A DnaJ-like Protein Homologous to the Yeast Co-chaperone Sis1 (TcJ6p) Is Involved in Initiation of Translation in Trypanosoma cruzi *

,

In Escherichia coli the DnaK, DnaJ, and GrpE genes encode heat shock proteins that are essential for growth at temperatures above 42°C (reviewed in Ref. 1). DnaJ stimulates the weak intrinsic ATPase activity of DnaK (2) and can also directly interact with specific substrates of the chaperone machinery (3). Various DnaJ homologs have been discovered in eukaryotic cells. Eukaryotic DnaJ-like proteins belong to the conserved heat shock protein 40 (Hsp40) 1 family and are involved in regulation of the Hsp70 molecular chaperones (DnaK-like proteins), mediating the biogenesis of proteins. Some Hsp40s may be considered to be true molecular chaperones in that they prevent aggregation by binding directly to unfolded polypeptide substrates (4,5). All DnaJ-like proteins contain a J domain of about 70 amino acids, generally N-terminal, which represents the signature of the family (6) and is required for interaction with the ATPase domain of Hsp70 (7). The basic mechanism of action of the Hsp70 proteins is sequential binding and release in an ATP-dependent manner of polypeptides in non-native conformations (protein folding). The subcellular compartmentalization of different Hsp70 members and their specific interactions with various DnaJ-like proteins allows these chaperones to be involved in a variety of protein biogenesis pathways including the assembly and disassembly of protein complexes, proteolysis, the translocation of proteins into organelles and translation initiation (reviewed in Ref. 8). The Saccharomyces cerevisiae genome data base (genome-www. stanford.edu/Saccharomyces) contains 20 putative Hsp40s homologous to E. coli DnaJ, but only half have been characterized and localized to major cellular compartments (4,8,9). The Hsp40 family has been divided into three distinct subgroups based on the presence of conserved functional domains in addition to the J domain (7,9). Class I Hsp40s have a glycinephenylalanine-rich (G-F) region followed by a cysteine-rich region (CRR domain), which forms a zinc finger motif with four repeats of a CXXCXGXG motif and a weakly conserved Cterminal domain. The archetypal E. coli DnaJ and the S. cerevisiae Ydj1 protein are class I Hsp40s. Class II Hsp40s include the S. cerevisiae Sis1 protein and its mammalian homolog, the co-chaperone Hdj1, and lack the CRR domain. Finally, in class III Hsp40s, only the conserved J domain is present. The J domain is thought to be involved in interactions with Hsp70s via its HPD tripeptide loop (10,11), whereas the G-F region is a critical determinant required for the specificity of S. cerevisiae Sis1p (9). In class I Hsp40 proteins, the zinc finger motif and the poorly conserved C-terminal domain have been shown to act as binding sites for proteins in denatured state (12,13). It has also been suggested that the J domain and G-F region may be sufficient for the basic functions of class I and class II Hsp40s in vivo, whereas the distal C-terminal regions are important in yeast in suboptimal growth conditions (9).
The unicellular protozoan, Trypanosoma cruzi, the etiological agent of Chagas' disease, has a complex life cycle where the parasite passes through three differentiation forms and two hosts, a reduviid insect vector (triatomine) and a mammalian host (14). During this cycle, the parasite encounters drastic environmental changes (e.g. in temperature and pH), and these changes are accompanied by major morphological changes. The epimastigotes (insect forms) replicate in the insect host midgut and differentiate to produce metacyclic trypomastigotes (by a process known as metacyclogenesis). These metacyclic trypomastigotes are non-proliferative and infectious forms released in the excreta of the bug. Although the insect feeds on the blood of the mammalian host, metacyclic trypomastigotes invade the bloodstream of the host and infect various cell types (mainly macrophages and muscle cells) in which the trypomastigotes differentiate to produce amastigotes (proliferative forms). Metacyclogenesis can be simulated in vitro using a chemically defined differentiation medium (15).
In T. cruzi, 5 members of the DnaJ-like family (TcDnaJ) have been partially characterized (16,17) and shown to have a highly conserved N-terminal J domain. The proteins tcj2, tcj3, and tcj4 belong to Hsp40s class I, whereas tcj1 and TcDJ1 belong to class III. Except for the putative mitochondrial DnaJlike protein, TcDJ1, the function and subcellular location of the other family members are still unknown. A highly conserved gene encoding a novel TcDnaJ was recently sequenced in five species of trypanosomatids (Trypanosoma b. brucei, Trypanosoma congolense, Trypanosoma vivax, T. cruzi, and Leishmania donovani) in a downstream region flanking the glucose transporter gene cluster (18). In T. cruzi, glucose transporters (TcrHT1) are encoded by a multigene family that is organized into several clusters, all contained within a genomic fragment no larger than 150 kb (18,19). In the T. cruzi CL strain, two clusters of TcrHT1 genes, separated by several dozen of kilobases, flank several duplicates of a 20-kb region composed of genes encoding an adenylate cyclase homologous to the T. b. brucei pESAG4 (20), a DnaJ protein homologous to human DNAJ1, and two small G-proteins homologous to human Rab1 and yeast YPT7, which belong to the Rab family. This kind of repetition in tandem of large stretches of DNA is characteristic of intrachromosomal amplifications, which are frequently observed in T. cruzi. Run-on experiments showed that in T. b. brucei the battery of genes flanking the THT (trypanosome hexose transporter) gene cluster are transcribed by an RNA polymerase II on a distance of at least 50 kb, suggesting that these genes belong to the same transcription unit (18). Similar large transcription units have been described in T. brucei that contained various genes grouped in an active expression site where they are co-transcribed by a polymerase I-like RNA polymerase (21). This is the case of the variant surface glycoprotein (VSG) genes and a number of co-expressed linked genes, which may be involved in optimization of parasite adaptation to different environmental conditions (22), suggesting that the parasite contains functional units similar to prokaryotic operons (23).
