Increased efficiency of homologous recombination ... - Semantic Scholar

Molecular & Biochemical Parasitology 153 (2007) 149–157

Increased efficiency of homologous recombination in Toxoplasma gondii dense granule protein 3 demonstrates that GRA3 is not necessary in cell culture but does contribute to virulence Mary Patricia J. Craver, Laura J. Knoll ∗ Department of Medical Microbiology and Immunology, University of Wisconsin-Madison, 1300 University Avenue, Madison, WI 53706, United States Received 10 January 2007; received in revised form 26 February 2007; accepted 28 February 2007 Available online 4 March 2007

Abstract Toxoplasma gondii possesses unique secretory organelles, which synchronously release proteins during and after invasion. One of these organelles, the dense granules, secrete proteins after invasion which are thought to be important in development of the parasite throughout all stages of its life cycle. Dense granule protein 3 (GRA3) is a 30 kDa protein localized to the intravacuolar network and parasitophorous vacuole membrane (PVM). Like many dense granule proteins, GRA3 has no homology to proteins with described functions. However, it has been hypothesized to be involved in nutrient acquisition for the parasite due to its localization on the PVM. To begin to investigate the importance of GRA3, the locus was disrupted by homologous replacement with a chloramphenicol resistance gene in a type II strain. Two GRA3 strains were obtained after two independent electroporations with efficiency greater than 80%. No differences between wild-type and GRA3 were detected in cell culture growth rate or bradyzoite formation. Location of other parasite dense granule proteins and association with host cell organelles were also not affected in GRA3. Interestingly, at an infectious dose approximately four-fold above the lethal dose 50% for wild-type parasites, all mice infected with GRA3-2 infected mice survived acute infection. Complementation of GRA3 expression in the GRA3-2 strain restored virulence to wild-type levels, and increased the virulence of the GRA3-1, confirming that the GRA3 protein plays a role during acute infection in a type II strain. © 2007 Elsevier B.V. All rights reserved. Keywords: Toxoplasma; GRA3; Dense granules

1. Introduction Toxoplasma gondii is an obligate intracellular protozoan parasite capable of infecting any warm-blooded animal. T. gondii has its sexual cycle within the feline intestinal epithelium, from which it is excreted as oocysts within the cat feces. During the asexual cycle in non-feline hosts, T. gondii disseminates as a fast replicating form called a tachyzoite that is responsible for acute infection symptoms. In response to some as yet unknown signal, tachyzoites slow their growth and differentiate into an encysted form called a bradyzoite that is responsible for chronic infection [4,18]. Despite its ability to reproduce sexually and its broad geographic range, T. gondii has a clonal population structure comprised principally of three lines [13]. Type I strains are

∗

Corresponding author. Tel.: +1 608 262 3161; fax: +1 608 262 8418. E-mail address: [email protected] (L.J. Knoll).

0166-6851/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.molbiopara.2007.02.013

highly virulent with a single viable parasite causing lethality to mice as tachyzoites overwhelm the spleen, lungs, and central nervous system (6–10 days post-infection). Type II strains are less virulent, requiring much higher numbers of parasites for a lethal dose during acute infection [12,30]. Interestingly, type I and type II strains differ genetically by only 1% [34]. Research is currently on going to determine genes important for acute virulence, and a few proteins have been identified through forward and reverse genetic strategies (CAT, GRA2, MIC1, MIC3, SAG3, CPSII, ROP16, ROP18) [5,8,10,11,21,27,35]. T. gondii is capable of invading any nucleated cell and replicates within a parasitophorous vacuole (PV) separate from the host endocytic pathway. This vacuole is encased by host cell microtubules, endoplasmic reticulum (ER), and mitochondria. T. gondii possesses three organelles which secrete proteins sequentially during invasion and development of the PV [3,7,32]. These organelles are the micronemes (secrete proteins for initial recognition and adhesion), rhoptries (secrete proteins during initial

150

M.P.J. Craver, L.J. Knoll / Molecular & Biochemical Parasitology 153 (2007) 149–157

