Differential Subcellular Localization of Members of the Toxoplasma gondii Small Heat Shock Protein Family

Abstract

The results of this study describe the identification and characterization of the Toxoplasma gondii α-crystallin/small heat shock protein (sHsp) family. By database (www.toxodb.org) search, five parasite sHsps (Hsp20, Hsp21, Hsp28, Hsp29, and the previously characterized Hsp30/Bag1) were identified. As expected, they share the homologous α-crystallin domain, which is the key characteristic of sHsps. However, the N-terminal segment of each protein contains unique characteristics in size and sequence. Most T. gondii sHsps are constitutively expressed in tachyzoites and fully differentiated bradyzoites, with the exception of Hsp30/Bag1. Interestingly, by subcellular localization we observed that T. gondii sHsps are located in different compartments. Hsp20 is located at the apical end of the cell, Hsp28 is located inside the mitochondrion, Hsp29 showed a membrane-associated labeling, and Hsp21 appeared throughout the cytosol of the parasites. These particular differences in the immunostaining patterns suggest that their targets and functions might be different.

The main group of heat shock proteins includes several classes of proteins differing in molecular weight, structure, and properties. All of them form a group, because they participate in the proper folding of proteins under normal and extreme conditions (19). Specifically, the subgroup of small heat shock proteins (sHsps) comprises diverse stress-inducible proteins with molecular masses ranging from 12 to 43 kDa. The main structure of sHsps consists of a conserved domain of approximately 90 amino acid residues in the C-terminal region and a more variable N-terminal sequence (11). This C-terminal conserved region is also found in the α-crystallin proteins, major proteins of the vertebrate eye lens that together with the sHsps make up the α-crystallin/sHsp superfamily (11).

sHsps have been found in almost all organisms studied (41). In many organisms, several members of the sHsp family are present in one cell compartment, suggesting functional diversity. Although their cellular role is not completely understood, many studies have demonstrated diverse in vivo functions for sHsps, including cytoskeleton protection (27, 30), modulation of the apoptotic process (6), and basic chaperoning activity (16, 22). Proper oligomerization and subunit exchange are critical, since perturbation of both processes, by introducing mutations in the N- or C-terminal sequence, led to loss of chaperone activity (21, 25, 29, 48). After substrate binding, sHsps associate into large aggregates, generally containing 9 to 24 monomers (21, 25, 29, 48). Also, an interesting function of sHsps concerns their transient expression during development and cell differentiation in a wide range of organisms, such as Caenorhabditis elegans (12), Drosophila melanogaster (18, 32, 33), Xenopus laevis (26), Mus musculus (17, 49), human (24), and some parasitic nematodes (20, 44).

Most organisms have just one or a few cytosolically localized sHsps. However, in some organisms the intracellular localization varied with the specific sHsp. As examples, in Drosophila melanogaster Hsp22 was localized to mitochondria, while Hsp23 and Hsp26 were found in the cytoplasm and Hsp27 accumulated in the nucleus (33). Moreover, in plants, which have many diverse sHsps, this group of proteins was found in mitochondria, cytosol, endoplasmic reticulum, and chloroplast (53).

Toxoplasma gondii, an obligate intracellular parasitic protozoan of the phylum Apicomplexa, is one of the most widely distributed protozoan parasites, infecting approximately one-third of the world's population and animals. In humans, the parasite presents two forms: tachyzoites and bradyzoites, both of them corresponding with the asexual cycle. The tachyzoite stage, which causes acute infection, is normally controlled by the onset of the specific immune response in the immunocompetent host. The parasite is able to persist in infected hosts by the differentiation to bradyzoite forms that are presumably protected by the formation of the cyst wall. The mechanisms that regulate the transition between these two stages are not clearly understood. However, the in vitro transition between the tachyzoite-to-bradyzoite stage has been associated with different stress inducers, such as temperature and pH (31, 54). Several studies have established an association between tachyzoite-to-bradyzoite conversion induced by stressors and the up-regulation of different Hsps (15, 51, 55).

Only one T. gondii sHsp (Bag1 or Hsp30) has been recognized so far. The formal name (Bag1) has been proposed on the basis of its differential expression in the bradyzoite stage. Expression of Bag1 protein is unaffected by heat shock stress but is induced early during parasite differentiation from tachyzoite into the latent bradyzoite form, and it appears to be a major protein in mature cyst tissue, localized in the cytoplasm of the parasite (4, 42). Vaccination with this protein enhanced protective immunity against challenge (35, 39). In order to investigate the role of Bag1, knockout mutants were obtained. However, despite its abundance in wild-type bradyzoites, suppression of this protein had no apparent effect on the ability of the parasites to differentiate into bradyzoite cysts. Moreover, no effect on the number, size, morphology, or survival was observed (5). In summary, Bag1 was not shown to be essential in cyst formation (5, 56).