In this study, we mapped a DNA fragment of 14.3 kb homologous to the downstream region of the glucose transporter gene cluster isolated from a T. cruzi genomic library. This gene unit differs from the repeat region of 20 kb previously described in the TcrHT1 locus by the presence of a pseudogene of the large trans-sialidase multigene family (reviewed in Ref. 24). To determine if these genes form a functional unit similar to the VSG polycistronic unit in T. brucei bloodstream forms, we decided to investigate the function of the novel TcDnaJ-encoded protein, termed TcJ6p, on the differentiation process in T. cruzi. Surprisingly, the trypanosomal co-chaperone displayed a high level of similarity to various eukaryotic class II Hsp40s (Caenorhabditis elegans protein homologous DnaJ, human DNAJ1 protein, Sis1p), whereas only 21-40% of sequence conservation was found with the other five members of the T. cruzi DnaJ-like family. The yeast Sis1p is the only class II Hsp40 protein that has been functionally characterized. This protein is required for the initiation of translation in S. cerevisiae (25). These observations suggested that the Sis1 and TcJ6 proteins might have similar functions. In this study, we used heterologous functional complementation and polysome sucrose gradient sedimentation to demonstrate that TcJ6p has a function similar to that of the Sis1 co-chaperone, which is essential for translation initiation in yeast.

Growth and Differentiation of Parasites and Isolation of Nucleic Acids
T. cruzi clone Dm28c (26) epimastigotes were cultured in liver infusion tryptose medium (27). In vitro metacyclogenesis of T. cruzi (Dm28c) was performed under chemically defined conditions as previously described (15). Recovering of parasites at different stages of in vitro differentiation was carried out as previously described (28). Total parasite DNA was extracted as described elsewhere (29). Total parasite RNA was prepared after the LiCl-urea method as previously described (30).

Gene Cloning and Genomic Mapping
A densely arrayed T. cruzi Dm28c EMBL3 Sau3A genomic library (29) was screened at high stringency with a probe corresponding to the 5Ј conserved region of a gene related to the sialidase family. The cDNA probe was obtained by reverse transcriptase polymerase chain reaction (RT-PCR). A reverse oligonucleotide, 5Ј-CCCTAAAGCAATTCTCT-CAGC-3Ј, corresponding to the 5Ј-conserved region of a sialidase-like gene was used as a primer using the Superscript II enzyme (Life Technologies, Inc.) as indicated by the manufacturer, on total RNA from the parasites pretreated by RQ1 RNase-free DNase (Promega). The first-strand cDNA was subsequently amplified by PCR according to the manufacturer's instructions (Life Technologies) using a forward miniexon oligonucleotide BamHI (5Ј-GCGGCGGATCCACAGTTTCTGTAC-TATATTG-3Ј (BamHI site is underlined) and the reverse oligonucleotide with the following cycle parameters: 94°C, 30 s; 52°C, 30 s; 72°C, 1 min for 3 cycles and 94°C, 30 s; 55°C, 30 s; 72°C, 1 min for 30 cycles. The purified PCR product was cloned and sequenced (GenBank TM data base accession number AY017371). All PCR reactions were performed on a PerkinElmer DNA Thermal Cycler (GenAmp PCR System 9600), and PCR products were separated in an agarose gel, purified on glass beads (Gene-clean, Bio 101) or directly purified on QIAquick column (Qiagen, Inc.), and subcloned into PCR2.1 or pCR-Blunt vectors (Invitrogen). Sequencing was performed manually using the Thermosequenase sequencing kit (Amersham Pharmacia Biotech) with [␣-33 P]ddNTPs (1500 Ci/mmol) as terminators, according to the manufacturer's instructions. Sequences were determined by sequencing on both strands of DNA. From 200,000 recombinant phages hybridized with the cDNA probe, 14 phages were isolated after three consecutive screenings. Analysis of DNA inserts of each phage were characterized by restriction enzyme mapping and by Southern blot hybridization with the same probe. Two types of recombinants were found based on SalI digestion pattern, and two of them, designed, respectively, cl263g (14.3 kb) and cl381 g (19.8 kb) were chosen for hybridization analysis. Positive bands of different sizes were subcloned from phages cl263g and cl381g in bacteriophage M13 (mp18 and mp19) or in plasmid pBlue-Script S/K ϩ (Stratagene, La Jolla, CA) to be sequenced. The Gen-Bank TM data base accession number of a 4.3-kb SalI cl263g genomic subclone containing the sialidase pseudogene, TcJ6, and 5Ј-coding region of Rab1 homolog is AF345336. The relative position of DNA fragments within phage subclone 4.3-kb cl263 genomic DNA was confirmed by sequencing extremities of each cloned fragment by PCR. The complete genetic map of cl263g and the relative positions of adenylate cyclase and small G-protein YPT7 homolog genes were determined with probes derived from the corresponding genes contained in the CosTcr1 cosmid of the CL strain, kindly donated by Dr. F. Bringaud and T. Baltz (18). The exact position of these open reading frames (ORFs) was confirmed by sequencing the 5Ј-and 3Ј-flanking regions of the 4.3-kb SalI fragment of cl263g. The 3Ј splice site of TcJ6 transcript was determined by RT-PCR using the forward mini-exon oligonucleotide BamHI and a reverse oligonucleotide specific to the 5Ј-end of TcJ6 gene, 5Ј-CGGGATGATACTTCAAAGGC-3Ј. The poly(A) addition site of TcJ6 transcript was determined by RT-PCR using a forward oligonucleotide specific to the 3Ј-end of TcJ6 gene, 5Ј-GCCTGCGTCTCTGAACGATGC-3Ј, with a reverse oligo-dT oligonucleotide BamHI, 5Ј-GCGGCG-GATCC(T) 18-3Ј (the BamHI site is underlined). The cycle conditions of both RT-PCRs were identical as described above.