formation of the PV), and dense granules (secrete after PV formation). Dense granules are named for their density when observed by transmission electron microscopy (TEM) [3,19]. Dense granules release proteins from both the apical and posterior ends of the parasite after fusion with the plasma membrane and the proteins are among the most abundantly transcribed in T. gondii [19,25]. Once released, the proteins are targeted to the vacuolar space (Cy-18, TgPI-1, TgPI-2, GRA1, 2), PV membrane (GRA3, 5, 7, 8, 10) or intravacuolar tubular network which is thought to be important for structure of the PV and nutrient acquisition by the parasite from the host cell (GRA2, 3, 4, 6, 7, 9, NTPase) [19]. In general, dense granule proteins are believed to be important in development of the parasite throughout all stages of the life cycle. This is due to the fact that these proteins are coordinately secreted. After invasion, they are localized within and on the PV, a vital structure for growth of the parasite within the host, and their transcription is tightly controlled during different life stages [19,28]. It is interesting to note that there is a higher abundance of dense granule organelles and proteins in foodborne coccidians (T. gondii, Neospora and Sarcocystis), pointing to a possible role of these organelles in cyst development [19]. Besides those proteins with known functional homology, several dense granule proteins have no homology to other proteins in the databases. Knock-out strategies have started to uncover the roles these proteins play in the parasites’ life cycle. GRA2, 4, 5, 6, and 7 have been disrupted in the type I (RH) strain background [7,20–22]. GRA2 knock-out parasites are the only ones described in the literature to have a virulence defect in mice. GRA2 knock-out parasites were also shown to be defective in the formation of the intravacuolar network. GRA6 and GRA2/GRA6 double knock-out mutants confirmed the role GRA2 plays in the initial formation of the network while GRA6 aids in its stabilization [20]. GRA2, 4, and 6 have also been shown to form a complex within the network and it is known that GRA3 and GRA5 are excluded from this complex [17]. GRA7 has been characterized as an important factor for the formation of Host Organelle-Sequestering Tubulo-structures (H.O.S.T.) which mediate import of host lysosomes for cholesterol acquisition by the parasite [7]. GRA3 is located on chromosome X, and it encodes a 30 kilodalton type I transmembrane containing protein localized to both the PVM and intravacuolar network. Its presence on the PVM, like other GRA proteins, points to a potential role in nutrient acquisition from the host and/or host organelle association. A recent publication noted a C-terminus ER retrieval motif present on the protein which may aid in the recruitment of ER to the PVM [14]. However, this is not expected since the GRA3 is predicted to be oriented with the C-terminus inside the PVM after secretion. Also, the type II form of GRA3 has a truncated motif due to polymorphisms in the C-terminus which shortens the protein by two amino acids. GRA3 transcription is on during tachyzoites but turned off during the bradyzoite stage indicating a possible role of the protein in development between these two stages [28]. Here, we create a GRA3 deficient type II parasite to begin a functional characterization of the GRA3 protein. The strategy used to obtain homologous recombinant clones resulted in the highest reported recombination

frequency for a gene lacking direct phenotypic selection in T. gondii. In vitro, no visible differences were observed between knock-out and wild-type parasites. However, in vivo one GRA3 deficient parasite strain showed a significant decrease in virulence, and complementation increased the virulence of both knock-out strains. 2. Methods 2.1. Parasites and cell culture T. gondii PRU HXGPRT (wild-type) and all clones obtained were serially passaged in human foreskin fibroblast (HFF) cells in Dulbecco’s modified eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 1% penicillin–streptomycin (Gibco), and 2 mM l-glutamine [8]. The 20 ␮M chloramphenicol (Cm) was also supplemented to strains with the Cm resistance cassette. 2.2. GRA3 targeting construct A knock-out construct (pMPC5) was designed to replace the GRA3 promoter region and start codons with a Cm resistance expression cassette (Fig. 1A). 3751 bp and 5859 bp 5 and 3 flanking regions, respectively, surrounded the disruption site and the entire disrupted locus was placed in pminiHXGPRT to allow for positive (Cm) and negative selection (6-thioxanthine) of clones [9,24]. Initially 5 -GCGGGTTGTTGCCTGTAGTC3 and 5 -GCCTGAGATTCCGCACATTT-3 amplified a central GRA3 genomic region that was digested and cloned into pminiHXGPRT SacI to NotI (pMPC3). A 5 region was amplified with 5 -CCCGGGTTGCTTCTGCAGACTCGTC-3 containing a SmaI site and 5 -GTAGGTGCAGGTGTCCCACTGT-3 . The product was digested SmaI to SacI and cloned into pMPC3 NaeI to SacI (pMPC4). pT/230-TUB/5CAT was digested with HindIII, blunted, and then cut with SacI to obtain the Cm resistance cassette [33]. A three part ligation included a StuI to NotI piece of GRA3 from pMPC3, the Cm resistance cassette, and pMPC4 as the vector backbone cut SacI to NotI. A 3 extension was added by amplifying the GRA3 genomic region with 5 -GTGGCAACTCGCATAAGCAGAT-3 and 5 -ACTAGTGTCCCTCTGAACGCCTGGTG-3 containing a SpeI site. The 3 extension was added to the above construct by sub-cloning to create pMPC5 (Fig. 1A). 2.3. Transfection and selection for GRA3 knock-out and complemented strains 1 × 107 PRU tachyzoites were electroporated with 25 or 50 ␮g of pMPC5 linearized with XhoI. A no DNA control was also used to monitor the effect of drug selections. After 4 h, the media was changed and supplemented with 20 ␮M Cm for positive selection of stable transformants. After 2 weeks, 340 ␮g/ml 6-thioxanthine (6-TX) was added to the Cm selection media for an additional week to select against randomly integrated clones and select for clones which underwent homologous recombination at the GRA3 locus. During this week, parasites were passed


151

Fig. 1. Targeted disruption of the GRA3 locus. (A) The schematic shows the GRA3 locus and construct used for double homologous recombination at the locus. Selectable markers CAT and HXGPRT are indicated as vertical and diagonal shaded boxes, respectively. The GRA3 ORF is indicated by a solid box. Enzyme sites shown illustrate the expected sizes of wild-type versus knock-out Southern bands. The 2.2kb NdeI to NotI probe for the Southern blot is indicated. (B) Antisense SAGE tags are present in the 3 part of the GRA3 locus and their approximate locations are indicated by three arrows.

