Email updates

Keep up to date with the latest news and content from BMC Genomics and BioMed Central.

Open Access Research article

A Plasmodium falciparum FcB1-schizont-EST collection providing clues to schizont specific gene structure and polymorphism

Isabelle Florent1*, Betina M Porcel2, Elodie Guillaume1, Corinne Da Silva2, François Artiguenave2, Eric Maréchal3, Laurent Bréhélin4, Olivier Gascuel4, Sébastien Charneau1, Patrick Wincker2 and Philippe Grellier1

Author Affiliations

1 FRE3206 CNRS/MNHN, USM504, Biologie Fonctionnelle des Protozoaires, RDDM, Muséum National d'Histoire Naturelle, Paris, France

2 UMR-CNRS 8030, Genoscope, Evry, France

3 UMR-CNRS 5168, CEA, INRA, Université Joseph Fourier, Grenoble, France

4 UMR-CNRS 5506, Laboratoire d'Informatique, de Robotique et de Micro-électronique de Montpellier, Université de Montpellier II, Montpellier, France

For all author emails, please log on.

BMC Genomics 2009, 10:235  doi:10.1186/1471-2164-10-235


The electronic version of this article is the complete one and can be found online at: http://www.biomedcentral.com/1471-2164/10/235


Received:11 September 2008
Accepted:19 May 2009
Published:19 May 2009

© 2009 Florent et al; licensee BioMed Central Ltd.

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Background

The Plasmodium falciparum genome (3D7 strain) published in 2002, revealed ~5,400 genes, mostly based on in silico predictions. Experimental data is therefore required for structural and functional assessments of P. falciparum genes and expression, and polymorphic data are further necessary to exploit genomic information to further qualify therapeutic target candidates. Here, we undertook a large scale analysis of a P. falciparum FcB1-schizont-EST library previously constructed by suppression subtractive hybridization (SSH) to study genes expressed during merozoite morphogenesis, with the aim of: 1) obtaining an exhaustive collection of schizont specific ESTs, 2) experimentally validating or correcting P. falciparum gene models and 3) pinpointing genes displaying protein polymorphism between the FcB1 and 3D7 strains.

Results

A total of 22,125 clones randomly picked from the SSH library were sequenced, yielding 21,805 usable ESTs that were then clustered on the P. falciparum genome. This allowed identification of 243 protein coding genes, including 121 previously annotated as hypothetical. Statistical analysis of GO terms, when available, indicated significant enrichment in genes involved in "entry into host-cells" and "actin cytoskeleton". Although most ESTs do not span full-length gene reading frames, detailed sequence comparison of FcB1-ESTs versus 3D7 genomic sequences allowed the confirmation of exon/intron boundaries in 29 genes, the detection of new boundaries in 14 genes and identification of protein polymorphism for 21 genes. In addition, a large number of non-protein coding ESTs were identified, mainly matching with the two A-type rRNA units (on chromosomes 5 and 7) and to a lower extent, two atypical rRNA loci (on chromosomes 1 and 8), TARE subtelomeric regions (several chromosomes) and the recently described telomerase RNA gene (chromosome 9).

Conclusion

This FcB1-schizont-EST analysis confirmed the actual expression of 243 protein coding genes, allowing the correction of structural annotations for a quarter of these sequences. In addition, this analysis demonstrated the actual transcription of several remarkable non-protein coding loci: 2 atypical rRNA, TARE region and telomerase RNA gene. Together with other collections of P. falciparum ESTs, usually generated from mixed parasite stages, this collection of FcB1-schizont-ESTs provides valuable data to gain further insight into the P. falciparum gene structure, polymorphism and expression.

Background

Malaria, the most devastating parasitic human disease, is due to infections by intracellular protozoan parasites belonging to the Plasmodium genus transmitted by Anopheles mosquitoes [1]. Four Plasmodium species are pathogenic to humans, with P. falciparum responsible for 90% of all reported cases of malaria, which causes 1.5 to 2.7 million deaths per annum [2]. No efficient vaccine is currently available, despite ongoing efforts over the last decades [3], and alternative drugs and targets are being investigated to fight the drug-resistant parasites that have emerged since the 1960s and are continuously spreading [4].

Deciphering of the P. falciparum genome in 2002 [5] revealed 5,300–5,400 genes, 60% of which were initially annotated as hypothetical, since no function could be ascribed to them based on sequence similarity. The PlasmoDB database http://www.plasmodb.org webcite gathers genomic and post-genomic data regarding P. falciparum and related species, and the last inventory (version 5.4) indicated 5,484 coding genes, 3,155 (~57%) of which were still annotated as hypothetical or hypothetical conserved (i.e. conserved throughout the Plasmodium genus). Determining gene structures is particularly difficult in the case of P. falciparum, not only because most genes are devoid of characterized orthologs on which gene models could be based, but also because of the very high A-T content of the genome, i.e. 80.6% on average [5]. Gene-coding predictions, based on several algorithms (PHAT, GeneFinder, GlimmerM, Hexamer) have however allowed models to be proposed for P. falciparum genes [6], but these gene models require experimental data to be validated.

We previously reported the construction of an EST library using highly synchronized P. falciparum parasites of the FcB1 strain (from Colombia) to isolate genes selectively expressed during merozoite morphogenesis [7]. The merozoite is the tiny (1 μm) free form of the parasite that is able to recognize, bind and then invade erythrocytes [8]. This very specialized cell displays a number of remarkable features, including a surface coat composed of highly polymorphic merozoite surface proteins (MSPs), some of which were shown to be essential for parasite invasion and survival [8,9]. The merozoite is also equipped with specialized organelles, such as micronemes, rhoptries and dense granules, devoted to invasion. For example, erythrocyte binding antigens, stored in micronemes, are released prior to invasion and allow host cell recognition, while rhoptry proteins stored in rhoptries are release later and are believed to participate in parasitophorous vacuole formation [8]. Shortly after invasion, internalized merozoites differentiate into rings, thus losing their shape, and specific invasive organelles and rings eventually differentiate into haemoglobin-degrading trophozoites, about 20 h after invasion [10]. Nuclear division takes place at about 36 h post-invasion, yielding schizont stages progressively containing up to 32 nuclei [10]. Merozoites are individualized around each nucleus in the very last hours of the erythrocytic cycle, just prior to their release from infected erythrocytes [10]. Protein synthesis, trafficking and organelle assembly to form mature merozoites take place mainly during the final 10–12 h of asexual development [11]. For these reasons, the FcB1-schizont-EST library was constructed by subtracting transcripts from highly synchronized late stages (42–48 h post-invasion) by transcripts from mixed remaining stages (0–40 h post-invasion) by suppression subtractive hybridization [7].

A pilot study of this library, limited to 50 clones, led to the identification of 40 genes over-expressed in schizont stages, including the well-known late-schizont/merozoite specific genes coding for EBA-175, CLAG/RhopH1, coronin, MSP1, MSP3, MSP6, myosin A, SERA and SERP proteins, which was in good agreement with the proposed stage specificity of this library [7]. Seven inserts randomly selected from this FcB1-schizont-EST library were used to probe cDNAs amplified from rings, early trophozoites, late trophozoites, early schizonts and late schizonts and all seven genes were consistently detected in samples corresponding to early and late schizonts as compared to a control gene expressed in all stages [7]. Molecular analysis of the PfDYN2 gene identified during this pilot study [PF10_0360] also confirmed its expression in late schizonts [12]. In addition to P. falciparum genes over-expressed in late schizont stages, this pilot study led to the identification of two genes (PF11_0494 and PFL0290w) whose EST sequences indicated intron/exon boundaries that differed from those previously predicted and four genes (CLAG/RhopH1/PFC0120w, MSP3/PF10_0345, PF13_0053 and PF14_0175) for which clear strain-dependent polymorphism was identified between FcB1 and 3D7 [7]. In this context, it was thus of interest to conduct a large-scale analysis of this FcB1-EST library. The expectations were: 1) to obtain an exhaustive collection of genes selectively over-expressed in late schizont stages; 2) to validate or invalidate P. falciparum gene models by aligning all ESTs with the genome sequence, and 3) to identify genes indicating protein polymorphism between the FcB1 and 3D7 strains. While writing this article, version 5.4 of PlasmoDB was released, taking into account three new sets of P. falciparum ESTs, namely those analyzed by Watanabe et al. [13] and Lu et al. [14] as well as the ESTs sequenced in the present work. This article presents and discusses the FcB1-schizont-EST data and its contribution to our knowledge of P. falciparum gene structure and polymorphism.