The aims of this work were to isolate and characterize the cDNAs encoding the sHsp family and to evaluate their expression and determine the subcellular localization as a first approach to establish the role of sHsps in the T. gondii parasite.

MATERIALS AND METHODS

Sequence analysis.

Database searches and sequence comparisons were performed using blastn, blastx, and BLAST two-sequence programs (www.ncbi.nlm.nih.gov/BLAST). Preliminary Toxoplasma genomic and/or cDNA sequence data were accessed via http://ToxoDB.org. Expressed sequence tag (EST) databases were also searches by using www.ncbi.nlm.nih.gov/BLAST. The Megalign (DNAstar) program was used for multiple alignment and sequence analysis. The Edit seq (DNAstar) program was used to find putative open reading frames. Hsp30/Bag1 was previously described (4, 42). Prediction of transmembrane helix location and topology was scanned through PredictProtein (Columbia University).

Parasites.

Strain PK tachyzoites (a clone derived from strain Me49 of T. gondii) were cultured in vitro in a human foreskin fibroblast (HFF) cell line monolayer, with Eagle's minimum essential media (Gibco) containing 1% fetal calf serum (Gibco). Tachyzoites were purified from an infected monolayer by filtration through 3-μm-pore-size polycarbonate filters (Nucleopore).

Me49 brain tissue cysts were obtained from C3H mice and homogenized in phosphate-buffered saline (PBS) and centrifuged at 1,800 rpm for 10 min at room temperature. The homogenate was enriched in cysts by three washes with Dextran 30% (Sigma) in PBS buffer. The cysts were quickly removed and stored in TRIzol Reagent (Invitrogen Life Technologies) at −80°C until futures analyses.

DNA sample preparation.

Purified tachyzoites were pelleted and resuspended in TES buffer (Tris-HCl [10 mM], EDTA [150 mM], NaCl [150 mM] [pH 7.4]) plus 1% sodium dodecyl sulfate (SDS) and 200 μg of proteinase K per ml. The samples were incubated for 2 h at 56°C. The DNA was extracted with equal volumes of phenol, chloroform-isoamyl alcohol (24:1), and ether and were precipitated with 2 volumes of ethanol. Finally, the DNA was centrifuged and resuspended in bidistilled water. The DNA concentration was determined at 260 nm in a Shimadzu PR-1 spectrophotometer.

sHsp amplification and expression analysis.

Total RNA was extracted using TRIzol Reagent (Invitrogen Life Technologies) according to the manufacturer's instructions. Two micrograms of total RNA from either tachyzoites or bradyzoites was reverse transcribed to cDNA with 200 U of Moloney murine leukemia virus reverse transcriptase (Invitrogen Life Technologies) and oligo(dT)_12-18 following the manufacturer's instructions. Total RNA from uninfected brain cells of mice and HFF were also isolated to be used as controls.

In order to amplify the different sHsps, reverse transcription-PCR (RT-PCR) was performed in a final volume of 30 μl containing 3 μl of 10× reaction buffer, 4 μl of a 25 mM MgCl₂, 1 μl of a 10 mM deoxynucleoside triphosphate mix, 1 μl of each 10 μM specific primer, 1.25 U of Taq polymerase (Invitrogen Life Technologies), and 1 μl of T. gondii cDNA. After an initial 5-min denaturing step at 95°C, 35 cycles of amplification were performed using a cycle profile of 94°C for 30 s, 55°C for 45 s, and 72°C for 45 s. After the last cycle, a final elongation was performed at 72°C for 10 min. Specific primers used are the following: Hsp20-ORF, CTGTCGTTATTTCTTCTTCA; Hsp20R, GGTACCATTCTCATTCCTCTGCGTCGT; Hsp21-ORF, TGCGCCGAAAGAATACAAGT; Hsp21R, GGTACCACTGCAGATCTGTGTCCGTGTG; Hsp28-ORF, CGCGCCACAACATTCCTCA; Hsp28R, GGTACCTCTCTGTGCGCCTGACTAC; Hsp29-ORF, AAGTTTCACATTCCGTTCG; Hsp29R, GGTACCCATTCACCGTTGCTGTGC (BamHI and KpnI sites are underlined). The resulting PCR product of the expected size was excised using a Qiaex II Gel Extraction kit (QIAGEN), cloned in pGEM-T vector (Promega), and sequenced employing the ABI PRISM Big Dye Terminator Cycle Sequencing Ready Reaction kit (Perkin Elmer Applied Biosystems). All sequences have been deposited in GenBank. Total genomic DNA isolated from PK strain was subjected to PCR amplification with the same primers and conditions used for the RT-PCR.