3 to 4 times with fresh drug added each time. Parasites were then cloned by limiting dilution in media containing both Cm and 6-TX. For complementation, 1 × 107 parasites from each knock-out strain were electroporated with 25 ␮g pMPC3 linearized with NotI and REMI with NotI was performed for each electroporation [2]. Parasites were stably selected and cloned by limiting dilution with 100 ␮g/ml mycophenolic acid and 50 ␮g/ml xanthine. 2.4. Southern blot analysis Genomic DNA was isolated and digested with NdeI. The resulting blots were probed with a random [32 P] labeled 2.2 kb probe from the locus cut NotI to NdeI (Fig. 1A). A shift up from a 3.5 kb NdeI band to a band of about 14 kb was expected if homologous recombination had occurred from the removal of two NdeI sites at the locus. Complements were selected based on the presence of other bands separate from the 14 kb band indicating incorporation of pMPC3.

2.5. Western blot analysis Reduced protein lysates containing 3 × 106 parasites were loaded into each lane of a 15% SDS-PAGE gel. The gel was transferred to Hybond nitrocellulose (Amersham) and probed with a 1:4000 rabbit polyclonal GRA3 and 1:5000 antirabbit horseradish peroxidase (HRP) antibody [1]. Signals were detected with ECL Western blotting detection reagents (Amersham). The blot was re-probed with 1:5000 polyclonal ␤-tubulin and 1:5000 anti-rabbit HRP to ensure equal loading of parasite proteins. 2.6. In vitro assessment of growth rate and bradyzoite development Parasites were syringe passaged at the 8–16 stage of growth for 2 consecutive passes prior to performing the growth curve to ensure healthy actively replicating parasites. After infecting HFF monolayers with 5 × 104 parasites for 3 h, the monolayers were washed 3 times with PBS and changed to DMEM

152


media described above with 10% or no serum. Three replicate coverslip plates were infected for 12, 24, and 36 h time points containing the knock-out and wild-type strains. Plates were fixed at their designated time points with 3% paraformaldehyde. An immunofluorescence assay was performed with serum from a chronically infected mouse to aid visualization of the parasites for counting. For each strain, 150 vacuoles were counted under each condition and time point. The mean vacuole size and standard deviations are reported from two independent experiments. The ability of wild-type and GRA3-1 and GRA3-2 to develop into bradyzoites in vitro was assessed. Coverslips with HFF monolayers were infected with recently lysed parasites. The media was switched after 3 h to RPMI1640 supplemented 1% FBS, 1% penicillin–streptomycin, and buffered with 50mM HEPES to pH 8 [29]. After 3 days at 37 ◦ C and ambient CO2 , monolayers were fixed with 3% paraformaldehyde. An immunofluorescence assay (IFA) dually stained bradyzoite cysts with 1:300 Dolichos biflorus (DB) lectin conjugated to fluorescein (Vector) and 1:1000 anti-BAG1 rabbit polyclonal antibody with 1:300 rhodamine anti-rabbit secondary antibody (Jackson). The percentage of complete cyst wall formation (homogeneous staining of PV) and PVs positive for BAG1 were calculated for 150 vacuoles for each strain in two independent experiments. The bar graphs represent the mean and standard deviation for two experiments. 2.7. Immunofluorescence microscopy Vero cells (6 × 104 ) were attached to coverslips for 24 h prior to infection with 4.5 × 104 parasites. After 24 h of growth, cells were either fixed with 3% paraformaldehyde or underwent mitochondria staining with Mitotracker Red CMXRos (Molecular Probes) prior to fixation [32]. Mitotracker stained cells were permeabilized and blocked with 3%BSA/.2% triton in PBS prior to co-staining with GRA2. Cells used for calnexin, mouse chronic serum, ␣-tubulin, and GRA7 staining underwent cold acetone permeabilization [31]. Then 3% BSA was used for blocking and staining. All washes were performed with either 0.2%triton/PBS or PBS alone depending upon the permeabilization method used. The following concentrations for primary antibodies were used, 1:300 polyclonal rabbit anti-GRA2 and anti-GRA7, 1:50 rabbit anti-calnexin, 1:300 mouse chronic serum. Alexa Fluor conjugated secondary antibodies (Molecular Probes) were used at 1:100 for calnexin staining and 1:200 for all others. Samples were mounted using VectaShield mounting media with DAPI (Vector Laboratories). Serial image stacks (0.2 micron Z-increment) were collected at 63× (PlanApo oil immersion 1.4 na) with OpenLabs 4.0 software (Improvision Inc.) on a motorized Zeiss Axioplan IIi equipped with a rear-mounted excitation filter wheel, a triple pass (DAPI/FITC/Texas Red) emission cube, differential interference contrast (DIC) optics, and a Hamamatsu ORCA-AG CCD camera. Fluorescence images were deconvolved and pseudocolored using Volocity 4.0.1 software (Improvision Inc.).