Results and discussion

FcB1-schizont-ESTs clustering on the P. falciparum genome

A total of 22,125 clones randomly picked from the library were sequenced, yielding 21,805 (98.5%) usable ESTs that were matched to the P. falciparum genome using the BLAST/est2genome method [15]. Three successive est2genome clustering analyses were performed. The first clustering (BLAST score > 700) was carried out using PlasmoDB version 4.4 http://www.plasmodb.org/plasmo/ webcite as a source of genomic data for P. falciparum. PlasmoDB version 4.4 provides, in addition to genomic sequences and gene annotations, various gene models built using algorithms such as PHAT, GeneFinder, GlimmerM and Hexamer [6]. The second clustering (BLAST score > 700) was achieved using PlasmoDB version 5.3, released on June 2007. It yielded very similar results except that alternative gene models were not accessible in this new PlasmoDB version, which prevented an in-depth analysis of some of the data (see further). These two clustering analyses allowed the matching of 19,459 ESTs (93.5%) into 328 distinct clusters, with a cluster being a group of overlapping ESTs (sharing at least 100 bp) matching a specific genomic locus (Figure 1). The number of ESTs per cluster ranged from 1 (see for example cluster_23, matching the hypothetical RESA-like gene PFB0085c) to thousands (see for example cluster_322, matching the rRNA unit on chromosome 5) (Additional file 1). The remaining unmatched 2,346 FcB1 ESTs that could correspond to gene fragments that diverge markedly between FcB1 and 3D7 strains, to genes present in FcB1 but absent from 3D7, or to genes encoded in the mitochondrial or apicoplast genomes, were clustered again on the P. falciparum genome (PlasmoDB version 5.3), thus lowering the BLAST score to 500. While 839 ESTs were rejected for being too small (< 60 nucleotides) or displaying very low complexity, 447 additional FcB1-ESTs could be significantly aligned with loci on the P. falciparum genome (3D7), revealing 24 additional clusters (Figure 1). The remaining unmatched ESTs were then aligned with protein sequences of the UniProt database [16] using the BLAST algorithm, revealing additional ESTs matching MSP1 variants of K1-type (210 ESTs on [UniProt:P04932] and 1 EST on [UniProt:A0SJF0/EMBL:DQ489588]), to Ebl-1 ([UniProt:Q8IEB6], 59 ESTs), the mitochondrial genome (160 ESTs on [EMBL:AJ276844]) and the tufA gene ([UniProt:Q25820], 1 EST) on the apicoplast genome [UniProt:X95276] (Figure 1). The results of the three est2genome clustering analyses on the 3D7 genome can be viewed at https://www.genoscope.cns.fr/secure-nda/projet_PU/cgi-bin/gbrowse/plasmodium/4.4 webcite and https://www.genoscope.cns.fr/secure-nda/projet_PU/cgi-bin/ggb/gbrowse/plasmodium/5.3 webcite respectively (*authorisation is required to access these website), and are discussed hereafter. Then results regarding FcB1-schizont-ESTs that did not match 3D7 genomic sequences but did match other P. falciparum sequences in the UniProt database are presented and discussed.

thumbnailFigure 1. Clustering strategy for the analysis of FcB1-schizont-ESTs. The first two clusterings performed using est2genome (BLAST score > 700) on the 3D7 genome (PlasmoDB versions 4.4 and 5.3) allowed clustering of 19,459 ESTs. By lowering the BLAST score to 500, 447 additional ESTs were clustered and mapped on the 3D7 genome (PlasmoDB vs 5.3). The remaining unmatched FcB1-schizont-ESTs were analysed by comparison with the UniProt database, revealing 270 additional ESTs matching MSP1 (K1 type) and Ebl-1, 160 ESTs matching the mitochondrial genome and 1 EST matching the apicoplast genome.

Additional file 1. Complete list of the 253 genomic loci of 3D7 spanned by FcB1-schizont-ESTs. (a, b, c and d) were downloaded from PlasmoDB (version 5.3) and correspond, respectively, to the gene accession numbers in PlasmoDB (a), their current description in PlasmoDB (b), the maximum expression time during the erythrocytic cycle according to the transcriptomic data by [17,19] (c) and the maximum expression stage according to transcriptomic data by [18] (d). (e, f) list the different loci according to their occurrence along each chromosome (from chromosome 1 to 14) and their approximate localisation on each chromosome, in kilobase pairs (f). The corresponding clusters (g) and the total number of ESTs (h) are also indicated. In column (i), we indicate whether the FcB1-EST data allowed us to confirm or modify gene models, or revealed some protein polymorphism and column (j) details these confirmations/modifications/indications. (NI) stands for "non-informative". In (k), Gene Ontology accession numbers, downloaded from GeneDG (genedb.org) are indicated. Genes previously identified in the pilot study [7] are indicated by "yes" in column (l). Color code: red is for functionally annotated genes, blue for putative genes, black for hypothetical genes, green for non-protein coding loci.

Format: XLS Size: 85KB Download file

This file can be viewed with: Microsoft Excel ViewerOpen Data

FcB1-schizont-EST clusters matching the P. falciparum 3D7 genome

A systematic analysis of the 352 clusters (328 + 24) matching the 3D7 genome was performed to identify and study the corresponding loci. These were distributed along the 14 P. falciparum chromosomes (Table 1) and corresponded to 243 coding genes (9254 ESTs, 326 clusters) and 8 non-protein-coding loci, namely: 4 distinct rRNA loci (10611 ESTs, 22 clusters) and 4 subtelomeric regions (41 ESTs, 4 clusters) (Additional file 1).

Table 1. Summary of the FcB1-schizont-EST distribution on the 14 P. falciparum chromosomes.

Protein coding genes

Among the 243 protein coding genes identified in this study, 60 (~25%) currently correspond to functionally annotated genes, 62 (~25%) to putative genes and 121 (~50%) to hypothetical genes (Table 1). Functionally annotated genes in PlasmoDB are genes whose annotations are supported by experimental data (molecular studies, biochemical characterizations, etc.) while putative genes are annotated based on significant similarities with functionally annotated genes in other species but lack experimental data in P. falciparum to support these annotations. Finally, hypothetical genes lack similarities with functionally annotated genes in the database and some of them rely solely on mathematical algorithms for identification. As expected, this new collection of 243 protein coding genes includes all 40 protein coding genes initially identified from the FcB1-schizont-EST library after the pilot study [7] (Additional file 1).