DNA and RNA Analysis
The procedures employed for Southern and Northern blot hybridizations are described elsewhere (30,32). Nucleic acids were blotted onto Hybond-N or Hybond-C Extra membranes (Amersham Pharmacia Biotech) and fixed following the manufacturer's instructions. Probes were labeled by nick translation (Amersham Pharmacia Biotech) or random priming (Roche Molecular Biochemicals) using [␣-32 P]dCTP (3000 Ci/ mmol) (Amersham Pharmacia Biotech) and purified through a Sephadex G-50 column. A small subunit ribosomal RNA probe used as a loading control in Southern analysis was obtained using the forward oligonucleotide, 5Ј-CGAACAACTGCCCTATCAGCC-3Ј, and the reverse oligonucleotide, 5Ј-CCTCATCTTTTTTTTTTGACATTGG-3Ј, as primers for RT-PCR on 5 g of T. cruzi total RNA (33). The conditions for RT-PCR reaction were identical, as described above but using the following cycle parameters: 94°C, 30 s; 55°C, 30 s; 72°C, 1 min for 30 cycles. The purified PCR fragment of 168 base pairs was cloned and sequenced. The ␣-tubulin probe was obtained from an incomplete cDNA of 1.3 kb sub-cloned in M13, kindly provided by Dr. Mônica Carreira (Instituto Oswaldo Cruz, Departamento de Bioquímica e Biologia Molecular). Sequence analysis was performed using the GCG/Wisconsin software (34).

Pulsed-field Gel Electrophoresis
Chromosome separation (2 ϫ 10 7 parasites/slot) was performed using the Pulsaphor system (Amersham Pharmacia Biotech) with 1.2% agarose gels in 0.5ϫ Tris-borate/EDTA electrophoresis buffer. The running conditions were as follows: pulses of 90 s at 50 V for 1 h followed by pulses of 90 s at 100 V for 1 h and at 150 V with a pulse time of 90 s ramped to 150 s at 4°C for 50 h.

Expression of a Truncated TcJ6 Protein in E. coli and Production of Polyclonal Antisera
To obtain a truncated form of TcJ6 E. coli recombinant protein (TcJ6⌬Jp), a BamHI site was created downstream of the N-terminal J domain to insert the TcJ6 ORF lacking the first 74 amino acids into the E. coli expression vector PQE8 (Qiagen, Inc.). The truncated ORF was amplified by PCR using a Vent DNA polymerase (New England Biolabs) on a genomic template using the forward oligonucleotide, 5Ј-AGG-GATCCAAGGGCGGCGTTCCCG-3Ј (where the BamHI site, underlined, corresponds to position 217 of TcJ6 gene), and the reverse oligonucleotide, 5Ј-GTAGATCAAAGCTTGTTTTAACAACTAAAACTGG-3Ј, where the HindIII site was created immediately upstream of stop codon (the HindIII site and the stop codon are underlined). After PCR amplification and treatment by a Taq polymerase to add 3Ј A-overhangs, the PCR fragment of 827 base pairs was subcloned into PCR2.1. The BamHI/HindIII fragment was subcloned into BamHI/HindIII-digested pQE8. The fusion construct was sequenced using flanking vector sequences as primers to ensure that the sequence derived from TcJ6 gene was in the correct orientation and in-frame with the 6ϫHis tag. Expression of the recombinant protein was induced in E. coli strain M15, and cell lysates were prepared and analyzed by SDS-PAGE. The fusion product was purified on a Ni 2ϩ -nitrilotriacetic acid affinity column according to the manufacturer's instructions (Qiagen, Inc.). Polyclonal anti-TcJ6⌬Jp antibodies were raised in rabbits immunized three times at 4-week intervals with 150 g of pure recombinant protein. Antibody depletion on yeast lysate was performed by serial incubations with yeast acetone powder. The IgGs were purified on protein A-Sepharose CL-4B (Amersham Pharmacia Biotech) and used for indirect immunofluorescence.

Preparation of Cellular Extracts and Immunoblotting
Yeast protein extracts were prepared as previously described (35). Total protein extracts of different stages of T. cruzi were prepared by washing the cells twice with PBS and resuspending them in 1ϫ Laemmli buffer (36) to a final concentration of 1 ϫ 10 7 cells/l. Proteins were separated by SDS-PAGE (36) and blotted onto Hybond-C Extra membranes (Amersham Pharmacia Biotech) following standard procedures. Cytosolic extracts of T. cruzi were prepared as described (37). Blots were probed with anti-TcJ6⌬Jp polyclonal antibody (1/10,000 dilution) and a secondary alkaline phosphatase-conjugated antibody (1/7,500 dilution; Promega) following the manufacturer's instructions (Promega).

Conditional Expression of TcJ6
Two different plasmids containing the GAL1-inducible promoter were used. A centromeric plasmid pRN93, termed YCP for yeast centromeric plasmid, a gift from C. W. Slayman from Yale University and a high-copy vector pYES2, termed YEP for yeast episomal plasmid, from Invitrogen. Complete TcJ6 ORF was amplified by PCR on genomic template using a forward primer with a HindIII and BamHI restriction sites immediately upstream of TcJ6 ATG, 5Ј-GGAAGCTTGGATCCTT-GTTGAAAACATGGG-3Ј (the HindIII and BamHI sites and TcJ6 ATG are underlined), and a reverse primer with a SacI restriction site at position ϩ11 of TcJ6 stop codon, 5Ј-CGAGCTCGTTTTAACAACTA-AAACTGG-3Ј (SacI site and stop codon are underlined). 1-kb HindIII/ SacI and BamHI/SacI DNA fragments were subcloned, respectively, into HindIII/SacI-digested pYES2 and BamHI/SacI-digested pRN93 vectors. Yeast were transformed with the resulting plasmids by the lithium acetate method (38). Total yeast RNA was isolated as described elsewhere (39).

