INTRODUCTION
Malaria caused by Plasmodium vivax is responsible for 100–300 million clinical cases per year worldwide (Price et al. Reference Price, Tjitr, Guerra, Yeung, White and Anstey2007; Greenwood et al. Reference Greenwood, Fidock, Kyle, Kappe, Alonso, Collins and Duffy2008). Although often considered ‘benign’ (at least in comparison to the disease caused by Plasmodium falciparum), severe P. vivax cases have been periodically reported (Price et al. Reference Price, Tjitr, Guerra, Yeung, White and Anstey2007). Thus efficient control tools are required against not only P. falciparum but also for P. vivax.
Erythrocyte invasion is an essential step in the life cycle of malaria parasites. The stages of the Plasmodium parasite that invade cells contain distinctive organelles at their apical end, termed micronemes, rhoptries, and dense granules, which release proteins during invasion. Recently, in another apicomplexan parasite Toxoplasma gondii, a microneme protein Apical Membrane Antigen 1 (TgAMA1) was found to form a complex with 3 rhoptry neck (RON) proteins: TgRON2, TgRON4 and TgRON5 during cell invasion (Alexander et al. Reference Alexander, Mital, Ward, Bradley and Boothroyd2005; Lebrun et al. Reference Lebrun, Michelin, El Hajj, Poncet, Bradley, Vial and Dubremetz2005; Boothroyd and Dubremetz, Reference Boothroyd and Dubremetz2008). Orthologues of these genes have been identified in the genomes of malaria parasite species and some have been characterized for their expression and complex formation. The P. falciparum orthologue of RON4, PfRON4, has been shown to associate with PfAMA1 (Alexander et al. Reference Alexander, Arastu-Kapur, Dubremetz and Boothroyd2006) and to be localized at the moving junction (Baum et al. Reference Baum, Tonkin, Paul, Rug, Smith, Gould, Richard, Pollard and Cowman2008). Complex formation between PfRON2 and PfAMA1 was also shown in P. falciparum (Cao et al. Reference Cao, Kaneko, Thongkukiatkul, Tachibana, Otsuki, Gao, Tsuboi and Torii2009). Protein expression at the apical end of the merozoite was confirmed for PfRON5 (Curtidor et al. Reference Curtidor, Patiño, Arévalo-Pinzón, Patarroyo and Patarroyo2011). When complex formation between RON2 and AMA1 was inhibited in P. falciparum, the parasite could not invade erythrocytes, suggesting that the complex formed by these proteins is essential for a successful invasion (Richard et al. Reference Richard, MacRaild, Riglar, Chan, Foley, Baum, Ralph, Norton and Cowman2010). Although the amino acid sequences of the RON2 and AMA1 proteins have diverged between species, the interaction between them has been maintained in both P. falciparum and T. gondii, which suggests functional conservation (Lamarque et al. Reference Lamarque, Besteiro, Papoin, Roques, Vulliez-Le Normand, Morlon-Guyot, Dubremetz, Fauquenoy, Tomavo, Faber, Kocken, Thomas, Boulanger, Bentley and Lebrun2011).
Molecules located on the merozoite surface are potential targets of host immune responses. Diversifying selection has been suggested to have been involved in the evolution of the microneme protein AMA1, suggesting that the genetic diversity of this gene has been maintained at the population level in order to evade host immunity (Polley and Conway, Reference Polley and Conway2001). Due to the fundamental role of RON2-AMA1 complex formation in host cell invasion, RON2 is, potentially, a target of invasion inhibitory antibodies and, therefore, for the development of anti-vaccines and/or drugs. Previously, an excess of non-synonymous substitutions over synonymous substitutions was detected in the gene encoding PfRON2 following the analysis of 6 culture-adapted parasite lines, suggesting that PfRON2 is subject to positive diversifying selection (Cao et al. Reference Cao, Kaneko, Thongkukiatkul, Tachibana, Otsuki, Gao, Tsuboi and Torii2009). Similar to PfRON2, PvRON2 was also reported to be expressed in late schizonts and is localized at the rhoptry neck in the merozoite (Arévalo-Pinzón et al. Reference Arévalo-Pinzón, Curtidor, Patiño and Patarroyo2011), but the degree of polymorphism of this protein within parasite populations has not been previously determined. In this study, we evaluated the diversity of the entire coding region of PvRON2 using parasites isolated from China in order to measure the degree of the polymorphism and to identify signatures of a positive diversifying selection on this protein.