Available transcriptomic profiles [17-19] were recovered for each of these 243 protein coding genes and plotted on graphs to check whether the stage specificity of the FcB1-schizont-library was in agreement with the results of these other transcriptomic studies (Additional file 2). Good concordance was observed, though the FcB1-schizont collection appeared to be slightly younger than expected, corresponding mostly to genes peaking at 37–42 h in the Bozdech transcriptomic study and in early schizonts in the Le Roch transcriptomic study. This shift may be due to differences in the time required to accomplish a full erythrocytic cycle in vitro (42 to 48 h), which is known to be strain dependent [19]. It may also be due to slight variations in morphological appreciation of the different developmental stages. Nonetheless, most genes were confirmed by both transcriptomic studies, with few exceptions (Additional file 1). The expression of about 80% of genes of the FcB1-schizont collection reported to peak in gametocytes according to the Le Roch transcriptomic studies, appeared to peak at 37–46 h in the Bozdech transcriptomic studies (Additional file 1). Finally, conflicting expression profiles may correspond to genes differentially regulated in FcB1 as compared to 3D7, as previously observed by Llinas et al. for a few genes of HB3 and Dd2 strains [19]. Therefore, the FcB1-schizont-EST collection appears to be of interest as a complementary dataset for studying merozoite morphogenesis.

Additional file 2. Stage distribution of protein coding genes spanned by FcB1-schizont-ESTs according to other microarray studies. Comparisons were performed with the transcriptomic data of Bozdech et al. [17] and Le Roch et al. [18], respectively. A: number of protein coding genes covered by FcB1-schizont-ESTs, according to their maximum expression time during P. falciparum development in erythrocytes, in hours [17]; B: number of protein coding genes covered by FcB1-schizont-ESTs, according to their maximum expression stage: 1 (early rings), 2 (late rings), 3 (early trophozoites), 4 (late trophozoites), 5 (early schizonts), 6 (late schizonts), 7 (merozoites), 8 (gametocytes) [18]. C and D correspond to the same data, while taking into consideration the total number of individual FcB1-schizont-ESTs per protein coding gene. Pf-iRBC, P. falciparum-infected red blood cells.

Format: PDF Size: 26KB Download file

This file can be viewed with: Adobe Acrobat ReaderOpen Data

The FcB1-schizont-EST collection contained well known merozoite specific proteins such as merozoite capping protein-1 (PF10_0268), merozoite surface proteins MSP1 (PFI1475w), MSP3 (PF10_0345), MSP4 (PFB0310c), MSP5 (PFB0305c), MSP6 (PF10_0346), MSP7 (PF13_0197), MSP7-like (PF13_0193) and MSP9/ABRA (PFL1365c), GLURP (PF10_0344), Pfemp3 (MAL13p1.405), EBA (MAL7P1.176) and EBA-181 (PFA0125c). Interestingly, we also identified four histone genes: PFF0860c (histone H2a), PFC0920w (histone H2a variant, putative), PFF0865w (histone H3), PFF0510w (histone H3, putative) reported to be involved in nucleosome structure (GO:0000786) and assembly (GO:0006334) as well as chromatin structure (GO:0000785) and assembly (GO:0031497). In addition to these four histone genes, 3 genes are also annotated in the Gene Ontology database by the term chromosome (GO:0005694): PFE0450w (chromosome condensation protein, putative), PF14_0316 (DNA topoisomerase II, putative) and PFL1930w (hypothetical protein conserved). As previously observed [7], we also identified several cytoskeleton associated proteins: myosin A (PF13_0233), myosin D (PFL1435c), coronin (PFL2460w), dynamin-1 (PF11_0465), dynamin-2 (PF10_0368) and formin-2 (PFL0925w). Several rhoptry proteins were also present in this FcB1-schizont dataset: 3 of the 5 known CLAG/RhopH1 (PFC0110w, PFC0120w, MAL7P1.229), RhopH2 (PFI1445w), RhopH3 (PFI0265c), RAP1 (PF14_0102), RAP3 (PFE0075c), RAMA (MAL7P1.208) and also PF14_0495, which is an ortholog of the TgRON2 protein secreted from the rhoptry neck portion in Toxoplasma gondii [20]. We also characterized the expression of MAL8P1.73, reported to be an ortholog of Ts4705, a T. gondii protein detected in rhoptry extracts [20], which is also known to bind AMA-1, like TgRON2 and TgRON4 [21].

To further characterize the cellular components, molecular functions and biological processes in which the genes of the FcB1-schizont dataset may be involved, GO-terms were downloaded from GeneDB (genedb.org). GO annotations were, however, available only for 159 of the 243 genes (see Additional file 1). We then used GOStat software [22]http://gostat.wehi.edu.au/cgi-bin/goStat.pl webcite to identify GO terms over-represented in the annotations of these 159 genes as compared to the complete P. falciparum genome, using a p-value threshold of 0.01 (see Table 2). Over-represented cellular components were actin cytoskeleton (GO:0015629), chromosome (GO:0005694), myosin complex (GO:0016459), nucleosome (GO:0000786) and rhoptry (GO:0020008). Similarly, over-represented molecular functions included actin-binding (GO:0003779), calcium ion binding (GO:0005509) and phospholipid binding (GO:0005543). Over-represented biological processes corresponded to cell division (GO:0051301), cytokinesis (GO:0000910), DNA packaging (GO:0006323), nucleosome assembly (GO:0006334) and entry into host cells (GO:0030260). These GO terms are in good agreement with our current knowledge of biological and molecular mechanisms that occur during merozoite morphogenesis. Conversely, a single GO term appeared to be under-represented in our study: the term defense response (GO:0006952) (see Table 2). In P. falciparum, this term has been attributed to var genes (see amigo.geneontology.org), which encode PfEMP1 surface proteins responsible for antigenic variation [23]. Since these var genes are known to be transcribed early during parasite development [23,24], the corresponding ESTs were not expected to be present in the FcB1-schizont library.

Table 2. GO term analysis of genes spanned by FcB1-schizont-ESTs.

Such a statistical analysis of GO-terms is still, however, limited by the number of P. falciparum proteins not yet annotated in the Gene Ontology database. For example, only two of the 10 rhoptry proteins that were identified in the FcB1-schizont list (see above) were annotated by the corresponding GO term (GO:0020008): RAP1 (PF14_0102) and RAP3 (PFE0075c).

Examination of FcB1-3D7 alignments

Although most FcB1-schizont ESTs do not span the full length of the corresponding genes, a detailed systematic comparative analysis between FcB1 EST sequences and corresponding 3D7 genomic sequences was performed to check the accuracy of the currently proposed gene models in PlasmoDB and also to identify protein polymorphism between these two strains.

Validation and modification of intron/exon boundaries

In most cases (~75%), FcB1-schizont-ESTs matched protein coding genes in the middle of described exons, without providing any relevant information for gene model validation. On several occasions (40 genes in total), FcB1-schizont-ESTs matched 5' or 3' borders of genes or spanned exon-exon boundaries. A gene by gene analysis of these cases allowed confirmation of one or several introns in 29 gene models (26 in Table 3, 3 in Table 4) and to propose modifications in 14 gene models (Table 4). Clustering of FcB1-schizont-ESTs in PlasmoDB (version 4.4), which displays alternative gene models, was particularly informative during this analysis and a few remarkable examples of gene model validations and gene model modifications are illustrated in Additional files 3 and 4, respectively. The most spectacular gene model correction was found for PFE0240w, whose FcB1-schizont-EST data provided evidence of four additional exons and an extended C-terminal end (Additional file 4).

Table 3. Gene models confirmed by FcB1-schizont-ESTs.

Table 4. Gene models modified by FcB1-schizont-ESTs.