Polysome Analysis
In Yeast-Cells overexpressing TcJ6p were grown to mid-logarithmic phase in minimal medium. 15 min before harvesting the cells, cycloheximide was added to a final concentration of 60 g/ml, and the polysome profile was analyzed in a 15-45% sucrose gradient as reported (40).
In T. cruzi-10 9 cells were incubated for 5 min at 28°C in liver infusion tryptose medium containing 100 g/ml cycloheximide. After centrifugation, the parasites were chilled and washed twice with icecold washing buffer (140 mM NaCl, 5 mM KCl, 2 mM MgCl 2 , 10 mM Hepes, pH 7.4, 100 g/ml cycloheximide, and 10 units/ml heparin) (41). The cells were lysed according to Brecht and Parsons (37) using the buffer system A and the preparation of cytoplasmic extracts was carried out as reported. The polysome profile was analyzed in a linear 15-50% sucrose gradient. After centrifugation, the gradients were harvested from the bottom using a pump system coupled to a fraction collector (Amersham Pharmacia Biotech) with a flow rate of 1.1 ml/min and a fraction size of 400 l. RNA isolation was carried out as reported (37). Protein fractions were precipitated by 12.5% trichloroacetic acid at Ϫ20°C, and washed twice with acetone, resuspended in Laemmli buffer (36), and analyzed after SDS-PAGE by immunoblotting.
Immunofluorescence Microscopy-Parasites were harvested by centrifugation and washed twice in PBS. Cells (5 ϫ 10 7 ml Ϫ1 ) were fixed by the addition of an equal volume of p-formaldehyde in PBS (w/v, 4%) and incubated at room temperature for 30 min. The fixed cells were washed twice with PBS and allowed to adhere to poly-L-lysine-coated glass coverslips. After blocking the cells with 3% bovine serum albumin, 5% milk, and methylamine 0.1% in PBS for 1 h at room temperature, the parasites were incubated with 4 g/ml polyclonal anti-TcJ6⌬Jp IgGs in PBS containing 1% milk and 0.05% Tween 20 for 2 h. The cells were washed with PBS and incubated with fluorescein isothiocyanate-labeled anti-rabbit IgG secondary antibody (1/250 dilution, Sigma) and propidium iodide (10 g/ml) in the same buffer as used for primary antibody for 30 min at room temperature. The cells were washed extensively, resuspended in anti-quench solution, and viewed and photographed using a laser confocal microscope equipped with a digital in situ imaging system.

RESULTS
Genomic Characterization of TcJ6 -A sequence derived from a 291-bp RT-PCR clone corresponding to the 5Ј-conserved re-gion of a gene related to the sialidase family (24) was used as a probe to screen a phage genomic library of T. cruzi. Two types of recombinant phage were found based on the SalI digestion pattern and contained 14.3-and 19.8-kb inserts; two of them designed, respectively, cl263g and cl381g, were selected for further analysis. The genomic clone cl263g was partially sequenced (Fig. 1A). Analysis of the complete nucleotide sequence of the sialidase-like gene, with sequence similarity extending over about 1 kb (boxed in Fig. 1A), revealed that this gene was interrupted by multiple stop codons in each frame and contained a small ORF (198 nucleotides). This ORF encodes a putative 66-amino acid polypeptide presenting similarities to other members of the gp85/sialidase family. However, the very small size of the polypeptide, the absence of the aspartate consensus essential for sialidase enzymatic activity, and the presence of stop codons suggest that this sialidase-like gene is probably a pseudogene, as already described for other members of the gp85/sialidase family (42). The nucleotide sequence of the 3Ј-flanking region of the sialidase pseudogene revealed the presence of two others ORFs encoding homologs of DnaJ, termed TcJ6 and Rab1. The nucleotide sequences of these genes were almost identical (99%) to that of the downstream flanking region of the TcrHT1 gene cluster, recently described by Bringaud et al. (18). A more detailed analysis of the flanking sequences identified other conserved copies present at the TcrHT1 locus, an ORF encoding an adenylate cyclase localized 5Ј to the sialidase pseudogene and an ORF localized 3Ј to the Rab1 gene and encoding another small G-protein homologous to yeast YPT7 (Fig. 1A). Thus, the 14.3-kb genomic fragment is highly homologous to the duplicated region flanking the glucose transporter gene cluster and may therefore be an alternative genomic rearrangement of the downstream-flanking region of the THT gene cluster in T. cruzi. The nucleotide sequence of the TcJ6 gene predicts an open reading frame of 1014 base pairs encoding a putative protein of 338 amino acids with a molecular mass of 36.5 kDa. The 3Ј splice site and the poly(A) addition site were determined by RT-PCR using a 5Ј reverse internal primer with a mini-exon-derived primer and a 3Ј forward internal primer with a oligo-dT primer, respectively. This analysis predicted a small 5Ј-untranslated region (UTR) of 20 nucleotides and a 3Ј-UTR of 191 nucleotides. Southern blotting with a restriction enzyme cutting outside the gene (PstI (P) in Fig. 1B) was used to determine the number of copies of the TcJ6 gene in the genome. Several copies were detected using a  probe spanning the entire ORF. We investigated whether these members of the new DnaJ-like subfamily of genes were, like other T. cruzi genes encoding Hsps, organized as direct tandem repeats by digestion with two restriction enzymes, one with a single site in the 5Ј extremity of the ORF (StuI) and one with a single site in the 3Ј extremity of the ORF (HincII). Low stringency hybridization of T. cruzi genomic DNA digested with StuI alone or with HincII alone to a probe covering the entire ORF gave no hybridizing