Correction
7 Nov 2013: Sato S, Sesay AK, Holder AA (2013) Correction: The Unique Structure of the Apicoplast Genome of the Rodent Malaria Parasite Plasmodium chabaudi chabaudi. PLOS ONE 8(11): 10.1371/annotation/f9f809fc-34b8-42c8-acf3-f8b2616a5f44. https://doi.org/10.1371/annotation/f9f809fc-34b8-42c8-acf3-f8b2616a5f44 View correction
Figures
Abstract
The apicoplast, a non-photosynthetic plastid of apicomplexan species, has an extremely reduced but highly conserved genome. Here, the apicoplast genome of the rodent malaria parasite Plasmodium chabaudi chabaudi (Pcc) isolate CB was characterized. Although the set of genes in the genome is identical, the copy number of some tRNA genes differs between Pcc and other Plasmodium species because the Pcc DNA has only one rRNA/tRNA gene cluster, which is normally duplicated in other species. The location of the duplicated trnR(ACG) and trnM implies that one of the duplicated clusters in the ancestral molecule has been lost due to an intramolecular recombination event. The Pcc DNA occurs in two isoforms with an internal inversion between them. The presence of a unique variant in the duplicated trnT gene suggests that the two isoforms are interconvertible. This is the first report of the complete nucleotide sequence of a Plasmodium apicoplast DNA.
Citation: Sato S, Sesay AK, Holder AA (2013) The Unique Structure of the Apicoplast Genome of the Rodent Malaria Parasite Plasmodium chabaudi chabaudi. PLoS ONE 8(4): e61778. https://doi.org/10.1371/journal.pone.0061778
Editor: Georges Snounou, Université Pierre et Marie Curie, France
Received: February 1, 2013; Accepted: March 13, 2013; Published: April 16, 2013
Copyright: © 2013 Sato et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by British Medical Research Council (File References U117532067 and U117584270). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The apicoplast is a secondary plastid of parasites belonging to the phylum Apicomplexa [1]–[3]. Although it lacks photosynthetic activity, the organelle is critical for survival and growth [4]. Like other non-photosynthetic plastids [5], the apicoplast has a much smaller genome than those of photosynthetic eukaryotes. The apicoplast genome first analyzed was that of the human malaria parasite Plasmodium falciparum (Pf), which is encoded in a circular DNA of 35 kb in size [6]. The Pf apicoplast genome encodes large and small subunit ribosomal RNAs and 25 different tRNA species. The genes of two rRNAs (rrl and rrs) are arranged head-to-head and form a cluster with nine tRNA genes. The cluster is duplicated to form an inverted repeat (IR). Proteins encoded by this tiny genome are mostly involved in either transcription or translation and their genes comprise two unidirectional clusters in the single copy region following the trnT gene at one end of each IR unit.
The Pf plastid DNA sequence reported by Wilson and colleagues was incomplete as it lacked a short sequence between the two IR units on the circular DNA [6]. Recently, Arisue and colleagues analyzed the corresponding part of apicoplast genomes of eight Plasmodium species other than Pf [7]. The results revealed that the gene repertoire, gene arrangement, and other structural attributes are strongly conserved in those species, except for a peculiarity in the plastid DNA of the rodent malaria parasite P. chabaudi chabaudi (Pcc) isolate AS. Plasmodium species other than Pcc had a gene specifying tRNA-Arg(ACG) (trnR(ACG)) between the genes for tRNA-Met (trnM) and tRNA-Val (trnV) in each IR unit. By contrast, the sequence between trnM and trnV in Pcc AS DNA was significantly divergent. An unusual trnR(ACG) (trnR(ACG)*) was predicted in this region, but the product of this putative gene cannot be tRNA-Arg(ACG) because the RNA is unlikely to have ACG as the anticodon.
The missing tRNA-Arg(ACG) is essential for decoding CGN codons that frequently occur in the Pcc plastid genome (Table 1). Consequently, unless the apicoplast imports the tRNA from the cytosol, the organellar genome should contain trnR(ACG). To clarify this, we analyzed the whole genomic DNA of Pcc isolate CB by high throughput sequencing (HTS). In conjunction with data obtained with Sanger sequencing, the HTS data suggest that the plastid genome of Pcc is unique in the apicoplast DNAs of Plasmodium spp.