2.8. Transmission electron microscopy Parasites were grown for 24 h in fibroblasts, trypsinized, and spun down at 200 × g for 5 min. Cells were washed once in phosphate buffer (100 mM, pH 7.2) then fixed in 2% paraformaldehyde/2.5% glutaraldehyde (Polysciences Inc.) in phosphate buffer for 1 h at 24 ◦ C. Cells were washed in phosphate buffer and post fixed in 1% osmium tetroxide (Polysciences Inc.) for 1 h. The cells were then rinsed extensively in dH2 O prior to en bloc staining with 1% aqueous uranyl acetate (Ted Pella Inc.) for 1 h. Following several rinses in dH2 O, samples were dehydrated in a graded series of ethanol and embedded in Eponate 12 resin (Ted Pella Inc.). Sections of 70–80 nm were cut, stained with uranyl acetate and lead citrate, and viewed on a JEOL 1200 EX transmission electron microscope. 2.9. In vivo analysis For acute infection, 6–7 week old CBA/J mice (NCI) were injected intraperitoneally (i.p.) with 4 × 105 parasites of each strain. To prepare parasites for injections, 10 mM HEPES was added to T25 infected flasks prior to syringe lysis. Parasites were spun at 1400 × g for 10 min and then resuspended in 500 ␮l of DMEM/10 mM HEPES for counting. After spinning, parasites were kept on ice for the duration of the procedure. Parasites were diluted to obtain a final concentration of 4 × 105 parasites/200 ␮l for injections in DMEM/10 mM HEPES. After injections, remaining parasites were diluted 1:10 for a plaque assay on HFFs to confirm equal numbers of viable parasites were injected. When mice were moribund (severely hunched and not moving) they were euthanized. The percentage of surviving mice during acute infection (to day 13) was calculated and plotted on a graph. Surviving mice were euthanized after 22 days. Mouse data is a compilation of 5 experiments with 4 mice per strain for each experiment (GRA3-2 had three mice for one experiment). The complemented strains were examined in 4 experiments with 4 mice each for a total of 16 mice per strain. Differences in virulence were determined to be significant using Fisher’s exact test, comparing the ratio of mice succumbing to infection from wild-type, GRA3, or complement parasites at day 13. 3. Results 3.1. Generation of GRA3 and complemented strains The goal of this study was to investigate the functional role of the GRA3 protein in a type II PRU strain. The initial strategy was to delete a 2.6 kb region of the GRA3 locus which spanned the entire open reading frame (ORF), promoter and untranslated regions (SacI to NotI region deleted in Fig. 1A). After four independent electroporations, a total of 108 clones were screened by Southern blot resulting in no knock-out clones (data not shown). Upon revisualization of the GRA3 locus, the presence of SAGE tags (http://www.toxodb.org/) indicated the


153

Fig. 2. Verification of GRA3. Panel (A) Southern blot of wild-type (WT), GRA3-1, GRA3-2, and complemented strains (GRA3-1/C1, GRA3-1/C2, GRA3-2/C1 and GRA3-2/C2). Arrows indicate the shift from 3.46 to about 14 kb in the knock-outs. Complemented strains GRA3-1/C1 and GRA3-1/C2 contain a single copy of the complement construct, while GRA3-2/C1 and GRA3-2/C2 contain multiple copies. Panel (B) Western blot with equally loaded parasites (␤-tubulin) shows the absence of GRA3 in the knock-out strains and restoration of expression of GRA3 protein in the complemented strains. Expression of GRA3 in the GRA3-1/C2 strain was consistently lower than the other complemented strains.

possibility of an antisense transcript overlapping the 3 end of this locus (Fig. 1B). To prevent disruption of the possible overlapping antisense transcript, an alternative strategy to delete the GRA3 protein alone was undertaken. This strategy deleted a 506 bp piece (SacI to StuI) that spans the promoter region and 72 bp past the second in frame ATG for GRA3 replacing the deletion with the positive selectable marker (Fig. 1A). A total of 34 clones from two separate electroporations were screened by Southern blot. This knock-out strategy was highly successful resulting in 81% and 83% clones with a disrupted GRA3 locus. One clone was selected from each electroporation and called GRA3-1 and GRA3-2 (see Fig. 2A for Southern blot). Disruption of GRA3 protein was confirmed by Western blot (Fig. 2B). Complemented strains were generated for GRA3-1 and GRA3-2 using the genomic construct pMPC3, which restores the entire endogenous GRA3 locus. Integration of the construct in stably selected clones was verified by Southern blot analysis (Fig. 2A). Two complement clones for each GRA3 strain were selected. GRA3-1 complement clones (GRA31/C1, GRA3-1/C2) have single inserts of the construct, while GRA3-2 complements (GRA3-2/C1, GRA3-2/C2) have multiple insertions of the construct (Fig. 2A). Expression of GRA3 protein is restored in all complement clones; however total GRA3 protein is consistently lower in GRA3-1/C2 (Fig. 2B). 3.2. In vitro growth rate To determine if the lack of GRA3 caused a growth defect in cell culture, the rate of division was assessed by counting the amount of parasites per vacuole at 12, 24 and 36 h time points. Disruption of GRA7, another dense granule protein localized to the PV membrane and important for nutrient acquisition, was found to have a greater growth defect in limiting serum conditions (2.5% serum) [7]. To investigate the possibility that GRA3 may also play a role in nutrient acquisition, concentrations of serum from 0 to 10% were used in a growth assay. No difference