fragment common to both digests. This suggests that the various copies of the gene encoding the TcJ6 protein are not organized in tandem repeats but are instead spread throughout the genome of T. cruzi. Southern blot hybridization of a pulse-field gel of three strains of T. cruzi showed that the gene copies encoding TcJ6p are located on a chromosome or a pair of homologous chromosomes of ϳ1,300 kb (Fig. 1C). In addition, probes for the sialidase pseudogene (Fig.  1C), Rab1, and TcrHT1 (data not shown), labeled the same set of chromosomes. These results suggest that, despite the difference in chromosome size between the various strains of T. cruzi, the TcJ6 gene copies are present on a subset of chromosomes of the megabase range that is conserved between various strains of T. cruzi. Altogether these data are consistent with the study by Bringaud et al. (18), showing the existence of various duplications of a large area characteristic of the 3Ј-end of the TcrHT1 gene cluster. TcJ6 Co-chaperone Homologous to Yeast Sis1p-Data base searches showed similarity between TcJ6 and E. coli DnaJ proteins (60% similarity for the first 76 amino acids, which define the J domain), and only 21-40% sequence similarity to the other five members of the T. cruzi family (tcj1-4 and TcDJ1) (Table I) (16,17). TcJ6p was most similar (ϳ40% similarity) to the putative cytosolic DnaJ protein tcj3. Similar characteristics can be observed for tcj1 and the putative mitochondrial TcDJ1, two class III Hsp40 members from the T. cruzi family lacking the GF domain and the zinc finger region. Surprisingly, the TcJ6 protein displays its highest score for sequence similarity with two eukaryotic class II Hsp40s, 54% similarity to a C. elegans hypothetical DnaJ-like protein (F54D5.8) and 50% similarity to the yeast Sis1 protein, which is required for normal initiation of translation in S. cerevisiae (Fig. 2) (25, 31). Particularly well conserved were the N-terminal sequence containing the highly conserved 70-amino acid J domain with its HPD tripeptide motif (boxed in Fig. 2) and the glycine-phenylalanine-rich "spacer" region of about 50 amino acids (G-F region), which is not followed by the zinc finger motif and is therefore a typical class II Hsp40. In the middle third of the sequence, Sis1p contains a striking glycine/methionine-rich sequence (G-M region), which in the case of TcJ6p, is replaced by a glycine-rich sequence (with a conservation of almost all the glycine residues of Sis1p). In a previous study (9), chimeric fusion proteins in which the various functional domains (J, G-F, G-M, C terminus) had been exchanged between the two yeast cytosolic proteins Sis1p and Ydj1p were used in experiments to rescue the ⌬sis1 yeast mutant. The G-F region was found to be essential for the function of Sis1p and redundant to the function of the G-M and C-terminal regions. These structural features suggest that the trypanosomal DnaJ may have a function similar to that of yeast Sis1p. In addition, analysis of the predicted amino acid sequence of TcJ6p suggests that, like Sis1p, the trypanosomal co-chaperone is cytosolic. Indeed, its hydrophilic character and the absence of both a detectable N-terminal putative peptide signal and a C-terminal CAAX motif (substrate for prenyl modification of some DnaJ-like proteins, which allows association with cell membranes) suggest that this protein is probably cytosolic.
TcJ6 Is Constitutively Expressed during the Parasite Development-The expression pattern of TcJ6 was determined during the in vitro metacyclogenesis of T. cruzi by both Northern blot analysis and immunoblotting (Fig. 3). Total RNA was extracted at various stages of the in vitro differentiation of T. cruzi epimastigotes and analyzed with probes for the TcJ6 gene and the nuclear gene encoding the small subunit ribosomal RNA, as a loading control for constitutively expressed genes (33). Analysis of the relative intensities of the bands corre-FIG. 2. Amino acid sequence homology between TcJ6 and yeast Sis1 proteins. The sequence alignment was performed using the GAP program of GCG (34). The bars indicate identity, and double and single dots refer to frequent and infrequent conservatives substitutions respectively. Gaps have been introduced to maximize the alignment and are represented by a dotted line. The conserved J and G-F domains, and the glycine/methionine-rich sequence are indicated by solid bars above the sequence alignment. The highly conserved HPD motif is boxed, and the glycine and phenylalanine residues of the G-F region and glycine residues of the G-M region are in bold.
sponding to the TcJ6 transcript and small subunit ribosomal RNA (ssrRNA) showed that the amount of the 1.4-kb mRNA increased during in vitro metacyclogenesis, reaching a maximum in metacyclic forms (Met in Fig. 3A and arrow). Similar results were obtained by semi-quantitative RT-PCR (data not shown). A larger RNA species (the asterisk in Fig. 3A) was also detected (sensitive to RNaseA-DNase-free, data not shown), reaching a maximum 24 h after the start of the differentiation.

FIG. 4. Analysis of TcJ6 expression in transformed S. cerevisiae. Panel A,
Northern blot analysis was carried out with 10 g of total RNA from CY732-YEPTcJ6 grown in the presence of glucose or galactose. Hybridization was carried out with a probe derived from the TcJ6 gene (probe A, Fig. 1). Panel B, immunoblot of protein extracts from CY732-YEPTcJ6 grown in the presence of glucose or galactose and from T. cruzi (Tc) epimastigotes (5 ϫ 10 6 cells), probed with an anti-TcJ6⌬Jp antibody depleted on yeast lysate.

FIG. 3. Analysis of the expression pattern of TcJ6.