Additional file 3. Remarkable examples of gene model validations. These examples are indicated by * in Table 3. For each example, the scale on top indicates kilobases along the chromosome (which is mentioned above the blue line), FcB1-schizont-ESTs are symbolized below (coloured boxes for exons, arrows for introns) while also indicating the cluster number and location (thin green line). All PlasmoDB gene models, as downloaded from vs 4.4, are indicated below (green boxes for exons, arrows for introns). The various confirmed gene models are: A: MAL13P1.103 (Chr13_08), hypothetical protein conserved (introns 1 to 4 and exons 1 to 5); B: PFB0815w (Chr02_15), Pf calcium-dependent protein kinase 1 (introns 1 to 4); C: PFC0120w (Chr03_02), cytoadherence linked asexual protein (introns 1 to 5); D: PFE1415w (Chr05_15), cell cycle regulator with zinc-finger domain, putative (introns 5 to 8); E, PFL0975w (Chr12_09), hypothetical protein conserved (introns 3 and 4 and end of the gene).

Format: PDF Size: 49KB Download file

This file can be viewed with: Adobe Acrobat ReaderOpen Data

Additional file 4. Remarkable examples of gene model modifications. These examples are indicated by * in Table 4. The representations are as indicated for Additional file 3. The various modified gene models are: A: PFA0630c, hypothetical protein, (Chr01_06), FcB1-schizont-EST data (cluster_149) is in agreement with chr1. genefinder_16r, chr1. glimmer 366 and chr1. phat_146 models; B: PF11_0194, hypothetical protein, (Chr11_08), FcB1-schizont-EST data (cluster_122) is in agreement with Chr11.genefinder_157r model; C: PFE0240w, hypothetical protein, conserved, (Chr05_05), FcB1-schizont-EST data (cluster_308) modifies gene for which no prediction was available. It indicates four additional exons and predicts a longer protein (172 aa versus 115 aa); D: PFI1565w, conserved protein, (Chr09_14), FcB1-schizont-EST data (cluster_81) is in agreement with chr9.glimmerm_973 and chr9.glimmerm_974 models for the end of the gene; E: PFL0290w, hypothetical protein, conserved, (Chr12_03), FcB1-schizont-EST data (cluster_59) suggests that intron 1 would be smaller. Intron 2, however, is confirmed.

Format: PDF Size: 40KB Download file

This file can be viewed with: Adobe Acrobat ReaderOpen Data

FcB1 versus 3D7 protein polymorphism for 21 protein coding genes

Although the FcB1-schizont-EST collection provides useful nucleotide sequence data to indicate single nucleotide polymorphism (SNP) specific to the FcB1 strain, these data (which will be available in PlasmoDB) are not discussed here. However, during the systematic comparative analysis of FcB1-schizont-ESTs versus 3D7 sequences, we identified 21 genes for which some protein polymorphism was observed between FcB1 and 3D7 encoded proteins (Table 5). As illustrated in Additional file 5 where these 21 protein alignments are displayed, in most cases, this protein polymorphism corresponded to a variable number of repeated elements or to amino acid variations in these repeat elements. Malaria protein polymorphism has been suggested to be one of the main strategies of the parasite to evade the host immune mechanism, and antigens that are under natural immune pressure tend to have a higher level of polymorphism [25,26]. These 21 genes therefore seem to be good candidates for encoding such antigens. However, according to PlasmoDB, only 4 of these 21 genes encode proteins harbouring putative signal peptides consistent with their exposure on cell surface: PF10_0177 (erythrocyte membrane-associated antigen), PF10_0345 (MSP3), PF10_0351 (hypothetical protein expressed in late schizogony) and PFL1385c (MSP9/ABRA). For the other proteins, these results raise questions about the reasons underlying such polymorphism or how these proteins may be exported either to the surface of the parasite or the surface of infected-red blood cells [27].

Table 5. Evidence of protein polymorphism between FcB1 and 3D7 strains.

Additional file 5. FcB1 versus 3D7 protein polymorphism. All 21 cases reported in Table 5 are illustrated. In each example, "FcB1" corresponds to the protein sequences deduced from a representative FcB1-schizont-EST and "3D7" to protein sequences deduced from the homologous 3D7 gene (EMBL/GenBank/DDBJ accession numbers indicated). Most sequence alignments were obtained using classical BLAST searches (default parameters) at NCBI http://blast.ncbi.nlm.nih.gov/Blast.cgi webcite. For Chr05_12, Chr12_15 and Chr13_12, sequence alignments were performed using the EMBOSS package available at EBI http://www.ebi.ac.uk/Tools/emboss/align/index.html webcite.

Format: PDF Size: 24KB Download file

This file can be viewed with: Adobe Acrobat ReaderOpen Data

3D7 non-protein coding loci covered by FcB1-schizont-ESTs

Due to the methodology used to build the FcB1-schizont-EST library, i.e. isolation of total RNA followed by selective conversion of polyA+ RNA into double strand cDNA by SMART-PCR [7], a very large number of sequences (10,611) corresponded to rRNA fragments, as previously observed [7]. These non-protein coding, rRNA-matching ESTs represent more that half of the sequenced ESTs, which is much higher than the 30%, estimated after the pilot study [7]. This may be due to the fact that all picked clones were randomly sequenced in this current high throughput study, while in the pilot study EST clones had been size selected prior to sequencing.

The rRNA gene organisation in the P. falciparum genome is very unusual as compared to other eukaryotes. Instead of having hundreds of repetitions of rRNA units in its genome, P. falciparum has seven complete and incomplete rRNA (18s-5.8s-28s) units on six of its 14 chromosomes and 3 copies of 5s rRNA genes on chromosome 14 [5] (Table 6). Some of these rRNA loci are known to be expressed in a developmentally regulated fashion, e.g. the two A-type rRNA units (on chromosomes 5 and 7) that are mainly expressed in human infection stages and the two S-type rRNAs (on chromosomes 11 and 13) that are expressed in insect infection stages [5,28]. The role played by other rRNA loci present in the P. falciparum genome is not entirely clear [5].

Table 6. Summary of the FcB1-schizont EST distribution on the different rRNA loci of P. falciparum.

Most non-protein coding FcB1-schizont-ESTs correspond to the two A-type rRNA units on chromosomes 5 and 7

In this study, a very large majority of the rRNA-matching ESTs clustered on both A-type rRNA units, located on chromosomes 5 and 7 (Table 6) and most matched both homologous A-type regions on chromosomes 5 and 7 (Additional file 6). However, a limited number of FcB1-schizont ESTs clustered on two atypical rRNA units, respectively located on chromosome 1 (626 ESTs) and chromosome 8 (2 ESTs) (Table 6). None of the FcB1-schizont EST clusters matched any of the two S-type rRNA loci, which was an interesting negative control of our data regarding the developmental stage of the studied parasite population.

Additional file 6. List of 26 FcB1-schizont-ESTs representative of the 9983 FcB1-ESTs matching rRNA loci on chromosomes 5 and 7. To limit this redundancy in the public databases, 26 sequences corresponding to the longest ESTs on each cluster were selected to be deposited in EMBL. (a) PU number of the EST; (b) size in bp; (c) corresponding rRNA element (5.8s, 18s, 28s or ITS1, internal transcribed spacer 1); (d, e, f) position of Hit on chromosome 5 (d), chromosome 7 (e) and chromosome 1 (f), respectively.

Format: PDF Size: 5KB Download file

This file can be viewed with: Adobe Acrobat ReaderOpen Data

FcB1-schizont-EST analysis provides evidence of the actual expression of two atypical rRNA loci located on chromosomes 1 and 8

A total of 626 ESTs matched chromosome 1 at the level of MAL1_28s, which is annotated as A-Type rRNA in PlasmoDB but described as being atypical (65% identity to A-type and 75% identity to S-type) by Gardner et al. [5]. The four longest ESTs matching this locus did not match elsewhere in the P. falciparum genome (Additional file 7). These results indicated that the MAL1_28s gene was indeed expressed in our experimental population of highly synchronized FcB1 parasites. Owing to the small number of truly specific ESTs corresponding to MAL1_28s, the expression level of this gene in the parasite population is obviously much lower than the level rRNA gene expression from typical Chr5 and Chr7 A-type loci (Table 6). While the physiological conditions allowing MAL1_28s rRNA expression remain to be determined, our experimental data clearly argue in favour of the actual expression of MAL1_28s RNA which, to the best of our knowledge, has not yet been documented.

