Infection, Genetics and Evolution 18 (2013) 247–256

Contents lists available at SciVerse ScienceDirect

Infection, Genetics and Evolution

journal homepage: www.elsevier .com/locate /meegid

Natural selection and population genetic structure of domain-I ofPlasmodium falciparum apical membrane antigen-1 in India

1567-1348/$ - see front matter � 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.meegid.2013.05.015

⇑ Corresponding author. Fax: +91 33 24614849.E-mail address: sanghamitrasg@yahoo.com (S. Sengupta).

Madhumita Basu a, Ardhendu Kumar Maji b, Mitashree Mitra c, Sanghamitra Sengupta a,⇑a Department of Biochemistry, University of Calcutta, 35, Ballygunge Circular Road, Kolkata 700019, West Bengal, Indiab Department of Protozoology, The Calcutta School of Tropical Medicine, 108, Chittaranjan Avenue, Kolkata 700 073, West Bengal, Indiac School of Studies in Anthropology, Pt. Ravishankar Shukla University, Raipur 492010, Chhattisgarh, India

a r t i c l e i n f o

Article history:Received 21 March 2013Received in revised form 16 May 2013Accepted 18 May 2013Available online 6 June 2013

Keywords:Plasmodium falciparumApical membrane antigen-1Genetic diversityKolkataIndiaProtein–protein interaction

a b s t r a c t

Development of a vaccine against Plasmodium falciparum infection is an urgent priority particularlybecause of widespread resistance to most traditionally used drugs. Multiple evidences point to apicalmembrane antigen-1(AMA-1) as a prime vaccine candidate directed against P. falciparum asexualblood-stages. To gain understanding of the genetic and demographic forces shaping the parasite sequencediversity in Kolkata, a part of Pfama-1 gene covering domain-I was sequenced from 100 blood samples ofmalaria patients. Statistical and phylogenetic analyses of the sequences were performed using DnaSP andMEGA. Very high haplotype diversity was detected both at nucleotide (0.998 ± 0.002) and amino-acid(0.996 ± 0.001) levels. An abundance of low frequency polymorphisms (Tajima’s D = �1.190, Fu & Li’sD⁄ and F⁄ = �3.068 and �2.722), unimodal mismatch distribution and a star-like median-joining networkof ama-1 haplotypes indicated a recent population expansion among Kolkata parasites. The high mini-mum number of recombination events (Rm = 26) and a significantly high dN/dS of 3.705 (P < 0.0001) inKolkata suggested recombination and positive selection as major forces in the generation and mainte-nance of ama-1 allelic diversity. To evaluate the impact of observed non-synonymous substitutions inthe context of AMA-1 functionality, PatchDock and FireDock protein–protein interaction solutions weremapped between PfAMA-1-PfRON2 and PfAMA-1-host IgNAR. Alterations in the desolvation and globalenergies of PfAMA-1-PfRON2 interaction complexes at the hotspot contact residues were observedtogether with redistribution of surface electrostatic potentials at the variant alleles with respect to refer-ent Pf3D7 sequence. Finally, a comparison of P. falciparum subpopulations in five Indian regional isolatesretrieved from GenBank revealed a significant level of genetic differentiation (FST = 0.084–0.129) withrespect to Kolkata sequences. Collectively, our results indicated a very high allelic and haplotype diver-sity, a high recombination rate and a signature of natural selection favoring accumulation of non-synon-ymous substitutions that facilitated PfAMA-1-PfRON2 interaction and hence parasite growth in Kolkataclinical isolates.

� 2013 Elsevier B.V. All rights reserved.

1. Introduction

Malaria is a serious public health problem in the tropics withestimated 216 million cases and 655,000 deaths in 2010 (WorldHealth Organization., 2011) Immunization with AMA-1 inducesantibodies that inhibit invasion, conferring protection in animals,mostly due to infections of the most virulent human malaria para-site, Plasmodium falciparum (Genton and Reed, 2007; Hu et al.,2008; Rodrigues et al., 2005). For effective control of this deadlydisease, vaccines are decisively needed. Immunization with differ-ent blood-stage antigens is shown to be protective in a number ofanimal models (Malkin et al., 2005; Narum et al., 2000; Stowers

et al., 2002). At present, the leading blood-stage vaccine candidatesare all proteins expressed during the invasion of the red blood cells(RBCs), either contained within the apical organelles or located onthe merozoite surface (Alaro et al., 2010). One of the current chal-lenges in developing vaccines targeting these antigens is the highlevel of genetic diversity among Plasmodium isolates in differentgeographical areas across the world. Highly polymorphic regionshave been observed in P. falciparum and Plasmodium vivax anti-genic surface proteins such as Circumsporozoite protein (CSP), Duf-fy-binding protein (DBP), Merozoite surface protein-1 (MSP-1),Apical membrane antigen-1 (AMA-1) and Thrombospondin relatedanonymous protein (TRAP) (Chenet et al., 2012). Furthermore, thepolymorphic regions are not evenly distributed across these anti-genic proteins (Escalante et al., 1998; Franks et al., 2003; Muet al., 2007; Polley and Conway, 2001; Volkman et al., 2007).

248 M. Basu et al. / Infection, Genetics and Evolution 18 (2013) 247–256

Protein–protein interactions (PPIs) serve as the centralmechanism in most cellular functions and it is essentially guidedby amino acid residues with different structural and functionalconstraints (Bonsor and Sundberg, 2011). Mutations in thesecontact residues may cause altered binding properties and thusdifferent amino acids will be subject to position specific selectivepressures (Hoberman et al., 2004). Until recently mostprotein–protein interactions have been proved experimentallyat the single amino acid level making them capable of capturingthe combinatorial effect of mutagenesis and selection atpopulation level (Munz et al., 2012; Nannemann et al., 2011).In this investigation, we study the genetic diversity of AMA-1not only to evaluate the processes maintaining thevariation but in addition to explain functionally the nature andpattern of amino acid substitutions observed in parasitepopulation.

The severe pathophysiological manifestations of malaria causedby P. falciparum are a direct consequence of the parasite’s blood stagereplication cycle, during which merozoites repeatedly invade, mul-tiply within, and destroy red blood cells (Woehlbier et al., 2010). Anumber of parasite proteins have been implicated in RBC invasionthat are comprised of (a) GPI-Anchored MSP family proteins(MSP1/2/4/10), (b) microneme proteins (AMA1, EBA-175, EBA-140/BAEBL, EBA-181/JESEBL), (c) peripheral surface proteins(MSP3/6/7, SERA3/4/5/6) and (d) rhoptry neck proteins (RON2/4etc.) (Cowman and Crabb, 2006). The exposure of P. falciparum mer-ozoites to low potassium ion concentrations as found in blood plas-ma leads to a rise in cytosolic calcium levels which triggers secretionof 175 kD erythrocyte binding antigen (EBA175) and apical mem-brane antigen-1 (AMA-1) to the merozoite surface (Singh et al.,2010). Subsequent interaction of EBA175 with glycophorin A (glyA),its receptor on erythrocytes, restores basal cytosolic calcium levelsand triggers release a set of rhoptry neck-derived parasite proteins(RON proteins) which associate with PfAMA1 at the moving junc-tion, a step crucial for RBC invasion (Alexander et al., 2006; Collinset al., 2009; Dutta et al., 2003; Lamarque et al., 2011; Tonkin et al.,2011). It has been demonstrated that antibodies directed againstAMA-1 at the time of presentation with malaria are strongly associ-ated with concurrent and future blood stage immunity, as measuredby the ability of the host to clear P. falciparum (Keh et al., 2012). AMA-1 recombinant vaccines based on P. falciparum 3D7 and FVO strainhave shown an excellent efficacy in Phase I trials (Malkin et al.,2005) and recently has been tested in Phase II clinical trial (Sagaraet al., 2009). The essential role of PfAMA-1 in parasite survival andhigh level of immunogenicity during natural infection in humanmakes this protein an attractive candidate for gene diversityanalysis.

AMA-1 is an 83-kDa type I integral membrane protein withan ectodomain organized in three domains (domain I-III) (Howellet al., 2003; Healer et al., 2005). It is synthesized late during thedevelopment of parasite schizonts and processed to a 66-kDaform that relocates to the surfaces of mature merozoites (Ban-nister et al., 2003; Howell et al., 2001). The domain-I ofPfAMA-1 is the most diverse region of this antigen (Escalanteet al., 2001; Marshall et al., 1996; Polley and Conway, 2001)and appears to be a major target of anti-AMA-1 protective anti-bodies (Dutta et al., 2003; Kocken et al., 2002; Mardani et al.,2012). To this end, the present study examines the level of ge-netic diversity of domain-I of P. falciparum ama-1 gene fromKolkata, West Bengal, and compares the diversity parameters be-tween different regions of India by collating this data with thoseavailable in GenBank. We also predict the 3D structure and theprotein–protein dockings of mutant PfAMA-1 using different insilico strategies to evaluate the impact of amino acid substitu-tions on the binding affinities of AMA-1 with its functionalpartners.