T. cruzi epimastigotes exponentially growing in complete medium (Epi) were subjected to a nutritional stress for 1 h (Str) and were then allowed to differentiate into infectious metacyclic cells (Met). 6 and 24 h after the start of in vitro differentiation (6H, 24H), most of the parasites had adhered to the plastic culture flask, and the process was complete within 96 h (28). Panel A, regulated expression of TcJ6 mRNA during metacyclogenesis in vitro and during a heat shock. 10 g of total RNA from each differentiation stage of T. cruzi were analyzed by Northern blotting with a TcJ6 genomic probe (StuI-HincII, 0.89 kb, probe B in Fig. 1, panel A) or a nuclear small subunit ribosomal RNA (ssrRNA) cDNA fragment. A similar Northern blot analysis was carried out using 10 g of total RNA from epimastigotes subjected to a heat shock at 37°C for 6 and 24 h (6H and 24 H) DNA probes corresponding to the flanking genes (adenylate cyclase, Rab1) recognized apart from their individual mRNAs the same high molecular weight transcripts, hence suggesting that this RNA is polycistronic (data not shown). The unusual detection of polygenic transcripts might result from the inhibition of processing of primary transcripts due to the stress treatment (acidic and nutritional) required for in vitro differentiation of T. cruzi, as already described for the tubulin unit of heat-shocked T. brucei cells in which were accumulated high molecular weight mRNA precursors containing both the ␣and ␤-tubulin-coding region (43). Epimastigotes subjected to a heat shock at 37°C for 6 h (6H in Fig. 3A) accumulated about 2-3 times more mRNA than did untreated epimastigotes (28°C). This probably reflects an increase in stability of the co-chaperone mRNA during heat shock, as reported for other trypanosomal Hsps (44).
To analyze protein expression during parasite differentiation, polyclonal rabbit antibodies were raised against a recombinant His-tagged TcJ6p lacking its J domain (⌬J in Fig. 3B) to prevent the cross-reactivity with other members of the TcDnaJ family. In Western blot analysis of T. cruzi cell extracts, this antiserum recognized a polypeptide with an apparent molecular mass of 36.5 kDa, which corresponds to the molecular mass predicted from analysis of the primary polypeptide sequence (Fig. 3B). This suggests that this protein is unmodified (not prenylated nor glycosylated). No major changes in protein level were observed during differentiation in vitro, and only a modest reduction was observed in metacyclic forms. Thus, the increase in abundance of steady-state RNA observed in metacyclic cells was not concomitant with an increase in the amount of protein. Heat shock at 37°C for 6 -24 h increased the amount of TcJ6 protein by about 2-fold, possibly due to a delay in translation of the mRNA that accumulated during the first 6 h of heat shock.
The antiserum detected a protein of similar electrophoretic mobility in other trypanosomatid species (T. b. brucei, Leishmania major) (Fig. 3C). The increase in molecular weight as observed in L. major is in agreement with the size of the protein predicted from the ORF of L. donovani (345 amino acids in L. donovani against 338 and 336 amino acids, respectively, in T. cruzi and T. b. brucei, 18). Therefore, the cross-reactivity of the anti-TcJ6⌬Jp antiserum with the DnaJ of other trypanosomatids species (T. b. brucei, L. major) corroborates a high epitope conservation of the DnaJ-like protein among these species, which share with TcJ6p more than 81% sequence similarity (18). In contrast, in yeast, although some short stretches of peptidic sequence are found conserved between Sis1p and TcJ6p (Fig. 2), the anti-TcJ6⌬Jp antiserum did not react with the protein immunoprecipitated Sis1p (data not shown). Cytoplasmic and membrane fractions of T. cruzi epimastigotes were probed with antiserum against TcJ6p, and it was found that the protein was associated exclusively to the cytosolic fraction (Cyt in Fig. 3C). Taken together these results demonstrate that the highly conserved trypanosomal cytosolic co-chaperone is constitutively expressed during metacyclogenesis in vitro.
TcJ6 Suppresses the Temperature-sensitive Phenotype of a SIS1 Mutant (sis1-85)-In yeast, SIS1 is essential for viability and encodes a DnaJ homolog required for normal initiation of translation. To determine whether overexpression of the trypanosomal DnaJ could substitute for Sis1 function, we used a yeast strain in which the SIS1 chromosomal copy has been deleted and contains a plasmid with a temperature-sensitive mutation, sis1-85 (which results in the absence of 22 amino acid residues (255-276) of 352 total residues of Sis1p). This strain is temperature-sensitive for growth (31). We transformed a CY732 strain containing the sis1-85 temperaturesensitive mutation with a high copy vector (pYES2 termed YEP for yeast episomal plasmid) or a CEN vector (pRN93 termed YCP for yeast centromeric plasmid) containing the TcJ6 gene under the control of a GAL promoter, which allows the induction of the heterologous protein in the presence of galactose but not of glucose. As a control we used a CY736 strain in which the SIS1 chromosomal copy was been deleted and contained the wild-type SIS1 on a centromeric plasmid. We checked that the trypanosomal DnaJ was synthesized correctly in yeast by Northern blot analysis of total RNA from the CY732-YEPTcJ6transformed strain grown in the presence of glucose or galactose. As shown in Fig. 4A, the expression of the TcJ6 mRNA was induced by galactose and tightly repressed in its absence. Immunoblotting detected TcJ6p in the presence of galactose FIG. 5. Functional complementation in S. cerevisiae. Panels A and B, growth phenotypes of wild-type (wt) strain CY736, sis1-85 strain (CY732), and CY732 transformed with insert-less vectors (CY732-YCP, CY732-YEP) or with vectors carrying the TcJ6 gene (CY732-YCPTcJ6, CY732-YEPTcJ6). The strains were streaked onto YPgal (1% yeast extract, 2% bacto-peptone, and 2% galactose) plates and grown for 4 days at 30°C or 39°C. Panel C and D, growth curves of the strains used in panels A and B. YPgal medium was inoculated with the various strains, which were grown at 30°C (solid but not in the presence of glucose (Fig. 4B). The fact that heterologous protein migrated in SDS-PAGE with an apparent molecular mass similar to the endogenous T. cruzi DnaJ (Tc in Fig. 4B) suggested