Additional file 7. Clustering of 626 FcB1-schizont-ESTs on the MAL1_28s gene of the chromosome 1 rRNA locus. A: view of the chromosome 1 locus (positions 473 k to 482 k) from PlasmoDB (version 4.4); B: view of the same locus on the Genoscope browser. Only the first, longest, 14 FcB1-ESTs are shown, the remaining 612 ESTs (corresponding to cluster_146 and cluster_147) are not shown here. The four largest ESTs (PU0AAA1YC08RM1, PU0AAA22YG18RM1, PU0AAA13YE12RM1, and PU0AAA44YM17RM1, indicated by *) are strictly specific to this chromosome 1 locus. Most of the smallest ESTs also clustered on homologous rRNA loci on chromosomes 5 and 7, MAL5_28s and MAL7_28s. (see Genoscope browser for details).

Format: PDF Size: 30KB Download file

This file can be viewed with: Adobe Acrobat ReaderOpen Data

The two ESTs matching chromosome 8 were mapped on a non-annotated area located between MAL8a_5.8s and PF08_temp1 (Additional file 8), with the latter also being annotated as rRNA encoding in PlasmoDB. Both are non-protein coding ESTs and do not match elsewhere in the genome. This strongly suggests that they are indeed encoded by this new locus of unknown function in which, interestingly, several SNPs are indicated (see PlasmoDB version 5.4). The FcB1-schizont-EST data therefore supports the transcription of a non-protein coding RNA at this new locus of chromosome 8.

Additional file 8. FcB1-schizont-ESTs matching the atypical rRNA locus on chromosome 8, downstream of MAL8a_5.8s and upstream of PF08_tmp1. A: view of the chromosome 8 locus (positions 1281 k to 1293 k) from PlasmoDB (version 4.4); B. view of the same locus on the Genoscope browser.

Format: PDF Size: 9KB Download file

This file can be viewed with: Adobe Acrobat ReaderOpen Data

Non-protein coding FcB1-schizont-ESTs matching sub-telomeric regions

Four clusters (cluster_304, Chr05_01; cluster_64, Chr08_01; cluster_98, Chr10_01 and cluster_188, Chr13_01) and one atypical EST (PU0AAA27YL11RM1, Chr07_15) matched sub-telomeric regions in the P. falciparum genome, very close to the chromosome ends (see Additional file 1). Notably, the 36 ESTs corresponding to cluster_304 (Chr05_01) also corresponded to cluster_188 (Chr13_01). An in-depth analysis of these ESTs, using the BLAST algorithm optimized for highly similar sequences (via NCBI BLAST server), indicated that they matched virtually all chromosome ends (90 to 93% identity at the nucleotide level, data not shown), in telomere-associated repeat element (TARE) regions [29,30]. To refine this analysis, these ESTs were aligned to the nucleotide sequence corresponding to a particularly well annotated telomeric end on chromosome 3 [EMBL:AL034560]. These FcB1-schizont-ESTs were then found to match the same region, likely corresponding to the TARE1 or TARE2 region, between telomere and R-CG7 segments [31] (Additional file 9).

Additional file 9. FcB1-schizont-ESTs matching sub-telomeric regions of chromosomes. A: schematic representation of P. falciparum sub-telomeric regions and consensus sequences reported for each structure, compiled from [29] and [30]. B: alignment of representative ESTs for Chr05_01 = Chr13_01 (cluster_304 and cluster_188, PU0AAA56YB23RM1), Chr10_01 (cluster_98, PU0AAA22YJ11RM1), Chr08_01 (cluster_64, PU0AAA57YH13RM2) and Chr07_15 (cluster_48, PU0AAA27YL11RM1) on the end of chromosome 3 (EMBL AL034560), shown as a typical example, These FcB1-schizont-ESTs are localized in TARE regions, between telomere and R-CG7 segments [31].

Format: PDF Size: 14KB Download file

This file can be viewed with: Adobe Acrobat ReaderOpen Data

Non-protein coding FcB1-schizont ESTs matching telomerase RNA

An independent analysis of the FcB1-schizont-EST library aimed at unravelling the genetic structure of the P. falciparum telomerase RNA (TR-RNA) revealed 5 FcB1 ESTs that mapped to the 5' end of the recently described telomerase RNA gene in a template binding region [32] (Additional file 10). It was then noted that four of these FcB1-schizont-ESTs also corresponded to the cluster 500_89 cluster, matching the hypothetical gene PFI0905w (see Additional file 1). The fact that these were non-protein coding ESTs indicates that they corresponded to the telomerase RNA gene rather than to cDNA fragments of the PFI0905w gene.

Additional file 10. FcB1-schizont-ESTs matching the template boundary element (TBE) and template regions of P. falciparum telomerase RNA. The sequence of P. falciparum telomerase RNA (Pf TR telomerase) was aligned with sequences of the five FcB1-schizont-ESTs (PU0AAA accessions numbers indicated) by using ClustalW showing extensive conservation. Template boundary element (TBE) and template regions, both located at the 5'-end of the P. falciparum telomerase RNA, are as defined by [32].

Format: PDF Size: 5KB Download file

This file can be viewed with: Adobe Acrobat ReaderOpen Data

In-depth analysis of FcB1-schizont-ESTs not matching the 3D7 genome reveals similarities to other variants of the MSP1 gene and to a paralog of EBA-140 (MAL13P1.60) and EBA-175 (MAL7P1.176)

The MSP1 gene codes for one of the most polymorphic proteins in P. falciparum. The analysis of FcB1-schizont-ESTs using est2genome allowed clustering of 139 ESTs with the N-terminal end of the MSP1 gene in PlasmoDB (PFI1475w, cluster_78 and cluster_79). However, additional FcB1-schizont-ESTs matching other MSP1 variants were discovered by comparing unmatched ESTs to the UniProt database using the BLAST algorithm (Figure 1). 210 of these FcB1-schizont-ESTs matched five partly overlapping regions of the K1-type MSP1 [UniProt:P04932] and 1 additional ESTs matched the C-terminal end of MSP1 in an Iranian isolate [EMBL:DQ489588] reported to be of K1-type [33]. Therefore, a total of 350 FcB1-schizont-ESTs spanned about 50% of the FcB1-MSP1 gene (see Additional file 11). Interestingly, in the pilot study of this library, we found that the MSP3 variant in FcB1 was also of K1-type [7].

Additional file 11. Coverage of the MSP-1 gene by FcB1-schizont-ESTs. MSP1 protein sequences (1720 amino acids for the 3D7-type (downloaded from PlasmoDB), 1630 amino acids for the K1-type [UniProt:P04932], 1720 amino acids for the Iranian isolate [UniProt: A0SJF0/EMBL:DQ489588] were aligned and are represented schematically, red segments corresponding to gaps. The regions covered by the 139 ESTs matching 3D7-MSP1 are indicated above the alignment, the regions covered by 210 ESTs matching K1-MSP1 below. The EST matching the MSP1 variant [UniProt:A0SJF0] (Iranian isolate of K1-type) is also represented. Cluster_78 (corresponding to 21 ESTs matching amino-acid positions 230 to 403 of 3D7-MSP1) and cluster_79 (corresponding to 118 ESTs matching the N-terminal end of the protein to position 260) are indicated as arrows. The remaining 210 ESTs matching K1-MSP1 are represented as five groups with no cluster names. For each group, the number of ESTs is indicated above each arrow and the boundaries below each arrow.