Results and Discussion
The whole genomic DNA of Pcc isolate CB was analyzed by high throughput sequencing (HTS) on the Illumina platform. Obtained reads covered the entire length of the previously partly sequenced plastid DNA of Pcc isolate AS [7] at an average depth of 50× without gaps (Table 2 and Figure S1A). The consensus sequence of HST reads had 42 differences from the 29,198 nt AS reference sequence (Table 3). Most of this variation within protein coding sequences was synonymous base substitution, but five SNPs and three deletions affect the amino acid sequences encoded by rpl23, rpl4, rps5, rpoC1, rpoC2A, rpoC2B and ORF91. Variations located within trnT or in intergenic regions surrounding trnM were not unique at each position, implying that these parts are duplicated in the Pcc plastid DNA.
We searched for reads encoding a regular tRNA-R(ACG) in our HTS data using the trnR(ACG) sequence of P. berghei (Pb), another rodent malaria parasite [7] as the reference (Figure S1B). The result clearly suggested that a previously unidentified trnR(ACG) exists in the Pcc CB DNA along with the peculiar trnR(ACG)* which Arisue et al. had found. Our data also suggested that both the regular and peculiar trnR(ACG) genes precede trnM, and share an identical intergenic sequence.
To further examine this finding, we next searched for HTS reads containing the trnM sequence (Figure S1C). The reads we identified were classified into two groups. The consensus of one group perfectly matched the reference sequence containing trnM (trnM-1). On the other hand, the consensus of the other group had several conflicts with the reference sequence outside the tRNA gene (trnM-2), despite the fact that this gene encodes an identical tRNA-Met. As suggested by our earlier search, the 3′ end of the regular trnR(ACG) appeared in front of trnM-2. These results confirmed that the Pcc plastid DNA has two distinct trnM genes specifying an identical tRNA-Met, and that the newly found trnM-2 adjoins the missing trnR(ACG).
The cluster of trnR(ACG) and trnM-2 was likely present in a region that escaped analysis by Arisue et al. Accordingly, to complete the partially determined sequence of the Pcc plastid DNA, we amplified the part between trnH and sufB by PCR, obtaining a product that appeared as a 6 kb band on electrophoresis in an agarose gel (Figure 1A). Although the rate should be low, PCR amplification may result in random introduction of base substitutions into the product. In order to mitigate any risk of introducing such errors into the product for sequence analysis, we avoided subcloning the PCR fragment and determined its nucleotide sequence directly. The sequencing data from each end of the DNA suggested that both trnH and sufB are preceded by the same sequence that encodes trnT and rrl (Figure 1B). The quality of the sequencing data from each side dropped abruptly beyond the 168th nucleotide from the predicted 3′ end of rrl. Further analysis revealed this was due to the co-occurrence of two different sequences beyond that point.
(A) Part of the Pcc CB plastid DNA was amplified by PCR from total parasite DNA with the primers 1095 and 1096 (see Table 4 for details of each primer). The product was fractionated on an agarose gel along with DNA size markers (M). DNA that appeared as a single band (*) was collected and analyzed further. (B) Alignment of trace data obtained for the 6 kb PCR product by PCR-direct sequencing with primers that anneal at an end (1095, 1096) or an internal position (1005, 853). Three parts of the alignment are presented in boxes with arrows representing the direction of the sequencing reaction starting from each primer; the name of an encoded gene is given with a horizontal arrow representing the direction of transcription. Non-coding sequence and ambiguous sequencing data due to the presence of two different sequences are highlighted with blue and red, respectively. Because rps4 and sufB share an identical sequence from the 1st to the 8th nucleotide of their coding sequence, the highlighted region in red starts at the 9th residue in the sequencing data from 1005 and 853. The C/T transition at position 28 of trnT, which is clearly identifiable in the sequencing data analyzed from inside (1005/853) but not in those from outside (1095/1096), is indicated with a vertical arrow. (C and D) Schematic representation of the PCR product. Selected genes in the region including the gene of the unusual tRNA-Arg(ACG) (trnR(ACG)*) and the 5' truncated rrl ('rrl) are indicated with color-coded thick arrows. The PCR product (horizontal thick black bar) amplified with primers 1095 and 1096 (red arrows) was a mixture of two DNA species. Because of the coexistence of two different types of molecule, the quality of sequencing data obtained (blue arrows) abruptly dropped at the end of the short IR sequence (highlighted with pink background) and gave a mixture of two sequences (dotted arrow). The sequence of two trnT genes (trnT-1 and -2) of Pcc CB is almost the same except for the variation at their 28th residue (T and C; circled). Each trnT is linked with its upstream gene but not with its downstream gene. Therefore the 28th residue of the gene was a mixture of C and T when the PCR product was sequenced from the outside toward the inside (C), but the residue was uniquely identified as either C or T when the same sample was sequenced from inside to outside (D). The length of each gene is not to scale with the others in this figure.