that TcJ6 protein is properly expressed in S. cerevisiae. We have therefore investigated whether the expression of TcJ6p could abolish the temperature-sensitive phenotype of the sis1-85 mutant (CY732 strain). As shown in Fig. 5A, at the permissive temperature (30°C) in the presence of galactose, the growth phenotype was the same for all strains, except for the sis1-85 strain transformed with YEPTcJ6, which grew more slowly (small colonies). At the nonpermissive temperature (39°C, Fig. 5B) in the presence of galactose, sis1-85 cells did not grow, but when they expressed TcJ6p, they displayed the same slow-growth phenotype as the control strain expressing the wild-type SIS1. Thus, the heterologous expression of the trypanosomal DnaJ gene complemented the sis1-85 strain at nonpermissive temperatures. The accentuated slow-growth phenotype of the mutant strain overproducing the TcJ6 protein (CY732-YEPTcJ6 in Fig. 5, A and B) may result from a cytotoxic effect of the accumulated protein during growth. We investigated this possibility by plotting the growth curves of the different strains in liquid culture. At the permissive temperature, cells overexpressing TcJ6p (CY732-YEPTcJ6) grown in galactose reached a plateau at a level about half that of the wild-type strain (CY736), expressing SIS1 or the sis1-85 mutant (CY732-YEP) transformed with the insert-less vector (Fig.  5C). This confirms that overexpression of TcJ6 protein limits yeast growth. As expected, at the nonpermissive temperature, sis1-85 mutant transformed with the insert-less vector did not grow (the asterisk in Fig. 5C). The growth of CY732 strain transformed with a low copy number plasmid carrying TcJ6 (CY732-YCPTcJ6 in Fig. 5D) was absolutely normal at 30°C and similar to that of strain CY732 transformed with the insert-less plasmid (comparable with the wild-type, CY736, in Fig. 5C). At a restrictive temperature, CY732-YCPTcJ6 grew slowly. The adaptation period (lag phase) was longer in these cells expressing low levels of TcJ6p than in cells overexpressing the co-chaperone from a multicopy plasmid. This suggests that complementation depends on the number of copies of TcJ6, with excessive expression limiting growth.
Close Association of TcJ6p with Ribosomal Subunits, 80 S Monosomes, and the Smaller Polysomes-In S. cerevisiae, a large fraction of Sis1p is associated with 40 S ribosomal subunits, 80 S initiation complexes, and smaller polysomes. At nonpermissive temperatures, the sis1-85 mutant rapidly accumulates high levels of 80 S ribosomes, and the amount of polysomes decreases. These data indicate that the Sis1 cochaperone is required for the normal initiation of translation (25). The complementation by TcJ6 of a SIS1 mutant suggests that the trypanosomal protein may also be associated with ribosomes. We investigated this possibility by sedimentation on a sucrose density gradient of cytoplasmic extracts of yeast cells overproducing the heterologous protein. Western blot analysis showed that, according to our fractionation procedure, the cytoplasmic trypanosomal protein sedimented throughout most of the gradient, including positions corresponding to ribosomal subunits, 80 S monosomes, and the smaller polysomes (Fig.  6A). However, a great proportion of the protein remained at the top of the sucrose gradient (T in Fig. 6A), which contains free cytosolic proteins. This may be due either to the presence of a large excess of overexpressed protein competing for binding factors required by the translation machinery (e.g. certain mammalian-like Hsp70 proteins containing RNA-binding sites that might be involved in the regulation of translation (45)) or to the trypanosomal co-chaperone having other physiological functions in the cytosol, such as a role in the regulation of protein degradation catalyzed by proteases through its chaperone activity, as has been demonstrated for Sis1p (46).
In T. cruzi, the cosedimentation of TcJ6 protein with ribosomes/small polysomes depended on the in vitro growth conditions. Indeed, the polysome profile of trypanosomes is developmentally regulated (37). In proliferating cells (epimastigotes in logarithmic growth phase) the secondary polysome peak was more pronounced in the gradient because of the dense loading of mRNA with ribosomes (Fig. 6C). In stationary cells, this peak decreased, whereas the peak of ribosomal subunits and monosomes increased because the cells were arrested in G 0 /G 1 phase (Fig. 6B). The sedimentation profile of the trypanosomal co-chaperone followed similar pattern changes. In epimastigotes in stationary phase (Fig. 6B), the sedimentation profile of the trypanosomal chaperone showed peaks corresponding to ribosomal subunits/80 S monosomes, whereas in exponentially growing parasites (Fig. 6C) it was spread over the smaller polysomes. As for the transformed yeast, a large excess of the protein was found associated with the top of the sucrose gradient (T in Fig. 6, B and C). Similarly, in yeast, the most important fraction of Sis1p and Ssap, which functions as the Hsp70 partner for translational function of Sis1p, is found associated to the cytoplasm (47). Whether this result confirmed the possibility that a cytosolic pool of TcJ6p might be involved in other biogenesis pathways, we cannot rule out that this cytosolic TcJ6p was stripped off the ribosomes during the harvest procedure. We noted that two additional upper bands with apparent molecular masses of around 40 and 45 kDa were associated with 80 S ribosomes (Fig. 6B, open arrow). The nature of these bands is still unknown, but they may correspond to hyperphosphorylated forms of the protein (see "Discussion").
Localization of Trypanosomal DnaJ by Indirect Immunofluorescence-The subcellular distribution of TcJ6p was determined by confocal microscopy at two stages of parasite differentiation, dividing epimastigotes and quiescent metacyclic trypomastigotes. In the case of epimastigotes (Fig. 7A), the staining was distributed throughout the cytosol of the cell and was preferentially concentrated in speckles in the peri-nuclear region (Fig. 7B). The overall distribution of the trypanosomal protein was similar in metacyclic cells, but more speckles were observed close to the kinetoplast and in the peri-nuclear region (Fig. 7C). The preferential TcJ6p location in the peri-nuclear region suggests a protein association with the endoplasmic reticulum.

