Structural and functional insights into the malaria parasite moving junction complex.

Abstract

Members of the phylum Apicomplexa, which include the malaria parasite Plasmodium, share many features in their invasion mechanism in spite of their diverse host cell specificities and life cycle characteristics. The formation of a moving junction (MJ) between the membranes of the invading apicomplexan parasite and the host cell is common to these intracellular pathogens. The MJ contains two key parasite components: the surface protein Apical Membrane Antigen 1 (AMA1) and its receptor, the Rhoptry Neck Protein (RON) complex, which is targeted to the host cell membrane during invasion. In particular, RON2, a transmembrane component of the RON complex, interacts directly with AMA1. Here, we report the crystal structure of AMA1 from Plasmodium falciparum in complex with a peptide derived from the extracellular region of PfRON2, highlighting clear specificities of the P. falciparum RON2-AMA1 interaction. The receptor-binding site of PfAMA1 comprises the hydrophobic groove and a region that becomes exposed by displacement of the flexible Domain II loop. Mutations of key contact residues of PfRON2 and PfAMA1 abrogate binding between the recombinant proteins. Although PfRON2 contacts some polymorphic residues, binding studies with PfAMA1 from different strains show that these have little effect on affinity. Moreover, we demonstrate that the PfRON2 peptide inhibits erythrocyte invasion by P. falciparum merozoites and that this strong inhibitory potency is not affected by AMA1 polymorphisms. In parallel, we have determined the crystal structure of PfAMA1 in complex with the invasion-inhibitory peptide R1 derived by phage display, revealing an unexpected structural mimicry of the PfRON2 peptide. These results identify the key residues governing the interactions between AMA1 and RON2 in P. falciparum and suggest novel approaches to antimalarial therapeutics.

Full text

Page 1

Structural and Functional Insights into the Malaria
Parasite Moving Junction Complex
Brigitte Vulliez-Le Normand1,2., Michelle L. Tonkin3., Mauld H. Lamarque4., Susann Langer3,
Sylviane Hoos5, Magali Roques4, Frederick A. Saul1,2, Bart W. Faber6, Graham A. Bentley1,2"*,
Martin J. Boulanger3"*, Maryse Lebrun4"*
1Unite ´ d’Immunologie Structurale, Institut Pasteur, Paris, France, 2URA 2185 CNRS, Paris, France, 3Department of Biochemistry & Microbiology, University of Victoria,
Victoria, British Columbia, Canada, 4UMR 5235 CNRS, Universite ´ de Montpellier 2, Montpellier, France, 5Plate-Forme de Biophysique des Macromole ´cules et de leurs
Interactions, Institut Pasteur, Paris, France, 6Department of Parasitology, Biomedical Primate Research Centre, Rijswijk, The Netherlands
Abstract
Members of the phylum Apicomplexa, which include the malaria parasite Plasmodium, share many features in their invasion
mechanism in spite of their diverse host cell specificities and life cycle characteristics. The formation of a moving junction
(MJ) between the membranes of the invading apicomplexan parasite and the host cell is common to these intracellular
pathogens. The MJ contains two key parasite components: the surface protein Apical Membrane Antigen 1 (AMA1) and its
receptor, the Rhoptry Neck Protein (RON) complex, which is targeted to the host cell membrane during invasion. In
particular, RON2, a transmembrane component of the RON complex, interacts directly with AMA1. Here, we report the
crystal structure of AMA1 from Plasmodium falciparum in complex with a peptide derived from the extracellular region of
PfRON2, highlighting clear specificities of the P. falciparum RON2-AMA1 interaction. The receptor-binding site of PfAMA1
comprises the hydrophobic groove and a region that becomes exposed by displacement of the flexible Domain II loop.
Mutations of key contact residues of PfRON2 and PfAMA1 abrogate binding between the recombinant proteins. Although
PfRON2 contacts some polymorphic residues, binding studies with PfAMA1 from different strains show that these have little
effect on affinity. Moreover, we demonstrate that the PfRON2 peptide inhibits erythrocyte invasion by P. falciparum
merozoites and that this strong inhibitory potency is not affected by AMA1 polymorphisms. In parallel, we have determined
the crystal structure of PfAMA1 in complex with the invasion-inhibitory peptide R1 derived by phage display, revealing an
unexpected structural mimicry of the PfRON2 peptide. These results identify the key residues governing the interactions
between AMA1 and RON2 in P. falciparum and suggest novel approaches to antimalarial therapeutics.
Citation: Vulliez-Le Normand B, Tonkin ML, Lamarque MH, Langer S, Hoos S, et al. (2012) Structural and Functional Insights into the Malaria Parasite Moving
Junction Complex. PLoS Pathog 8(6): e1002755. doi:10.1371/journal.ppat.1002755
Editor: Meg Phillips, University of Texas Southwestern Medical Center, United States of America
Received December 16, 2011; Accepted May 1, 2012; Published June 21, 2012
Copyright: ? 2012 Vulliez-Le Normand 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 Canadian Institutes of Health Research (CIHR) grant MOP82915 to MJB and ANR-2009 MIEN-0222 grant to ML and GAB.
MJB is a CIHR New Investigator and a Michael Smith Foundation for Health Research Scholar. MLT is supported by the Alexander Graham Bell Natural Sciences and
Engineering Research Council Scholarship. MHL is supported by an ANR-2009 grant. FVO 25-545 protein and the different strains of PfAMA1 domains I and II were
produced with funding of the European (Malaria) Vaccine Initiative. The funders had no role in study design, data collection and analysis, decision to publish, or
preparation of the manuscript.
Competing Interests: I have read the journal’s policy and have the following conflicts: Patent as a conflict of interest (The content of this manuscript has been
the object of a European Patent Application filed on May 6, 2011 under the application number EP 11305540.4).
* E-mail: graham.bentley@pasteur.fr (GAB); mboulang@uvic.ca (MJB); mylebrun@univ-montp2.fr (ML)
. These authors contributed equally to this work.
" These authors also contributed equally to this work.
Introduction
Plasmodium spp., and P. falciparum in particular, are devastating
global pathogens that place nearly half the human population at
risk to malaria, leading to more than 250 million cases yearly and
over one million deaths [1]. The success of the malaria parasite
can be attributed to its intracellular lifestyle, invading host cells
both in liver and blood stages. Invasion of red blood cells is an
active process involving a moving junction (MJ), which is formed
by intimate contact between erythrocyte and parasite membranes
and is thought to be coupled to the parasite’s actin-myosin motor
[2,3]. A number of merozoite antigens, either exposed on the
surface or stored in secretory organelles, play a role in the invasion
process [4]. One of these is Apical Membrane Antigen 1 (AMA1),
a type-one transmembrane protein secreted from the micronemes
to the merozoite surface and present at the MJ [5,6]. AMA1 is
highly conserved in the Plasmodium genus [6] and, moreover, in the
Apicomplexa phylum to which Plasmodium belongs [7,8], suggest-
ing a common functional role in diverse host cell invasion
scenarios. In the apicomplexan organism Toxoplasma gondii, the
receptor for AMA1 was shown to be Rhoptry Neck Protein 2
(RON2), a component of the parasite-derived RON protein
complex that is secreted into the host cell during invasion and
integrated into the host cell membrane [9,10]. This interaction
was subsequently confirmed in P. falciparum as well [11,12].
Apicomplexans thus provide both receptor and ligand to drive
active invasion.
In many malaria-endemic regions, P. falciparum has become
resistant to classic drugs, such as chloroquine, and is rapidly
developing resistance to recently introduced drugs. Since both
PLoS Pathogens | www.plospathogens.org1June 2012 | Volume 8 | Issue 6 | e1002755