To analyze the sHsp expression, 2 μg of total RNA from either tachyzoites or bradyzoites was reverse transcribed as described below. RT-PCR analysis was performed with the same PCR condition used for sHsp amplification. Specific primers for expression analysis used are the following: Hsp20F, GGATCCATGAGTTGCTGTGGCGGTAC; Hsp20R, GGTACCATTCTCATTCCTCTGCGTCGT; Hsp21F, GGATCCATGGCATCCCCAACCTCAGC; Hsp21R, GGTACCACTGCAGATCTGTGTCCGTGTG; Hsp28F, GGATCCATGAGCCGCCGAGATCAC; Hsp28R, GGTACCTCTCTGTGCGCCTGACTAC; Hsp29F, GGATTCATGGCAGACTCAAGCGG; Hsp29R, GGTACCCATTCACCGTTGCTGTGC. BamHI and KpnI sites are underlined. Bradyzoite (Hsp30/Bag1)- and tachyzoite (Sag1/P30)-specific genes were included as controls. The primer set used for Bag1/Hsp30 was forward, CTATTTGGATGAAGTAAG; reverse, CGCTGATTTGTTGCTTTG. The primer set for Sag1 gene was forward, GCGCGGATCCATGGTCACGGTGACAGTA; and reverse, GCGCAAGCTTTCACGCGACACAAGCTGCGA.

Semiquantitative RT-PCR assay.

To analyze the sHsp expression during heat stress, PK strain parasites were grown under tachyzoite (37°C) or heat stress (42°C) conditions for 2 h. The RT-PCR semiquantitative analysis, using the same set of primers described above for expression assays, was performed for Hsp20, Hsp21, Hsp28, and Hsp29 genes. RNA samples (2 μg) were reverse transcribed using Moloney murine leukemia virus reverse transcriptase (Invitrogen) and oligo(dT)_12-18 following the manufacturer's instructions. Each set of primers was tested in a range from 18 to 40 cycles, using a 1:5 dilution of a cDNA derived from each sample, in order to know the number of cycles where the product accumulation was in the linear phase of the curve. For each set of primers, the following specific annealing temperatures and cycle numbers were determined: Hsp20, annealing of 59°C and 36 cycles, giving a 607-bp PCR product; Hsp21, annealing of 62°C and 29 cycles, giving a 588-bp PCR product; Hsp28, annealing of 62°C and 29 cycles, giving a 759-bp PCR product; and Hsp29, annealing of 59°C and 35 cycles, giving an 804-bp PCR product. To ensure that equal amounts of cDNA from each parasitic stage were being compared, the primers derived from the T. gondii α-tubulin gene (TubF, 5′-CGACGGTGGGGTCCAAAT-3′; and TubR, 5′-GAGCTCTTCTGCCTGGAA-3′) were used as controls. For α-tubulin, PCR conditions were annealing temperature of 58°C and 32 cycles, giving a 304-bp PCR product. All PCR products for each set of primers were resolved using the same 1% agarose gel and then stained, soaking the gel in a 1% ethidium bromide solution in order to avoid differences between gels and staining. Absolute integrated optical density of each band was measured with the program Gel-Pro Analyzer version 1.3 (Media Cybernetics). All PCRs were determined in duplicate.

Expression and purification of recombinant T. gondii sHsp genes.

The cloned open reading frames (ORFs) of the four T. gondii sHsps were subcloned into the prokaryotic expression vector pRSET A (Invitrogen Life Technologies) downstream and in frame with a sequence that encodes an N-terminal fusion peptide (6×His). Recombinant Hsp28 (rHsp28) corresponds to a truncated version (from Met25 to the end, amino acid 277). PQE31-Bag1 plasmid was kindly provided by Mariana Matrajt (University of Vermont). BamHI and KpnI restriction sites were engineered into the 5′ and 3′ primers, respectively. Escherichia coli BL21DE3 was freshly transformed with the expression plasmids. Cultures were grown to an optical density at 600 nm of 0.6 before protein expression was induced by the addition of isopropyl thio-β-d-galactoside (0.2 mM). After overnight incubation at 30°C, the cells were harvested and purified using a commercial nickel-nitrilotriacetic acid column (QIAGEN) in native conditions according to the manufacturer's instructions.