MATERIALS AND METHODS
Ethics statement
The study was approved by the Institutional Review Board (IRB00004221) of Jiangsu Institute of Parasitic Diseases, Wuxi, China. Questionnaire surveys, physical examination and laboratory work were conducted after the purpose of the study had been explained to participants, who were given the right to withdraw from the study at any time without consequences. Written informed consent was obtained from each participant.
Bioinformatics methods
A BLASTP search was performed against P. vivax (Sal-I line) and Plasmodium knowlesi (H strain) database in PlasmoDB using the PfRON2 amino acid sequence as a query, and RON2 orthologues, PvRON2 (PVX_117880) and PkRON2 (PKH_125430), were identified (Carlton et al. Reference Carlton, Adams, Silva, Bidwell, Lorenzi, Caler, Crabtree, Angiuoli, Merino, Amedeo, Cheng, Coulson, Crabb, Del Portillo, Essien, Feldblyum, Fernandez-Becerra, Gilson, Gueye, Guo, Kang'a, Kooij, Korsinczky, Meyer, Nene, Paulsen, White, Ralph, Ren, Sargeant, Salzberg, Stoeckert, Sullivan, Yamamoto, Hoffman, Wortman, Gardner, Galinski, Barnwell and Fraser-Liggett2008; Pain et al. Reference Pain, Böhme, Berry, Mungall, Finn, Jackson, Mourier, Mistry, Pasini, Aslett, Balasubrammaniam, Borgwardt, Brooks, Carret, Carver, Cherevach, Chillingworth, Clark, Galinski, Hall, Harper, Harris, Hauser, Ivens, Janssen, Keane, Larke, Lapp, Marti, Moule, Meyer, Ormond, Peters, Sanders, Sanders, Sargeant, Simmonds, Smith, Squares, Thurston, Tivey, Walker, White, Zuiderwijk, Churcher, Quail, Cowman, Turner, Rajandream, Kocken, Thomas, Newbold, Barrell and Berriman2008; Aurrecoechea et al. Reference Aurrecoechea, Brestelli, Brunk, Dommer, Fischer, Gajria, Gao, Gingle, Grant, Harb, Heiges, Innamorato, Iodice, Kissinger, Kraemer, Li, Miller, Nayak, Pennington, Pinney, Roos, Ross, Stoeckert, Treatman and Wang2009). The transmembrane (TM) domains were predicted by TopPred 0.01 (von Heijne, Reference von Heijne1992), SOSUI v1.11 (Hirokawa et al. Reference Hirokawa, Boon-Chieng and Mitaku1998), HMMTOP v2.0 (Tusnády and Simon, Reference Tusnády and Simon2001), TMpred (Hofmann and Stoffel, Reference Hofmann and Stoffel1993), and TMHMM v2.0 (Krogh et al. Reference Krogh, Larsson, von Heijne and Sonnhammer2001) web-based algorithms. Repeat sequences and domains were predicted by XSTREAM 1.73 (Newman and Cooper, Reference Newman and Cooper2007).