2. Materials and methods

2.1. Study site and sample collection

The study was conducted in Kolkata, West Bengal from India.Malaria transmission is seasonal (from May to July and then againSeptember–November) in Kolkata with a predominance of P. falci-parum infection and the proportion is in the range of 50–70 (Joshiet al., 2007). Peripheral blood (2–3 ml) was drawn from mildsymptomatic malaria patients from the malaria clinic attached toThe Calcutta School of Tropical Medicine, Kolkata, India betweenOctober 2008 and November 2009. The study protocol for clinicalspecimens used in this research has been approved by the institu-tional ethical committees of the University of Calcutta and The Cal-cutta School of Tropical Medicine, India. All study participants havegiven the written informed consent prior to sample collection.Blood samples positive for Plasmodium falciparum infection, deter-mined by Giemsa-stained thick smears were only considered forthis study and patients were categorized according to WHO (WorldHealth Organization., 2006) protocols. Detailed demographic infor-mation for the full set of samples has been presented elsewhere(Basu et al., 2010).

2.2. DNA extraction, amplification and purification of PCR productsand sequencing

Genomic DNA from Plasmodium falciparum infected blood sam-ples was extracted from QIAamp Blood Midi kit (Qiagen, Hilden,Germany) according to manufacturer’s protocol. Pfama-1 gene spe-cific oligonucleotides (Pfama-1 F.P: 50-CTGGAACTCAATATA-GACTTC-30 and Pfama-1 R.P: 50-GAAAAAGTTTGCCCTAGAAAGAA-30) were designed from P. falciparum genomic DNA (3D7 strain:GenBank accession number U65407.1) to amplify a 505 bp frag-ment encompassing domain-I of the gene by polymerase chainreaction. The thermal cycler (Applied Biosystems� GeneAmp�

PCR System 9700) was conditioned for an initial denaturation for1 min at 94 �C, followed by 40 cycles of denaturation at 94 �C for30 s, annealing at 58 �C for 30 s, elongation at 72 �C for 1 minand a final extension at 72 �C for 5 min. The amplicon of 505 bpharboring the hypervariable region (HVR) of Pfama-1 gene wasthen purified and sequenced. After amplification, the PCR productswere analyzed by agarose gel electrophoresis and purified by QIA-quick Gel Extraction kit (Qiagen, Hilden, Germany) as per the man-ufacturer’s instructions. The purified amplicons were thensequenced and the PCR protocol was programmed for initial dena-turation at 94 �C for 30 s followed by 25 cycles of denaturation at94 �C for 10 s, holding 5 s at 50 �C and 4 min at 60 �C and finallystored at 4 �C. Sequencing was performed in both direction, usingthe forward and reverse Pfama-1 primers described earlier usingBig Dye ver 3.1 dye terminator technology and ran on ABI Prism3100 Genetic Analyzer (Applied Biosystems, Foster City, CA).

2.3. Sequence alignment and data analysis

For sequence alignment, initially P. falciparum 3D7 ama-1nucleotide sequence (PF11_0344) located in chromosome 11 fromposition 1,293,856 to 1,295,724 was extracted from PlasmoDBdatabase (http://plasmodb.org/plasmo/). Raw sequence files weremanually edited for removing ambiguous reads and signal noiseand a stretch of 449 bp sequence has been derived from 100 P. fal-ciparum isolates which was deposited to Genbank database (http://www.ncbi.nlm.nih.gov/genbank/). The sequence data was alignedusing ClustalW (http://www.ebi.ac.uk/Tools/msa/clustalw2/) forthe nucleotide and the corresponding amino acid changes andthe extent of genetic variability was compared with previously re-

M. Basu et al. / Infection, Genetics and Evolution 18 (2013) 247–256 249

ported prototypic alleles of Pf3D7. To compare the sequence iden-tity NCBI BLAST (http://blast.ncbi.nlm.nih.gov/) analysis was alsoperformed with all sequences with respect to Pf3D7. Nucleotidemismatches and statistical analyses were carried out using DnaSPversion 4.0 (Rozas and Rozas, 1999; Rozas et al., 2003). Variantsites with quality score >30 in both forward and reverse sequenceswere included in the analyses. The phylogenetic and molecularevolutionary analyses were conducted using MEGA version 3.1(Kumar et al., 2004) (http://www.megasoftware.net/). These soft-wares were used to translate the nucleotide sequences into aminoacid codes and the number of synonymous and non-synonymousamino acid substitutions were calculated and taken intoconsideration.

Nucleotide sequences analyzed in this report have been submit-ted to the GenBank database under accession numbers KC476551-KC476650. Apart from Kolkata parasite isolates, previously pub-lished Pfama-1 sequences from five different geographical regionsfrom India viz., Kamprup (Assam), Cuttack (Orissa), Andaman &Nicobar (A & N), Ghaziabad and Aligarh (UP) and Panjim (Goa)were collected from GenBank database (http://www.ncbi.nlm.nih.-gov/genbank/) with accession numbers EF413088-EF413170 (Garget al., 2007). Sequences from Rajasthan (Rajesh et al., 2008) withaccession no. EF543164- EF543168) and from Delhi (Escalanteet al., 2001) with accession no. AY016428, 31, 34, 37, 39 werepooled with UP isolates generating a total of 19 isolates altogetherfrom North India (NI) for further comparative sequence diversityanalysis.

To understand the genetic architecture of Pfama-1 gene, severalgenetic diversity measures were estimated from 100 nucleotide se-quences encompassing the domain-I of the gene using DnaSP v4.0.Similar analyses were also performed with the subsets sampledfrom five different parts of India (Assam, Orissa, Andaman & Nico-bar, North India and Goa) after retrieving the sequences from pub-lic database. These analyses included (i) a description of thenumber of segregating (S) sites (ii) observed nucleotide diversityper site between any two sequences assuming that the sample israndom (p); (iii) average number of pairwise nucleotide differ-ences within population (k); (iv) number of haplotypes (H) andhaplotype diversity (Hd).

The estimation of recombination in Pfama-1 gene was per-formed upon the alignment of the 100 sequences by DnaSP in orderto calculate the minimum number of recombination events (Rm)that have occurred along the sequence and to estimate the recom-bination parameter, R. This statistic incorporates the effective pop-ulation size and probability of recombination between adjacentnucleotides per generation. The r2 index of linkage disequilibrium(LD) were also estimated across all polymorphic sites (using DnaSPsoftware), and the relationship between LD (in terms of r2) and dis-tance between nucleotide sites of each pairwise comparison wasplotted. The statistical significance (P < 0.05) of each pairwise testof LD on these haploid data was evaluated by v2 test after Bonfer-roni corrections.

Tajima’s D (Tajima, 1989; 1996) and Fu & Li’s (Fu and Li, 1993)statistics were applied to test the neutral theory of evolution usingDnaSP. Positive values for Tajima’s D correspond to positive selec-tion whereas negative values correspond to negative or purifyingselection (Garg et al., 2007; Tachida, 2000). The ratio of non-synon-ymous to synonymous substitutions (dN/dS) is widely used as anindicator of the action of natural selection in gene sequences. Anexcess of non-synonymous relative to synonymous mutations isa clear signal of positive selection, whereas a lack of non-synony-mous relative to synonymous polymorphisms suggests negativeor purifying selection imposed by functional constraint. Theserates of substitution within species were estimated using themethod of (Nei and Gojobori, 1986) with the Jukes and Cantor cor-rection (Jukes and Cantor, 1969) as implemented in the MEGA 3.1

software. P values and standard error was determined by 500 boot-strap replications and the rates were compared with the Z-test ofselection (MEGA 3.1).

Phylogenetic tree was constructed for haplotypes generatedfrom Kolkata isolates by neighbor-joining (NJ) method with boot-strap confidence and Kimura 2-parameter distance matrix (Kim-ura, 1980) in MEGA v3.1. The distribution of individual (aminoacid) haplotypes derived from Kolkata was also drawn using Net-work program (Bandelt et al., 1999). Network analysis originatedas a mathematical tool to understand social relationships and hasbeen used to study the transmission of infectious diseases (Fried-man and Aral, 2001). In population genetics, it has been adaptedto explore relationships among sequences by linking haplotypesthat are identical at a predefined proportion of polymorphic sites(Bull et al., 2008; Barry et al., 2009). As each haplotype (node)may have multiple connections (edges), this analysis has the po-tential to define not only distinct clusters or subgroups of highlyrelated sequences but also the relationships among them. Further-more, this program can identify the location of less frequently ob-served admixed haplotypes in the network, which may representnovel recombinants.

The intra- and inter-population genetic diversity were calcu-lated using the Arlequin software package version 3.5 (Excoffieret al., 2005) (http://cmpg.unibe.ch/software/arlequin3), and thedistance between populations was measured by the fixation index(FST). FST is the ratio of the sum of the estimated variances of thehaplotype frequencies due to differences in haplotypes in differentpopulations and those due to differences among the different pop-ulations to the total variance in haplotype frequency (Hudson et al.,1992). FST varies between 0 (each defined population with thesame haplotype mix), and 1 (each defined population with totallydifferent haplotypes).