Format: PDF Size: 6KB Download file

This file can be viewed with: Adobe Acrobat ReaderOpen Data

Comparison of FcB1-schizont-ESTs with the UniProt database also revealed 59 ESTs matching [UniProt:Q8IEB6]. These ESTs partly spanned the C-terminal end of the protein (amino acids 1646 to 2188) with 93 to 100% identity. [UniProt:Q8IEB6] is annotated as Ebl-1 from the strain 3D7 but, in fact, two P. falciparum proteins appear as paralogs of [UniProt:Q8IEB6]: EBA-175 (MAL7P1.176, BLAST e-value = 6.7 e-83) and EBA-140 (MAL13P1.60, BLAST e-value = 5.7e-82). We thus believe that these ESTs correspond to a FcB1 gene belonging to the EBL family, known to be involved in invasion [34].

A few FcB1-schizont ESTs mapped on organellar genomes

FcB1-schizont-ESTs matching the mitochondrial genome [UniProt:AJ276844] were mapped at the level of the three described genes: putative cytochrome oxidase III [UniProt:Q9MDY3], putative cytochrome oxidase I [UniProt:Q02766] and putative cytochrome b [UniProt:Q02768]. The single EST matching the apicoplast genome [UniProt:X95276], corresponded to the tufA gene [UniProt:Q25820]. This latter result, based on a single EST in the whole library corresponding to transcription of the apicoplast genome, which is known to occur just prior to the transcription of merozoite specific genes [17], provides an interesting control of the proposed stage specificity of this FcB1-EST library.

Conclusion

Extensive analysis of the FcB1-schizont-EST library, which was constructed to be enriched in transcripts over-expressed in the last hours of the parasite development in erythrocytes, provided valuable information regarding genes expressed in the late stages of the parasite development. Concerning the stage specificity, both the pilot study including experimental controls [7] and the current work provide support that the genes spanned by these ESTs were actually expressed late during the parasite development, even though in many cases this conclusion was based on comparative analysis with other transcriptomic studies [17-19]. The fact that a single EST (out of a total of 21,805) matched the apicoplast genome, further indicates that all genes corresponding to this FcB1-schizont are expressed later, at least in the FcB1 strain. The transcriptomic profiles determined by [17] and [18] for the 243 genes of this FcB1-schizont collection indicate, however, that these genes may be expressed slightly earlier than initially expected: 37–42 h (early schizogony) rather than 42–48 h (late schizogony). This observation is in line with the fact that we identified 10 genes coding for rhoptry proteins and only two coding for micronemal proteins (EBA (MAL7P1.176) and EBA-181 (PFA0125c)). Indeed, it has been shown that rhoptry proteins are expressed prior to micronemal proteins during merozoite morphogenesis [11]. For all of these reasons, we believe that this FcB1-schizont collection is a truly appropriate dataset to focus on early steps of merozoite morphogenesis after apicoplast transcription. The 121 proteins currently annotated as hypothetical or hypothetical conserved are therefore favourable candidates involved in this morphogenesis. One shortcoming is, however, the fact that in several cases no additional experimental data is currently available to confirm the expression of these candidates. Additional data and molecular analyses are thus required to confirm the involvement of these candidates in merozoite morphogenesis as, for example, was done in the case of PfDYN2 [12]. But, interestingly, the total number of genes identified in the FcB1-schizont dataset (243) is consistent with the number of genes proposed by [17] as corresponding to merozoite specific genes.

An interesting outcome of this study is that intron/exon boundaries were validated in 29 genes, revised gene models were proposed for 14 genes and FcB1 versus 3D7 protein polymorphisms were identified for 21 genes. Although this information was mostly collected at the gene fragment rather than full-length gene level, it represents a valuable contribution, along with similar data by others [13,14], to gain greater insight into P. falciparum gene structure and polymorphism.

The most unexpected results concerned ESTs corresponding to non-protein coding RNAs present in this FcB1-schizont collection. Although these ESTs are limited in number, especially as compared to the very high number of ESTs matching the well known A-type rRNA loci, they provided indications on the actual expression of a 28s rRNA gene located on chromosome 1 and on the expression of a non-protein coding transcript between MAL8a_5.8s and Pf08_tmp1 genes. ESTs providing this information do not cluster elsewhere in the P. falciparum genome, so they are strictly specific to these two mainly uncharacterized loci. The actual physiological conditions under which these loci are expressed are, however, totally unknown. A second small series of 41 non-protein coding ESTs provided information regarding the actual transcription of TARE regions, likely corresponding to TARE-1 or TARE-2. To the best of our knowledge, this has not been shown before. Finally, 5 ESTs matched the 5' end of P. falciparum telomerase RNA. Rather than indicating that this RNA is expressed, which has recently been reported [32], this result raises the question as to the structure and expression of PFI0905w, located at the very same genomic locus.

The methodology used to built the FcB1-schizont-EST library, i.e. suppressive subtractive hybridization [35], which was selected to ensure stage specificity, has some technical disadvantages. The first is the very high redundancy in the ESTs produced, which is directly due to the PCR-amplification step. The second is the fact that these ESTs are necessarily digested by an endonuclease to be subtracted (here using RsaI) and are therefore of small or very small size. Consequently, most genes identified by this strategy were only partly spanned by FcB1-schizont-ESTs. These shortcomings did not affect the analyses in the pilot study [7] and were therefore not expected to limit the present work. Nevertheless, the high throughput analysis of the FcB1-schizont-EST library affected the full-length coverage of gene loci by EST clusters and was biased by EST multicopies. For example, in the case of MSP1 and based on the results of the pilot study, we were expecting to yield enough ESTs to span the entire gene, but, instead, we obtained a large number of highly redundant ESTs at fragmented locations. Despite this low full-length gene coverage, the in-depth analysis of this FcB1-schizont-EST collection should help, together with data by others [13,14], to gain further insight into P. falciparum gene structure and polymorphism. In the present work, which involved a high-throughput analysis of this FcB1-schizont-EST library, a few unexpected gene expression features were also discovered, such as the actual transcription of atypical rRNA loci and subtelomeric regions.

Methods

Library construction

The library construction, by suppressive subtractive hybridization, was previously published [7]. Briefly, the chloroquine-resistant P. falciparum FcB1/Colombia strain was cultured and synchronised using standard methods. Two specific populations were isolated: a highly synchronized late stage population corresponding to late schizont/merozoite stages (42–48 h post-invasion, with schizonts containing at least 4 nuclei) and a reference population containing a homogeneous distribution of rings, trophozoites and early schizonts (containing at most 3 nuclei). Parasites were isolated from these two populations using a 0.2% saponin/1 × PBS treatment and total RNA was purified, treated by RQ1 RNase-free DNase (Promega) and converted to high quality cDNAs using the SMART (switch mechanism at the 5'-end of reverse transcript) PCR cDNA synthesis methodology, as recommended (Clontech). Subtracted cDNA populations were then generated using the PCR-Select procedure [35] according to the manufacturer's instructions (Clontech), with slight modifications [7]. Digestion with RsaI yielded cDNA fragments with an average size of ~580 bp. These subtracted cDNAs fragments (i.e. ESTs), corresponding to transcripts over-expressed in late schizont/merozoite stages, were subsequently ligated into pT-Adv (Clontech) and transformed into Escherichia coli TOP10F' competent cells (Clontech).

A total of 22,125 randomly picked clones were sequenced by the rolling circle sequencing procedure, using universal primers, at Genoscope. Raw sequences were treated to mask the various primers used to construct the subtracted library (SMART primer, NP1 and NP2 primers, Clontech), small sequences were removed yielding 21,805 reads, 87.2% of which read through the complete insert. Correspondence between EMBL/GenBank/DDBJ accession numbers, PU accession numbers and clusters described in this current work are provided in Additional file 12.