To confirm this finding, we sequenced the same PCR product from the inside outwards toward each end. As expected, trnT was identified directly after rrl in the sequence determined from rrl. On the other hand, the sequence from rrs contained trnI followed by the cluster of trnR(ACG) and trnM-2. After trnM-2, there was a 168 nucleotide 3′-end sequence of rrl followed by trnT. The sequencing data from both rrs and rrl were identical from the beginning of the rrl sequence except for one nucleotide at the 28th position of trnT (see below), and each sequence turned into a mixture of two sequences beyond the eighth residues of rps4/sufB. These results suggest that the PCR product was a mixture of two fragments that have the same internal sequence in the opposite direction between rps4 and sufB (Figure 1C and D). Hence Pcc plastid DNA does have an IR, but unlike other Plasmodium spp., each unit of the IR is extremely short – only 288 bp between the 168th positions from the 3′ end of rrl ('rrl) and the eighth positions of rps4/sufB.
In summary, using PCR to amplify the region between rps4 and sufB we obtained a mixture of two DNA fragments in each of which rrs, rrl and several tRNA genes appear in the opposite direction. This indicates there are two molecular forms of Pcc CB plastid DNA. Each form has the same RNA/tRNA gene cluster only once, in contrast to all other Plasmodium species where it is duplicated. The direction of the RNA/tRNA gene cluster is unique to each form, and we tentatively name the DNA molecule in which rrs is encoded on the same strand as rps4 as form A, whereas the other form which has rrs on the other strand to rps4 as form B (Figure 2).
Schematic diagram of the two forms of the plastid DNA of Pcc CB. Each gene is represented with a box color-coded in white (one specifying a protein), blue (rRNA gene), yellow (tRNA gene) or red (a pseudogene). The colour of the name of each gene represents the transcription direction (black, clockwise; red, counter-clockwise). The bulge connecting two parts of the trnL(UAA) coding sequence indicates the intron. The nucleotide sequences of the complete Pcc CB plastid DNA in forms A (A) and B (B) were deposited to the DDBJ/EMBL/GenBank databases with the accession numbers HF563595 and HF563596, respectively.
Curiously, we found a one-nucleotide difference between the two units of the short IR of Pcc CB. The difference, a C/T transition at the 28th position of trnT, probably emerged after the separation of the Pcc isolates CB and AS, because the corresponding variation was not found in trnT of Pcc AS [7]. Direct sequencing data for the PCR product unambiguously showed that this variation is tightly linked with the upstream gene (rrl or 'rrl). By contrast, the nature of the residue at the site seemed independent of the downstream sequence (Figure 1). These observations imply that the variation in trnT emerged in either of the two forms only once and spread to the other form due to an intramolecular recombination event that occurred between the shorter sequences bound by the variable site and the eighth residue of rps4/sufB.
Our detailed analysis has confirmed that the Pcc plastid genome encodes the same set of tRNA species as other Plasmodium species. But, as the IR in the DNA molecule is exceptionally short, the copy number of each tRNA gene is distinct from that of other Plasmodium. The library for HTS analysis was prepared with a protocol that omits PCR amplification [8], therefore the number of reads matching each point of the reference is expected to reflect the relative abundance of the template in the starting material. The number of reads with the characteristic variation suggested that the ratio of the trnR(ACG)*-trnM-1 and the trnR(ACG)-trnM-2 gene clusters was 1∶1 (Figure S1D). This indicates that each tRNA gene is present at the same copy number in the Pcc apicoplast genome. In addition, counting the reads having each variant within those matching the trnT sequence (Figure S1E), we estimate the ratio between the two trnT genes as 1∶1.