2.4. Protein–protein interaction: PatchDock & FireDock analyses

In order to identify the strongly associated functional partnerslinked to PfAMA-1, STRING version 9.05 (http://string-db.org/)was used that serves as a comprehensive database of all knownand predicted protein interactions (Franceschini et al., 2013). Theseprotein–protein interactions include direct (physical) and indirect(functional) associations and are derived from four sources: (1)genomic context, (2) high throughput experiments, (3) conservedco-expression and (4) text mining from previous knowledge(Szklarczyk et al., 2011). STRING quantitatively integrates interac-tion data from these sources for a large number of organisms, andtransfers information between these organisms where applicable.The algorithm works iteratively trying to position the nodes apartfrom each other with a ‘‘preferred distance’’ which is proportionalto the String global score (Szklarczyk et al., 2011). So, interactingproteins with higher global score have more chances to end upin the same cluster.

In order to validate STRING-based structural prediction ofPfAMA-1 interaction network, a geometry based molecular dock-ing program, PatchDock (http://bioinfo3d.cs.tau.ac.il/PatchDock/)(Shatsky et al., 2004; Schneidman-Duhovny et al., 2005) was em-ployed to elucidate the nature of protein–protein connections. Thisalgorithm initially computes the overall molecular surface of theprotein followed by finding of the geometric patches (concave,convex and flat surface pieces) in the protein (Hu et al., 2000).The selection of best patches was performed retaining the ‘‘hotspot’’ areas and as a final point, these patches were matched withthe patches from another query protein based on hybrid of theGeometric Hashing (Wolfson and Rigoutsos, 1997) and Pose-Clus-tering matching techniques (Stockman, 1987). The bad complexeswere discarded with unacceptable penetrations of the atoms of thereceptor to the atoms of the ligand. Top 20 solutions were auto-

250 M. Basu et al. / Infection, Genetics and Evolution 18 (2013) 247–256

matically generated from PatchDock each having the geometricshape complementarity score, complex interface area, desolvationenergy or atomic contact energy and the actual rigid transforma-tion files (Schneidman-Duhovny et al., 2005). Each candidate wassubsequently refined by restricted interface side-chain rearrange-ment and by soft rigid-body optimization using Fast InteractionREfinement in molecular DOCKing (FireDock) (http://bio-info3d.cs.tau.ac.il/FireDock/) (Andrusier et al., 2007). Followingrearrangement of the side-chains, the relative position of the dock-ing partners was refined by Monte Carlo minimization of the bind-ing score function and are ranked on the basis of this score whichincludes global energy parameter (Zhang et al., 1997), softened vander Waals interactions, partial electrostatics and additional estima-tions of the binding free energy (Andrusier et al., 2007). In thisstudy, domain-I of wild type and mutant PfAMA-1 polypeptides(2Z8 W chain A. pdb) were examined for their differential bindingaffinities towards a series of parasite and host proteins [PfRON2(3ZWZ. pdb) and host IgNAR (2Z8 W chain C. pdb)] necessary forgeneration of moving junction complex and in erythrocyte inva-sion process.

For homology modeling, wild type P. falciparum (3D7) AMA-1(2Z8 W: chain A), PfRON2 (3ZWZ. pdb) and host IgNAR (2Z8 Wchain C. pdb) protein files were extracted from RCSB protein databank (www.rcsb.org/) and .pdb files were generated for all the mu-tant proteins using SwissModel Workspace (www.swissmod-el.expasy.org/workspace/). Three dimensional configurationswere fitted with the mutant AMA-1 polypeptides harboring variantamino acids in domain-I using Swiss Pdb-Viewer (www.spdbv.vi-tal-it.ch/) software [Swiss Institute of Bioinformatics (Basel, Swit-zerland)] and compared with respect to the wild type P.falciparum 3D7 AMA-1 sequence. The root mean square deviation(RMSD) and localized surface charge distribution of the wild typeand variant amino acid residues were determined by the SPDB-Viewer (www.spdbv.vital-it.ch/) and GRASP2 softwares (http://wiki.c2b2.columbia.edu/honiglab_public/index.php/Soft-ware:GRASP2), respectively.

3. Results

3.1. Sequence diversity and population structure of Kolkata clinicalisolates

The main goal of this study was to analyze the genetic diversityof P. falciparum field isolates from Kolkata, West Bengal based onthe hypervariable domain-I of ama-1 gene sequences and to disen-tangle the footprints of natural selection and demographic forcesresponsible for maintaining the observed diversity. The assessmentof a 449 bp sequence spanning 468–917 bp region of the ama-1gene in 100 P. falciparum clinical isolates from Kolkata (NCBI Gen-Bank accession no: KC476551-KC476650) resulted in 121 variantsites, of which 94 were bi-allelic and remaining 27 positions weretri-(25) and tetra-(2) allelic. Majority of the variant sites (83 of121) were parsimony informative. Nucleotide diversity was esti-mated from (i) the total number of segregating sites, (ii) the aver-age number of nucleotide differences between two sequencesrandomly chosen from the population and (iii) the average numberof mismatches between two sequences Table 1. The estimate of p(p = 0.034 + 0.001) was less than hw (hw = 0.039 + 0.01) resulting ina negative Tajima’s D statistic (�1.190) and statistically significantFu and Li’s D⁄ (�3.069, P < 0.05) and F⁄ (�2.722, P < 0.05) estimatesindicating an abundance of low frequency variants (Table 1). Theaverage number of mismatches between any two sequences was14.974 and the mismatch distribution yielded a unimodal patternwith very low raggedness (0.0016) which was suggestive of a re-cent population expansion (Fig. 1).

Ninety-two distinct haplotypes were derived from nucleotidesequence data (Table S1), 86 of which were singleton, five haplo-types were found twice while Hap 21 was present in three copiesin the data. Hap 21 differed from the reference Pf3D7 ama-1 hap-lotype at 15 nucleotide positions. Haplotype diversity (Hd) wasestimated to be 0.998 ± 0.002. The phylogenetic relationshipamong the haplotypes was demonstrated using a neighbor joiningtree (Fig. 1). The median-joining network among the parsimonyinformative haplotypes (n = 58) displayed a star-like phylogeny,characteristic of population expansion. Estimates of recombinationrate was deduced using minimum number of recombinationevents (Rm) which could be defined as the number of recombina-tion events parsimoniously inferred from a sample of sequences(Hudson and Kaplan, 1985). A very high minimum number recom-bination events (Rm = 26) from Kolkata was observed compared toother regional isolates (Table 1) corroborating with the low linkagedisequilibrium profile in the region. Out of 4371, 94 (0.02%) and 25(0.006%) pairwise comparisons displayed statistically significant r2

estimate before and after Bonferroni correction respectively. Theregression line in the r2 vs. nucleotide position plot declined shar-ply starting from a r2 value of 0.2 (Fig. 1).

One hundred and twenty-one segregating sites resulted in 67non-synonymous amino acid changes and 87 haplotypes whichwere denoted as KH1-KH87. The position and type of amino acidalterations were shown in Table S2. Of these 67 polymorphic ami-no acid positions, 38 sites (56.7%) were dimorphic, 18 (26.9%) weretrimorphic, 5 (7.5%) sites were tetramorphic and 6 of them (8.9%)showed more than four allele changes. Thirty-one percent of thepolymorphic sites (21 out of 67) had a variant allele frequency>10% and 29 (43.3%) of them were multi-allelic in nature. Majority(53.7%) of the variant positions were polar to polar residue changeswhile 19.4% of the alterations were polar to hydrophobic in nature.16.4% and 7.5% of the variations show nonpolar to nonpolar andnonpolar to polar amino acid changes respectively. The most pre-valent haplotype KH10 had a frequency of 5% in Kolkata parasitepopulation (Table S2) and it differed in 11 amino acid positions(N162H, G172E, Y175D, E187K, E197H, H200D, K206E, I225N,D242Y, K243D and I282K) from the reference Pf3D7 sequence.

3.2. Test of neutrality

To test the neutral theory of evolution, the rates of synonymousand non-synonymous mutations and the ratio of non-synonymousto synonymous substitutions (dN/dS) were computed using MEGAversion 4. A significantly higher estimate of non-synonymous sub-stitutions over number of non-synonymous sites(dN = 0.189 + 0.039) was obtained compared to that of synonymoussubstitutions over number of synonymous sites(dS = 0.051 + 0.014) yielding a very high dN/dS ratio of 3.705 (Z-test:P < 0.0001) (Table 1). The excess of non-synonymous polymor-phisms in domain-I of AMA-1 suggested a departure from neutralevolution for this part of Pfama-1 gene. To understand how thenon-synonymous changes might influence the functionality ofAMA-1, the protein–protein interactions of AMA-1 with relevantparasite and host proteins were investigated.