DISCUSSION
The TcJ6 Gene Is Highly Conserved in Trypanosomatids-A novel member of the T. cruzi DnaJ-like gene family was found together with a sialidase pseudogene and genes encoding proteins homologous to an adenylate cyclase and two small Gproteins in a 14.3-kb genomic DNA fragment highly homologous to the 3Ј region of the glucose transporter gene cluster (18). Southern blot analysis detected at least four copies of TcJ6, which, unlike the Hsp60, Hsp70, and Hsp90 genes of T. cruzi (48 -50), were not arranged in tandem repeats but were instead dispersed on a single chromosome in the megabase range. The situation was different in T. b. brucei and L. donovani, in which two allelic copies of this gene were present (18). This may be accounted for by intrachromosomal amplification of large genomic fragments, typical in T. cruzi, even in the absence of drug selection, and resulting in extensive variations in the genome size of T. cruzi strains and clones (51). It has been postulated that the organization of the genes flanking the glucose transporter gene cluster has been highly conserved during evolution, because these genes may have related functions (18). In this respect, analysis of expression of the genes detected in the cl263g clone has shown that the transcript level of adenylate cyclase and TcJ6 genes increases in metacyclic cells ( Fig. 3A and data not shown). This increase in the mRNA levels of adenylate cyclase and TcJ6 genes observed in metacyclic cells might be a result of processing of the primary transcripts into individual steady state mRNAs, occurring by trans-splicing and polyadenylation, which have been accumulated during the differentiation process. However, at the protein level, the amount of adenylate cyclase increases during metacyclogenesis, but this is clearly not the case for TcJ6p ( Fig.  3B and data not shown). These observations reveal no obvious functional coupling between the various genes of the unit.
TcJ6p, a New Member of the Class II Hsp40 Subfamily-TcJ6p belongs to the class II Hsp40s, since it has the highly conserved N-terminal J domain and lacks the cysteine-rich region but displays an extensive G-F region. Functional complementation of a class II Hsp40, the yeast Sis1p, and association with ribosomal subunits and translating ribosomes demonstrated that the physiological role of TcJ6p is similar to that of Sis1p. Do all the class II subfamily members have functions similar to Sis1p? A very important piece of evidence was provided by gene swapping experiments with the building domains (J, G-F, G-M, C terminus) of two yeast cytosolic proteins, Ydj1p (class I) and Sis1p (class II), in experiments of rescue of ⌬sis1 yeast mutant. This work showed that although J domains are interchangeable, G-F regions, which are essential for the function of Sis1p, are not (9). Thus, the G-F region is specifically required to discriminate between the function of Ydj1p and Sis1p and not the J domain, as was previously thought. Analysis of the sequence of the trypanosomal DnaJ of L. donovani reveals a very interesting feature; the presence of a G-M-rich extension (GGMPGGMPG), which is absent from the other trypanosomatids analyzed (18) and is very similar to the yeast G-M motif, is repeated twice in Sis1p (GMGGMPG-GMGGMHGGMGGMPGG) (31). In this respect, it is noteworthy that the molecular mass of the L. major chaperone is close to the 37.5-kDa of Sis1p (see Fig. 3C).
Does TcJ6p Behaves as a Heat Shock Protein?-Although the amount of Sis1p has been shown to double after a heat shock from 23 to 39°C (31), our data provide no clear evidence that heat shock from 28 to 37°C results in significant overproduction of the trypanosomal co-chaperone. After 6 h of heat shock, we detected a 2-fold enrichment of TcJ6 transcript, which was restored to normal after 24 h. This probably reflects an increase in the stability of co-chaperone mRNA during heat shock, as already observed for Hsp83 from Leishmania amazonensis (44). Because the presence of wild-type Sis1p in the sis1-85 mutant was found to repress the overexpression of Sis1-85 protein (52), we are currently investigating whether TcJ6p could work as a transcriptional regulator in yeast by down-regulating expres- sion of Sis1p via its cis-element present in the sis1-85 mutant.
Involvement of TcJ6p in Translation Initiation-Functional complementation of the sis1-85 mutant at nonpermissive temperatures and association with ribosomal subunits and 80 S monosomes are consistent with a direct role for the trypanosomal co-chaperone in translation initiation (25). In most eukaryotes, global changes in translation occur mostly at the level of initiation (53). In T. brucei, Brecht and Parsons (37) found that the low level of translation in quiescent cells was probably due to a decrease in translation initiation. In T. cruzi, the polysome profiles in the stationary and logarithmic growth phases of epimastigote forms led to the same conclusion. In arrested cells, the blocking of elongation should lead to a depletion of ribosomal subunits, whereas blocking of initiation should lead to the accumulation of these subunits. We observed an accumulation of ribosomal subunits and 80 S monosomes with a small secondary peak of polysomes for cells in stationary phase and the opposite profile in proliferating cells in logarithmic growth phase, which showed a large secondary polysome peak. The preferential association of a fraction of TcJ6p with 80 S monosomes in arrested cells, accounting for less than 10% of total cytosolic protein, might be regulated by phosphorylation of the trypanosomal DnaJ. Indeed, the most likely scenario to explain the detection in Western analysis of two slower-migrating forms (around 40 and 45 kDa) is that they represent differentially phosphorylated forms. In yeast, a band shift of similar magnitude has been reported for the Mcm1 transcription factor (30 kDa, unphosphorylated form), which is present in different multiple phosphorylated isoforms (to 45 kDa) (54). TcJ6p contains five potential phosphorylation sites that are conserved in other trypanosomatids (18). In yeast, Sis1p is phosphorylated (10% of total cytosolic Sis1p (31)), suggesting that the regulation by phosphorylation and dephosphorylation of Sis1p might mediate the association/dissociation of a protein-specific complex of the translation machinery. In trypanosomes, the possible phosphorylation of TcJ6p might be responsible for its association with the initiation complex 80 S. This association is probably indirect and mediated by some partner chaperones containing RNA binding domains (Hsp70s) involved in regulatory processes such as mRNA degradation and/or translation (45).