Page 2

AMA1 and RON2 are specific to Apicomplexa and essential for
invasion, interruption of the AMA1-RON2 interaction presents an
ideal new target for the design and development of inhibitors. This
is supported by the recent observation that the invasion-inhibitory
peptide R1 [13,14] blocks interaction between AMA1 and the
RON complex in P. falciparum [15], but due to the polymorphism
of AMA1, the effectiveness of this peptide inhibitor is limited to a
subset of parasite isolates. Interestingly, R1 does not prevent apical
contact but no formation of a functional MJ ensues from this event
[15].
Crystal structures of PfAMA1 in complex with invasion-
inhibitory antibodies [16,17] have implicated a hydrophobic
groove on Domain I (DI) of PfAMA1 as being critical for function.
The topological nature of the PfAMA1 groove [18] is conserved in
P. vivax AMA1 [19] and T. gondii AMA1 [20], and contains a
number of residues that are conserved or semi-conserved across
Plasmodium species, as well as other members of Apicomplexa [21],
suggesting that it contributes to the receptor-binding site of
AMA1. This was recently confirmed by the crystal structure of
TgAMA1 in complex with a synthetic peptide, TgRON2sp, which
inserts in the groove of TgAMA1 [22].
Here, we report the crystal structure of the complex formed
between PfAMA1 and peptide segments of PfRON2, which,
together with our previous structural results on the TgAMA1-
TgRON2 co-structure [22], highlights a conserved, crucial
interaction in apicomplexan host cell invasion. Functional
characterization of hot-spot residues driving AMA1-RON2
complex formation leads to a deeper understanding of key
interactions occurring at the MJ of P. falciparum and reveals the
molecular basis of cross-strain reactivity while preserving specific-
ity for the species. We also describe the crystal structure of
PfAMA1 in complex with the invasion-inhibitory peptide R1 [14],
and show that this peptide presents an intriguing structural
mimicry of PfRON2. Collectively, our results provide an
important structural basis for designing cross-strain reactive
molecules that inhibit invasion by P. falciparum.
Results
PfRON2sp specifically binds to PfAMA1
From the 67-residue construct, PfRON2-5, that we previously
showed to have affinity for PfAMA1 [11], and guided by the
TgAMA1-TgRON2sp structure [22], we synthesized two analo-
gous PfRON2 peptides: PfRON2sp1 (residues 2021–2059; num-
bering from the initiation methionine in PF14_0495), and
PfRON2sp2 (residues 2027–2055). Significantly, there is no
polymorphism in this sequence among P. falciparum isolates. Both
constructs incorporate a disulfide-bound b-hairpin loop proposed
to be critical in complex formation [22] while PfRON2sp2 is
truncated at both the N- and C-termini (Fig. 1A). Since the
extracellular region of PfRON2 is non-polymorphic, we deter-
mined the affinity of both peptides for PfAMA1 by Surface
Plasmon Resonance (SPR) measurements using the 3D7, CAMP,
FVO and HB3 proteins to explore the possible effects of AMA1
polymorphisms. Theaffinity
PfRON2sp1 is 25-fold higher than for PfRON2sp2 (Fig. 1B to
E, Table 1), highlighting a moderate, yet influential, role for the N-
and C-terminal tails. Interestingly, KDvalues for the PfRON2sp
peptides showed no significant variation in binding to PfAMA1
from the four strains.
PfRON2sp1 and PfRON2sp2 were co-crystallized with the first
two ectoplasmic domains (DI, DII) of recombinant PfAMA1 3D7
or CAMP strains, respectively. The co-structure of PfAMA1 3D7
PfRON2sp1 (PDB entry code 3ZWZ) was refined to 2.2 A˚
resolution, while PfAMA1 CAMP PfRON2sp2 (PDB entry code
3SRI) was refined to 1.6 A˚resolution (Tables 2, 3). The two co-
structures overlay with a root mean square deviation (rmsd) of 0.81
A˚in 304 Ca positions, and the two peptides alone overlay with a
rmsd of 0.34 A˚ over the complete length of the modeled
PfRON2sp2 (25 Ca) (Fig. 2A). These data confirm that the
reduced affinity of PfRON2sp2 is due to the truncated N- and C-
termini. Since PfRON2sp1 is more biologically relevant than its
truncated counterpart, it is used for the following analyses unless
otherwise noted.
PfRON2sp1, traced from Thr2023 to Leu2058, includes a
disulfide bridge between Cys2037 and Cys2049 and makes several
direct contacts with PfAMA1 (Fig. S1), resulting in a total buried
surface area of 3154 A˚2(1441 A˚2for PfAMA1 and 1713 A˚2for
PfRON2sp1). Overall, the binding paradigm established by
TgAMA1-TgRON2sp [22] is maintained, with an N-terminal
helix seated at one end of the AMA1 receptor-binding groove and
extended through an ordered coil to a disulfide-closed b-hairpin
loop, generating a U-shaped conformation (Fig. 2A). Similarly,
exposing a functional receptor-binding groove on AMA1 requires
displacement of the extended non-polymorphic DII loop, which
adopts a disordered state (not modeled between Lys351 to
Ala387); this region is stabilized by DI in apo PfAMA1 (Fig. 2B).
Intriguingly, the backbone of the N-terminal helix and additional
coil of PfRON2sp1 (2024-QQAKDIGAG-2032) overlays remark-
ably well with a section of the apo PfAMA1 DII loop (360-
YEKIKEGFK-368) (rmsd,0.4 A˚), which also includes a helical
region (Fig. 2B - box 1). Three water molecules buried by the DII
loop in the apo form are retained in the receptor-bound state and
facilitate a network of hydrogen bonds that bridge PfAMA1 DI to
either the DII loop or PfRON2sp in apo PfAMA1 or the receptor
complex, respectively (Fig. 2C). The majority of intermolecular
contacts are formed by the segment Lys2027-Met2042 of
PfRON2sp1. An influential residue on PfRON2 appears to be
Arg2041, a residue specific to the P. falciparum species, located at
the tip of the b-hairpin with its guanidyl group fitting snugly into a
preformed pocket of PfAMA1 (Fig. 2D).
of
PfAMA1from 3D7 for
Author Summary
Malaria arises from infection of erythrocytes by single-cell
parasites belonging to the genus Plasmodium, the species
P. falciparum causing the most severe forms of the disease.
The formation of a moving junction (MJ) between the
membranes of the parasite and its host cell is essential for
invasion. Two important components of the MJ are Apical
Membrane Antigen 1 (AMA1) on the parasite surface and
the Plasmodium rhoptry neck (RON) protein complex that
is translocated to the erythrocyte membrane during
invasion. The extra-cellular region of RON2, a component
of this complex, interacts with AMA1, providing a bridge
between the parasite and its host cell that is crucial for
successful invasion. The parasite thus provides its own
receptor for AMA1 and accordingly this critical interaction
is not subject to evasive adaptations by the host. We
present atomic details of the interaction of PfAMA1 with
the carboxy-terminal region of RON2 and shed light on
structural adaptations by each apicomplexan parasite to
maintain an interaction so crucial for invasion. The
structure of the RON2 ligand bound to AMA1 thus
provides an ideal basis for drug design as such molecules
may be refractory to the development of drug resistance in
P. falciparum.
Structural Insights into the Moving Junction
PLoS Pathogens | www.plospathogens.org2 June 2012 | Volume 8 | Issue 6 | e1002755

Page 3

R1 occupies the PfRON2sp-binding site on PfAMA1
The invasion-inhibitory peptide R1, comprising 20 residues
(VFAEFLPLFSKFGSRMHILK) [14], has been shown by nuclear
magnetic resonance (NMR) to bind to the PfAMA1 hydrophobic
groove, but this study gave little structural detail of the interaction
[15]. We therefore crystallized PfAMA1 3D7 (DI and II) with R1
to compare with the PfRON2 complex. Surprisingly, two
molecules of R1 are bound to PfAMA1, which we denote
respectively as the major peptide (R1-major), lying deeply in the
binding groove, and the minor peptide (R1-minor), lying above
R1-major and making fewer contacts with PfAMA1 (Fig. 3 and
Table S1). Several solvent molecules bridge directly between
Figure 1. Surface Plasmon Resonance studies of peptides PfRON2sp1 and PfRON2sp2 binding to recombinant PfAMA1 from
multiple strains reveal that PfRON2sp1 has a consistently higher affinity. (A) PfRON2sp1 (orange) and PfRON2sp2 (grey) represent peptides
of PfRON2 (green). SP, signal peptide. TMD, putative transmembrane domain. (B). Sensorgrams showing PfRON2sp1 (analyte) binding to PfAMA1 3D7
(immobilized). The PfRON2sp1 concentrations are indicated for each curve (nM). (C). Sensorgrams showing PfRON2sp2 (analyte) binding to PfAMA1
CAMP (immobilized), with PfRON2sp2 concentrations indicated. (D, E). Variation percentage of bound sites (deduced from the steady-state response)
with respect to analyte concentration (D, PfRON2sp1; E, PfRON2sp2) obtained from binding to immobilized recombinant PfAMA1 from strains 3D7
(shown in B), CAMP (shown in C), FVO and HB3. The derived apparent equilibrium dissociation constants KDare given in Table 1.
doi:10.1371/journal.ppat.1002755.g001
Structural Insights into the Moving Junction
PLoS Pathogens | www.plospathogens.org3June 2012 | Volume 8 | Issue 6 | e1002755