was seen in the GRA3 growth rate in vitro, even with no serum supplementation (Fig. 3A). 3.3. In vitro bradyzoite development Since dense granule proteins are thought to play a role in development of the PV in tachyzoites as well as bradyzoites, the ability of the wild-type and GRA3 parasites to differentiate was assessed by performing a 3 day in vitro switch. Two markers of bradyzoite development were visualized by immunofluorescence; the cyst wall protein CST1, and the bradyzoite specific heat shock protein BAG1 [36,37]. There is no difference between wild-type and GRA3 parasites in their ability to differentiate in vitro (Fig. 3B). 3.4. Immunofluorescence microscopy of T. gondii and host cell markers GRA3 is present on both the intravacuoler network membranes and PVM. To assess if other dense granule proteins localization to these areas was disrupted in the absence of GRA3, an IFA was performed using antibodies against GRA2 and GRA7. GRA2 was selected because of its location in the proposed network complex within the intravacuolar network. GRA7 was selected due to its presence on the PVM and network membranes. GRA2 was not expected to be defective in GRA3-1 due to the absence of GRA3 in the network protein complex [17]. Both GRA2 and GRA7 localization do not appear to be affected in GRA3-1 (Fig. 4B and C). Host cell organelle association with the PVM was investigated in both wild-type and GRA3-1 parasites. GRA3’s presence on the PVM surface has led to hypotheses about the function of GRA3 in host cell organelle association, specifically the ER [14]. Also, host cell microtubules wrapped around the PVM are known to play a role in nutrient acquisition by the parasite [7]. To investigate the role of GRA3 in host organelle association, host cell ER, ␣-tubulin, and mitochon-

154


3.6. In vivo studies of GRA3 parasites To assess the role of GRA3 during infection, we performed i.p. inoculation of mice with 4 × 105 tachyzoites: a dose shown previously to be approximately four-fold above the lethal dose 50% (LD50 ) for wild-type parasites [23]. Sixty percent of mice infected with GRA3-1 died during acute infection (6–13 days), lower, but not significantly, from the 75% of mice that died during acute infection of wild-type parasites (Fig. 6A). Interestingly, all GRA3-2 infected mice survived acute infection (Fig. 6B). To confirm that the absence of GRA3 protein was responsible for the reduction of virulence in GRA3-2, we examined the virulence of the GRA3-2 complemented strains in mice. Infections of mice with GRA3-2/C1 and GRA32/C2 (Fig. 6B) restored virulence by lethality during acute infection (p < 0.0001). Complementation of the GRA3-1 strain showed a significant increase (p < 0.005) in virulence in at least one of the complemented strains (GRA3-1/C1, Fig. 6A). As noted previously, the GRA3-1/C2 consistently had lower expression of GRA3 by Western blot, than the other complemented strains, which may explain the non-significant increase in virulence by that strain. Examination of the in vivo bradyzoite cysts from the acute infection survivors by DB-FITC staining of brains showed all mice were infected and had wild-type size and shape of cysts. 4. Discussion

Fig. 3. In vitro characterization of GRA3. Panel (A) growth curve showing average parasite numbers per vacuole at 12, 24, and 36 h. Closed shapes indicate strains grown in DMEM with 10% serum, open shapes indicate strains grown in DMEM with no serum. For each strain, 150 vacuoles were counted with the mean and standard deviation reported from two independent experiments. Panel (B) three day in vitro bradyzoite switch of GRA3-1, GRA3-2, and wildtype. IFA detected cyst wall completeness and BAG1 positive vacuoles as a measure of bradyzoite development. For each strain, 150 vacuoles were counted in two separate experiments. The mean of both experiments is reported with the standard deviation.

dria were observed by IFA. All three markers appear to be unchanged in GRA3-1 parasites compared to wild-type parasites (Fig. 4A–C). 3.5. Transmission electron microscopic analysis of GRA3 To visualize the intravacuoler network, PVM, and host cell organelle association at a higher magnification, we performed Transmission Electron Microscopy (TEM) on wild-type and GRA3-1 parasites. TEM showed that the intravacuoler network membranes and PVM appear unchanged in GRA3-1 parasites (Fig. 5). TEM also confirmed that host cell mitochondria and ER association is preserved in GRA3-1 parasites (Fig. 5). Taken together, these cell culture analyses demonstrate no apparent in vitro defects in GRA3 parasites.