Source of Plasmodium vivax
P. vivax-infected blood samples were collected from 13 symptomatic patients who had acquired the infections in Henan province in 2009 (HN02, HN19, HN29, HN40, HN44, HN49, HN80, HN89, HN90, HN114, HN133, HN159, and HN160; only the 3′ region was anlaysed for HN89), 6 patients from Anhui province in 2007 and 2008 (BB0703, BB0704, BB0706, GS0702, HY0701, and P4; only the 5′ region was analysed for P4), and 2 patients from Jiangsu in 2009 (JS10 and JS17; the 5′ region was analysed for JS10 and the 3′ region was anlaysed for JS17). Infections were diagnosed by examination of Giemsa's solution-stained thick blood films. Finger-prick blood samples obtained from patients were preserved by spotting onto Whatman 3MM filter papers, naturally dried and kept at −20 °C until DNA extraction, while venous blood samples obtained from patients were preserved with EDTA anticoagulant and kept in liquid nitrogen prior to DNA extraction.
DNA isolation, polymerase chain reaction (PCR) amplification and sequencing
DNA was extracted from blood spotted on filter papers and venous blood samples using a QIAmp DNA micro-kit (QIAGEN, Germany) following the manufacturer's protocol. Extracted DNA samples were stored at −30°C until use. The oligonucleotide primers were designed based on the nucleotide (nt) sequence of P vRON2 (Sal-I line, PVX_117880) in the PlasmoDB. Two independent DNA fragments were PCR-amplified; the first one contains 5′ region encoding N-terminal amino acid positions (aa) 1–924 with the primers PvRON2-F0 (CTCTCTGGATTCACATCTCTCCGTTC) and PvRON2-R3 (CTTTACGTTCAGTCCTTCTTCTGAC) and the second one contains 3′ region encoding C-terminal aa 841–2203 with the primers PvRON2-F2 (GGATACACGCCTATCATCGTAAAGTATGAC) and PvRON2-R2 (GCACACCTGGACGAAGCTTTCTTCC). After excluding primer sequences, 2 DNA fragments share 224 nt sites. Oligonucleotides used in this study are summarized in Table 1.
Oligonucleotides used in this study
PCR amplification was performed with KOD -Plus-DNA polymerase (Toyobo, Japan) in a 20 μl final reaction. Amplification conditions for the first fragment was as follows: 1 cycle of 2 min at 94°C, followed by 40 cycles of 15 sec at 92°C, 15 sec at 55 °C and 3 min at 68°C, with a final 10 min extension step at 68°C. The second fragment was amplified as above except 45 cycles for the second step and 4 min 30 sec for the extension time were used. PCR products were visualized on a 1% agarose gel under UV transillumination. The specific band was cut out from the gel and purified using a Wizard SV Gel and PCR Clean-Up system (Promega, Madison, WI, USA). After the addition of an A cohesive end using Taq DNA polymerase (Fermentas), PCR products were cloned into pGEM-T easy vector (Promega). Plasmids prepared using a Wizard Plus SV Minipreps DNA purification system (Promega) were sequenced using an ABI PRISM®310 genetic analyzer (Applied Biosystems, Foster City, CA). Six clones were selected for each DNA fragment and final sequences were determined from at least 3 identical clones to avoid potential error during PCR amplification and sequencing. Sequence data have been deposited to the GenBank™/EMBL/DDBJ databases.
Sequence alignment and analysis
The PvRON2 open reading frame nt sequences were aligned with that retrieved from the genome database (Sal-I) using the CLUSTAL W program (Thompson et al. Reference Thompson, Higgins and Gibson1994) with manual correction; Nucleotide diversity (π) and its standard error (s.e.) were computed using the Jukes and Cantor method in MEGA 4.0 after removing insertion/deletions (indels). The mean number of synonymous substitutions per synonymous site (d S) and non-synonymous substitution per non-synonymous site (d N), and their standard errors were computed using the Nei and Gojobori method (Nei and Gojobori, Reference Nei and Gojobori1986) with Jukes and Cantor correction, implemented in MEGA 4.0. The statistical difference between d S and d N was tested using a Fishers' exact test. Positive selection was also evaluated using the McDonald-Kreitman test with PkRON2 nt sequence (McDonald and Kreitman, Reference McDonald and Kreitman1991).