Page 4

PfAMA1 and R1-major. As in the PfAMA1-PfRON2sp complex-
es, the N-terminus of R1-major binds to a region of PfAMA1 that
becomes exposed after displacement of the DII loop.
R1-major makes several direct contacts with PfAMA1 (113
interatomic distances,3.8 A˚), including 19 hydrogen bonds and a
salt bridge between the amino group of Lys-P11 (R1 peptide
residues numbers are prefixed by P) and the Asp227 carboxylate
group of PfAMA1 (Table S1A). Contacts made by R1-minor to
PfAMA1 are fewer (26 contacts,3.8 A˚) and include only five
hydrogen bonds (Table S1B). Interactions between R1-major and
R1-minor are maintained by a total of 24 interatomic contacts,
including three hydrogen bonds (Table S1C). In total, 3025 A˚2of
molecular surface is buried between PfAMA1 and the two
peptides, with R1-major contributing about 75% to this area.
The buried surface between R1-major and R1-minor is 563 A˚2,
reflecting the smaller number of close interatomic contacts
between these two components.
Since the structure of the PfAMA1 3D7-R1 complex revealed
two bound peptide molecules, binding measurements of R1 to
PfAMA1 3D7 were made by isothermal titration calorimetry (ITC)
to examine the stoichiometry (Fig. S2). The measured KD of
145 nM is comparable with previous measurements by SPR [13]
and the deduced stoichiometry was 1:1 over the peptide
concentrations used. This implies that the second binding site in
the crystal structure (R-minor) has an affinity that could not be
determined under the experimental conditions used for ITC but
can be estimated to be at least 10-fold weaker than the major site.
R1 mimicry of PfRON2
While R1-major follows the general contour of the receptor-
binding groove, it does so in a linear rather than the U-shaped
conformation adopted by PfRON2sp1 (Fig. 4A). R1-minor
occupies a similar region in space as the second strand of the
PfRON2sp b-hairpin, contacting the same DI loop of PfAMA1 but
running in the opposite direction to form a parallel two-stranded
b-sheet with the major peptide (Fig. 4A). Portions of R1-major
exhibit structural similarity to PfRON2, displaying a 1.2 A˚rmsd in
the twelve Ca positions (PfRON2sp1, Ala2031 to Met2042; R1-
major, Phe-P5 to Met-P16) (Fig. 4A). Moreover, sequence
alignment based on the structural superposition reveals a
remarkable similarity between the central regions of the two
ligands; the segments Ala2031-Met2042 of PfRON2 and Phe-P5–
Met-P16 of R1 have five identical amino acids and two
conservative differences (Fig. 4B). R1-major residue Arg-P15
contributes the most contacts to PfAMA1 and is positioned within
the same pocket of PfAMA1 as PfRON2 Arg2041 (Fig. 4A - box 3)
where it maintains six of the seven hydrogen bonds observed for
PfAMA1-PfRON2sp. Interestingly, while PfRON2 mimicry is
observed in the cystine loop-binding region (Phe2038/Phe-P12 to
Arg2041/Arg-P15), R1-major establishes clear anchor points in
the hydrophobic groove different from PfRON2; Phe-P2 and Phe-
P5 brace the peptide N-terminus in the region exposed by
displacement of the DII loop, with Phe-P5 occupying the pocket
left vacant by Phe367 of PfAMA1 (Fig. 4A - box 1).
PfAMA1 Polymorphisms at positions 175 and 225 are
determinant for the 3D7 specificity of R1
R1 is strain specific, binding to PfAMA1 from the 3D7 (cognate
antigen) and D10 strains, but with much reduced affinity to the
HB3 or W2mef proteins, as determined by ELISA [14] or SPR
[13] measurements (recapitulated in Table S2). In contrast,
PfRON2sp1 bound to all the PfAMA1 proteins tested (Table 1)
with a higher affinity than for R1 peptide. Consistent with these
values, PfRON2sp1 displayed a higher capacity to inhibit red cell
invasion by P. falciparum 3D7 than the R1 peptide (Fig. 5).
Moreover, PfRON2sp1 shows cross-strain inhibition of invasion as
expected from its biological function (Table 1), contrasting with
the more restricted strain specificity of R1 (Fig. 5, Table S2) [14].
The PfAMA1 3D7-R1 crystal structure shows that three
polymorphic residues (175, 224 and 225) contact R1-major (Table
S2). The 224 polymorphism, Met/Leu, is conservative and since
contacts are formed by the main chain only, this should not affect
R1 specificity. The 3D7 and D10 antigens both carry Tyr175 and
Ile225; for the W2mef and HB3 antigens, residue 175 is Tyr and
Asp, respectively, and residue 225 is Asn in both. Thus,
polymorphisms at positions 225 and possibly 175 appear to be
determinant for the 3D7 specificity of R1 at the major peptide-
binding site (Table S2A). R1-minor contacts polymorphic residue
230, which is Lys in all strains studied (Table S2B). As our data
suggest a weak affinity for this binding site, however, it is unlikely
that this polymorphism has a significant effect on the specificity for
R1. We examined these polymorphisms further using the mutant
PfAMA1 Dico3 [23], which differs only at residue 175 for the
3D7-contacting residues (Table S2A), and a 3D7 mutant with the
substitution Ile225Asp, which we call 3D7mut. The equilibrium
KD, determined from the SPR steady-state responses to R1
binding, was 15.261.9 mM for 3D7mut and 22.363.3 mM for
Dico3, showing a reduction in affinity of over 200-fold with respect
to the native 3D7 antigen (Fig. 6, Table S2C). This affinity is
comparable to that observed for HB3 and W2mef [13] (recapit-
ulated in Table S2), and confirms that both Tyr175 and Ile225 are
important for the strain-specific recognition of R1. Tyr175,
located at the tip of a flexible DI loop that is solvent-exposed in
the apo antigen [18], becomes buried by R1-major and forms a
hydrogen bond to this ligand via the phenol group. Ile225 is also
buried by R1-major, forming a pair of hydrogen bond via its main
chain to the R1-major main chain.
Hot spots driving specific PfAMA1-PfRON2 complex
formation
Guided by the similarities between the PfRON2sp and R1 co-
structures, and the conservation of key contact residues (Fig. 7A),
we probed the functional importance of a subset of PfRON2
residues by testing the binding to BHK-21 cells expressing
PfAMA1 of GST-PfRON2-5 fusion proteins carrying single
alanine mutations at: Pro2033 (aligns structurally with Pro7 of
peptide R1, which was shown to be critical for binding [24]),
Phe2038 (interacts with invariant residue Phe183 in the hydro-
phobic groove and aligns structurally with Phe12 of R1), Arg2041
(extensive contacts with PfAMA1 and structurally equivalent to
Arg-P15 of R1) and Pro2044 (the peptide bond Ser2043-Pro2044
Table 1. Apparent equilibrium dissociation constants KD(nM)
for the binding of peptides PfRON2sp1 and PfRON2sp2 to
AMA1 from different strains of P. falciparum.
Strain PfRON2sp1PfRON2sp2
3D720.366.3520674
CAMP14.663.8 165642
FVO 9.263.080615
HB318.364.6 6806180
Independent experiments were performed at least three times and the values
represent the mean 6 SD.
doi:10.1371/journal.ppat.1002755.t001
Structural Insights into the Moving Junction
PLoS Pathogens | www.plospathogens.org4June 2012 | Volume 8 | Issue 6 | e1002755