Additional file 12. EMBL/GenBank/DDBJ AC numbers. Correspondence between EMBL/GenBank/DDBJ AC numbers (1st column), PU numbers (2nd column) and corresponding clusters or UniProt accession numbers, as described in the text (3rd column), with nme standing for "non matched EST". [UniProt:A0SJF0] corresponds to Iranian-MSP1, K1-type, [EMBL:AJ276844]) corresponds to the P. falciparum mitochondrial genome, [UniProt:P04932] corresponds to MSP1 variants of K1-type, [UniProt:Q8IEB6] corresponds to Ebl-1, [UniProt:X95276] corresponds to the P. falciparum apicoplast genome.

Format: XLS Size: 1.3MB Download file

This file can be viewed with: Microsoft Excel ViewerOpen Data

Alignment of FcB1-schizont EST sequences on the 3D7 P. falciparum genome

The P. falciparum genomic data was imported from PlasmoDB http://www.plasmodb.org webcite. In this current work, we used PlasmoDB versions 4.4 and 5.3. We used a two-step strategy to align EST sequences on the P. falciparum genome. As a first step, BLAST [36] was used to align microsatellite repeat-masked EST sequences and genomic sequences using the following settings: W = 20, X = 8, match score = 5, mismatch score = -4. The sum the HSP (high-scoring pair) scores was then calculated for each possible location, and the location with the highest score was then retained if the sum of scores was more than 700. Once the location of the transcript sequence was determined, the corresponding genomic region was extended by 5 kb on each side. Transcript sequences were then realigned on the extended region using EST_GENOME [15] (mismatch 2, gap penalty 3) to define transcript exons [37]. These transcript models were fused by a single linkage clustering approach whereby transcripts from the same genomic region sharing at least 100 bp are merged [38].

FcB1-schizont-EST sequences have been released to the EMBL/GenBank/DDBJ under the accession numbers [EMBL:CU657981] to [EMBL:CU672219].

The clustering of FcB1-schizont-ESTs on the P. falciparum genome (PlasmoDB version 4.4) can be viewed at: https://www.genoscope.cns.fr/secure-nda/projet_PU/cgi-bin/gbrowse/plasmodium/ webcite (authorisation required) [39]. The clustering of the FcB1-schizont-ESTs on the P. falciparum genome (PlasmoDB version 5.3) can be viewed at: https://www.genoscope.cns.fr/secure-nda/projet_PU/cgi-bin/ggb/gbrowse/plasmodium/ webcite. (authorisation required)

Abbreviations

AMA-1: apical membrane antigen; bp: base pair(s); BLAST: basic local alignment sequence tool; cDNA: complementary DNA(s); CDS: coding sequence(s); CLAG: cytadherence linked antigen; EBA: erythrocyte binding antigen; EBL: erythrocyte binding like; EST(s): expressed sequence tag(s); GLURP: glutamate rich protein; GO: Gene Ontology; MSP: merozoite surface antigen; ORF: open reading frame(s); PCR: polymerase chains reaction; PfEMP1/3: P. falciparum erythrocyte membrane protein 1/3; PHAT: pretty handy annotation tool; RAMA: Rhoptry associated membrane antigen; RAP: rhoptry-associated protein; RESA: ring-infected erythrocyte surface antigen; Rhop: Rhoptry; RON: rhoptry neck; rRNA: ribosomal RNA; SERA: serine repeat antigen; SERP: serine repeat protein; SMART-PCR: switch mechanism at the 5'-end of reverse transcript-polymerase chain reaction; SNP: single nucleotide polymorphism; SSH: suppressive subtracting hybridization; TARE: telomere-associated repeat element; TR-RNA: telomerase RNA.

Authors' contributions

IF project design, collaborator contacts, analyses of data alignment, interpretation of results, manuscript, figures and tables writing; BMP P. falciparum genomic and EST data recovery, treatment, est2genome clustering, data formatting and graphical interface set-up; EG alignment data analyses and interpretations, participation to text and figures writing; CDS P. falciparum genomic and EST data treatment, est2genome clustering, UniProt BLAST searches, data formatting for EMBL submissions and publication, contribution to figures; FA bioinformatics supervision, P. falciparum genomic and EST data recovery and treatment, est2genome clustering, data formatting for EMBL submissions and publication; EM alignment data analyses, interpretation of results, participation to text and figures writing; LB and OG GO-term recovery and statistical analysis, participation to text and figures writing; SC subtracted library construction; PW: sequencing supervision, P. falciparum genomic and EST data recovery and treatment, est2genome clustering, data formatting for EMBL submissions and publication; PG: project design, interpretation of results. All authors read and approved the manuscript.

Acknowledgements

We thank Béatrice Segurens and Sébastien Haffray for library plating at Genoscope, Odile Rogier for help with the data processing and analysis at Genoscope. This work was made possible by a collaboration with Genoscope and by financial support of both the Agence Nationale de la Recherche (PlasmoExplore project ANR-06-CIS6-MDCA-14-01) and the French Ministry of Research.

References

  1. Tuteja R: Malaria – an overview.

    Febs J 2007, 274(18):4670-4679. PubMed Abstract | Publisher Full Text OpenURL

  2. Sachs J, Malaney P: The economic and social burden of malaria.

    Nature 2002, 415:680-685. PubMed Abstract | Publisher Full Text OpenURL

  3. Greenwood BM, Fidock DA, Kyle DE, Kappe SH, Alonso PL, Collins FH, Duffy PE: Malaria: progress, perils, and prospects for eradication.

    J Clin Invest 2008, 118(4):1266-1276. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  4. Yeh I, Altman RB: Drug Targets for Plasmodium falciparum : a post-genomic review/survey.

    Mini Rev Med Chem 2006, 6(2):177-202. PubMed Abstract | Publisher Full Text OpenURL

  5. Gardner MJ, Hall N, Fung E, White O, Berriman M, Hyman RW, Carlton JM, Pain A, Nelson KE, Bowman S, et al.: Genome sequence of the human malaria parasite Plasmodium falciparum.

    Nature 2002, 419(6906):498-511. PubMed Abstract | Publisher Full Text OpenURL

  6. Bahl A, Brunk B, Crabtree J, Fraunholz MJ, Gajria B, Grant GR, Ginsburg H, Gupta D, Kissinger JC, Labo P, et al.: PlasmoDB: the Plasmodium genome resource. A database integrating experimental and computational data.

    Nucleic Acids Res 2003, 31(1):212-215. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  7. Florent I, Charneau S, Grellier P: Plasmodium falciparum genes differentially expressed during merozoite morphogenesis.

    Mol Biochem Parasitol 2004, 135(1):143-148. PubMed Abstract OpenURL

  8. Cowman AF, Crabb BS: Invasion of red blood cells by malaria parasites.

    Cell 2006, 124(4):755-766. PubMed Abstract | Publisher Full Text OpenURL

  9. Sanders PR, Kats LM, Drew DR, O'Donnell RA, O'Neill M, Maier AG, Coppel RL, Crabb BS: A set of glycosylphosphatidyl inositol-anchored membrane proteins of Plasmodium falciparum is refractory to genetic deletion.

    Infect Immun 2006, 74(7):4330-4338. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  10. Bannister LH, Hopkins JM, Fowler RE, Krishna S, Mitchell GH: A brief illustrated guide to the ultrastructure of Plasmodium falciparum asexual blood stages.