GRA3 has successfully been disrupted in the type II strain of T. gondii with an increased efficiency of 81–83%. In general, the frequency of obtaining homologous recombination at loci lacking direct phenotypic selection or screening in T. gondii is low. In RH, the frequency has been reported to be as high as 20% with the disruption of MIC6, and in the type II PLK strain BAG1 was disrupted at a frequency of 16% [26,37]. However, the frequency of obtaining homologous recombinant clones is usually lower. The reason for the increased frequency of recombination at the GRA3 locus may be explained by a couple reasons. First of all, the positive and then negative selection strategy utilized here is a powerful tool for obtaining homologous recombinant clones. Most published strategies for obtaining knock-out clones utilize a single positive selection prior to screening, although one report used a similar positive and then negative selection described in this paper [20]. Secondly, we used large fragments of genomic DNA flanking the positive selectable marker to enhance homologous recombination. A significant increase in recombination occurred when the total flanking DNA was doubled from 8 to 16 kb in another study [9]. Therefore, the combination of positive and negative selection as well as large flanking DNA is important for efficient gene disruption in T. gondii. An additional reason for the increase in recombination rate is that the GRA3 locus itself may favor homologous recombination. During the tachyzoite stage of growth, histones are acetylated at the GRA3 locus and transcription is high ([28], and Northern blot analysis, data not shown). This may make the locus more accessible to homologous recombination. However, it is interesting to note that recombination did not occur with a con-


155

Fig. 4. Parasite and host cell markers are unchanged in GRA3-1. We examined parasite (green) and host (red) markers in wild-type (left three panels) and GRA31 strains (right three panels). Row (A) uses chronic serum to label the parasites (CS), and anti-Calnexin antibodies to label the host ER (Calnex). Row (B) uses anti-GRA2 to strain the parasite intravacuolar network and Mitotracker to label the mitochondria (Mito). In row (C) anti-GRA7 was used to visualize the PVM and network membranes, and anti-␣-tubulin marked the host microtubules (␣-tub).

struct containing a larger segment of the locus disrupted from SacI to NotI. Besides the SAGE tags, the presence of a histone deacetylase gene (HDAC3), near the 3 end of the GRA3 locus may have affected the recombination ability of the larger deletion construct. HDAC3 is likely an essential protein for T. gondii

due to its role in controlling global gene expression in the parasite during different developmental stages [28]. Although the GRA3 locus may be contributing to the increased frequency of homologous recombination, we have seen greater than 50% recombination rates using the above described selection strat-

Fig. 5. Transmission electron microscopy of GRA3-1 and wild-type. The top panels show the presence of host ER and mitochondria (M) at the PVM in both wild-type (left two panels) and GRA3-1 (right two panels). The bottom panels show the presence of an intact intravacuolar network (IVN) within both GRA3-1 and wild-type vacuoles. Scale bars are 200 nm.

156


Fig. 6. Disruption of GRA3 allows survival of mice during acute infection. Panel (A) shows the percent survival of mice infected with wild-type (WT), GRA3-1, and its two complemented strains (GRA3-1/C1 and GRA3-1/C2). In panel (B) the percent survival of mice infected with wild-type (WT), GRA3-2, and its two complemented strains (GRA3-2/C1 and GRA3-2/C2) is shown. Experiments in panels A and B were performed at the same time, and wild-type is shared between the two panels. Asterisks with solid line are for wild-type, open triangles with dotted line are for GRA3, while solid squares and circles with dotted line are complement infected mice. The graph shows up to day 13, surviving mice were euthanized after 22 days. The results are from five experiments of four mice each for the wild-type and GRA3 strains (GRA3-2 only had three mice for one experiment), and four experiments of four mice each for the complemented strains. One GRA3-2 mouse died day 21 during the course of one experiment. Using Fisher’s exact test, the increase in virulence of GRA3-1/C1 compared to GRA3-1 is statistically significant at p < 0.005. The difference in survival of GRA3-2 versus wild-type or complement infected mice is statistically significant at p < 0.0001.

egy with another, less actively transcribed gene (Van and Knoll, unpublished results). In this study, we show that disruption of one of the more abundant proteins in T. gondii, GRA3, has no phenotype differences in cell culture compared to wild-type, however, it has a slight virulence defect in mice. Although GRA3-1 does not have a significant difference in virulence like GRA3-2 compared to wild-type, restoration of wild-type GRA3 protein increased virulence in both knock-out strains. All strains were maintained and passed identically during the above experiments, which should minimize any effects of serial passage on virulence. However,