Antibody production and immunoblot analysis.

Two C3H mice were immunized with 10 μg of purified recombinant rHsp20 in complete Freund's adjuvant and boostered every 2 weeks with three successive injections with 10 μg of recombinant protein in incomplete Freund's adjuvant (Sigma). rHsp21, rHsp28, rHsp29, and rHsp30/Bag1 were used to produce rabbit polyclonal antibodies. Rabbits were immunized with 250 μg of each sHsp emulsified with Freund's complete adjuvant (Sigma). Three booster doses were given at intervals of 15 days with the same amount of antigen and Freund's incomplete adjuvant (Sigma)

Recombinant protein and parasite extracts were analyzed by Western blot. Protein extracts from uninfected HFF cells were also assayed as a negative control. Equal numbers of parasites and recombinant proteins amount were loaded onto SDS-12% polyacrylamide gel electrophoresis gels performed under reducing conditions, electrophoresed, and transferred to nitrocellulose membrane as previously described (15). Nonspecific binding sites were blocked with 5% nonfat dried milk in Tris-buffered saline (TBS) containing 0.05% Tween 20 (TBS-TM), and the membranes were then incubated for 1 h at room temperature with the rabbit anti-Hsp21, anti-Hsp28, anti-Hsp29, and anti-Hsp30 polyclonal antibodies and the mouse anti-Hsp20 polyclonal antibody, all diluted in TBS-TM (1:1,000). The membranes were washed with TBS-TM and incubated with alkaline phosphatase-conjugated anti-rabbit or anti-mouse secondary antibodies diluted in TBS-TM (1:5,000) (Santa Cruz Biotechnology). The membranes were developed using nitroblue tetrazolium-5-bromo-4-chloro-3-indolylphosphate reagent (Promega).

Indirect immunofluorescence assays (IFA).

Tachyzoites were inoculated into host HFF cell monolayers and grown onto coverslips for 48 h. Bradyzoites of Me49 strain were obtained from brain of infected C3H mice. Briefly, the brain homogenate was washed three times with Dextran 30% (Sigma) in PBS buffer. Encysted bradyzoites were placed into coverslips and air dried and then freed after digestion with trypsin (0.05%) in EDTA (0.53 mM). Bradyzoites and host cells containing tachyzoites were fixed in 4% formaldehyde for 30 min and permeabilized with 0.2% Triton X-100 in PBS for 10 min. After washing with PBS, cells were blocked with 10% bovine serum albumin and incubated with the appropriate dilution of each primary antibody (rabbit anti-Hsp21, anti-Hsp28, anti-Hsp29, anti-Hsp30, and mouse anti-Hsp20) for 1 h. Monoclonal antibody anti-P34 or anti-P30/SAG1 has been used as a bradyzoite or tachyzoite stage marker, respectively. Polyclonal mouse anti-T. gondii H2A.Z histone was used for nucleus labeling. Following incubation, cells were washed three times with PBS and then incubated with the corresponding secondary antibodies (1:200) using fluorescein isothiocyanate-conjugated goat anti-rabbit or anti-mouse (green color) antibody or rhodamine isothiocyanate-conjugated goat anti-rabbit (red color) antibody (Jackson ImmunoResearch Laboratories, Inc.). Following antibody labeling, coverslips were washed three times and mounted in Fluoromont G (Southern Biotechnology Associates) and viewed using a Nikon Eclipse E600 microscope. Green and red fluorescence were recorded separately, and the images were merged using the Adobe Photoshop program. To specifically stain the mitochondrion of the parasite, the cells were incubated with MitoTraker 10 nM (Molecular Probes) 30 min before fluorescence microscope observation.

Nucleotide sequence accession numbers. The nucleotide sequence accession numbers for Hsp20, Hsp21, Hsp28, and Hsp29 have been deposited in GenBank under accession numbers AY644771, AY756061, AY650281, and AY721614, respectively.