Page 5

is cis and is thus important for the b-hairpin conformation).
Consistent with the structure, mutation of Arg2041 to Ala
abrogated binding to PfAMA1 (Fig. 7B). Similar effects were
observed with Pro2044, Phe2038 and Pro2033 mutations, the
latter also shown to be a key residue in the TgAMA1-TgRON2
interaction [22].
Similarly, a subset of key PfAMA1 residues was also chosen for
mutation: Phe183 (an invariant residue that contributes to the
hydrophobic groove and that interacts with Phe2038 of PfRON2 via
aromatic interactions), Asn223 (which makes important polar interac-
tions with PfRON2), residue 225 (a polymorphic residue that
contributes many contacts to PfRON2 in the structure both the
CAMP (Asn225) and 3D7 (Ile225) complexes), Tyr234 (which makes
polar contacts to Arg2041 of PfRON2) and Tyr251 (which has been
suggested by previous studies to be important [12,25]). A clear role for
Phe183 in the PfAMA1-PfRON2 complex formation was evident
when expressed on the surface of BHK-21 cells and tested for their
ability to bind GST-PfRON2-5 fusion protein (Fig. 7C). A less
pronounced role of Tyr234 was observed and none for the remaining
residues, including Tyr251. Although these conclusions differ from
those of others [12,25], these results are consistent with the limited
contacts shown by this residue in the structures and with our earlier
findings on the TgAMA1-TgRON2 interaction, where the equivalent
TgAMA1 residue, Tyr230, had a minimal effect on the binding.
Table 2. Crystallographic parameters, data collection statistics and refinement summary.
PfAMA1 3D7-PfRON2sp1PfAMA1 CAMP -PfRON2sp2PfAMA1 3D7-R1
Spacegroup P21
P21
P212121
a, b, c (A˚) 70.15, 38.26, 70.7570.72, 38.14, 72.08 38.32, 144.32, 145.64
a, b, c (deg.)90, 99.73, 90 90, 97.72, 90 90, 90, 90
Wavelength (A˚)0.97950.9537 0.9791
Resolution range (A˚)45.41-2.1046.97-1.60 40.28-2.15
(2.21-2.10)(1.69-1.60) (2.25-2.15)
Measured reflections109520 153050156625
Unique reflections2204148207 42798
Redundancy5.0 (5.0)3.2 (3.2) 3.7 (2.5)
Completeness (%)100.0 (100.0) 94.9 (92.8) 95.3 (75.7)
I/s(i) 8.7 (3.2)12.7 (1.7) 13.3 (2.2)
Rmerge
0.140 (0.470)0.056 (0.618)0.075 (0.485)
Values in parenthesis are for the last resolution shell.
doi:10.1371/journal.ppat.1002755.t002
Table 3. Refinement statistics.
PfAMA1 3D7-PfRON2sp1PfAMA1 CAMP-PfRON2sp2PfAMA1 3D7-R1
Resolution (A˚)34.87–2.10 (2.15-2.10) 35.04-1.60 (1.64-1.60) 37.06-2.15 (2.15-2.21)
Rcryst/Rfree
0.164/0.201 (0.202/0.241) 0.176/0.195 (0.230/0.247)0.171/0.214 (0.215/0.249)
No. of atoms
Protein A/B/C/D/E/F 2377/2592309/1902375/2385/157/60/135/77
Solvent226265450
Glycerol30N/AN/A
B-values (A˚2)
Protein A/B/C/D/E/F17.3/29.9 27.5/48.336.4/40.5/50.6/77.8/61.5/92.6
Solvent28.7 37.646.1
Glycerol39.1N/A N/A
r.m.s. deviation from ideality
Bond lengths (A˚) 0.0150.0100.010
Bond angles (deg.)1.52 1.051.10
Ramachandran statistics
Most favoured 97.6%96.7%96.3%
Allowed2.4% 3.3%3.7%
Disallowed0.0% 0.0%0.0%
doi:10.1371/journal.ppat.1002755.t003
Structural Insights into the Moving Junction
PLoS Pathogens | www.plospathogens.org5June 2012 | Volume 8 | Issue 6 | e1002755

Page 6

Discussion
The structure of PfAMA1 in complex with the extracellular
region of its receptor PfRON2 and the accompanying functional
analysis reveal atomic details of the interaction between two key
partners at the MJ. The binding site on PfAMA1 includes the
hydrophobic groove and a region that becomes exposed by
displacement of the flexible DII loop from its apo conformation.
Comparison of residues from both components at the PfAMA1-
PfRON2 interface with those of other apicomplexan homologs
underscores the separate co-evolution of the receptor-ligand pair
in members of the phylum.
The DII loop displays a strong propensity for mobility in P.
falciparum [16,18] and P. vivax AMA1 structures [19], particularly
at its N- and C-terminal extremities (weak or absent electron
density); the central region of the DII loop is more structured and
Figure 2. Structure of PfAMA1 complexed with PfRON2-derived peptides. (A) Top - Co-crystal structures of PfAMA1 (blue surface) with
PfRON2sp1 (orange) and PfRON2sp2 (grey), show a disulfide-anchored U-shaped conformation in the apical groove of PfAMA1. Bottom - Electron
density map (orange) for PfRON2sp1 contoured at 1.0 s, highlighting well ordered density from the N-terminal helix, through the cystine loop, to the
C-terminal coil. (B) Notable changes in the structure of PfAMA1 between the apo structure (green; PDB ID 1Z40) and the PfAMA1-PfRON2sp1 co-
structure (blue-orange) as observed from a side view. Box 1 - The DII loop of apo PfAMA1 is ejected from the apical groove during binding to
PfRON2sp1, leaving room for the PfRON2sp1 N-terminal helix to occupy the space vacated by the DII loop helix. Box 2 - The b-strands of the
PfRON2sp1 cystine loop order a PfAMA1 surface loop, generating a contiguous three-stranded b-sheet. (C) In the region of the PfRON2sp1 N-terminal
helix, there is notable structural mimicry to the PfAMA1 apo DII loop, including several conserved residues, and a conserved hydrogen bonding
network incorporating three buried water molecules. (D) Arg2041, specific to P. falciparum, fits snugly into a deep pocket in the surface of PfAMA1
and is stabilized through a complex network of seven hydrogen bonds.
doi:10.1371/journal.ppat.1002755.g002
Structural Insights into the Moving Junction
PLoS Pathogens | www.plospathogens.org6June 2012 | Volume 8 | Issue 6 | e1002755

Page 7

stabilized by contacts with DI, and is better defined in some of
these AMA1 structures. Here, we show that the DII loop is
displaced by PfRON2sp, as well as by the R1 peptide. In T. gondii,
the DII loop is 14 residues shorter than in the Plasmodium orthologs
and appears less mobile [20] but nonetheless is readily displaced
by TgRON2sp [22]. Flexibility may therefore have an important
functional role: it protects a significant portion of the binding site
in apo AMA1 against the host’s immune response but can be
readily displaced to extend the hydrophobic groove for effective
binding to RON2. The anti-PfAMA1 invasion-inhibitory mono-
clonal antibody 4G2, which binds to the N- and C-termini of the
DII loop [19], probably prevents its displacement for effective
binding to PfRON2. The absence of polymorphisms in the DII
loop in spite of immune targeting of this region underlines its
important functional role [21].
We have previously demonstrated an evolutionary constraint on
the AMA1–RON2 interaction within apicomplexan parasites [11].
Our functional analysis of the TgAMA1-TgRON2sp co-structure
suggested that the cystine loop initially anchors the receptor to the
hydrophobic groove, causing expulsion of the DII loop to promote
interaction throughout the entire binding site [22]. Comparison of
the TgAMA1-TgRON2sp and PfAMA1-PfRON2sp co-structures
reveals that the cystine loop, while conserved across the two
genera, is the most divergent region within the RON2 (Fig. 8). The
separate co-evolution of the AMA1-RON2 pair in Apicomplexa is
clearly illustrated by the difference between the cystine loop
conformations of PfRON2sp and TgRON2sp. In particular, this
allows Arg2041 to access the specific PfAMA1 pocket (Fig. 8),
where it participates in an intricate network of polar interactions.
From mutagenesis, we have demonstrated a crucial role of
Figure 3. Structure of PfAMA1 complexed with R1 peptide. (A). The co-crystal structure of PfAMA1 (blue surface) with R1 reveals two bound
peptides, R1 major (yellow) and R1 minor (purple). (B). Detailed analysis of interactions at the PfAMA1–R1-major, PfAMA1–R1-minor, and R1-major–
R1-minor interfaces. Surface representation of PfAMA1 (blue), with R1-major (yellow) and R1-minor (purple) shown as cartoons. Box 1 – R1-major
anchors its N-terminus to PfAMA1 through 3 backbone hydrogen bonds. Box 2 – the central region of the PfAMA1 apical groove is occupied by R1-
major through both hydrophobic and polar interactions. Box 3 – R1-minor forms most of its anchor points to PfAMA1 through the apical loops and
does not contact the base of the groove, which is occupied by R1-major. Panel 4 – Backbone hydrogen bonds between R1-minor and R1-major
generate a b-sheet, while R1-major is further pinned to the PfAMA1 groove through 3 hydrogen bonds. Panel 5 – R1-major integrates into PfAMA1
with the use of an arginine knob-in-hole interaction stabilized by 6 hydrogen bonds, which is also exploited byPfRON2sp.
doi:10.1371/journal.ppat.1002755.g003
Structural Insights into the Moving Junction
PLoS Pathogens | www.plospathogens.org7 June 2012 | Volume 8 | Issue 6 | e1002755