3.3. Impact of non-synonymous changes on the interaction betweenPfAMA-1 and its functional partners

The functional interacting partners of Plasmodium falciparumAMA-1 were selected using STRING (version 9.05) database whichrevealed a strong association between PfAMA-1 and PfRON2 (inter-action score = 0.999). In humans, a crucial step in malaria patho-genesis is the formation of a moving junction (MJ) complexbetween the membranes of the invading apicomplexan parasiteand the host cell (Besteiro et al., 2011). The MJ contains two key

Table 1Population specific measures of DNA sequence polymorphisms at domain I of ama-1 in six Indian P. falciparum isolates.

Region N Hw p + SD Tajima’s D Fu & Li’s D⁄ Fu & Li’s F⁄ r Rm S (SE) N (SE) dS (SE) dN (SE) dN/dS

Kolkata 100 0.039 0.034 ± 0.001 �1.190 –3.069⁄ �2.722⁄ 0.0016 26 20.925±1.190 63.075 ± 1.137 0.051 + 0.014 0.189 + 0.039 3.705⁄Assam 22 0.020 0.024 ± 0.001 0.827 0.092 0.219 0.0088 9 21.000 ± 1.154 63.000 ± 1.192 0.072 ± 0.025 0.235 ± 0.044 3.263⁄Orissa 24 0.021 0.025 ± 0.001 0.767 0.137 0.389 0.0088 11 21.303 ± 1.081 62.697 ± 1.093 0.070 ± 0.026 0.245 ± 0.036 3.500⁄A & Na 19 0.022 0.022 ± 0.002 0.093 0.206 0.200 0.0448 6 21.488 ± 1.184 62.512 ± 1.209 0.078 ± 0.031 0.225 ± 0.045 2.880⁄North India 19 0.021 0.024 ± 0.001 0.490 0.339 0.111 0.0218 9 20.860 ± 1.184 63.140 ± 1.266 0.044 ± 0.020 0.208 ± 0.045 4.720⁄Goa 9 0.015 0.018 ± 0.003 1.057 0.929 1.076 0.1636 6 21.194 ± 1.184 62.806 ± 1.183 0.058 ± 0.028 0.195 ± 0.051 3.360⁄

N; number of isolates, hw; Watterson’s h, p; pairwise nucleotide diversity; Tajima’s D; allele frequency distribution of nucleotide sequence data based on the differencebetween two estimators of h, r; Raggedness Index, Rm; minimum number of recombination events, N and S, average numbers of non-synonymous and synonymous sitesrespectively, dS; number of synonymous substitutions over number of synonymous sites; dN, number of non-synonymous substitutions over number of non-synonymoussites; SE, standard error computed using the Nei-Gojobori method with Jukes-Cantor correction. SE was estimated using the bootstrap method with 500 replicationsimplemented in MEGA4.0.⁄Indicates statistically significant P value (<0.05).

a A & N = Andaman & Nicobar Islands.

Fig. 1. (A) The mismatch distribution pattern in Kolkata isolates considering all the segregating sites. Red line indicates the expected distribution pattern and blue curveindicates the mismatch distribution observed in Kolkata Pfama-1 population. The raggedness statistic was denoted in the plot. (B) Linkage disequilibrium plot: r2 estimatewas plotted across the nucleotide distance considering all segregating sites. The red dots indicate the significant v2 statistic and blue circles represent non-significant values.The relation between LD and nucleotide distance was shown in a form of regression line, indicating a high meiotic recombination rate leading to the generation of many newalleles. (C) Circular phylogenetic tree of P. falciparum ama-1 (nucleotide) haplotypes constructed with Kolkata samples using MEGA version 3.1. (D) Median Joining Networkfor PfAMA-1 Kolkata isolates: colored (yellow) circles represent ama1 (polypeptide) haplotypes and lines signify mutational steps connecting the haplotypes. KH denotesKolkata haplotypes. A star-like network of AMA-1 haplotypes indicated a recent population expansion among Kolkata parasites. (For interpretation of the references to colourin this figure legend, the reader is referred to the web version of this article.)

M. Basu et al. / Infection, Genetics and Evolution 18 (2013) 247–256 251

parasite components: the micronemal surface protein Apical Mem-brane Antigen 1 (AMA1) and its receptor, the Rhoptry Neck Protein(RON) complex (Lamarque et al., 2011). In particular, RON2, atransmembrane component of the RON complex, interacts directlywith AMA1 (Tonkin et al., 2011; Vulliez-Le Normand et al., 2012).

To investigate the probable impact of amino acid residue changeson PfAMA-1-PfRON2 interaction affinities, a subset of PfAMA1mutations which were previously reported as ‘‘hot spots’’ drivingspecific PfAMA1-PfRON2 complex formation was selected. Theseresidues included G172E, Y175D, E187K, N223H, M224I, I225N,

252 M. Basu et al. / Infection, Genetics and Evolution 18 (2013) 247–256

K230Q and Y251C (Collins et al., 2009; Vulliez-Le Normand et al.,2012). To gain understanding of the conformational contributionof non-synonymous changes on the PfAMA-1-ligand interactions,the computationally derived models of mutant PfAMA-1 weredocked separately onto crystal structures of RON2. The interactionaffinity between two proteins was scored by evaluating the pro-tein–protein docking solutions derived from PatchDock and Fire-Dock web-servers. PatchDock is a geometry-based moleculardocking algorithm which performs coarse global search for feasibleorientations yielding good molecular shape complementarity. ThePatchDock algorithm divides the Connolly dot surface representa-tion of the molecules into concave, convex and flat patches to gen-erate candidate transformations which were further evaluated by ascoring function that considers both geometric fit and atomicdesolvation energy and finally, an RMSD (root mean square devia-tion) clustering is applied to the candidate solutions to discardredundant ones (Shatsky et al., 2004; Schneidman-Duhovnyet al., 2005). Since PatchDock treats proteins as rigid bodies, theinteraction solutions from PatchDock were refined using FireDockwhich optimizes side-chain conformations and rigid-body orienta-tion (Andrusier et al., 2007) allowing selection of the top-rankedmodel with minimum global energy. The majority of the interac-tion solutions derived for PfAMA-1 Kolkata variants at positionsG172E, Y175D, E187K, I225N, K230Q and Y251C yielded higherarea of contacts and lower desolvation (ACE) and the global ener-gies compared to the that of wild-type Pf3D7-PfRON2 interactionsuggesting an alteration of inter-atomic contacts due to the abovenon-conservative changes of PfAMA-1 alleles (Table 2). In addition,remarkable alterations of surface electrostatic potential (Fig. 2) androot mean square deviation (RMSD) were also noted. Our studythus indicated a possible structural adaptation of PfAMA-1 proteininduced by the non-synonymous mutations. Notably the frequencyof some of the variant alleles including G172E, Y175D, E187K andI225N found in the Kolkata Pf population were 0.34, 0.85, 0.42 and0.83, respectively (Table 2) implicating a signature of directionalselection in PfAMA-1 gene.

To review the possible impact of the non-synonymous changeson the antigenic behavior of AMA-1, protein–protein interactionaffinities between PfAMA-1 and host IgNAR were evaluated uponemploying PatchDock and FireDock. Except for Y251 position,top-ranked docking solutions of all non-synonymous changespresent in the hydrophobic cleft showed increased requirementof desolvation energy for transferring atoms from water to theprotein’s interior environment (Table 3). An appraisal of the fre-quency distribution of 12 major RON2 contact sites in AMA-1hydrophobic cleft showed that three of the positions (183, 208and 252) harbored synonymous changes while the frequency ofthe remaining sites was <5% in the Kolkata Pf-population exceptfor M190I (frequency = 0.17) (Table 3) indicating a strong selec-tion pressure exerted by host immune responses on this ligandbinding site.

Table 2Best interaction solutions of PfAMA-1 mutant alleles with PfRON2 from PatchDock and Fi

Docking estimates Pf3D7⁄ 172E 175D 1

Variant allelefrequency

– 0.34 0.85 0

PatchDock Score 12580 13228 12116 1Area 1614.5 1638.9 1825.6 1ACEa �239.0 �468.6 �396.4 �

FireDock Global Energy �17.40 �84.68 �22.59 �RMSDb – 0.53 Å 0.41 Å 0

⁄Amino acids at positions: G172, Y175, E187, N223, M224, I225, K230 and Y251 were pa ACE = atomic contact energy.b RMSD = root mean square deviation.

