INTRODUCTION

Top Abstract Introduction Materials & Methods Results & Discussion References

Partial cDNA sequencing to generate expressed sequence tags (ESTs) is being used at present for the fast and efficient obtainment of a detailed profile of genes expressed in various tissues, cell types, or developmental stages (1). Genome projects have taken advantage of EST studies because ESTs represent a particular type of sequence-tagged sites useful for the physical mapping of genomes (24). ESTs can serve the same purpose as sequence-tagged sites, with the additional bonus of pointing directly to expressed genes.

One of the most interesting applications of the EST database (dbEST) is gene discovery (6). A significant development with important implications in this field has been the enormous growth of the dbEST (5). Novel genes can be found by querying the dbEST with a protein or DNA sequence. Among a number of recent examples of findings made by following this approach, a new member of the human Ly-6 family was detected (10) and 66 human ESTs were identified and mapped based on their resemblance to 66 Drosophila genes (3).

In 1994, the Special Programme for Research and Training in Tropical Diseases of the World Health Organization launched an initiative to analyze the genomes of the parasites Filaria, Schistosoma, Leishmania, Trypanosoma brucei, and Trypanosoma cruzi. Five networks were established, with the aims of (i) gaining significant knowledge on the molecular biology of these parasites; (ii) identifying new genes and their products which could be used to design new drugs, to speed up vaccine development, and to improve diagnosis; and (iii) sharing material and expertise and providing an information system that is accessible globally to researchers in the field (32).

T. cruzi is the agent of the American trypanosomiasis, Chagas' disease, for which there is neither a definitive chemotherapeutic treatment nor a vaccine being tested at present. This parasite has a complex life cycle in the Triatomine insect vector (epimastigote and metacyclic trypomastigote parasite stages) and in the mammalian host (the bloodstream trypomastigote and the intracellular amastigote stages). Thus, the expression of a number of stage-specific genes might be related to the different environments and requirements of each parasite stage. Given these facts, and as part of the T. cruzi genome project (32), we have started a project on gene discovery through EST sequencing. A total of 1,949 ESTs were sequenced from a normalized epimastigote cDNA library of the parasite clone (CL Brener) selected for this genome project (31). Their analysis revealed that the putative functions of about 18.4% of the ESTs might be deduced by sequence comparison with genes from other organisms, while about 67% have no sequence homologies in the databases and thus might represent some T. cruzi-specific sequences.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results & Discussion References

cDNA library. Poly(A)⁺ RNA isolated from CL Brener epimastigotes was used to construct a directional cDNA library in the plasmid vector pT7T318D with a modified polylinker, which consists of the restriction sites for SfiI, EcoRI, SnaBI, BamHI, PacI, NotI, and HindIII placed between the T7 and T3 promoters (7). This reduced polylinker was necessary for the efficiency of the subsequent normalization procedure. Normalization was done by partial reassociation kinetics and hydroxyapatite chromatography, whereby the excess of abundant cDNA clones was removed (7). Further details of the construction and characterization of the normalized library will be described elsewhere. Around 23,040 clones were randomly picked and plated in 384-well microplates in the laboratory of Ulf Pettersson (Uppsala, Sweden).

Nucleotide sequencing. Aliquots (1 to 2 µl) of each clone from 384-well microplates were grown overnight at 37°C in 3 ml of 2xTY containing 100 µg of ampicillin per ml (26). The template DNA for the sequencing reaction was prepared from 1.5 ml of culture by an alkaline lysis method with minor modifications (26), followed by a polyethylene glycol 8000 precipitation. The amount of isolated DNA template was estimated on a 1.0% agarose gel by comparison to serial dilutions of pBluescript II KS(+) (Stratagene). Sequencing reactions were performed in a Genius thermal cycler (Techne) by using a Dye Terminator Cycle Sequencing Ready Reaction Kit with AmpliTaq DNA polymerase (FS enzyme) (Applied Biosystems) according to the protocols supplied by the manufacturer and were analyzed in an ABI prism 377 sequencer (Applied Biosystems). Single-pass sequencing was performed on each template with T7 primer, and sequences longer than 100 bases were further analyzed. The ESTs were edited to remove vector sequences from 5' ends and to remove unreliable data from the 3' ends by using the program Factura (Perkin-Elmer).

Sequence analysis. The sequences were compared against the National Center for Biotechnology Information (NCBI) nonredundant protein database by using the program BLASTx (2) on the BLAST network service at NCBI. Sequences that did not match sequences in the protein databases were further analyzed by searching for similarities at the nucleotide level by using the BLASTn program against the nonredundant nucleotide sequence database.