Page 8

Arg2041 in complex formation (Fig. 7B). Moreover, this region of
the cystine loop also appears to play an influential role in species
selectivity as superposition of PvAMA1 structure [19] onto
PfAMA1-PfRON2sp shows that Arg2041 would be sterically
hindered at the interface but Thr, the equivalent residue in
PvRON2 from P. vivax, can be accommodated (Fig. 9A). This
accounts for our prior observation that the original 67-residue
segment of PfRON2 does not bind to PvAMA1 [11].
An additional feature of the PfRON2sp cystine loop region is
the presence of a cis peptide bond between Ser2043 and Pro2044;
the Ser-Pro-Pro segment contributes negligible buried surface area
but is important for maintaining the b-hairpin conformation for
efficient complex formation. Sequence alignment reveals that the
Pro duo (Pro2044–Pro2045) is preserved in all analyzed Plasmodium
species (Fig. 8A) and is thus likely important for specific
recognition of AMA1. We propose that it provides necessary
internal structure at the tip of the cystine loop and places the
disulfide bond in the proper orientation to brace the AMA1-
RON2 interaction. The influential role of Pro2044 is confirmed by
mutagenesis where substitution with Ala, which would disfavor the
cis peptide bond, abrogates PfAMA1-PfRON2 binding (Fig. 7B).
While T. gondii does not share the conserved proline pair, its
cystine loop is two residues shorter (Fig. 8A), which mirrors the
narrower groove of TgAMA1. Altogether, the overall U-shape
Figure 4. Structural mimicry of PfRON2 by peptide R1 in binding to PfAMA1. (A) Top (left) and end-on (right) views of PfAMA1-PfRON2sp1
(orange cartoon) overlayed on PfAMA1-R1-major (yellow)/R1-minor (purple), show that the PfAMA1 groove is capable of accepting only PfRON2sp1
or the two R1 peptides at one time. Box 1 shows that Phe-P5 of R1 mimics Phe367 of the DII loop, while boxes 2 and 3 highlight spatial conservation
of a phenylalanine anchor at the center of the groove, and a knob-in-hole interaction incorporating the peptide Arg-P15. R1-major is shown in yellow,
PfRON2sp1 in orange and apo PfAMA1 in green. (B). Comparison of the R1 and PfRON2sp1 sequences reveals five identical (red) and two similar (blue)
residues.
doi:10.1371/journal.ppat.1002755.g004
Structural Insights into the Moving Junction
PLoS Pathogens | www.plospathogens.org8 June 2012 | Volume 8 | Issue 6 | e1002755

Page 9

architecture of RON2 in complex with AMA1 appears to be
remarkably well maintained within apicomplexan parasites but
specific features are clearly visible in the cystine loop of PfRON2
and TgRON2, highlighting how a receptor-ligand complex has
evolved to maintain a common and crucial event in the biology of
these parasites.
Although the PfAMA1-PfRON2 interface is highly conserved,
five polymorphic residues of PfAMA1 contact the non-polymor-
phic PfRON2sp [26]. Of these, however, only residue 225 (Asn/
Ile) varies significantly. The remaining polymorphisms should not
affect binding as they involve main chain contacts only (residues
172, 174, 187 and 224). Our study allows a detailed structural
assessment of polymorphism at residue 225 since complexes with
PfAMA1 from the 3D7 (Ile225) and CAMP (Asn225) strains were
determined. The 3D7 and CAMP orthologs both maintain two
hydrogen bonds between the main chain of residue 225 and
PfRON2 Thr2039. However, Ile225 presents a deep pocket to
Arg2041 with apolar contacts formed between the aliphatic
regions of these two side chains, while Asn225 presents a shallower
pocket to Arg2041 with the Asn225 amide group stacking against
the guanidyl group. Nonetheless, our binding studies by SPR show
no significant difference in the affinity of these two PfAMA1
homologs for PfRON2sp2. Sequence variations at PfRON2-
interacting positions, 172(Glu/Gly), 187(Glu/Asn) and 225 (Ile/
Asn) are represented by the strains 3D7, CAMP, FVO and HB3
that we have analyzed by SPR; the very similar KDconstants,
ranging from approximately 10 to 20 nM, confirm that these exert
little effect in the strength of the interaction.
Peptide R1 shows a more restricted specificity as it binds
strongly to the cognate 3D7 and closely related D10 antigens but
only weakly to orthologs that do not carry the same polymorphic
amino acids at position 175 or 225 (Table S2). Tyr175 in PfAMA1
3D7 makes a hydrogen bond to the main chain of R1-major but,
as this residue is located in a flexible loop with some freedom to
adapt to the PfAMA1-R1 interface, it is unclear why the Asp175
polymorphism leads to reduced affinity. In the case of Ile225 of
PfAMA1 3D7, the main chain forms two hydrogen bonds to the
main chain of R1-major but the preference of R1 for the Ile225
polymorphism remains unexplained as it contrasts with PfRON2sp
where main chain hydrogen bonds are also formed by both Ile225
(3D7) and Asn225 (CAMP) to the main chain of PfRON2. This
emphasizes that specificity differences may present subtleties that
are difficult to decipher. Here, the crystal structure of R1 in
complex with the 3D7mut (Ile225Asn) and Dico3 (Tyr175Asp)
mutants of PfAMA1 would provide invaluable insights into this
question. Taken together, these results highlight that unlike the
natural ligand PfRON2, R1, which was selected by phage display,
is highly susceptible to polymorphisms.
R1 exhibits a close structural similarity to PfRON2, with the
major/minor peptide pair displaying a similar boomerang form as
PfRON2, binding to the same region of PfAMA1 and following the
same generalcontourofthe binding-site groove.Ourstructuraldata
show that binding of R1-minor is dependent upon prior binding of
R1-major as it lies above the latter in the binding groove and makes
fewer contacts to PfAMA1. This, indeed, is consistent with the ITC
Figure 5. Highly potent cross-strain inhibition of red blood cell
invasion of PfRON2sp1. Comparison of PfRON2sp1 and R1 peptides
(concentrations 0.2 to 20 mM) in inhibiting red blood cell invasion by P.
falciparum 3D7 or HB3 highlights the higher inhibitory efficiency and
cross-strain reactivity of PfRON2sp1. Parasitemia of control infected red
blood cells (IRBC) 16 hours post-invasion was used as the 100%
invasion reference. Means (6 SD for N=3) are shown.
doi:10.1371/journal.ppat.1002755.g005
Figure 6. Surface Plasmon Resonance studies of peptide R1
binding to PfAMA1 mutants 3D7mut and Dico3. (A). Left -
sensorgrams, showing R1 (analyte) binding to PfAMA1 3D7mut
(immobilized). R1 concentrations are indicated for each curve (mM).
Right - the variation in percentage of bound sites (deduced from the
steady-state response) with respect to analyte concentration. (B). Left -
sensorgrams, showing R1 (analyte) binding to Dico3 (immobilized), with
R1 concentrations indicated. Right - the variation in percentage of
bound sites (deduced from the steady-state response) with respect to
analyte concentration. The equilibrium dissociation constant KDderived
from the steady state binding curves is 15.2 mM for 3D7mut and
22.3 mM for Dico3.
doi:10.1371/journal.ppat.1002755.g006
Structural Insights into the Moving Junction
PLoS Pathogens | www.plospathogens.org9 June 2012 | Volume 8 | Issue 6 | e1002755