RESULTS
Signal peptide, transmembrane and repeat regions of PvRON2 protein sequence
The full-length PvRON2 protein consists of 2203 residues with a putative signal peptide sequence at its N-terminus at aa 1–17. Among 5 web-based tests, 4 algorithms predicted that PvRON2 was a transmembrane protein bearing (potentially) 1–3 TM domains (Fig. 1 and Table 2). The putative TM domain towards the C-terminus (TM3; aa 2087–2109 by TMHMM) was predicted by 4 algorithms; however, other TM domains, TM1 (aa 1123–1143 by TopPred) and TM2 (aa 2015–2035 by TopPred), were only predicted by TMPred and TopPred. Even though we were unable to unambiguously determine the exact number of TM domains, the presence of at least 1 TM suggests that PvRON2 is an intergral membrane protein.
Prediction of PvRON2 transmembrane domains using different algorithm
Amino acid polymorphic sites of PvRON2. SP and TM indicate putative signal peptide (amino acid positions 1–17) and predicted transmembrane regions, respectively. The light grey bar indicates indel. Asterisks indicate polymorphic sites.
Two tandem repeat (TR) sequences were detected in the PvRON2 amino acid sequence by XSTREAM 1.73. An 8 times repeated 11 amino acid sequence ‘G(A/S)(E/D)(G/R)KGYG(P/T)(Y/D)G’ was located at aa 258–345, and a 5 amino acid second TR ‘(Q/E/K/N/H)(G/D)G(H/L/Y/P)G’ was located at aa 541–650 (after Sal-I sequence). Based on the second TR number, PvRON2 could be classified into 3 types; type 1 consisting of Sal-I and VCG-1 had 22 repeats, type 2 consisting of 8 China isolates had 21 repeats, and type 3 consisting of 11 China isolates had 23 repeats (Fig. 2).
Three patterns of repeat polymorphism of PvRON2 second tandem repeat region.
Polymorphism in the PvRON2 sequence
Ten non-synonymous nt substitutions were found among 20 PvRON2 sequences at nt positions 1241, 1675, 2146, 2456, 2477, 3182, 4390, 4840, 4988, and 6555 (Table 3, Fig. 2). One synonymous substitution was found at nt 3444. An indel polymorphism due to a difference in repeat number was also found at aa 541–650 (after Sal-I sequence) as described above (Fig. 2). Arévalo-Pinzón et al. (Reference Arévalo-Pinzón, Curtidor, Patiño and Patarroyo2011) sequenced PvRON2 cDNA sequence from Colombian VCG-1 isolate and found that there were 2 non-synonymous nucleotide substitutions at nt 1241 (T to G) and 1811 (C to A) (numbers are after Sal-I sequence) and 1 insertion of a nucleotide triplet ‘GAA’ between nt 1485 and 1486 of Sal-I sequence in comparison with the Sal-I reference line. A substitution at nt 1241 was found in all 19 China isolates in this study; however, a substitution at nt 1811 and an indel both found in the VCG-1 isolate were not found in these 19 Chinese isolates.
Nucleotide and amino acid polymorphism of PvRON2
a Nucleotide and amino acid numbering are after Sal-I line. Syn and Nsyn indicate synonymous substitution and non-synonymous substitution, respectively.