Nucleotide sequence accession numbers. EST sequence data has been deposited in the dbEST with the following accession numbers: AA867894 to AA867980, AA882519 to AA883010, AA890742 to AA891021, AA908031 to AA908158, AA926379 to AA926628, AA952317 to AA952754, AA958023 to AA958272, and AA960728 to AA960749.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials & Methods Results & Discussion References

A normalized cDNA library was used to reduce considerably the number of high- and intermediate-abundance sequences and to maximize the chances of finding new genes through random sequencing (28). A total of 1,994 clones were randomly selected, and the 5' ends of the inserts were sequenced. After deletion of vector sequences and unreliable data, an average length of 420 bases per clone was obtained and used for database searches. Sequence similarities identified by the BLAST programs were considered statistically significant with a Poisson P value of 10⁵. Among the 1,994 sequences, 31 contained no insert and 14 exhibited homology with rRNA and were excluded from further analysis.

We first estimated the redundancy of our data on the basis of the redundancy of homology with sequences in the databases. A total of 644 ESTs were identified by homology with 398 different genes in the databases, representing a calculated level of redundancy of 27.9%. As shown in Fig. 1, data were classified according to the number of matches (hits) per gene. Among the 644 ESTs, 357 appeared more than once (redundant EST group), representing 111 putative genes, and 287 appeared only once. The most frequently represented genes in the library were those encoding histone H2A (accession no. gnl|PID|e290647) and histone H3 (gi|442456), which appeared 21 and 12 times, respectively (Fig. 1B). In contrast to the case for other organisms, histone transcripts in trypanosomatids are polyadenylated (19). Since the clones were picked from a normalized library, the redundancy of a cDNA clone should not be thought to represent the expression level of the gene.

View larger version (28K): [in this window] [in a new window]   FIG. 1.   Level of redundancy of ESTs that matched sequences in the NCBI nonredundant databases. (A) Percentage of ESTs with the indicated number of matches to the same gene. (B) Genes with five or more hits. The analysis was performed on a total of 644 ESTs.

On the basis of database searches, the 1,949 EST sequences were classified into four groups, as shown in Table 1. About 18.7 and 14.3% matched sequences from trypanosomatids and from other organisms, respectively. About 67% did not have a database match and thus might represent T. cruzi-specific genes. The percentage of ESTs with matches was somewhat higher (33%) than that obtained in other EST studies of protozoan parasites (11, 16, 20).

Further analyses of our data were performed by taking into account only nonredundant ESTs. That is, when more than one EST showed homology to a gene annotated in the databases, only one EST was considered in the analysis.

ESTs with predicted or known functions were classified into putative cellular roles (4). The proportion of ESTs in each role category is shown in Fig. 2. Of the 398 nonredundant ESTs analyzed, the largest number (23.3%) was related to protein synthesis; other categories include sequences related to metabolism (7.9%), protein destination (8.2%), transcription (4.7%), and energy (3.7%). Interestingly sequences related to cell surface proteins accounted for 10.9% of the analyzed ESTs (the second-largest category of known functions). It is well known that T. cruzi has a large number of surface proteins belonging to at least two main families: the mucin gene family and the superfamily of surface antigens.

View larger version (49K): [in this window] [in a new window]   FIG. 2.   Functional classification of T. cruzi ESTs, showing the proportions of predicted genes according to their putative biological functions. A total of 398 nonredundant ESTs having a P value of 105 were classified into 13 categories.

The mucin gene family, for which a minimum of 484 genes has been estimated (15), is composed of two groups of genes, as defined by their central domains. One group contains genes having a variable number of tandem repeats, whereas genes in the second group have nonrepetitive sequences (14). Six ESTs matched members of the mucin gene family; one matched members belonging to the former group (TENS0234), whereas the other five ESTs matched different members belonging to the second group of genes (TENS0206, TENS0592, TENS1868, TENS0163, and TENS1740).

The superfamily of surface antigens is composed of hundreds of members that can be grouped into four families (groups I to IV) based on their similarities (9, 13).

Several ESTs showed significant matches to members belonging to group II, which comprises the so-called GP85 surface glycoproteins (TENS0211, TENS0203, TENS0196, TENS0182, TENS0142, TENS0215, TENS1365, TENS0190, TENS0229, TENS1292, and TENS0222). Interestingly, the top-ranking sequences of the BLAST searches corresponding to the last two ESTs matched the sequences coding for amastigote surface protein-2 and -1, respectively, which have recently been described as the first trans-sialidase (TS) superfamily members preferentially expressed in the amastigote stage (21, 27). In contrast, members of group I (which contains some members that express TS activity), group III, and group IV were hit by only one EST each (TENS0149, TENS0779, and TENS1235, respectively).