Page 10

measurements that show a stoichiometry of 1:1, indicating a weaker
affinity for the minor peptide-binding site. R1-major is thus favored
as the principle inhibitor of the interaction with PfRON2, but this
does not preclude a contribution by the minor peptide-binding site
at high peptide concentrations.
Therapeutic strategies aimed at inhibiting the interaction
between PfAMA1 and PfRON2 should be very effective in
treating malaria as they address a critical phase in the life cycle of
the parasite and, importantly, should not be compromised by
polymorphism since the PfAMA1-PfRON2 interface is highly
conserved. Our results provide a structural basis for designing
inhibitors against the most virulent malaria parasite. The
PfRON2sp1 peptide used in this study has a very high affinity to
PfAMA1 and is very efficient at inhibiting invasion. Moreover, in
contrast to the less strongly binding peptide R1, PfRON2sp1 is not
strain specific. Structural details of the PfAMA1-PfRON2 inter-
action offer the possibility to design molecules with the desired
specific inhibitory properties by in silico screening and structural
validation. The binding of PfRON2 Arg2041 to a specific pocket
on PfAMA1 could be a critical target region. Indeed, the
Figure 7. Mutations of PfAMA1 and PfRON2-5 reveal residues critical for high affinity interaction. (A) Interface between PfAMA1 and
PfRON2sp1 shown in open-book presentation. Residues of both components that were mutated are labeled. (B). Binding characteristics of
recombinant GST-PfRON2-5 mutants to dissect hot-spot residues in PfRON2. PfAMA1-expressing BHK-21 cells were incubated with 10 mg/ml of
PfRON2 or mutated proteins (GST-fusion proteins), washed and the binding of recombinant PfRON2 fragment was revealed with anti-GST antibody.
PfAMA1 was detected with mAb F8.12.19, which recognizes extracellular Domain III. (C). Binding consequences of PfAMA1 mutations. Mutated
versions of PfAMA1 were expressed on the surface of BHK-21 cells and incubated with wild-type PfRON2 recombinant proteins at 10 and 1 mg/ml.
doi:10.1371/journal.ppat.1002755.g007
Structural Insights into the Moving Junction
PLoS Pathogens | www.plospathogens.org 10 June 2012 | Volume 8 | Issue 6 | e1002755

Page 11

important role played by Arg-P15 at the PfAMA1-R1 interface
closely mirrors the equivalent interaction in the PfAMA1-
PfRON2sp complexes and, interestingly, the same pocket is
occupied by Arg and Lys in PfAMA1 complexes with the invasion
inhibitory antibodies IgNAR [17] and 1F9 [16], respectively
(Fig. 9B). Phe2038 (corresponding to Phe-P12 in R1) is also a key
residue, as its substitution by Ala affected binding. The importance
of this sub-site is further highlighted by the concomitant loss in
affinity when Phe183 (with which it interacts) was mutated in
PfAMA1. Collectively, these data provide a firm basis for
designing molecules with optimal inhibitory properties to treat
malarial infection.
Materials and Methods
Recombinant protein production
(i) Baculovirus insect cell expression: A synthetic codon-
optimized gene encoding DI and DII of PfAMA1 3D7 [27]
(residues 104–438; numbering based on the initiation methionine,
PF11_0344) (GenScript)was subclonedintoamodified
pAcGP67B vector (Pharmingen) for expression in insect cells
using established protocols [20]. Final yield of recombinant
protein was approximately 3 mg per L of culture.
(ii) P. pastoris expression: Synthetic genes were optimized for
PfAMA1 coding of residues 97–442, from strains 3D7 (Genbank
accession number U33274), CAMP (accession number M34552)
and HB3 (accession number U33277). Potential N-glycosylation
sites were mutated and genes were cloned EcoRI-KpnI in the
pPicZalpha A vector (Invitrogen), resulting in an 11-residues
sequence extension followed by myc-epitope and hexa-His tags at
the C-terminus), expressed in P. pastoris, and purified as described
[28]. Yield after purification was approximately 20 mg per L of
culture. PfAMA1 FVO (residues 25–545, no tags, accession
number AJ277646) was produced as described before [29]. The
DiCo3 protein was modified compared to the published protein
[23]; it includes the PfAMA1 FVO prodomain (amino acids 25–
96) and one additional mutation to minimize proteolytic cleavage
Lys376–.Arg (B. Faber, unpublished results). The PfAMA1
3D7mut (Ile225–.Asn, residues 25–545, no tags) mutant was
generated by site-directed mutagenesis (Genscript) and produced
in P. pastoris in a similar fashion to the native protein [29].
Peptide synthesis
A 39-residue peptide corresponding to residues 2021 to 2059 of
PfRON2 (PfRON2sp1) was synthesized by Kinexus (Vancouver,
Canada) and disulfide cyclized. Lyophilized PfRON2sp1 was
solubilized in 100% DMSO and subsequently diluted in HBS
(20 mM HEPES pH 7.5, 150 mM NaCl) for use in co-crystalli-
zation and functional studies. Peptides PfRON2sp2 (residues 2027
to 2054) and R1 were synthesized by PolyPeptide (Strasbourg,
France) and solubilized in 3.5% DMSO for subsequent use.
Crystallization and data collection
Crystals of PfAMA1 3D7 PfRON2sp1 were grown in 30%
PEG400, 100 mM Tris-HCl pH 8.5, 200 mM tri-sodium citrate
dihydrate and the protein (5 mg/mL final concentration)
incubated with PfRON2sp1 (1:2 molar excess). A crystal in
cryoprotectant buffer was flash cooled at 100 K and diffraction
data were collected on beamline 9-2 at SSRL (Stanford
Synchrotron Radiation Laboratory, Stanford, US). Crystals of
PfAMA1 CAMP PfRON2sp2 were obtained in 20% PEG 4000,
0.1 M Tris/HCl pH 8.6, 0.1 M sodium acetate and 20%
isopropanol and the protein (6.4 mg/mL final concentration)
incubated with PfRON2sp2 (1:5 molar excess). Diffraction data
were collected from a crystal in cryoprotectant buffer at 100 K on
beamline ID29 at European Synchrotron Radiation Facility
(Grenoble, France). Crystals of PfAMA1 3D7 R1 were obtained
in 15% PEG 4000, 0.1 M Tris/HCl pH 8.5, 0.1 M sodium
acetate and 10% isopropanol and the protein (5.4 mg/mL final
concentration) incubated with R1 (1:6 molar excess). Diffraction
data were collected at 100 K on beamline PROXIMA 1 at
SOLEIL (St. Aubin, France).
Data processing, structure solution and refinement
Diffraction data were processed using Imosflm [30] or XDS
[31] and Scala [32] in the CCP4 suite of programs [33].
Crystallographic parameters and data collection statistics are
given in Table 2. Initial phases were obtained by molecular
replacement using PHASER [34] or AMoRe [35] with the
unliganded PfAMA1 structure (PDB 1Z40). Tracing of the
PfRON2 and R1 peptides, and addition of solvent molecules,
was performed manually in COOT [36] and refinement was
performed with Refmac5 [37] or autoBUSTER (Global Phasing
Ltd, Cambridge, UK). A summary of refinement statistics is given
Figure 8. The RON2 cystine loop governs specificity. (A).
Alignment of RON2 sequences truncated to correlate PfRON2sp1 with
RON2 sequences from the following accession numbers: TgRON2 -
TGME49_100100, NcRON2 - NCLIV_064620, PfRON2 - PF14_0495,
PvRON2 - PVX_117880, PyRON2 – PY_06813, BbRON2 (BBOV_I001630).
(B). Overlay of TgRON2sp (green; PDB ID 2Y8T) onto PfAMA1-PfRON2sp
(blue-orange) shows that both peptides adopt a helix/coil/cystine loop/
coil architecture in the AMA1 groove, with the highest divergence
localized to the cystine loop (black arrow). (C). Electrostatic surface
renderings of PfAMA1 (left) and TgAMA1 (right), with the secondary
structure of the RON2 binding partner and residues defining the base of
cystine loop shown, illustrates that both interactions are highly
complementary, but highly genus specific.
doi:10.1371/journal.ppat.1002755.g008
Structural Insights into the Moving Junction
PLoS Pathogens | www.plospathogens.org11June 2012 | Volume 8 | Issue 6 | e1002755