RESULTS

Sequence analysis. BLAST analysis of a T. gondii expressed sequence tag (EST) database retrieved five different open reading frames (ORFs) encoding proteins within the α-crystallin domain related to the previously characterized sHsps of Arabidopsis thaliana (46). One of them was Hsp30/Bag1, the only previously characterized T. gondii sHsp (4, 5, 42, 56). Based on sequences retrieved from EST databases, specific primers matching putative 5′- and 3′-untranslated regions were designed, and cDNA sequences were obtained after RT-PCR and subsequently cloned and sequenced. Based on nucleotide sequence analysis their ORFs were deduced, showing that they encode four small heat shock proteins called Hsp20, Hsp21, Hsp28, and Hsp29 (607, 588, 831, and 804 bp, respectively) (Fig. 1).

FIG. 1.

A. ClustalW sequence alignment of T. gondii sHsps. The alignment includes the open reading frame of Hsp30/Bag1 (X82213) and the four cloned T. gondii sHsps: Hsp20 (AY644771), Hsp21 (AY756061), Hsp28 (AY650281), and Hsp29 (AY721614). Black shading indicates amino acid residues that are conserved in three or more sHsps. Intron positions are indicated with asterisks. Amino acids in italics and underlined show the place in the gene where the intron is located. The α-crystallin domain is underlined. The conserved A-x-x-x-n-G-v-L motif inside the α-crystallin domain and the typical I/VxI/V motif located in the C-terminal extension are marked. B. Schematic representation of T. gondii sHsp intron/exon structure. Numbers above the schemes refer to the amino acid position where introns are located. Relative intron positions are annotated by a rhombus, circle, or triangle in the presence of phase 0, phase 1, or phase 2 introns, respectively.

The genomic version of each sHsp was retrieved from the Toxoplasma genome project (www.toxodb.org). Hsp20, Hsp21, Hsp28, Hsp29, and BAG1/Hsp30 genes would be located on chromosomes VIII, XI, V, IX, and VI, respectively. By comparison analysis between genomic and cDNA sequences, Hsp21, Hsp29, and BAG1/Hsp30 genes showed 4 exons and 3 introns, the Hsp20 gene presented 3 exons and 2 introns, and the Hsp28 gene was intronless (Fig. 1B). These introns contained the typical GT:AG consensus splicing signal. No putative alternative splicing signal was contemplated in this study. In concordance with the database information, Hsp20, Hsp21, Hsp29, and BAG1/Hsp30 genomic DNA amplification products had larger molecular sizes than the corresponding cDNA amplification products, suggesting the presence of introns in these genes (data not shown). Only Hsp28 amplification showed the same size of genomic and cDNA amplification products (data not shown).

Multiple alignment of the five sHsps present in the T. gondii genome (Fig. 1A) confirmed that the α-crystallin domain is the only recognizably homologous region in all sHsps. Several structural motifs characteristic of sHsps were also present in T. gondii sHsps. The conserved A-x-x-x-n-G-v-L motif inside the α-crystallin domain and the typical I/VxI/V motif located in the C-terminal extension (7, 10, 43) were observed in all members of the T. gondii sHsp family (Fig. 1A). The N-terminal sequence of all sHsps did not share homology with any of other known sHsps prodomains. Analysis of the primary structure of the Hsp29 molecule revealed the presence of unusual stretches rich in threonine residues in the N-terminal region and a putative transmembrane domain located between 125 and 146 amino acids. The predicted Hsp28 amino acid sequence showed a methionine-rich region towards the N-terminal sequence.

With respect to the α-crystallin domain, Hsp21, Hsp29, and Bag1/Hsp30 genes contained a single phase-one intron located at the same position as Homo sapiens Hsp27. The Hsp20 gene also showed an intradomain phase-one intron, but it was in a new position (Fig. 1B).

Stage-specific RNA expression and heat stress induction on the sHsp family.

RT-PCR assay was performed in tachyzoite and bradyzoite stages in order to know the expression profile of each T. gondii sHsp (Fig. 2). In all cases the amplification yielded a product of the expected size for each sHsp. Tachyzoite- and bradyzoite-specific genes (SAG1 and Hsp30/Bag1 genes, respectively) were included as controls. Figure 2 shows that mRNA for Hsp20, Hsp21, and Hsp29 are present in both stages of the parasite, whereas Hsp28 could only be amplified in the tachyzoite stage.

FIG. 2.

A. T. gondii sHsp expression in tachyzoite and bradyzoite stages. RT-PCR assays with the four cloned T. gondii sHsps: Hsp20, Hsp21, Hsp28, and Hsp29. Bradyzoite (Hsp30/Bag1)- and tachyzoite (Sag1/P30)-specific genes were included as controls. PCR products were electrophoresed on 1% agarose gels and stained with ethidium bromide. B. T. gondii sHsp expression during heat stress treatment. Semiquantitative RT-PCR assays were performed using total mRNA of tachyzoites grown under normal (37°C) or stress (42°C) conditions for 2 h. The housekeeping α-tubulin gene was used as a control. Figures are representative of two independent experiments with similar results. T, tachyzoite; B, bradyzoite; C-, control.

