stressful conditions, such as those that may trigger stage conversion. However, once conversion to bradyzoites has occurred, levels of ZFP1 decrease, indicating that this protein may be less important for bradyzoite maintenance. There are reports in other systems where, similar to our observation, genes involved in regulating development are present at the onset of stage conversion but decrease once conversion is under way (9, 15, 16). We did not identify any obvious difference between WT and TBD-6 in ZFP1 protein levels under the limited tachyzoite and bradyzoite conditions tested. Whether there are subtle changes in protein levels, kinetics, or localization of ZFP1 in the differentiation mutants is not clear at present. Not unexpectedly, there appears to be redundancy in the bradyzoite induction pathway. Alternate pathways that regulate stage conversion (independent of ZFP1) clearly exist, since harsher or prolonged stresses are able to drive TBD-6 to convert to bradyzoites at WT levels.
Hence, our study indicates that the forward genetic approach can identify genetic regulators of tachyzoite to bradyzoite stage conversion. Understanding the developmental biology of the parasite will aid greatly in the development of treatment and preventive options.
This study was supported by a Career Award in the Biomedical Sciences by the Burroughs Wellcome fund to U.S. and a grant from the NIAID (AI41014) to J.C.B.
We gratefully acknowledge the help of Mark Gilbert for assistance with fluorescence-activated cell sorting analysis, Seon-Kyeong Kim for the gift of the SRS9 antibody, David Roos for the ptubFNRYFP construct, and Peter Bradley for the pNotI-HA-Gra3′ construct. We extend a special thanks to all of the members of the Singh and Boothroyd labs for helpful suggestions and discussions.
Editor: J. F. Urban, Jr.