3.4. Comparison of sequence diversity between Kolkata and otherIndian isolates

To understand whether geographic distance contributes to ge-netic differentiation of P. falciparum population, a comparativeanalysis of sequence diversity was also performed by retrieving456 bp parasite sequence data from different regions of India fromGenBank database with accession numbers: [EF413088-EF413170](Garg et al., 2007), [EF543164- EF543168] (Rajesh et al., 2008) and[AY016428, 31, 34, 37, 39] (Escalante et al., 2001). The regions in-cluded both high (for example; Assam, Orissa and Andaman & Nic-obar Islands) and low (Goa and North India) transmission areas.The comparison of Pfama-1 sequences of different regions was per-formed to determine nucleotide diversity estimates, recombina-tion parameters (Rm), the mean number of synonymoussubstitutions per synonymous sites (dS) and non-synonymous sub-stitutions per non-synonymous sites (dN) (Table 1). Total numberof variant sites in Assam, Orissa, A& N, North India and Goa were24, 27, 23, 19 and 16 respectively. Unlike Kolkata samples, the esti-mate of p was higher that of hw resulting in positive Tajima’s D forall five regions. However, the ratio of dN/dS was > 1 indicating a sig-nature of positive selection in other Indian isolates as well. To as-sess the level of genetic differentiation between the populations,polypeptide haplotypes were constructed from all the samplesfrom Kolkata, Assam, Orissa, A & N, Goa and North India Pfama-1sequences and Wright’s fixation index (FST) was estimated fromDnaSP v4.0. The number of haplotypes observed in Assam, Orissa,A & N, Goa and North India isolates were 58, 20, 22, 14, 6 and 19respectively (Table S3). A very strong genetic differentiation wasnoted between Kolkata and other regional parasite populationsdue to large number of private haplotypes in Kolkata (Table 4).The minimum number of recombination events (Rm) was the high-est in Kolkata (26) compared to other regions such as Assam (9),Orissa (11), Andaman & Nicobar Islands (6), North India (9) andGoa (6). In summary, the comparison of Pfama-1 sequences amongsix regional isolates in India demonstrated that the parasite diver-sity in Kolkata was markedly distinct with a significant excess lowfrequency polymorphisms caused differential recombination, ge-netic and demographic history in the samples.

4. Discussion

In the present study, we have combined a comparative evolu-tionary analysis of ama-1 domain-I gene sequence, with the dock-ing approaches to infer protein–protein interactions betweenAMA-1 and RON2 and IgNAR to dissect the relative impact ofrecombination, selection and demographic history on the observedsequence diversity of the Kolkata P. falciparum parasite population.An analysis of a 449 bp sequence of Pfama-1 gene in 100 clinicalisolates of Kolkata revealed very high haplotype diversity. A totalnumber of 121 nucleotide (haplotype diversity = 0.998) and 67

reDock analyses.

87K 223H 224I 225N 230Q 251C

.42 0.03 0.02 0.83 0.04 0.02

2230 12474 11758 12950 14334 13446835.8 1388.9 1891.5 2038.2 1725.6 1738.4449.8 �338.1 19.89 �418.4 �490.8 �415.185.50 �57.23 -21.62 �22.88 �83.78 �50.23.53 Å 0.01 Å 0.01 Å 0.59 Å 0.11 Å 0.53Å

resent in wild type Pf3D7 AMA-1 sequence.

Fig. 2. Homology modeling and surface charge distribution pattern of wild-type and variant PfAMA-1 polypeptide: PDB files were generated using SwissModel Workspacewith amino acid substitutions at codon positions 172, 175, 187, 223, 224, 225, 230 and 251. (A-H) display surface electrostatic potential maps for the referent Plasmodiumfalciparum 3D7 strain (2Z8 W: chain A) and mutated polypeptides harboring the non-synonymous substitutions using GRASP2. Black and white arrowheads indicate thepolymorphic position in the polypeptide.

Table 3Best interaction solutions of PfAMA-1 mutant alleles with host IgNAR from PatchDock and FireDock analyses.

Docking Estimates Pf3D7⁄ 169G 176F 187K 190I 201L 202C 224I 225N 251C 273L

Variant allele frequency – 0.02 0.03 0.42 0.17 0.05 0.01 0.02 0.83 0.02 0.02PatchDock Score 10960 14438 12452 13564 12190 12734 12830 12738 13334 10790 13280

Area 1161.6 2535.7 2100.6 1914.5 1719 1806.6 1904.1 1625.8 1874 1170.7 1797.7ACEa �161.4 176.3 148.6 66.28 26.3 183.2 69.83 193.2 159.3 �175.5 194.61

FireDock Global Energy �103.3 �15.03 �13.59 �4.81 �1.22 �2.44 �11.45 �26.97 �13.7 �94.51 �1.89

⁄ Amino acids at positions: V169, L176, E187, M190, F201, Y202, M224, I225, Y251 and M273 were present in wild type Pf3D7 AMA-1 sequence.a ACE = atomic contact energy.

M. Basu et al. / Infection, Genetics and Evolution 18 (2013) 247–256 253

non-synonymous amino acid (haplotype diversity = 0.996) changesyielded 92 and 87 unique haplotypes respectively. Similar level ofstrain variability was also detected in other endemic regions. Forexample, among the 50 Thai isolates sequenced, there were 27haplotypes (Polley et al., 2003) while 45 and 78 haplotypes werederived from 51 Nigerian and 129 costal Kenyan sequences (Polleyand Conway, 2001). In contrast, a study in Papua New Guinea(PNG) have found only 27 haplotypes within domain I in 168AMA1 alleles sampled (Cortes et al., 2003) which is considerablylower than that observed within domain I in Kenyan or Nigerianparasite isolates (Polley and Conway, 2001). The largest populationsample dataset from Mali reported 214 unique haplotypes was re-ported from 506 samples (Takala et al., 2009). The differences inhaplotype diversity are likely to reflect variances in malaria trans-

mission, as has been observed with other polymorphic loci wherehigher transmission is generally associated with increased diver-sity (Babiker et al., 1997; Schoepflin et al., 2009) and was due com-bined effects of a higher effective population size andrecombination rate in addition to natural selection. A similar trendof high haplotype diversity has also been observed from studiesconducted in India. The extent of genetic polymorphisms at do-main-I of PfAMA-1 was shown to be higher amongst the isolatesfrom Kolkata (this study), Assam, Orissa and A & N (high malariatransmission areas) as compared to those from UP and Goa isolates(low malaria transmission areas) (Garg et al., 2007). In summary,analysis of the AMA-1 sequences have demonstrated that the levelof diversity amongst Indian P. falciparum isolates was comparableto those reported for African isolates but was relatively higher than

Table 4Pairwise FST between populations.

No. ofhaplotypes

Shared haplotypes

Kolkata(58)

Assam(20)

Orissa(22)

A & N(14)

North India(19)

Goa(6)

FST estimateKolkata – 16 – – 4 4Assam 0.084⁄ – 4 2 2 2Orissa 0.129⁄ 0.013 – 3 2 –A & Na 0.097⁄ 0.054⁄ 0.047 – 3 –North India 0.092⁄ 0.006 0.005 0.049 – 6Goa 0.094⁄ 0.016 0.014 0.077⁄ 0.004 –

⁄Indicates P < 0.05.a A & N = Andaman & Nicobar Islands.

254 M. Basu et al. / Infection, Genetics and Evolution 18 (2013) 247–256

the Asian and South American populations (Escalante et al., 2001;Garg et al., 2007; Polley and Conway, 2001; Polley et al., 2003;Schoepflin et al., 2009).

Nucleotide diversity parameters were estimated based on num-ber and frequency of segregating sites. Since, estimate of p de-pends on the frequencies of variants, the negative Tajima’s D andFu & Li’s D⁄ and F⁄ estimates observed in the data indicates a pre-ponderance of low frequency alleles which could be due to purify-ing selection or recent population expansion (Hartl and Clark,1997) since both of these affect the number of segregating sitesand h strongly. The abundance of low frequency alleles in KolkataP. falciparum population yields a smooth unimodal mismatch dis-tribution which is again a signature of population expansion. How-ever, it is noteworthy that to disentangle the effects of selectionand demography unequivocally, analysis of neutral markers shouldbe included. In areas with high transmission intensity, recombina-tion between genetically distinct parasite clones (out-crossing)generates novel parasite variants with clinically important pheno-types such as virulence, drug resistance or immune evasion (Au-burn et al., 2012). The strong recombination activity (Rm = 26)and high haplotype diversity observed in the data indicates recom-bination as a major force for generating newer alleles.