To analyze the sHsp expression during heat stress treatment, PK strain tachyzoites were grown at 37°C and 42°C for 2 h. The transcription profile was examined using semiquantitative RT-PCR (Fig. 2B). To ensure that equal amounts of cDNA from each parasitic stage were being compared, mRNA from the housekeeping gene α-tubulin was also assayed in parallel trials. The densitometric analysis showed an induction of Hsp20, Hsp21, Hsp28, and Hsp29 expression in stressed tachyzoites compared to unstressed parasites (Fig. 2B).

Western blot analysis.

Recombinant protein (rHsp) and polyclonal antiserum samples for each sHsp were obtained. Western blot analysis of whole tachyzoite homogenate and rHsp purification extracts showed bands of the expected size (20 to 30 kDa; putative monomers) as well as other slower migrating bands (Fig. 3). The specificity of each anti-recombinant sHsp (rsHsp) was demonstrated by the lack of cross-reactivity among them and against proteins from uninfected human fibroblasts (data not shown). Western blot analysis of whole wild-type E. coli homogenate and E. coli expressing the rsHsp confirmed that all reactive bands are only present in recombinant bacteria (data not shown).

FIG. 3.

Western blot analysis with rabbit anti-rHsp21, anti-rHsp28, anti-rHsp29, anti-rHsp30, or mouse anti-rHsp20 polyclonal antibody. Protein extracts from whole PK tachyzoite homogenates (Tachy) and recombinant sHsps purified under native conditions (rHsp) were analyzed. All lanes contain 5 μg of total protein. Migration of size markers is indicated (in kilodaltons).

Subcellular localization.

In order to determine the subcellular localization of T. gondii sHsps during tachyzoite and bradyzoite stages, indirect immunofluorescence assays (IFAs) were carried out. In tachyzoite and bradyzoite stages, Hsp20 antiserum gave a crescent-shaped pattern at the apical end of the cell, labeling mostly a small area at the conoid of the parasite (Fig. 4B and E). Moreover, in some bradyzoites a membrane-associated labeling was also observed (Fig. 4H).

FIG. 4.

Indirect immunofluorescence localization of Hsp20 in intracellular tachyzoite (Tachy) (B) and bradyzoite (Brady) (E and H) stages detected with mouse polyclonal anti-rHsp20. CP, phase-contrast images. Note the targeting of Hsp20 at the apical region in the merged labeled image (merged) in tachyzoites (C) and bradyzoites (F) and the membrane-associated labeling in bradyzoite stages (I).

Using the anti-Hsp21 antibody, the signal in tachyzoite and bradyzoite stages appeared throughout the cytosol of parasites but not in the nucleus, as can be observed in the merged image of α-Hsp21 and α-histone (Fig. 5).

FIG. 5.

Indirect immunofluorescence localization of Hsp21. Tachyzoites (upper panel) and bradyzoites (lower panel) were stained with rabbit polyclonal antibody anti-Hsp21 and mouse polyclonal antibody anti-H2A.Z histone to evidence the cytosolic localization of Hsp21. Monoclonal antibodies anti-P30 and anti-P34 were assayed as tachyzoite and bradyzoite markers, respectively. PC, phase-contrast images. Note that histone labeling in bradyzoite parasites is less uniform than the label observed in tachyzoites, which may be due to H2A.Z being a special histone in the H2A family.

The localization of Hsp28 displayed a staining pattern inside the mitochondrion of the parasite that was similar to that observed with the MitoTraker marker (Fig. 6). In agreement with the RT-PCR results, Hsp28 was not detected by IFA in bradyzoites obtained from brains of infected mice (Fig. 6). However, it was observed as a faint label when bradyzoites were fully differentiated in vitro, in contrast to the intense signal observed in the tachyzoite stage (data not shown).

FIG. 6.

Indirect immunofluorescence localization of Hsp28. Tachyzoites (upper panel) and bradyzoites (lower panel) were assayed with rabbit polyclonal anti-rHsp28. To specifically stain the mitochondrion of the parasite, the cells were incubated with a MitoTraker marker. Colocalization of Hsp28 and mitotracker staining is observed in the merged image in the tachyzoite stage. Note the lack of reactivity of rabbit polyclonal anti-Hsp28 in the bradyzoite stage. As a control for tachyzoite (Tachy) and bradyzoite (Brady) stages, monoclonal anti-P30 and anti-P34 antibodies were assayed, respectively. Ph, phase-contrast images.