Consequently, the only duplicate tRNA genes that encode (practically) the same tRNA species in the Pcc apicoplast genome are trnM and trnT; generally, seven other tRNA genes (trnA, trnI, trnL(UAG), trnN, trnR(ACG), trnR(UCU), and trnV) are also duplicated in the apicoplast genome of other species of Plasmodium [6], [7]. The presence of trnI directly upstream the trnR(ACG)-trnM-2 cluster suggests that the uniqueness of the tRNA gene copy numbers in the Pcc genome was caused by an intramolecular recombination event, probably between trnV, which precedes trnR(ACG) in the IR unit of Plasmodium plastid DNA, and trnI (Figure 3). The resultant DNA molecule would have had only one copy of rrs in front of the remaining single-copy trnI. Subsequent truncation of the 5' region of rrl in the affected IR unit and degradation of the trnR(ACG) to trnR(ACG)* in the other IR unit would have given rise to the present forms of the Pcc plastid DNA.
An intramolecular recombination event probably between trnI and trnV caused partial deletion of one of the two IR units in the ancestral DNA molecule (thick black arrow). The chimeric trnI/trnV in the intermediate molecule (1) was lost probably because it was a pseudogene. Subsequent truncation of the 5' region of rrl in the affected IR unit (thick blue arrow) and degradation of the trnR(ACG) in the other IR unit (thick red arrow) results in the Pcc plastid DNA. The order of these two events as well as which intermediate molecule (2) was generated in this process, are unknown. Finally, differentiation between trnM-1 and -2 as well as between trnT-1 and -2 occurred and these resulted in the Pcc CB plastid DNA. Switching between forms A and B could have happened at any point of this process, and may happen frequently as suggested by the fact that the apparent molar ratio between the two forms is 1∶1. Genes are color-coded as in panels A and B, and those changing/changed are indicated with a green circle.
The ratio between the two forms of Pcc plastid DNA is difficult to determine directly, but the coverage of reference sequences by the HTS reads provided an estimate. The ratio between the average coverage of a nuclear chromosome and the plastid DNA was about 1∶2 (Table 2). This suggests that the average copy number of the plastid DNA in an apicoplast is 2, given that prior to schizogony the parasite cell has only one nucleus and one apicoplast, and each nucleus contains only one haploid set of chromosomes. If this estimate is correct, one molecule each of the two forms of DNA is likely co-present with the other in the organelle. How the two forms can co-exist keeping their ratio at 1∶1 is unknown. One attractive possibility is that both forms are concatenated into a heterodimeric molecule at some point of the cell cycle (Figure S2), although there is as yet no experimental evidence supporting such a form’s physical existence in the Pcc plastid. It is also possible that there are two types of parasites in the population each with exclusively one or other form of plastid DNA molecule. Whichever is the case, it is noteworthy that the Pf apicoplast at the ring-stage seems to contain only 1 or few copies of the plastid DNA [9].
Conclusions
We conclude that, unlike other Plasmodium spp., the Pcc plastid DNA has only one copy of the rRNA/tRNA gene cluster. The DNA is present in two different forms A and B that share identical sequence except for the opposite direction of the rRNA/tRNA gene cluster between rps4 and sufB. The estimated ratio between forms A and B is 1∶1 and the two forms seem to be interchangeable via an intramolecular recombination in the small region between the 28th nucleotide of trnT and the eighth nucleotide of rps4/sufB. The coincidental presence of two trnTs with a C/T transition at the 28th position in the Pcc CB plastid DNA molecule suggests either that intramolecular recombination occurs frequently or that recombination events are very rare and the variants have been fixed since one event generated the two forms. Whichever is the case, the plastid DNA of Pcc is unique amongst the apicoplast DNAs of Plasmodium spp.
Materials and Methods
Parasite DNA
The purified genomic DNA of P. chabaudi chabaudi CB was a gift from Jean Langhorne group in the Division of Parasitology at MRC-NIMR, UK. Briefly, the DNA was obtained from parasite infected mice using a protocol reviewed and approved by the Ethical Review Panel of the MRC-NIMR and approved and licensed by the UK Home Office as governed by law under the Animals (Scientific Procedures) Act 1986 (Project license 80/1904). The animals were handled in strict accordance with the “Code of Practice Part 1 for the housing and care of animals (21/03/05)” available at http://www.homeoffice.gov.uk/science-research/animal-research/.