Evidence for natural selection, positive or negative, on geneencoding antigens indicate variation of functional constraints. Asignificantly higher dN/dS ratio in Kolkata observed in the data isindicative of positive selection, presumably due to host immunepressure. To gain an in depth understanding of the impact ofAMA1-amino acid changes, genetic variability data was subjectedto in silico molecular docking studies. During the course of infec-tion, the commitment of Plasmodium merozoites to invade redblood cells is marked by the formation of a junction between themerozoite and the RBC with the coordinated induction of the par-asitophorous vacuole (Cowman et al., 2012). Recent proteomeanalysis of the closely-related apicomplexan parasite, Toxoplasmagondii has revealed a panel of novel proteins (RON2, RON4, andRON5) located at the neck portion of the rhoptries which interactswith the microneme protein Apical Membrane Antigen 1 (AMA-1)to form a moving junction complex. Co-immunoprecipitation ofPfRON2 with PfAMA1 in T. gondii and P. falciparum substantiatessuch complex formation (Besteiro et al., 2009; Lamarque et al.,2011). The recent solutions of the crystallographic structures of do-mains-I and -II of P. falciparum strain 3D7 AMA-1 revealed that anextended cleft served as a ligand-binding site in the AMA-1-anti-body interaction (Bai et al., 2005; Nuttall et al., 2004). The baseof the cleft is rich in solvent-exposed hydrophobic side chains,formed by residues V169, L176, F183, M190, Y202, V208, M224,Y251, I252, M273, L357, and F367 and potentially most importantcontact is mediated by Y251 through its aromatic ring interactingwith the VNAR residues Tyr94 and Tyr96 (Henderson et al., 2007).The strong functional association between RON2 and AMA-1 was

also detected from STRING database, in this study. We selectedthe major contact amino acids on AMA-1 protein that mediate itsinteraction with RON2 and IgNAR from the protein–protein inter-action studies to conduct a comparative docking analysis (Hender-son et al., 2007; Lamarque et al., 2011). The binding free energy ofthe protein–ligand complex in water environment primarily de-pends on two components: binding enthalpy and solvation/desolv-ation free energy (Kolar et al., 2011). Sites with low desolvationenergy values (negative values) are predicted to be involved ininteractions exhibiting a favorable energy change (Casasoli et al.,2009). The top-ranked docking solutions for mutant residues thatinteract with RON2 show a lowering of atomic contact energy(desolvation energy) compared to referent Pf3D7 sequences, whilethose with IgNAR have been found to associate with large energychanges. We predict that the non-synonymous changes observedin Kolkata parasite pool favor the PfAMA-1-RON2 interactionwhich is essential for invasion and subsequent parasite growth inerythrocytes. Our analysis is based on solvation component of freeenergy which is a well accepted target function in structural stud-ies aiming at quantitative understanding of the energetics of vari-ous biological processes (Honig and Nicholls, 1995; Shoichet andKuntz, 1991; Zhang et al., 1997), however post-docking molecularand biophysical studies are necessary to establish our prediction.

Taken together our data shows that domain-I of Pfama-1 exhib-its very high level of allelic diversity in Kolkata isolates presumablydue to high recombination rate in the molecule. The strong geneticdifferentiation between Kolkata and other regional isolates is dueto large number of private haplotypes present in Kolkata. Themajority of the amino acid residues that mediate the contact be-tween PfAMA-1 and RON2 harbor non-synonymous substitutionsfavoring efficient RBC invasion which in turn may result in anexpansion of parasite population. The nucleotide diversity param-eters reinforces above prediction. Future co-crystallographic andstructural biology studies using isogenic parasite lines with differ-ent AMA-1 alleles responsible for merozoite attack is necessary toconfirm our in silico predictions. Furthermore, evaluation of param-eters including multiplicity of infection, their relative proportionsand genetic divergence which influence out-crossing risk is essen-tial in different epidemiological settings before selecting any para-site surface antigens as vaccine targets.

Acknowledgements

The authors are grateful to all study participants for their coop-eration. This work has been supported by the funding from Depart-ment of Science and Technology, New Delhi (SERC Fast TrackScheme: SR/FTP/L-56/2005 dated 25.04.2006) and UniversityGrants Commission (Major Research Project-F. No 33-232/2007(SR) dated 13.03.2008). We also thank CAS (UGC), DST-FIST,and IPLS (DBT) for providing some of the instrument facilities atthe Department of Biochemistry, University of Calcutta. MB is sup-ported by pre-doctoral fellowships from UGC-SAP-RFSMS schemeand CSIR.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.meegid.2013.05.015.

References

Alaro, J.R., Lynch, M.M., Burns Jr., J.M., 2010. Protective immune responses elicitedby immunization with a chimeric blood-stage malaria vaccine persist but arenot boosted by Plasmodium yoelii challenge infection. Vaccine 28, 6876–6884.

Alexander, D.L., Arastu-Kapur, S., Dubremetz, J.F., Boothroyd, J.C., 2006. Plasmodiumfalciparum AMA1 binds a rhoptry neck protein homologous to TgRON4, a

M. Basu et al. / Infection, Genetics and Evolution 18 (2013) 247–256 255

component of the moving junction in Toxoplasma gondii. Eukaryot Cell 5,1169–1173.

Andrusier, N., Nussinov, R., Wolfson, H.J., 2007. FireDock: fast interactionrefinement in molecular docking. Proteins 69, 139–159.

Auburn, S., Campino, S., Miotto, O., Djimde, A.A., Zongo, I., Manske, M., Maslen, G.,Mangano, V., Alcock, D., MacInnis, B., Rockett, K.A., Clark, T.G., Doumbo, O.K.,Ouedraogo, J.B., Kwiatkowski, D.P., 2012. Characterization of within-hostPlasmodium falciparum diversity using next-generation sequence data. PLoSOne 7, e32891.

Babiker, H.A., Lines, J., Hill, W.G., Walliker, D., 1997. Population structure ofPlasmodium falciparum in villages with different malaria endemicity in eastAfrica. Am. J. Trop. Med. Hyg. 56, 141–147.

Bai, T., Becker, M., Gupta, A., Strike, P., Murphy, V.J., Anders, R.F., Batchelor, A.H.,2005. Structure of AMA1 from Plasmodium falciparum reveals a clustering ofpolymorphisms that surround a conserved hydrophobic pocket. Proc. Natl.Acad. Sci. USA 102, 12736–12741.

Bandelt, H.J., Forster, P., Rohl, A., 1999. Median-joining networks for inferringintraspecific phylogenies. Mol. Biol. Evol. 16, 37–48.

Bannister, L.H., Hopkins, J.M., Dluzewski, A.R., Margos, G., Williams, I.T., Blackman,M.J., Kocken, C.H., Thomas, A.W., Mitchell, G.H., 2003. Plasmodium falciparumapical membrane antigen 1 (PfAMA-1) is translocated within micronemes alongsubpellicular microtubules during merozoite development. J. Cell Sci. 116,3825–3834.

Barry, A.E., Schultz, L., Buckee, C.O., Reeder, J.C., 2009. Contrasting populationstructures of the genes encoding ten leading vaccine-candidate antigens of thehuman malaria parasite, Plasmodium falciparum. PLoS One 4, e8497.

Basu, M., Maji, A.K., Chakraborty, A., Banerjee, R., Mullick, S., Saha, P., Das, S.,Kanjilal, S.D., Sengupta, S., 2010. Genetic association of Toll-like-receptor 4 andtumor necrosis factor-alpha polymorphisms with Plasmodium falciparum bloodinfection levels. Infect. Genet. Evol. 10, 686–696.

Besteiro, S., Dubremetz, J.F., Lebrun, M., 2011. The moving junction of apicomplexanparasites: a key structure for invasion. Cell Microbiol. 13, 797–805.

Besteiro, S., Michelin, A., Poncet, J., Dubremetz, J.F., Lebrun, M., 2009. Export of aToxoplasma gondii rhoptry neck protein complex at the host cell membrane toform the moving junction during invasion. PLoS Pathog. 5, e1000309.

Bonsor, D.A., Sundberg, E.J., 2011. Dissecting protein-protein interactions usingdirected evolution. Biochemistry 50, 2394–2402.

Bull, P.C., Buckee, C.O., Kyes, S., Kortok, M.M., Thathy, V., Guyah, B., Stoute, J.A.,Newbold, C.I., Marsh, K., 2008. Plasmodium falciparum antigenic variation.Mapping mosaic var gene sequences onto a network of shared, highlypolymorphic sequence blocks. Mol. Microbiol. 68, 1519–1534.

Casasoli, M., Federici, L., Spinelli, F., Di Matteo, A., Vella, N., Scaloni, F., Fernandez-Recio, J., Cervone, F., De Lorenzo, G., 2009. Integration of evolutionary anddesolvation energy analysis identifies functional sites in a plant immunityprotein. Proc. Natl. Acad. Sci. USA 106, 7666–7671.

Chenet, S.M., Tapia, L.L., Escalante, A.A., Durand, S., Lucas, C., Bacon, D.J., 2012.Genetic diversity and population structure of genes encoding vaccine candidateantigens of Plasmodium vivax. Malar J. 11, 68.

Collins, C.R., Withers-Martinez, C., Hackett, F., Blackman, M.J., 2009. An inhibitoryantibody blocks interactions between components of the malarial invasionmachinery. PLoS Pathog 5, e1000273.

Cortes, A., Mellombo, M., Mueller, I., Benet, A., Reeder, J.C., Anders, R.F., 2003.Geographical structure of diversity and differences between symptomatic andasymptomatic infections for Plasmodium falciparum vaccine candidate AMA1.Infect Immun. 71, 1416–1426.