Positive diversifying selection on the PvRON2 gene
Nucleotide sequences of PvRON2 from 19 China isolates and the Sal-I isolate were aligned and evaluated for the signature of positive selection. Nucleotide diversity (π) and its standard error of 5′ region (nt 3–2667) and 3′ region (nt 2668–6609) were 0·00023±0·00010 and 0·00024±0·00011, respectively (Table 4). A significant excess of non-synonymous substitutions over synonymous substitutions was not detected for both regions (Hughes and Nei, Reference Hughes and Nei1988). To further evaluate the signature of positive selection on PvRON2, we performed a McDonald-Kreitman test using the P. knowlesi RON2 homologue. For the McDonald-Kreitman test, nucleotide sequences around indels and repeat regions were excluded from the analysis. The PvRON2 sequences used for the analysis were nt 4–663, 676–771, 1027–1065, 1072–1206, 1360–1482, 1489–1623, 1954–2472, and 2482–6609 (Sal-I line). The PkRON2 sequences used for the analysis were nt 4–663, 679–774, 997–1035, 1045–1179, 1303–1425, 1435–1569, 1768–2286, and 2302–6426. For the 5′ region, PvRON2 nt 4–2747 and PkRON2 nt 4–2567 were used. For the 3′ region, PvRON2 nt 2558–6609 and PkRON2 nt 2488–6426 were used. A significant excess of non-synonymous substitutions over synonymous substitutions was detected among P. vivax sequences by comparing the fixed differences between P. vivax and P. knowlesi for the 3′ region (P < 0·02), suggesting that positive diversifying selection has acted on this region (Table 5). This analysis did not detect positive selection on the 5′ region.
Nucleotide diversity of PvRON2 gene
n, number of sequences analysed.
nt, nucleotide position.
Sites, nucleotide sites analysed excluding sequence around indels.
π, pairwise nucleotide diversity.
N and S, average numbers of non-synonymous and synonymous sites.
d N, number of non-synonymous substitutions over number of non-synonymous substitutions sites.
d S, number of synonymous substitutions over number of synonymous substitutions sites.
s.e., standard error computed using the Nei-Gojobori method with Jukes-Cantor correction using the bootstrap method with 500 replications. The numbers of synonymous (Sd) and non-synonymous (Nd) differences were calculated by Nei-Gojobori method. Nucleotide positions 1864–1938 were excluded from the analysis for 5′ region to obtain an reliable alignment. Nucleotide (nt) number is after Sa1-I reference line. P value indicates the d N is significantly greater than d S.
n.s. indicates not significant (P>0·05).
The McDonald-Kreitman test of selection for PvRON2 gene
n, number of Plasmodium vivax sequences analysed.
nt, nucleotide position.
Sites, nucleotide sites analysed. Nucleotide positions 4–2667 and 2668–6609 of PvRON2 were used for the 5′ and 3′ regions, respectively after excluding sequence around indels and repeat region to obtain a reliable alignment. Nucleotide sequences used for the analysis were nucleotide positions 4–663, 676–771, 1027–1065, 1072–1206, 1360–1482, 1489–1623, 1954–2472, and 2482–6609 for PvRON2 (after Sal-I reference line) and nucleotide positions 4–663, 679–774, 997–1035, 1045–1179, 1303–1425, 1435–1569, 1768–2286, and 2302–6426 for P. knowlesi RON2.
Syn and Nsyn indicate synonymous substitution and non-synonymous substitution, respectively. Fisher's exact test (one-tailed) was used.
P value indicates that Nsyn is significantly greater than Syn.
n.s. indicates not significant (P > 0·05).
DISCUSSION
The present study describes the polymorphism in the nucleotide sequences of PvRON2, a homologue of PfRON2, in samples collected from P. vivax malaria patients in China. A McDonald-Kreitman test detected positive selection on the 3′ region. Thus, we propose that PvRON2 polymorphism has evolved under positive diversifying selection, possibly mediated by host immune pressure. A significant excess of d N over d S was also detected for PfRON2, strengthening the case for RON2 as a potential target of the host immunity (Cao et al. Reference Cao, Kaneko, Thongkukiatkul, Tachibana, Otsuki, Gao, Tsuboi and Torii2009). However, because the Z-test used to evaluate the difference between d N and d S required a larger number of nucleotide differences than the detected one used for the data set presented here, the detected significant excess of d N over d S on the 5′ region of PfRON2 might be a false positive, thus further studies are required for PfRON2.