Double immunofluorescence assays employing anti-Hsp29 antibodies and a monoclonal antibody directed against anti-P30, a surface protein, indicated a membrane-associated localization of this protein in tachyzoite and bradyzoite stages (Fig. 7). Tachyzoites showed a heterogeneous membrane-associated labeling in which one of the extremes of the parasites has no label, while bradyzoites also showed a nonuniform membrane association; however, in this stage both ends of the parasite lack the label (Fig. 7). These differences were observed in all parasites observed from the three different experiments.

FIG. 7.

Indirect immunofluorescence localization of Hsp29. Tachyzoites were stained with monoclonal anti-P30, mouse polyclonal anti-T. gondii H2A.Z, and rabbit polyclonal anti-Hsp29 antibodies to evidence the membrane-associated labeling of Hsp29. Bradyzoites were labeled with monoclonal antibody anti-P34 as a control and rabbit polyclonal anti-Hsp29. Overlays of the doubly stained tachyzoites (Tachy) and bradyzoites (Brady) are shown in the merged Hsp29/P30 and Hsp29/P34 images, respectively. PC, phase-contrast images.

DISCUSSION

To our knowledge, the results of this study describe for the first time the identification and characterization of the α-crystallin/sHsp family of an Apicomplexa parasite. The T. gondii sHsp family was identified by database homology search and cloned. As expected, they share the homologous α-crystallin domain, which is the key characteristic of sHsps. However, an N-terminal segment, generally involved in oligomerization and chaperone function (21), contains unique characteristics in size and sequence. Because of the lack of homology in the N-terminal region, it could be supposed that T. gondii sHsps are implicated in different functions. Interestingly, by subcellular localization we observed that T. gondii sHsps are located in different cell compartments. These particular differences in the immunostaining patterns can also suggest that their targets and functions might be different.

By comparison, analysis determined that only Hsp28 was an intronless gene. The other four sequences have only one intron inside the α-crystallin domain, the only recognizably homologous region comparable with other organisms. Interestingly, Hsp21, Hsp29, and Hsp30 contained a single intron located in the same position as Homo sapiens Hsp27, suggesting a close evolutionary relationship between these groups of genes. The Hsp20 gene has an intradomain intron in a new position, suggesting that a double event of loss-gain of an intron has occurred in this gene.

Most T. gondii sHsps appear to be constitutively expressed in tachyzoites and bradyzoites, with the exception of Hsp28, which seems to be predominantly expressed in the tachyzoite stage, and Hsp30/Bag1, which is expressed only in the bradyzoite stage (4, 42). In addition, all T. gondii sHsps studied here were induced after heat shock treatment, at least in the tachyzoite stage. Increased expression of sHsps has been reported in several stress conditions, such as heat shock, ischemia, hyperthermia, osmotic stress, and chemical toxicity (8, 13). In the future, it would be interesting to analyze up-regulation of T. gondii sHsps in different stress conditions.

Oligomerization into large complexes is also a distinctive feature of sHsps (28). Western blot analysis of whole tachyzoite extract and rHsps showed several bands that migrate slower than their expected molecular size and specifically react with the appropriate anti-sHsp antibodies. These bands could be associated with oligomeric forms resistant to SDS and β-mercaptoethanol treatment. Biswasi and Das (3) showed that bovine α_L-crystallin retained some of its oligomeric structure in the presence of low concentrations of SDS. A preliminary analysis of gel filtration using recombinant T. gondii sHsps purified under native conditions showed the presence of approximately 12- to 20-mer forms (unpublished data). The presence of these oligomeric structures in principle supports the former interpretation. However, it must be noted that the size of the oligomer forms obtained by gel filtration do not correspond to the bands observed in the Western blot. Further analysis should be done to determine oligomer formation as well as their size and stability.

Although most organisms have just one or a few cytosolically localized sHsps, it is interesting to note that T. gondii has at least four different sHsps localized in different cell compartments, a feature that has also been observed in plants, which have many diverse sHsps localized in mitochondria, cytosol, endoplasmic reticulum, and chloroplast (53). By IFA, it was demonstrated that Hsp21 is found in the cytosol, the same subcellular localization as Bohne et al. (4) described for Hsp30/Bag1. Experiments with knockout mutants showed that Hsp30/Bag1 is not essential for cyst formation (5, 56). Probably, this lack of phenotype may reflect redundancy of gene function in the organism. In view of the fact that Hsp21 shares the same subcellular localization as Hsp30/Bag1, it is possible that Hsp21 may be able to partially compensate for Hsp30/Bag1 function in knockout mutants.