Cowman, A.F., Berry, D., Baum, J., 2012. The cellular and molecular basis for malariaparasite invasion of the human red blood cell. J. Cell Biol. 198, 961–971.

Cowman, A.F., Crabb, B.S., 2006. Invasion of red blood cells by malaria parasites. Cell124, 755–766.

Dutta, S., Haynes, J.D., Moch, J.K., Barbosa, A., Lanar, D.E., 2003. Invasion-inhibitoryantibodies inhibit proteolytic processing of apical membrane antigen 1 ofPlasmodium falciparum merozoites. Proc. Natl. Acad. Sci. USA 100, 12295–12300.

Escalante, A.A., Grebert, H.M., Chaiyaroj, S.C., Magris, M., Biswas, S., Nahlen, B.L., Lal,A.A., 2001. Polymorphism in the gene encoding the apical membrane antigen-1(AMA-1) of Plasmodium falciparum. X. Asembo Bay Cohort Project. Mol.Biochem. Parasitol. 113, 279–287.

Escalante, A.A., Lal, A.A., Ayala, F.J., 1998. Genetic polymorphism and naturalselection in the malaria parasite Plasmodium falciparum. Genetics 149, 189–202.

Excoffier, L., Laval, G., Schneider, S., 2005. Arlequin (version 3.0): an integratedsoftware package for population genetics data analysis. Evol. Bioinform. Online1, 47–50.

Franceschini, A., Szklarczyk, D., Frankild, S., Kuhn, M., Simonovic, M., Roth, A., Lin, J.,Minguez, P., Bork, P., von Mering, C., Jensen, L.J., 2013. STRING v9.1: protein–protein interaction networks, with increased coverage and integration. NucleicAcids Res. 41, D808–815.

Franks, S., Baton, L., Tetteh, K., Tongren, E., Dewin, D., Akanmori, B.D., Koram, K.A.,Ranford-Cartwright, L., Riley, E.M., 2003. Genetic diversity and antigenicpolymorphism in Plasmodium falciparum: extensive serological cross-reactivity between allelic variants of merozoite surface protein 2. InfectImmun. 71, 3485–3495.

Friedman, S.R., Aral, S., 2001. Social networks, risk-potential networks, health, anddisease. J. Urban Health 78, 411–418.

Fu, Y.X., Li, W.H., 1993. Statistical tests of neutrality of mutations. Genetics 133,693–709.

Garg, S., Alam, M.T., Das, M.K., Dev, V., Kumar, A., Dash, A.P., Sharma, Y.D., 2007.Sequence diversity and natural selection at domain I of the apical membraneantigen 1 among Indian Plasmodium falciparum populations. Malar J. 6, 154.

Genton, B., Reed, Z.H., 2007. Asexual blood-stage malaria vaccine development:facing the challenges. Curr. Opin. Infect Dis. 20, 467–475.

Hartl, D.L., Clark, A.G., 1997. Principles of Population Genetics, 3rd ed. IE-MACMILLAN UK.

Healer, J., Triglia, T., Hodder, A.N., Gemmill, A.W., Cowman, A.F., 2005. Functionalanalysis of Plasmodium falciparum apical membrane antigen 1 utilizinginterspecies domains. Infect Immun. 73, 2444–2451.

Henderson, K.A., Streltsov, V.A., Coley, A.M., Dolezal, O., Hudson, P.J., Batchelor, A.H.,Gupta, A., Bai, T., Murphy, V.J., Anders, R.F., Foley, M., Nuttall, S.D., 2007.Structure of an IgNAR-AMA1 complex: targeting a conserved hydrophobic cleftbroadens malarial strain recognition. Structure 15, 1452–1466.

Hoberman, R., Klein-Seetharaman, J., Rosenfeld, R., 2004. Inferring propertyselection pressure from positional residue conservation. Appl. Bioinform. 3,167–179.

Honig, B., Nicholls, A., 1995. Classical electrostatics in biology and chemistry.Science 268, 1144–1149.

Howell, S.A., Well, I., Fleck, S.L., Kettleborough, C., Collins, C.R., Blackman, M.J., 2003.A single malaria merozoite serine protease mediates shedding ofmultiple surface proteins by juxtamembrane cleavage. J. Biol. Chem. 278,23890–23898.

Howell, S.A., Withers-Martinez, C., Kocken, C.H., Thomas, A.W., Blackman, M.J.,2001. Proteolytic processing and primary structure of Plasmodium falciparumapical membrane antigen-1. J. Biol. Chem. 276, 31311–31320.

Hu, J., Chen, Z., Gu, J., Wan, M., Shen, Q., Kieny, M.P., He, J., Li, Z., Zhang, Q., Reed, Z.H.,Zhu, Y., Li, W., Cao, Y., Qu, L., Cao, Z., Wang, Q., Liu, H., Pan, X., Huang, X., Zhang,D., Xue, X., Pan, W., 2008. Safety and immunogenicity of a malaria vaccine,Plasmodium falciparum AMA-1/MSP-1 chimeric protein formulated inmontanide ISA 720 in healthy adults. PLoS One 3, e1952.

Hu, Z., Ma, B., Wolfson, H., Nussinov, R., 2000. Conservation of polar residues as hotspots at protein interfaces. Proteins 39, 331–342.

Hudson, R.R., Kaplan, N.L., 1985. Statistical properties of the number ofrecombination events in the history of a sample of DNA sequences. Genetics111, 147–164.

Hudson, R.R., Slatkin, M., Maddison, W.P., 1992. Estimation of levels of gene flowfrom DNA sequence data. Genetics 132, 583–589.

Joshi, H., Valecha, N., Verma, A., Kaul, A., Mallick, P.K., Shalini, S., Prajapati, S.K.,Sharma, S.K., Dev, V., Biswas, S., Nanda, N., Malhotra, M.S., Subbarao, S.K., Dash,A.P., 2007. Genetic structure of Plasmodium falciparum field isolates in easternand north-eastern India. Malar J. 6, 60.

Jukes, T., Cantor, C., 1969. Mammalian Protein Metabolism, in: HN, M. (Ed.),Evolution of protein molecules. Academic Press, New York, pp. 21–132.

Keh, C.E., Jha, A.R., Nzarubara, B., Lanar, D.E., Dutta, S., Theisen, M., Rosenthal, P.J.,Dorsey, G., Nixon, D.F., Greenhouse, B., 2012. Associations between antibodiesto a panel of Plasmodium falciparum specific antigens and response to sub-optimal antimalarial therapy in Kampala, Uganda. Plos One 7, e52571.

Kimura, M., 1980. A simple method for estimating evolutionary rates of basesubstitutions through comparative studies of nucleotide sequences. J. Mol. Evol.16, 111–120.

Kocken, C.H., Withers-Martinez, C., Dubbeld, M.A., van der Wel, A., Hackett, F.,Valderrama, A., Blackman, M.J., Thomas, A.W., 2002. High-level expression ofthe malaria blood-stage vaccine candidate Plasmodium falciparum apicalmembrane antigen 1 and induction of antibodies that inhibit erythrocyteinvasion. Infect Immun. 70, 4471–4476.

Kolar, M., Fanfrlik, J., Hobza, P., 2011. Ligand conformational and solvation/desolvation free energy in protein-ligand complex formation. J. Phys. Chem. B115, 4718–4724.

Kumar, S., Tamura, K., Nei, M., 2004. MEGA3: integrated software for molecularevolutionary genetics analysis and sequence alignment. Brief Bioinform. 5, 150–163.

Lamarque, M., Besteiro, S., Papoin, J., Roques, M., Vulliez-Le Normand, B., Morlon-Guyot, J., Dubremetz, J.F., Fauquenoy, S., Tomavo, S., Faber, B.W., Kocken, C.H.,Thomas, A.W., Boulanger, M.J., Bentley, G.A., Lebrun, M., 2011. The RON2–AMA1interaction is a critical step in moving junction-dependent invasion byapicomplexan parasites. Plos Pathog. 7, e1001276.

Malkin, E.M., Diemert, D.J., McArthur, J.H., Perreault, J.R., Miles, A.P., Giersing, B.K.,Mullen, G.E., Orcutt, A., Muratova, O., Awkal, M., Zhou, H., Wang, J., Stowers, A.,Long, C.A., Mahanty, S., Miller, L.H., Saul, A., Durbin, A.P., 2005. Phase 1 clinicaltrial of apical membrane antigen 1: an asexual blood-stage vaccine forPlasmodium falciparum malaria. Infect Immun. 73, 3677–3685.

Mardani, A., Keshavarz, H., Heidari, A., Hajjaran, H., Raeisi, A., Khorramizadeh, M.R.,2012. Genetic diversity and natural selection at the domain I of apicalmembrane antigen-1 (AMA-1) of Plasmodium falciparum in isolates from Iran.Exp. Parasitol. 130, 456–462.

Marshall, V.M., Zhang, L., Anders, R.F., Coppel, R.L., 1996. Diversity of the vaccinecandidate AMA-1 of Plasmodium falciparum. Mol. Biochem. Parasitol. 77, 109–113.