Three TM domains were predicted by different bio-informatics tools, TM3 was identified by 4 algorithms, but TM1 and TM2 were only supported by 2 algorithms. This difference may be due to the algorithm used in the webware; algorithms that basically rely on hydrophobicity tend to predict TM1 and TM2 in addition to T3, but algorithms that put more weight on amino acid characteristics, such as TMHMM, HMMTOP and SOSUI, did not detect TM1 and TM2. As this protein is secreted from the rhoptry neck and inserted into the host cell membrane, the predicted TM domains are probably not anchored in the parasite membrane prior to discharge, thus there would be a mechanism to prevent the exposure of these putative TM domains to the environment. They could, for example, be buried in the globular domain before discharge, but be exposed during host cell invasion. Such constraints may make TM1 and TM2 difficult to predict. Performing similar analyses and localization studies of RON2 in T. gondii, the following membrane topology of RON2 was proposed; the N-terminal segment before TM2 is located on the cytosolic side, the region between TM2 and TM3 or the segment after TM2 is exposed outside of the cell, and TM1 was actually not a true TM domain but peripherally associated with the host cell membrane by a hydrophobic interaction (Lamarque et al. Reference Lamarque, Besteiro, Papoin, Roques, Vulliez-Le Normand, Morlon-Guyot, Dubremetz, Fauquenoy, Tomavo, Faber, Kocken, Thomas, Boulanger, Bentley and Lebrun2011). Based on this model, all polymorphic sites of PvRON2 detected in this study locate in the region that is not exposed to the outside of the erythrocyte. It is not clear how the host immune system may recognize these sites during the asexual stage cycle. However, one possible explanation is that PvRON2 is released into the hepatocyte during sporozoite invasion where it may be targeted by host cellular immunity. RON2 expression in the sporozoite has previously been reported in the rodent malaria parasite Plasmodium berghei (Tufet-Bayona et al. 2009).
The RON2-AMA1 interaction was found to be strong and species specific (Lamarque et al. Reference Lamarque, Besteiro, Papoin, Roques, Vulliez-Le Normand, Morlon-Guyot, Dubremetz, Fauquenoy, Tomavo, Faber, Kocken, Thomas, Boulanger, Bentley and Lebrun2011). No significant binding of a homologous region from the rodent malaria parasite P. yoelii RON2 with PfAMA1 was observed, suggesting evolutionary constraints within Plasmodium species for the RON2-AMA1 interaction (Srinivasan et al. Reference Srinivasan, Beatty, Diouf, Herrera, Ambroggio, Moch, Tyler, Narum, Pierce, Boothroyd, Haynes and Miller2011). The region between TM2 and TM3 was proposed to be located on the erythrocyte surface and was shown to bind to AMA1 in both T. gondii and P. falciparum (Tyler and Boothroyd, Reference Tyler and Boothroyd2011; Lamarque et al. Reference Lamarque, Besteiro, Papoin, Roques, Vulliez-Le Normand, Morlon-Guyot, Dubremetz, Fauquenoy, Tomavo, Faber, Kocken, Thomas, Boulanger, Bentley and Lebrun2011). Recently, hot spot residues in the binding of TgAMA1-TgRON2 were identified (Tonkin et al. Reference Tonkin, Roques, Lamarque, Pugniére, Douguet, Crawford, Lebrun and Boulanger2011), which corresponds to aa 2035 to 2073 of PvRON2. There were no amino acid substitutions in this region among 20 isolates, supporting the conserved and potentially crucial role of this interaction.
ACKNOWLEDGEMENT
We would like to thank Key Lab. on Technology for Parasitic Disease Prevention and Control, Ministry of Health, Jiangsu Provincial and Key Lab. on Molecular Biology of Parasites for providing the necessary equipment and assistance.
FINANCIAL SUPPORT
This work was partly supported by the National Basic Research Program (973 Program) in China (2007CB513100), and the National Natural Science Foundation of China (No. 30700695, and 30872214).