this does not rule out that differences between the two knock-out strains occurred during their passage which altered their virulence independent of the GRA3 disruption. Nevertheless, we believe GRA3 plays a minor role in acute virulence for type II strains due to the increase in virulence when both strains are complemented with wild-type levels of GRA3. Interestingly, analysis of type I and II GRA3 genes shows that there are a large amount of polymorphisms between the two different alleles. These polymorphisms result in one synonymous and 20 nonsynonymous amino acid changes between the type I and II GRA3 proteins (data not shown). The rate of polymorphisms in all three lineages of T. gondii is between 1 and 3% at the amino acid level while GRA3 has a rate of 9.4% between the type I and II lineages [6]. In fact, the GRA3 monoclonal antibody derived after an RH infection does not react with the type II GRA3 protein because of a single amino acid change indicating a difference in antigenicity between the two proteins [16]. Nonsynonymous changes in proteins may be driven by adaptive immune pressures as shown in Plasmodium falciparum and several other pathogens [15]. These changes indicate GRA3 may be under a positive selective pressure, and could point to a particular role the protein plays in strain specific host cell interaction or immune system evasion. In the future, it would be interesting to look at a type I derived GRA3 knock-out to see if there are any differences in virulence between the two strains. Overall, we have combined strategies used in the field previously to disrupt GRA3 at the highest frequency reported for a gene in T. gondii. We hope this strategy will aid future researchers in the investigation of genes lacking direct phenotypic selection to further unravel the functions of other genes in T. gondii. Acknowledgements We sincerely thank Jay Bangs for the use of his microscope, Wandy Beatty and Darcy Gill for the electron microscopy, David Sibley for the use of his lab during our EM studies, ToxoDB (http://www.toxodb.org/), and the following individuals for antibodies: Louis Weiss (BAG1), Keith Joiner (GRA3), Isabelle Coppens (GRA7, host (-tubulin), Anthony Sinai (host Calnexin), and David Sibley (GRA2, ␤-tubulin). This research was supported by the National Institutes of Health (NIH), Award A1054603 (L.J.K.), Regional Center of Excellence V ‘Great Lakes’ Award 1-U54-AI-057153 (L.J.K.) and NIH National Research Service Award T32 AI007414 (M.P.J.C.). References [1] Bermudes D, Dubremetz JF, Achbarou A, Joiner KA. Cloning of a cDNA encoding the dense granule protein GRA3 from Toxoplasma gondii. Mol Biochem Parasitol 1994;68:247–57. [2] Black M, Seeber F, Soldati D, Kim K, Boothroyd JC. Restriction enzymemediated integration elevates transformation frequency and enables cotransfection of Toxoplasma gondii. Mol Biochem Parasitol 1995;74:55–63. [3] Black MW, Boothroyd JC. Lytic cycle of Toxoplasma gondii. Microbiol Mol Biol Rev 2000;64:607–23. [4] Buzoni-Gatel D, Werts C. Toxoplasma gondii and subversion of the immune system. Trends Parasitol 2006;22:448–52.

M.P.J. Craver, L.J. Knoll / Molecular & Biochemical Parasitology 153 (2007) 149–157 [5] Cerede O, Dubremetz JF, Soete M, et al. Synergistic role of micronemal proteins in Toxoplasma gondii virulence. J Exp Med 2005;201:453–63. [6] Cleary MD, Singh U, Blader IJ, Brewer JL, Boothroyd JC. Toxoplasma gondii asexual development: identification of developmentally regulated genes and distinct patterns of gene expression. Eukaryot Cell 2002;1:329–40. [7] Coppens I, Dunn JD, Romano JD, et al. Toxoplasma gondii sequesters lysosomes from mammalian hosts in the vacuolar space. Cell 2006;125:261–74. [8] Ding M, Kwok LY, Schluter D, Clayton C, Soldati D. The antioxidant systems in Toxoplasma gondii and the role of cytosolic catalase in defence against oxidative injury. Mol Microbiol 2004;51:47–61. [9] Donald RG, Carter D, Ullman B, Roos DS. Insertional tagging, cloning, and expression of the Toxoplasma gondii hypoxanthine-xanthine-guanine phosphoribosyltransferase gene. Use as a selectable marker for stable transformation. J Biol Chem 1996;271:14010–9. [10] Dzierszinski F, Mortuaire M, Cesbron-Delauw MF, Tomavo S. Targeted disruption of the glycosylphosphatidylinositol-anchored surface antigen SAG3 gene in Toxoplasma gondii decreases host cell adhesion and drastically reduces virulence in mice. Mol Microbiol 2000;37:574–82. [11] Fox BA, Bzik DJ. De novo pyrimidine biosynthesis is required for virulence of Toxoplasma gondii. Nature 2002;415:926–9. [12] Gavrilescu LC, Denkers EY. IFN-gamma overproduction and high level apoptosis are associated with high but not low virulence Toxoplasma gondii infection. J Immunol 2001;167:902–9. [13] Grigg ME, Bonnefoy S, Hehl AB, Suzuki Y, Boothroyd JC. Success and virulence in Toxoplasma as the result of sexual recombination between two distinct ancestries. Science 2001;294:161–5. [14] Henriquez FL, Nickdel MB, McLeod R, et al. Toxoplasma gondii dense granule protein 3 (GRA3) is a type I transmembrane protein that possesses a cytoplasmic dilysine (KKXX) endoplasmic reticulum (ER) retrieval motif. Parasitology 2005;131:169–79. [15] Hughes AL. Circumsporozoite protein genes of malaria parasites (Plasmodium spp.): evidence for positive selection on immunogenic regions. Genetics 1991;127:345–53. [16] Kong JT, Grigg ME, Uyetake L, Parmley S, Boothroyd JC. Serotyping of Toxoplasma gondii infections in humans using synthetic peptides. J Infect Dis 2003;187:1484–95. [17] Labruyere E, Lingnau M, Mercier C, Sibley LD. Differential membrane targeting of the secretory proteins GRA4 and GRA6 within the parasitophorous vacuole formed by Toxoplasma gondii. Mol Biochem Parasitol 1999;102:311–24. [18] Lyons RE, McLeod R, Roberts CW. Toxoplasma gondii tachyzoitebradyzoite interconversion. Trends Parasitol 2002;18:198–201. [19] Mercier C, Adjogble KD, Daubener W, Delauw MF. Dense granules: are they key organelles to help understand the parasitophorous vacuole of all apicomplexa parasites? Int J Parasitol 2005;35:829–49. [20] Mercier C, Dubremetz JF, Rauscher B, et al. Biogenesis of nanotubular network in Toxoplasma parasitophorous vacuole induced by parasite proteins. Mol Biol Cell 2002;13:2397–409.