Page 12

in Table 3. All molecular representation figures were generated in
the PyMOL Molecular Graphics System, version 1.2r3pre,
Schro ¨dinger, LLC. Coordinates and structure factors have been
deposited in the Protein Data Bank with the following entry codes:
PfAMA1-PfRON2sp1,3ZWZ;
PfAMA1-R1, 3SRJ.
PfAMA1-PfRON2sp2,3SRI;
Binding studies by SPR
SPR measurements were made with a Biacore 2000 instrument
(Biacore AB). AMA1 proteins diluted in 10 mM sodium acetate
pH 4.5 for 3D7, CAMP, HB3 and FVO strains, or pH 4.0 for
3D7mut and Dico3, were covalently immobilized by an amine-
coupling procedure on CM5 sensor chips (GE Healthcare). The
reference flow cell was prepared by the same procedure in absence
of protein. Binding assays were performed at 25uC in PBS and
0.005% Tween 20 by injecting a series of peptide (PfRON2sp1
and PfRON2sp2 on 3D7, CAMP, HB3 and FVO, and R1 on
3D7mut and Dico3) concentrations at a constant flow rate of
5 mL/min. A heterologous peptide was used to verify the absence
of non-specific binding. Peptide dissociation was realized by
injecting the running buffer, and the surface was regenerated by
injecting glycine/HCl pH 1.5 followed by SDS 0.05%. Control
flow cell sensorgrams were subtracted from the ligand flow cell
sensorgrams and averaged buffer injections were subtracted from
analyte sensorgrams. For peptide R1, steady-state signals (Req)
were obtained directly from the plateau region of the sensorgrams,
while for PfRON2sp peptides, estimated values of Req were
obtained by extrapolation from the experimental curves since the
association phase did not reach a final equilibrium state. All
calculations were made using the BIAevaluation 4.2 software
(BIAcore AB). The saturation curves obtained by plotting Req
versus the peptide concentration were fitted with a steady-state
Figure 9. The Arg knob-in-hole interaction is critical for species selectivity and interaction with invasion inhibitory antibodies and
peptides. (A). Left - A cut-away surface of PfAMA1 (blue), reveals that Arg2041 of PfRON2sp1 (orange) integrates deeply into a well-defined pocket.
Right - However, no analogous pocket is observed in PvAMA1 (grey; PDB ID 1W8K). (B). Peptides and antibodies known to be invasion inhibitory for P.
falciparum occupy the key Arg binding site, as shown by orthogonal views of the PfAMA1-PfRON2sp1 co-structure (blue-orange) overlayed with the
mAb 1F9 co-structure (1F9, green; PDB ID 2Q8B), IgNAR14l-1 co-structure (IgNAR, purple; PDB ID 2Z8V), and R1 co-structure (R1, yellow; reported
here).
doi:10.1371/journal.ppat.1002755.g009
Structural Insights into the Moving Junction
PLoS Pathogens | www.plospathogens.org 12June 2012 | Volume 8 | Issue 6 | e1002755

Page 13

model to obtain the Rmax and the apparent equilibrium
dissociation constants, KD. To normalize the response for the
different ligands, these curves were reported as the percentage of
bound sites (ratio Req/Rmax) versus the analyte concentration..
Isothermal calorimetry.
ITC measurements were made
using a ITC200calorimeter (MicroCal). PfAMA1 3D7 (P. pastoris)
and peptide R1 were diluted in PBS to final concentrations of
0.6 mM and 55 mM, respectively. PfAMA1 3D7 (initial volume
200 mL) was titrated at 25uC by consecutive injections of the
peptide R1 (2 mL aliquots at 3 min intervals). Raw data were
normalized and corrected for the heat of dilution of R1 in PBS.
Binding stoichiometry was determined by fitting the final data to a
1:1 interaction model using the Origin7 software (OriginLab).
P. falciparum cultures and invasion assays
The P. falciparum cell cultures and the invasion assays were
performed as described previously [11]. Briefly, highly synchro-
nized P. falciparum 3D7 and HB3 schizonts (1.5% hematocrit, 1.5%
parasitemia) were incubated with R1 or PfRON2sp1 peptides.
Blood smears were collected 16 hours post-invasion and used for
ring-stage parasites counting. The results presented are represen-
tative of three independent experiments, each performed in
triplicate.
Transient transfection experiments and cell binding
assays
Cell binding assays using PfAMA1-expressing BHK-21 cells and
recombinant GST-PfRON2-5 fusion proteins were performed as
previously described [11]. Although not quantitative, this cell-
binding assay truly reflects the interaction between AMA1 and
RON2 as we carefully checked all the experimental steps as well as
the image recording as described below. Transfections were carried
out using Lipofectamine Reagent (Invitrogen) as instructed by the
manufacturer with 36105BHK-21 cells grown on coverslips for
24 h in 6 well plates. Cells were grown for an additional 24 h post-
transfection before subsequent analysis. Expression and correct
folding of PfAMA1 (and the mutants) at the host cell surface was
verified by IFA performed with or without permeabilisation, using
antibodies either specific to the cytoplasmic tail (anti-myc tag) or
specific to the extracellular ectodomain of PfAMA1 (mouse mAb
F8.12.19 [38]). For binding assays, coverslips from a same
transfection experiment were washed in HBSS (Invitrogen) before
addition of recombinant PfRON2-5 wild type or mutants diluted in
HBSS at 10, 1 or 0.1 mg/ml. Coverslips incubated with GST were
systematically used as a control. After five washes in PBS to remove
unbound protein, cells were fixed in 4% PAF and further processed
for IFA as described above [11]. The binding characteristics of
RON2 (anti-GST labelling) on the PfAMA1 mutant were only
considered valid when its signal was identical to that of wild type
PfAMA1. All other micrographs were obtained with a Zeiss
Axiophot microscope equipped for epifluorescence. Adobe photo-
shop (Adobe Systems, Mountain View, CA) was used for image
processing. Matching pairs of images were recorded with the same
exposure time and processed identically.
The PfAMA1 and GST-PfRON2 mutated constructs were
generated by site directed mutagenesis using Quickchange II XL
protocol (Stratagene).
Supporting Information
Figure S1
PfAMA1-PfRON2sp1 interface. (A). Open-book surface rep-
resentation of PfAMA1 (left) and PfRON2sp1 (right) showing the
extensive involvement of residues from both molecules in forming
a complex interface. Residues involved in hydrogen bonding are
coloured blue, while residues contributing significant buried
surface area (BSA.20 A˚2for PfAMA1, .5 A˚2for PfRON2sp1)
are colored green. (B). Table of residues involved in hydrogen
bonding at the PfAMA1- PfRON2sp1 interface (left) and residues
contributing significant buried surface area (right), as calculated by
PISA (http://www.ebi.ac.uk/msd-srv/prot_int/pistart.html).
Polymorphic residues of PfAMA1 are shown in blue.
(PPTX)
Detailed analysis of interactions at the
Figure S2
R1 binding to PfAMA1 3D7.
(PPTX)
Isothermal titration calorimetry of peptide
Table S1
the PfAMA1-R1 crystal structure. (A). Polar contacts between
PfAMA1 3D7 and R1-major (column 1), and buried surface areas
of individual residues of PfAMA1 3D7 (column 2) and R1-major
(column 3). Salt bridges are indicated in bold. (B). Polar contacts
between PfAMA1 3D7 and R1-minor (column 1), and buried
surface areas of individual residues of PfAMA1 3D7 (column 2)
and R1-minor (column 3). (C). Polar contacts between R1-major
and R1-minor (column 1), and buried surface areas of individual
residues of R1-major (column 2) and R1-minor (column 3).
Polymorphic residues of PfAMA1 are shown in blue.
(PPTX)
Polar interactions and buried surface areas in
Table S2
peptide R1. (A). Polymorphic residues contacting R1-major
showing the sequence for strains analyzed using ELISA (*) [14],
SPR (+) [13] and in this study using SPR (u). (B). Polymorphic
residues contacting R1-minor, showing the sequence for strains as
presented in (A). (C). Binding to PfAMA1, classified as strong (s) or
weak (w) for the studies presented in (A) and (B).
(PPTX)
Polymorphic residues of PfAMA1 contacting
Table S3
(PPTX)
Primers used in this study.
Acknowledgments
We thank Roberto Rodriguez-Garcia and Marjolein van der Eijk for
technical assistance, Patrick England for helpful discussions and the staff of
the European Synchrotron Radiation Facility (Grenoble, France) and
SOLEIL (St. Aubin, France), for providing facilities for diffraction
measurements and for assistance.
Author Contributions
Conceived and designed the experiments: ML MJB GAB. Performed the
experiments: BVLN MLT MHL SL SH MR FAS MJB GAB. Analyzed
the data: BVLN ML MHL MLT MJB GAB. Contributed reagents/
materials/analysis tools: BWF. Wrote the paper: BVLN ML MLT MJB
GAB.
References
1. WHO (2010) Guidelines for the treatment of malaria. 2nd edition. Geneva:
World Health Organization Press.
2. Aikawa M, Miller LH, Johnson J, Rabbege J (1978) Erythrocyte entry by
malarial parasites. A moving junction between erythrocyte and parasite. J Cell
Biol 77: 72–82.
3. Baum J, Papenfuss AT, Baum B, Speed TP, Cowman AF (2006) Regulation
of apicomplexan actin-based motility. Nat Rev Microbiol 4: 621–
628.
4. Cowman AF, Crabb BS (2006) Invasion of red blood cells by malaria parasites.
Cell 124: 755–766.
Structural Insights into the Moving Junction
PLoS Pathogens | www.plospathogens.org 13 June 2012 | Volume 8 | Issue 6 | e1002755