    Parasitol Today 2000, 16(10):427-433. PubMed Abstract | Publisher Full Text OpenURL

  11. Margos G, Bannister LH, Dluzewski AR, Hopkins J, Williams IT, Mitchell GH: Correlation of structural development and differential expression of invasion-related molecules in schizonts of Plasmodium falciparum.

    Parasitology 2004, 129(Pt 3):273-287. PubMed Abstract OpenURL

  12. Charneau S, Bastos IM, Mouray E, Ribeiro BM, Santana JM, Grellier P, Florent I: Characterization of PfDYN2, a dynamin-like protein of Plasmodium falciparum expressed in schizonts.

    Microbes Infect 2007, 9(7):797-805. PubMed Abstract | Publisher Full Text OpenURL

  13. Watanabe J, Suzuki Y, Sasaki M, Sugano S: Full-malaria 2004: an enlarged database for comparative studies of full-length cDNAs of malaria parasites, Plasmodium species.

    Nucleic Acids Res 2004, (32 Database):D334-338. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  14. Lu F, Jiang H, Ding J, Mu J, Valenzuela JG, Ribeiro JM, Su XZ: cDNA sequences reveal considerable gene prediction inaccuracy in the Plasmodium falciparum genome.

    BMC Genomics 2007, 8:255. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  15. Mott R: EST_GENOME: a program to align spliced DNA sequences to unspliced genomic DNA.

    Comput Appl Biosci 1997, 13(4):477-478. PubMed Abstract OpenURL

  16. Uniprot C: The universal protein resource (UniProt).

    Nucleic Acids Res 2008, (36 Database):D190-195. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  17. Bozdech Z, Llinas M, Pulliam BL, Wong ED, Zhu J, DeRisi JL: The Transcriptome of the Intraerythrocytic Developmental Cycle of Plasmodium falciparum.

    PLoS Biol 2003, 1(1):E5. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  18. Le Roch KG, Zhou Y, Blair PL, Grainger M, Moch JK, Haynes JD, De La Vega P, Holder AA, Batalov S, Carucci DJ, et al.: Discovery of gene function by expression profiling of the malaria parasite life cycle.

    Science 2003, 301(5639):1503-1508. PubMed Abstract | Publisher Full Text OpenURL

  19. Llinas M, Bozdech Z, Wong ED, Adai AT, DeRisi JL: Comparative whole genome transcriptome analysis of three Plasmodium falciparum strains.

    Nucleic Acids Res 2006, 34(4):1166-1173. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  20. Bradley PJ, Ward C, Cheng SJ, Alexander DL, Coller S, Coombs GH, Dunn JD, Ferguson DJ, Sanderson SJ, Wastling JM, et al.: Proteomic analysis of rhoptry organelles reveals many novel constituents for host-parasite interactions in Toxoplasma gondii.

    J Biol Chem 2005, 280(40):34245-34258. PubMed Abstract | Publisher Full Text OpenURL

  21. Alexander DL, Mital J, Ward GE, Bradley P, Boothroyd JC: Identification of the moving junction complex of Toxoplasma gondii : a collaboration between distinct secretory organelles.

    PLoS Pathog 2005, 1(2):e17. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  22. Beissbarth T, Speed TP: GOstat: find statistically overrepresented Gene Ontologies within a group of genes.

    Bioinformatics 2004, 20(9):1464-1465. PubMed Abstract | Publisher Full Text OpenURL

  23. Kyes SA, Kraemer SM, Smith JD: Antigenic variation in Plasmodium falciparum : gene organization and regulation of the var multigene family.

    Eukaryot Cell 2007, 6(9):1511-1520. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  24. Kyes S, Pinches R, Newbold C: A simple RNA analysis method shows var and rif multigene family expression patterns in Plasmodium falciparum.

    Mol Biochem Parasitol 2000, 105(2):311-315. PubMed Abstract | Publisher Full Text OpenURL

  25. Hisaeda H, Yasutomo K, Himeno K: Malaria: immune evasion by parasites.

    Int J Biochem Cell Biol 2005, 37(4):700-706. PubMed Abstract | Publisher Full Text OpenURL

  26. Rich SM, Ferreira MU, Ayala FJ: The origin of antigenic diversity in Plasmodium falciparum.

    Parasitol Today 2000, 16(9):390-396. PubMed Abstract | Publisher Full Text OpenURL

  27. Tonkin CJ, Pearce JA, McFadden GI, Cowman AF: Protein targeting to destinations of the secretory pathway in the malaria parasite Plasmodium falciparum.

    Curr Opin Microbiol 2006, 9(4):381-387. PubMed Abstract | Publisher Full Text OpenURL

  28. Waters AP: The ribosomal RNA genes of Plasmodium.

    Adv Parasitol 1994, 34:33-79. PubMed Abstract OpenURL

  29. Figueiredo LM, Pirrit LA, Scherf A: Genomic organisation and chromatin structure of Plasmodium falciparum chromosome ends.

    Mol Biochem Parasitol 2000, 106(1):169-174. PubMed Abstract | Publisher Full Text OpenURL

  30. Marty AJ, Thompson JK, Duffy MF, Voss TS, Cowman AF, Crabb BS: Evidence that Plasmodium falciparum chromosome end clusters are cross-linked by protein and are the sites of both virulence gene silencing and activation.

    Mol Microbiol 2006, 62(1):72-83. PubMed Abstract | Publisher Full Text OpenURL

  31. O'Donnell RA, Freitas-Junior LH, Preiser PR, Williamson DH, Duraisingh M, McElwain TF, Scherf A, Cowman AF, Crabb BS: A genetic screen for improved plasmid segregation reveals a role for Rep20 in the interaction of Plasmodium falciparum chromosomes.

    Embo J 2002, 21(5):1231-1239. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  32. Chakrabarti K, Pearson M, Grate L, Sterne-Weiler T, Deans J, Donohue JP, Ares M Jr: Structural RNAs of known and unknown function identified in malaria parasites by comparative genomics and RNA analysis.

    Rna 2007, 13(11):1923-1939. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  33. Mehrizi AA, Zakeri S, Salmanian AH, Sanati MH, Djadid ND: Plasmodium falciparum: sequence analysis of the gene encoding the C-terminus region of the merozoite surface protein-1, a potential malaria vaccine antigen, in Iranian clinical isolates.

    Exp Parasitol 2008, 118(3):378-385. PubMed Abstract | Publisher Full Text OpenURL

  34. Iyer J, Gruner AC, Renia L, Snounou G, Preiser PR: Invasion of host cells by malaria parasites: a tale of two protein families.

    Mol Microbiol 2007, 65(2):231-249. PubMed Abstract | Publisher Full Text OpenURL

  35. Diatchenko L, Lukyanov S, Lau YF, Siebert PD: Suppression subtractive hybridization: a versatile method for identifying differentially expressed genes.

    Methods Enzymol 1999, 303:349-380. PubMed Abstract OpenURL

  36. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ: Basic local alignment search tool.

    J Mol Biol 1990, 215(3):403-410. PubMed Abstract | Publisher Full Text OpenURL

  37. Castelli V, Aury JM, Jaillon O, Wincker P, Clepet C, Menard M, Cruaud C, Quetier F, Scarpelli C, Schachter V, et al.: Whole genome sequence comparisons and "full-length" cDNA sequences: a combined approach to evaluate and improve Arabidopsis genome annotation.

    Genome Res 2004, 14(3):406-413. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  38. Porcel BM, Delfour O, Castelli V, De Berardinis V, Friedlander L, Cruaud C, Ureta-Vidal A, Scarpelli C, Wincker P, Schachter V, et al.: Numerous novel annotations of the human genome sequence supported by a 5'-end-enriched cDNA collection.

    Genome Res 2004, 14(3):463-471. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  39. The Plasmodium Genome Resource [http://www.plasmodb.org/] webcite