High Throughput Sequencing (HTS)
Eight micrograms of total genomic DNA of Plasmodium chabaudi chabaudi isolate CB (Pcc CB) obtained from the blood of infected mice was fragmented in a microTUBE with an Adaptive Focused Acoustics fiber (Covaris, Woburn, Massachusetts) using an S2 focused ultra-sonicator (Covaris) set at 5% duty cycle; intensity 4; 200 cycles/burst for 90 s. In order to minimize the undesirable bias that affects the sequence analysis of extremely A/T-rich DNA such as that of Plasmodium spp. [8], we processed the fragmented DNA using a NEXTflex PCR-Free DNA Sequencing Kit (BIOO Scientific, Austin, Texas) and NEXTflex PCR-Free barcode 1 (BIOO Scientific) following the protocol provided with the kit. The processed DNA was applied to an agarose gel on E-Gel iBase (Life Technologies, Carlsbad, California) and fragments that ran as 400–500 bp were collected to form the sequence library for further analysis. After QC test and quantification, the library of DNA was analyzed on a HiSeq 2000 sequencer (Illumina, San Diego, California) and paired-end sequencing data of 100 nt/read were collected following the standard protocol. The raw data were deposited in the European Nucleotide Archive (ENA) under the accession number ERP002313.
HTS reads were aligned on reference sequences using CLC Genomic Workbench software package (CLC Bio, Aarhus, Denmark) either to obtain the consensus sequence and/or count the number of matching reads.
PCR Direct Sequencing
To minimize errors accumulating during steps of molecular cloning, part of the nucleotide sequence of Pcc CB plastid DNA that we were unable to obtain simply by analyzing the HTS data was determined by PCR direct sequencing. Pcc CB plastid DNA containing the rRNA gene cluster was amplified from total genomic DNA by PCR with oligonucleotides 1095 (annealing to trnH; see Table 4) and 1096 (sufB) as below: 20 ng of Pcc CB total genomic DNA was added to 80 µL of reaction mixture containing x1 concentration of Reaction Buffer IV (Thermo Fisher Scientific, Waltham, Massachusetts), MgCl2 (2 mM), dNTP (0.4 mM each) and primers (250 nM each). The mixture was divided into two and each 40-microliter aliquot was dispensed into a 0.2 ml PCR tube. After adding 2 U of KAPA Taq DNA polymerase (KAPA Biosystems, Cape Town, South Africa), the PCR reaction was carried out in a thermal cycler with initial denaturation at 94°C for 2 min followed by 40 cycles of (denature at 94°C for 10 s, annealing at 53°C for 30 s, extension at 63°C for 270 s). To monitor the reaction, 1 µL was analyzed by electrophoresis on an agarose gel. To the remaining sample in each tube, 8 U of FastAP thermosensitive alkaline phosphatase (Thermo Fisher Scientific) and 80 U of NxGen Exonuclease I (Lucigen, Middleton, Wisconsin) were added, and each mixture was incubated at 37°C for 20 min, before both enzymes were inactivated at 80°C for 15 min. The nucleotide sequence of the PCR products was determined by Beckman Coulter Genomics (Essex, UK) using Sanger sequencing methodology and the oligonucleotides listed in the Table 4 as primers.
Codon Usage Analysis
Codons present in all annotated protein CDS (total of 31) in the apicoplast genome were counted for the two Pcc isolates, AS (AB649423.1) [7] and CB (HF563595/HF563596) (this work), and compared with those of P. falciparum C10 (X95275.2/X95276.2) [6] for which decoding tRNA species have been proposed [10]. The rpoC2 ( = rpoD) gene, in which a frame shift has been proposed [6], was regarded as two separate protein CDS (rpoC2A and rpoC2B).
Estimation of Relative Copy Numbers
Relative copy number between each nuclear chromosome and the apicoplast was estimated from the average coverage of the entire length of each DNA by HTS reads in Table 2. Relative copy number for the trnR(ACG)*-trnM-1 gene cluster and the trnR(ACG) -trnM-2 gene cluster (Fig. S1D) and between trnT-1 and trnT-2 (Fig. S1E) was estimated from the number of HTS reads containing characteristic variants. Because the reference sequences were short, each read was mapped on it as a single-end read. As the distance between variant sites 1 and 2/3 (Fig. S1D) was longer than 100 bp, there were no reads containing all three variant sites. However, analysis of paired end reads (Fig. S1C) suggested that those variants were exclusively linked together. Thus the count of each cluster was obtained as the sum the numbers of reads with variation at either site 1 or sites 2/3.