Mu, J., Awadalla, P., Duan, J., McGee, K.M., Keebler, J., Seydel, K., McVean, G.A., Su,X.Z., 2007. Genome-wide variation and identification of vaccine targets in thePlasmodium falciparum genome. Nat. Genet. 39, 126–130.

Munz, M., Hein, J., Biggin, P.C., 2012. The role of flexibility and conformationalselection in the binding promiscuity of PDZ domains. Plos Comput. Biol. 8,e1002749.

256 M. Basu et al. / Infection, Genetics and Evolution 18 (2013) 247–256

Nannemann, D.P., Birmingham, W.R., Scism, R.A., Bachmann, B.O., 2011. Assessingdirected evolution methods for the generation of biosynthetic enzymes withpotential in drug biosynthesis. Future Med. Chem. 3, 809–819.

Narum, D.L., Ogun, S.A., Thomas, A.W., Holder, A.A., 2000. Immunization withparasite-derived apical membrane antigen 1 or passive immunization with aspecific monoclonal antibody protects BALB/c mice against lethal Plasmodiumyoelii yoelii YM blood-stage infection. Infect Immun. 68, 2899–2906.

Nei, M., Gojobori, T., 1986. Simple methods for estimating the numbers ofsynonymous and non-synonymous nucleotide substitutions. Mol. Biol. Evol. 3,418–426.

Nuttall, S.D., Humberstone, K.S., Krishnan, U.V., Carmichael, J.A., Doughty, L.,Hattarki, M., Coley, A.M., Casey, J.L., Anders, R.F., Foley, M., Irving, R.A., Hudson,P.J., 2004. Selection and affinity maturation of IgNAR variable domains targetingPlasmodium falciparum AMA1. Proteins 55, 187–197.

Polley, S.D., Chokejindachai, W., Conway, D.J., 2003. Allele frequency-based analysesrobustly map sequence sites under balancing selection in a malaria vaccinecandidate antigen. Genetics 165, 555–561.

Polley, S.D., Conway, D.J., 2001. Strong diversifying selection on domains of thePlasmodium falciparum apical membrane antigen 1 gene. Genetics 158, 1505–1512.

Rajesh, V., Singamsetti, V.K., Vidya, S., Gowrishankar, M., Elamaran, M., Tripathi, J.,Radhika, N.B., Kochar, D., Ranjan, A., Roy, S.K., Das, A., 2008. Plasmodiumfalciparum: genetic polymorphism in apical membrane antigen-1 gene fromIndian isolates. Exp. Parasitol. 119, 144–151.

Rodrigues, M.H., Rodrigues, K.M., Oliveira, T.R., Comodo, A.N., Rodrigues, M.M.,Kocken, C.H., Thomas, A.W., Soares, I.S., 2005. Antibody response of naturallyinfected individuals to recombinant Plasmodium vivax apical membraneantigen-1. Int. J. Parasitol. 35, 185–192.

Rozas, J., Rozas, R., 1999. DnaSP version 3: an integrated program for molecularpopulation genetics and molecular evolution analysis. Bioinformatics 15, 174–175.

Rozas, J., Sanchez-DelBarrio, J.C., Messeguer, X., Rozas, R., 2003. DnaSP, DNApolymorphism analyses by the coalescent and other methods. Bioinformatics19, 2496–2497.

Sagara, I., Ellis, R.D., Dicko, A., Niambele, M.B., Kamate, B., Guindo, O., Sissoko, M.S.,Fay, M.P., Guindo, M.A., Kante, O., Saye, R., Miura, K., Long, C., Mullen, G.E.,Pierce, M., Martin, L.B., Rausch, K., Dolo, A., Diallo, D.A., Miller, L.H., Doumbo,O.K., 2009. A randomized and controlled Phase 1 study of the safety andimmunogenicity of the AMA1-C1/Alhydrogel + CPG 7909 vaccine forPlasmodium falciparum malaria in semi-immune Malian adults. Vaccine 27,7292–7298.

Schneidman-Duhovny, D., Inbar, Y., Nussinov, R., Wolfson, H.J., 2005. PatchDock andSymmDock: servers for rigid and symmetric docking. Nucleic Acids Res. 33,W363–W367.

Schoepflin, S., Valsangiacomo, F., Lin, E., Kiniboro, B., Mueller, I., Felger, I., 2009.Comparison of Plasmodium falciparum allelic frequency distribution in differentendemic settings by high-resolution genotyping. Malar J. 8, 250.

Shatsky, M., Dror, O., Schneidman-Duhovny, D., Nussinov, R., Wolfson, H.J., 2004.BioInfo3D: a suite of tools for structural bioinformatics. Nucleic Acids Res. 32,W503–W507.

Shoichet, B.K., Kuntz, I.D., 1991. Protein docking and complementarity. J. Mol. Biol.221, 327–346.

Singh, S., Alam, M.M., Pal-Bhowmick, I., Brzostowski, J.A., Chitnis, C.E., 2010. Distinctexternal signals trigger sequential release of apical organelles duringerythrocyte invasion by malaria parasites. Plos Pathog. 6, e1000746.

Stockman, G., 1987. Object recognition and localization via pose clustering. J.Comput. Vision Graphics Image Process. 40, 361–387.

Stowers, A.W., Kennedy, M.C., Keegan, B.P., Saul, A., Long, C.A., Miller, L.H., 2002.Vaccination of monkeys with recombinant Plasmodium falciparum apicalmembrane antigen 1 confers protection against blood-stage malaria. InfectImmun. 70, 6961–6967.

Szklarczyk, D., Franceschini, A., Kuhn, M., Simonovic, M., Roth, A., Minguez, P.,Doerks, T., Stark, M., Muller, J., Bork, P., Jensen, L.J., von Mering, C., 2011. TheSTRING database in 2011: functional interaction networks of proteins, globallyintegrated and scored. Nucleic Acids Res. 39, D561–568.

Tachida, H., 2000. Molecular evolution in a multisite nearly neutral mutation model.J. Mol. Evol. 50, 69–81.

Tajima, F., 1989. Statistical method for testing the neutral mutation hypothesis byDNA polymorphism. Genetics 123, 585–595.

Tajima, F., 1996. The amount of DNA polymorphism maintained in a finitepopulation when the neutral mutation rate varies among sites. Genetics 143,1457–1465.

Takala, S.L., Coulibaly, D., Thera, M.A., Batchelor, A.H., Cummings, M.P., Escalante,A.A., Ouattara, A., Traore, K., Niangaly, A., Djimde, A.A., Doumbo, O.K., Plowe,C.V., 2009. Extreme polymorphism in a vaccine antigen and risk of clinicalmalaria: implications for vaccine development. Sci. Trans. Med. 1, 2ra5.

Tonkin, M.L., Roques, M., Lamarque, M.H., Pugniere, M., Douguet, D., Crawford, J.,Lebrun, M., Boulanger, M.J., 2011. Host cell invasion by apicomplexan parasites:insights from the co-structure of AMA1 with a RON2 peptide. Science 333, 463–467.

Volkman, S.K., Sabeti, P.C., DeCaprio, D., Neafsey, D.E., Schaffner, S.F., Milner Jr., D.A.,Daily, J.P., Sarr, O., Ndiaye, D., Ndir, O., Mboup, S., Duraisingh, M.T., Lukens, A.,Derr, A., Stange-Thomann, N., Waggoner, S., Onofrio, R., Ziaugra, L., Mauceli, E.,Gnerre, S., Jaffe, D.B., Zainoun, J., Wiegand, R.C., Birren, B.W., Hartl, D.L., Galagan,J.E., Lander, E.S., Wirth, D.F., 2007. A genome-wide map of diversity inPlasmodium falciparum. Nat. Genet. 39, 113–119.

Vulliez-Le Normand, B., Tonkin, M.L., Lamarque, M.H., Langer, S., Hoos, S., Roques,M., Saul, F.A., Faber, B.W., Bentley, G.A., Boulanger, M.J., Lebrun, M., 2012.Structural and functional insights into the malaria parasite moving junctioncomplex. Plos Pathog. 8, e1002755.

Woehlbier, U., Epp, C., Hackett, F., Blackman, M.J., Bujard, H., 2010. Antibodiesagainst multiple merozoite surface antigens of the human malaria parasitePlasmodium falciparum inhibit parasite maturation and red blood cell invasion.Malar J. 9, 77.

Wolfson, H.J., Rigoutsos, I., 1997. Geometric hashing: an overview. IEEE Comput. Sci.Eng. 11, 263–278.

World Health Organization., 2006. Guidelines for the treatment of malaria. WHO/HTM/MAL/2006.1108.

World Health Organization., 2011. World Malaria Report 2011.Zhang, C., Vasmatzis, G., Cornette, J.L., DeLisi, C., 1997. Determination of atomic

desolvation energies from the structures of crystallized proteins. J Mol Biol 267,707–726.