Page 14

5. Riglar DT, Richard D, Wilson DW, Boyle MJ, Dekiwadia C, et al. (2011) Super-
resolution dissection of coordinated events during malaria parasite invasion of
the human erythrocyte. Cell Host Microbe 9: 9–20.
6. Waters AP, Thomas AW, Deans JA, Mitchell GH, Hudson DE, et al. (1990) A
merozoite receptor protein from Plasmodium knowlesi is highly conserved and
distributed throughout Plasmodium. J Biol Chem 265: 17974–17979.
7. Gaffar FR, Yatsuda AP, Franssen FF, de Vries E (2004) Erythrocyte invasion by
Babesia bovis merozoites is inhibited by polyclonal antisera directed against
peptides derived from a homologue of Plasmodium falciparum apical membrane
antigen 1. Infect Immun 72: 2947–2955.
8. Hehl AB, Lekutis C, Grigg ME, Bradley PJ, Dubremetz JF, et al. (2000)
Toxoplasma gondii homologue of plasmodium apical membrane antigen 1 is
involved in invasion of host cells. Infect Immun 68: 7078–7086.
9. Besteiro S, Michelin A, Poncet J, Dubremetz JF, Lebrun M (2009) Export of a
Toxoplasma gondii rhoptry neck protein complex at the host cell membrane to
form the moving junction during invasion. PLoS Pathog 5: e1000309.
10. Tyler JS, Boothroyd JC (2011) The C-Terminus of Toxoplasma RON2 Provides
the Crucial Link between AMA1 and the Host-Associated Invasion Complex.
PLoS Pathog 7: e1001282.
11. Lamarque M, Besteiro S, Papoin J, Roques M, Vulliez-Le Normand B, et al.
(2011) The RON2-AMA1 interaction is a critical step in moving junction-
dependent invasion by apicomplexan parasites. PLoS Pathog 7: e1001276.
12. Srinivasan P, Beatty WL, Diouf A, Herrera R, Ambroggio X, et al. (2011)
Binding of Plasmodium merozoite proteins RON2 and AMA1 triggers
commitment to invasion. Proc Natl Acad Sci U S A 108: 13275–13280.
13. Harris KS, Casey JL, Coley AM, Karas JA, Sabo JK, et al. (2009) Rapid
optimization of a peptide inhibitor of malaria parasite invasion by comprehen-
sive N-methyl scanning. J Biol Chem 284: 9361–9371.
14. Harris KS, Casey JL, Coley AM, Masciantonio R, Sabo JK, et al. (2005) Binding
hot spot for invasion inhibitory molecules on Plasmodium falciparum apical
membrane antigen 1. Infect Immun 73: 6981–6989.
15. Richard D, MacRaild CA, Riglar DT, Chan JA, Foley M, et al. (2010)
Interaction between Plasmodium falciparum apical membrane antigen 1 and the
rhoptry neck protein complex defines a key step in the erythrocyte invasion
process of malaria parasites. J Biol Chem 285: 14815–14822.
16. Coley AM, Gupta A, Murphy VJ, Bai T, Kim H, et al. (2007) Structure of the
malaria antigen AMA1 in complex with a growth-inhibitory antibody. PLoS
Pathog 3: 1308–1319.
17. Henderson KA, Streltsov VA, Coley AM, Dolezal O, Hudson PJ, et al. (2007)
Structure of an IgNAR-AMA1 complex: targeting a conserved hydrophobic cleft
broadens malarial strain recognition. Structure 15: 1452–1466.
18. Bai T, Becker M, Gupta A, Strike P, Murphy VJ, et al. (2005) Structure of
AMA1 from Plasmodium falciparum reveals a clustering of polymorphisms that
surround a conserved hydrophobic pocket. Proc Natl Acad Sci U S A 102:
12736–12741.
19. Pizarro JC, Vulliez-Le Normand B, Chesne-Seck ML, Collins CR, Withers-
Martinez C, et al. (2005) Crystal structure of the malaria vaccine candidate
apical membrane antigen 1. Science 308: 408–411.
20. Crawford J, Tonkin ML, Grujic O, Boulanger MJ (2010) Structural
characterization of apical membrane antigen 1 (AMA1) from Toxoplasma gondii.
J Biol Chem 285: 15644–15652.
21. Chesne-Seck ML, Pizarro JC, Vulliez-Le Normand B, Collins CR, Blackman
MJ, et al. (2005) Structural comparison of apical membrane antigen 1
orthologues and paralogues in apicomplexan parasites. Mol Biochem Parasitol
144: 55–67.
22. Tonkin ML, Roques M, Lamarque MH, Pugniere M, Douguet D, et al. (2011)
Host cell invasion by apicomplexan parasites: insights from the co-structure of
AMA1 with a RON2 peptide. Science 333: 463–467.
23. Remarque EJ, Faber BW, Kocken CH, Thomas AW (2008) A diversity-covering
approach to immunization with Plasmodium falciparum apical membrane antigen 1
induces broader allelic recognition and growth inhibition responses in rabbits.
Infect Immun 76: 2660–2670.
24. Lee EF, Yao S, Sabo JK, Fairlie WD, Stevenson RA, et al. (2011) Peptide
inhibitors of the malaria surface protein, apical membrane antigen 1:
identification of key binding residues. Biopolymers 95: 354–364.
25. Collins CR, Withers-Martinez C, Hackett F, Blackman MJ (2009) An inhibitory
antibody blocks interactions between components of the malarial invasion
machinery. PLoS Pathog 5: e1000273.
26. Cao J, Kaneko O, Thongkukiatkul A, Tachibana M, Otsuki H, et al. (2009)
Rhoptry neck protein RON2 forms a complex with microneme protein AMA1
in Plasmodium falciparum merozoites. Parasitol Int 58: 29–35.
27. Gupta A, Bai T, Murphy V, Strike P, Anders RF, et al. (2005) Refolding,
purification, and crystallization of apical membrane antigen 1 from Plasmodium
falciparum. Protein Expr Purif 41: 186–198.
28. Kocken CH, Withers-Martinez C, Dubbeld MA, van der Wel A, Hackett F,
et al. (2002) High-level expression of the malaria blood-stage vaccine candidate
Plasmodium falciparum apical membrane antigen 1 and induction of antibodies that
inhibit erythrocyte invasion. Infect Immun 70: 4471–4476.
29. Faber BW, Remarque EJ, Kocken CH, Cheront P, Cingolani D, et al. (2008)
Production, quality control, stability and pharmacotoxicity of cGMP-produced
Plasmodium falciparum AMA1 FVO strain ectodomain expressed in Pichia
pastoris. Vaccine 26: 6143–6150.
30. Leslie A (1992) Recent changes to the MOSFLM package for processing film
and image plate data. . Joint CCP4 + ESF-EAMCB Newsletter on Protein
Crystallogr 26: 239858.
31. Kabsch W (2010) Xds. Acta Crystallogr D Biol Crystallogr 66: 125-132.
32. Evans P (2006) Scaling and assessment of data quality. Acta Crystallogr D Biol
Crystallogr 62: 72–82.
33. Winn MD, Ballard CC, Cowtan KD, Dodson EJ, Emsley P, et al. (2011)
Overview of the CCP4 suite and current developments. Acta Crystallogr D Biol
Crystallogr 67: 235–242.
34. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, et al.
(2007) Phaser crystallographic software. J Appl Crystallogr 40: 658–674.
35. Navaza J (2001) Implementation of molecular replacement in AMoRe. Acta
Crystallogr D Biol Crystallogr 57: 1367–1372.
36. Emsley P, Cowtan K (2004) Coot: model-building tools for molecular graphics.
Acta Crystallogr D Biol Crystallogr 60: 2126–2132.
37. Murshudov GN, Vagin AA, Dodson EJ (1997) Refinement of macromolecular
structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystal-
logr 53: 240–255.
38. Igonet S, Vulliez-Le Normand B, Faure G, Riottot MM, Kocken CH, et al.
(2007) Cross-reactivity studies of an anti-Plasmodium vivax apical membrane
antigen 1 monoclonal antibody: binding and structural characterisation. J Mol
Biol 366: 1523–1537.
Structural Insights into the Moving Junction
PLoS Pathogens | www.plospathogens.org 14June 2012 | Volume 8 | Issue 6 | e1002755