157

[21] Mercier C, Howe DK, Mordue D, Lingnau M, Sibley LD. Targeted disruption of the GRA2 locus in Toxoplasma gondii decreases acute virulence in mice. Infect Immun 1998;66:4176–82. [22] Mercier C, Rauscher B, Lecordier L, et al. Lack of expression of the dense granule protein GRA5 does not affect the development of Toxoplasma tachyzoites. Mol Biochem Parasitol 2001;116:247–51. [23] Mordue DG, Scott-Weathers CF, Tobin CM, Knoll LJ. A patatinlike protein protects Toxoplasma gondii from degradation in activated macrophages. Mol Microbiol 2007;63:482–96. [24] Pfefferkorn ER, Bzik DJ, Honsinger CP. Toxoplasma gondii: mechanism of the parasitostatic action of 6-thioxanthine. Exp Parasitol 2001;99:235– 43. [25] Radke JR, Behnke MS, Mackey AJ, et al. The transcriptome of Toxoplasma gondii. BMC Biol 2005;3:26. [26] Reiss M, Viebig N, Brecht S, et al. Identification and characterization of an escorter for two secretory adhesins in Toxoplasma gondii. J Cell Biol 2001;152:563–78. [27] Saeij JP, Boyle JP, Coller S, et al. Polymorphic secreted kinases are key virulence factors in toxoplasmosis. Science 2006;314:1780–3. [28] Saksouk N, Bhatti MM, Kieffer S, et al. Histone-modifying complexes regulate gene expression pertinent to the differentiation of the protozoan parasite Toxoplasma gondii. Mol Cell Biol 2005;25:10301–14. [29] Schwarz JA, Fouts AE, Cummings CA, Ferguson DJ, Boothroyd JC. A novel rhoptry protein in Toxoplasma gondii bradyzoites and merozoites. Mol Biochem Parasitol 2005;144:159–66. [30] Sibley LD, Mordue DG, Su C, Robben PM, Howe DK. Genetic approaches to studying virulence and pathogenesis in Toxoplasma gondii. Philos Trans R Soc Lond B Biol Sci 2002;357:81–8. [31] Sinai AP, Paul S, Rabinovitch M, Kaplan G, Joiner KA. Coinfection of fibroblasts with Coxiella burnetti and Toxoplasma gondii: to each their own. Microbes Infect 2000;2:727–36. [32] Sinai AP, Webster P, Joiner KA. Association of host cell endoplasmic reticulum and mitochondria with the Toxoplasma gondii parasitophorous vacuole membrane: a high affinity interaction. J Cell Sci 1997;110(Pt 17):2117– 28. [33] Soldati D, Boothroyd JC. A selector of transcription initiation in the protozoan parasite Toxoplasma gondii. Mol Cell Biol 1995;15:87– 93. [34] Su C, Evans D, Cole RH, et al. Recent expansion of Toxoplasma through enhanced oral transmission. Science 2003;299:414–6. [35] Taylor S, Barragan A, Su C. A secreted serine-threonine kinase determines virulence in the eukaryotic pathogen Toxoplasma gondii. Science 2006;314:1776–80. [36] Zhang YW, Halonen SK, Ma YF, Wittner M, Weiss LM. Initial characterization of CST1, a Toxoplasma gondii cyst wall glycoprotein. Infect Immun 2001;69:501–7. [37] Zhang YW, Kim K, Ma YF, et al. Disruption of the Toxoplasma gondii bradyzoite-specific gene BAG1 decreases in vivo cyst formation. Mol Microbiol 1999;31:691–701.