The results reported above show that several ESTs have significant matches to trypomastigote- and amastigote-expressed members of the TS superfamily. Although these molecules are stage-specific proteins not present at detectable levels in the epimastigote stage, this result might be expected for trypanosomatids. Unlike transcriptional gene regulation in other organisms, gene regulation in these parasites takes place mainly by posttranscriptional mechanisms (23), even for the expression of stage-specific proteins (29). Thus, it is possible that a low level of trypomastigote- and amastigote-specific mature mRNAs coding for these proteins is present at the epimastigote stage, even though the encoded proteins are absent. Another possibility is that these cDNAs are derived from contaminating metacyclic trypomastigote forms (estimated to be at about 1%) present in the epimastigote culture.

We next organized the EST data set according to matches to the NCBI nonredundant databases. Table 2 lists all significant matches to non-T. cruzi entries in GenBank sorted according to matches to the "other trypanosomatids" and "other organisms" categories. In cases where several entries from various species had significant scores, only the top-ranking score is given. A complete (including matches to T. cruzi) and updated listing of matches to known sequences present in GenBank can be found at our laboratory home page (http://www.iib.unsam.edu.ar/genomelab/tcruzi/5ests.html). A detailed analysis of the putative genes identified is not within the scope of this work and will certainly be done by interested researchers in the field. However, a number of interesting matches with sequences from other organisms were observed. Among them are several proteins identified in other trypanomatids, including several metabolic enzymes (TENS1285, TENS1439, TENS1345, and TENS1204); a homolog to a recently described TRACK (receptor for activated C kinase) in T. brucei rhodesiense (TENS1408); a cyclophilin A (TENS0472); a nucleic acid-binding protein (homolog to the universal minicircle binding protein) (TENS1943); and a homolog to GP63-3 (TENS1942), a metalloprotease originally found in Leishmania and recently described for T. brucei rhodesiense (17). This protein seems to play an important role in the invasion (30) and survival (12) of the leishmanial parasites within the macrophage and has not been detected previously in T. cruzi. This result emphasizes the efficacy of the EST approach, which has allowed us to identify a gene potentially important in the host-parasite interplay.

Other ESTs matched known proteins in other organisms, including TATA-binding protein-interacting protein 49 (TENS1944), serine/threonine protein kinase (TENS1391), serine/threonine protein phosphatase 2b catalytic subunit (calcineurin) (TENS1318), phosphorylation-regulatory protein HP-10 (TENS1945), meiotic spindle formation proteins (TENS1395, and TENS1293), mitotic centromere-associated kinesin (TENS1375),  and p112 proteosome subunits (TENS1289 and TENS1377), DNAJ protein (TENS1338), ADP-ribosylation factor (TENS1946), a probable cell division control protein (TENS1370), several RAS-related proteins (TENS1644, -1947, -1948, and -0394), translation initiation factor 5A (TENS1949), a negative regulatory factor of a transcriptional activator (TENS1941), enolases (TENS1381 and -1274), and a phosphoinositide-specific phospholipase C (TENS1347). Interestingly this last EST showed significant matches to phosphatidylinositol-specific phospholipases C from different organisms and did not show any significant match either to an already-reported T. cruzi glycosylphosphatidylinositol-specific phospholipase C (PID|e329378) or to glycosylphosphatidylinositol-specific phospholipases from other trypanosomatids, suggesting the presence of at least two different enzymes in T. cruzi. Some of the sequences mentioned above have also been identified in a recently published paper (8).

Several ESTs had strong matches with hypothetical, probable, or putative proteins (Table 2), many of them derived from genome sequencing projects for different organisms (mouse, human, Drosophila, yeast, and Arabidopsis, etc.). Although statistically significant similarities do not necessarily mean that these putative proteins actually exist, some of the highly significant matches might indicate that they are indeed real proteins conserved during evolution. Obviously, further sequence analysis and biochemical work are needed to distinguish among these and other possible alternatives.

Until the budget for the complete sequencing of the T. cruzi genome is available, a reasonable accomplishment will be the identification of a large proportion of the gene content in T. cruzi. This might be done by EST or genomic sequencing (18) in the near future. The next step in the short run would be the analysis of the data and the development of new approaches both for the identification of targets for chemotherapy and for vaccine development. Given the difficulties in the treatment of parasitic diseases and the frequent appearance of mutants resistant to chemotherapeutic agents among some protozoa such as Plasmodium and Leishmania (22, 25), gene discovery might be a cost-efficient way to contribute to the eradication of these diseases, which mostly affect developing countries.