Supporting Information
Figure S1.
High throughput sequence analysis of Plasmodium chabaudi chabaudi CB plastid DNA. HTS reads of a library prepared from Pcc CB total genomic DNA were aligned on the plastid DNA sequence of Pcc isolate AS [7](A), P. berghei plastid DNA sequence containing trnR(ACG) [7](B) or Pcc AS plastid DNA sequence around trnM (C), the consensus of the two trnR(ACG)-trnM gene clusters in the Pcc CB plastid DNA (D) or Pcc AS trnT (E), using CLC Genomics Workbench (CLC BIO, Aarhus, Denmark). Reads that matched the reference sequence as a pair are shown in blue, whereas single matching reads are either in green (matching forward) or red (matching reverse). At the end of some reads where the sequence does not match the reference is shown in paler color, and internal residues that are different from those in the reference sequence are highlighted. Gaps are indicated by a space (A) or “–” (B–D). The position of each variation representing the type of gene cluster/gene is indicated on top of (D) and (E). Only reads with a significant high quality at the site of variation (highlighted) were counted, and the results are given in a table (inset in D and E).
https://doi.org/10.1371/journal.pone.0061778.s001
(PDF)
Figure S2.
Hypothetical heterodimeric form. Genes are color-coded as in Figure 2. Homologous recombination at any part of Pcc CB plastid DNA between one molecule each of forms A and B (Figure 2) will generate a unique heterodimeric form (Form C). Although such a form's presence has not been confirmed, it is possible that the Pcc plastid DNA occurs in this form at some stage of parasite development (see text). Note that this form has two units of a gene cluster (boxed) containing the same set of genes (except that trnR(ACG) has been degraded to trnR(ACG)*) in the same order as the IR unit of regular Plasmodium plastid DNA.
https://doi.org/10.1371/journal.pone.0061778.s002
(PDF)
Acknowledgments
This paper is dedicated to the memory of Dr. Don Williamson. We thank Jean Langhorne group (MRC-NIMR, UK) for providing the purified genomic DNA of P. chabaudi chabaudi CB, and Dr. Iain Wilson for discussion.
Author Contributions
Conceived and designed the experiments: SS. Performed the experiments: SS AKS. Analyzed the data: SS AKS. Contributed reagents/materials/analysis tools: SS AKS AAH. Wrote the paper: SS AAH.
References
- 1. Köhler S, Delwiche CF, Denny PW, Tilney LG, Webster P, et al. (1997) A plastid of probable green algal origin in Apicomplexan parasites. Science 275: 1485–1489.
- 2. McFadden GI, Reith ME, Munholland J, Lang-Unnasch N (1996) Plastid in human parasites. Nature 381: 482.
- 3. Sato S (2011) The apicomplexan plastid and its evolution. Cell Mol Life Sci 68: 1285–1296.
- 4. Yeh E, DeRisi JL (2011) Chemical rescue of malaria parasites lacking an apicoplast defines organelle function in blood-stage Plasmodium falciparum. PLoS Biol 9: e1001138.
- 5. Green BR (2011) Chloroplast genomes of photosynthetic eukaryotes. Plant J 66: 34–44.
- 6. Wilson RJM, Denny PW, Preiser PR, Rangachari K, Roberts K, et al. (1996) Complete gene map of the plastid-like DNA of the malaria parasite Plasmodium falciparum. J Mol Biol 261: 155–172.
- 7. Arisue N, Hashimoto T, Mitsui H, Palacpac NM, Kaneko A, et al. (2012) The Plasmodium apicoplast genome: conserved structure and close relationship of P. ovale to rodent malaria parasites. Mol Biol Evol 29: 2095–2099.
- 8. Kozarewa I, Ning Z, Quail MA, Sanders MJ, Berriman M, et al. (2009) Amplification-free Illumina sequencing-library preparation facilitates improved mapping and assembly of (G+C)-biased genomes. Nat Methods 6: 291–295.
- 9. Williamson DH, Preiser PR, Moore PW, McCready S, Strath M, et al. (2002) The plastid DNA of the malaria parasite Plasmodium falciparum is replicated by two mechanisms. Mol Microbiol 45: 533–542.
- 10. Preiser P, Williamson DH, Wilson RJM (1995) tRNA genes transcribed from the plastid-like DNA of Plasmodium falciparum. Nucleic Acids Res 23: 4329–4336.