The results obtained by IFA indicate that Hsp28 protein appears to be targeted to the mitochondrion of the parasite. Sequence analysis showed that Hsp28 contains stretches of methionine towards the N-terminal region, which is a remarkable characteristics of angiosperm chloroplast sHsps (53). In T. gondii, very little is known about the respective functions of mitochondria in tachyzoite and bradyzoite forms. However, it was reported that specific inhibition of the mitochondrial electron transport chain by drugs also promotes conversion of tachyzoites into encysted bradyzoites (1, 50, 51). Therefore, they suggested that active mitochondria may not be required at the bradyzoite stage, and the active import and proper assembly of multimeric proteins only takes place in the tachyzoite stage. This is in concordance with the difficulty of detecting Hsp28 in the bradyzoite stage, suggesting that its expression in this stage could be less required than it is in tachyzoite stages. In T. gondii, mitochondria contain nuclear-encoded heat shock proteins such as Hsp10 and Hsp60 which function selectively as mitochondrial chaperones in the posttranslational assembly of multimeric proteins (45, 51). Mitochondrial chaperones are essential for the biogenesis of mitochondria and therefore for all possible functions of these organelles (2). For all these reasons, similar to other molecular chaperones described, Hsp28 is an ideal target for investigations concerning the possible role of mitochondria in the two major developmental stages, tachyzoite and encysted bradyzoite stages of T. gondii.

Our data provide clear evidence that Hsp29 protein is associated to the membrane of the parasite, in the tachyzoite as well as in the bradyzoite stage. The membrane localization of sHsps is unusual, but some cases could be observed. As examples, α-crystallin from the vertebrate lens, Hsp17 from Cyanobacterium synechocystis, and HspB2 from human heart and skeletal muscles can be associated also with plasma membrane, thylakoid membrane, and mitochondria membrane, respectively (9, 40, 52). The presence of sHsps in membranes can regulate membrane fluidity and preserve membrane integrity during thermal fluctuations (52). In our case, it is not possible to distinguish if Hsp29 is located inside the plasma membrane or associated with others membrane proteins, for example, the subpellicular cytoskeleton network (37). In this sense, molecular chaperones that facilitate formation of the cytoskeleton elements influence their function and provide protection for their constituent proteins, providing an indispensable services to the cell that have been described previously (30). Two situations are described in the literature where the properties of actin are modulated by sHsps: the inhibition of actin polymerization, where sHsps act as capping proteins, and the actin cytoskeleton protection (34, 38). On the other hand, sequence analysis of Hsp29 revealed some peculiar characteristics, such as the presence of threonine-rich regions in the N-terminal region. Threonine could be phosphorylated and/or N- and O-glycosylated. Further studies are necessary to know the role and importance of threonine stretches in T. gondii.

Surprisingly, Hsp20 antibodies brightly label the apical region of T. gondii tachyzoites and bradyzoites, resembling the shape pattern observed for apical membrane antigen 1 (TgAMA1) (23). In addition, a dispersed antigen labeling along the plasma membrane was observed in some bradyzoites. The apical region of the protozoan Apicomplexa phylum corresponds to a highly complex site of exocytic activity. It involves not only several specialized secretor organelles, believed to be important for host invasion, but also a complex cytoskeleton (36).

Proteins localized to the surface of the parasite, either in the membrane or associated to other membrane proteins, are of particular interest in view of their potential role in the invasion of host cells (14, 47). In light of this and the data presented here, it may be asked why such membrane association is a property for some proteins of this family. The further exploration of this issue may bring unexpected findings in the understanding of the full physiological functions of Hsp29 and Hsp20 in T. gondii. Due to their subcellular localization, it could be suggested that T. gondii Hsp20 and Hsp29 are potential drug targets to control toxoplasmosis in patients with active infection.

The functions of T. gondii sHsps are still unknown. Their different localization is of particular interest, as it shows that the five sHsps of T. gondii have distinct locales within the cell. It is tempting, then, to suggest that these five structurally similar proteins have functions in different intracellular compartments, arguing that they serve important roles in survival. Further investigations will be directed to detailed comparison of the effect of wild-type sHsps and their mutants to finally elucidate the particular role of each sHsp in the parasite.