Host Cell Invasion by Apicomplexan Parasites: Insights from the Co-Structure of AMA1 with a RON2 Peptide

See allHide authors and affiliations

Science  22 Jul 2011:
Vol. 333, Issue 6041, pp. 463-467
DOI: 10.1126/science.1204988


Apicomplexan parasites such as Toxoplasma gondii and Plasmodium species actively invade host cells through a moving junction (MJ) complex assembled at the parasite–host cell interface. MJ assembly is initiated by injection of parasite rhoptry neck proteins (RONs) into the host cell, where RON2 spans the membrane and functions as a receptor for apical membrane antigen 1 (AMA1) on the parasite. We have determined the structure of TgAMA1 complexed with a RON2 peptide at 1.95 angstrom resolution. A stepwise assembly mechanism results in an extensive buried surface area, enabling the MJ complex to resist the mechanical forces encountered during host cell invasion. Besides providing insights into host cell invasion by apicomplexan parasites, the structure offers a basis for designing therapeutics targeting these global pathogens.

Apicomplexan parasites cause widespread disease, including malaria (Plasmodium) and toxoplasmosis (Toxoplasma). The success of these obligate intracellular pathogens is largely due to their mechanism of active invasion (1) involving the coordinated secretion of proteins from the microneme and rhoptry secretory organelles (2). In an intriguing process, the parasite provides both ligand and receptor to enable host cell invasion (3).

The molecular basis of active invasion by apicomplexans relies on secretion of a rhoptry neck RON) complex, consisting of at minimum RONs 2, 4, and 5, into the host cell. RON2 is integrated into the host plasma membrane, and RONs 4 and 5 are localized to its cytoplasmic face (4). A surface-exposed ectodomain of RON2 serves as the receptor (5, 6) for the micronemal-secreted apical membrane antigen 1 (AMA1) (710) displayed on the parasite cell surface. The assembled AMA1-RON complex forms a ringlike structure known as the moving junction (MJ), which migrates from the anterior to the posterior of the parasite (11), leading to internalization of the parasite into a parasitophorous vacuole (PV). In addition to linking the parasite and host cell membranes, the MJ acts as a diffusion barrier, excluding host transmembrane proteins from the PV (12) and thereby preventing degradation of intracellular parasites (13).

AMA1 and RONs 2, 4, and 5 are conserved across Apicomplexa (4, 1417), suggesting a broadly conserved host-cell invasion mechanism. In support of this prediction, we have recently shown that the AMA1-RON2 interaction is equally important for the invasive process by Toxoplasma gondii and Plasmodium falciparum (5). Structural studies of TgAMA1 (18), P. vivax AMA1 (19), and PfAMA1 (20) revealed a conserved hydrophobic trough in the ectodomain that displays structural hallmarks of a receptor binding surface and thus may be involved in coordinating the binding of RON2. Here, we report a structural and functional characterization of TgAMA1 in complex with a complementary TgRON2 peptide.

To refine the 65-residue region of TgRON2 capable of binding TgAMA1 (TgRON2-2) (5), we generated deletion constructs targeting both its N termini (TgRON2-2 D1) and C termini (TgRON2-2 D2) (Fig. 1A). A qualitative cellular binding assay showed that both deletion constructs were capable of binding TgAMA1 to the same extent (Fig. 1B); however, surface plasmon resonance (SPR) and invasion studies revealed participation of the N-terminal region of TgRON2-2, whereas the C-terminal region appeared to be dispensable (Fig. 1C) (21). On the basis of these observations, we synthesized a 37–amino acid peptide (TgRON2 synthetic peptide, TgRON2sp) corresponding to residues 1297 to 1333. Because a pair of conserved cysteines present in the central region may endow the peptide with crucial structure, TgRON2sp was disulfide-cyclized. TgRON2sp was active in binding TgAMA1 in cellular binding assays (Fig. 1B), host-cell invasion inhibition assays at nanomolar concentrations (Fig. 1D) [SPR determined equilibrium dissociation constant (Kd) of 1.42 × 10−8 M; fig. S1A], and competition enzyme-linked immunosorbent assays (ELISA) against TgRON2-2 (21) (fig. S1B).

Fig. 1

Defining the TgRON2 recognition sequence for TgAMA1. (A) TgRON2-2 deletion constructs; TMD, transmembrane domain (25). (B) Immunofluorescence microscopy images of TgRON2 constructs labeling TgAMA1 in micronemes of fixed and permeabilized intracellular parasites using an antibody against glutathione S-transferase (labeling of TgRON2-2) and an antibody against ectodomain B3.90 (labeling of TgAMA1). Scale bars indicate 5 μm; FITC, fluorescein isothiocyanate. (C) SPR analysis of TgRON2 constructs binding to TgAMA1 (top) presented as relative % binding. Immunofluorescence-based measurement showing the ability of TgRON2 constructs to inhibit T. gondii invasion of human foreskin fibroblast cells (bottom). Note the statistically significant effect observed relative to TgRON2-2. (D) TgRON2sp displays similar invasion inhibition properties to TgRON2-2. Student’s t test; **P < 0.01; ***P < 0.001; means  ±  SD for N = 3.

To structurally characterize the complex, we crystallized TgAMA1 bound to TgRON2sp and refined the structure to a resolution of 1.95 Å (Fig. 2A and table S1) (21). Clear electron density in the apical groove of TgAMA1 enabled accurate tracing of the entire TgRON2sp sequence (Fig. 2B). TgRON2sp bound in a U-shaped conformation, resulting in a total buried surface area of 3765 Å2 (Fig. 2B, fig. S2, and table S2) with significant shape complementarity [complexation significance score of 1.00 (22)]. These features likely endow the MJ with the ability to withstand the sheer mechanical force associated with the parasite moving through the constricted MJ ring.

Fig. 2

TgRON2sp induces conformational change in TgAMA1. (A) Overview of the essential TgAMA1-TgRON2 complex that links apicomplexan parasites and their target host cells. Proposed extracellular location of TgRON2 C terminus (denoted by c) (6). (B) Top view of TgRON2sp (green) bound in the TgAMA1 apical groove (purple). Electron density map of TgRON2sp shown in green and contoured at 1.0σ (center) reveals a well-ordered peptide that adopts a U-shaped conformation. (C) Top view of TgAMA1 showing the 29-Å displacement of the DII loop from its ordered position in the apo form (orange) to its TgRON2sp bound form (purple). To highlight displacement of the DII loop, TgRON2sp is not shown. (D) Electrostatic renderings of the TgAMA1 apical surface when the DII loop is ordered in the absence of TgRON2sp (left, dotted line highlights position of DII loop) and when it is displaced by TgRON2sp (middle, TgRON2sp not shown in order to visualize the newly exposed basic surface in the TgAMA1 apical groove). TgRON2sp is displayed in an open book view with respect to the apical groove of TgAMA1. The acidic surface on TgRON2sp electrostatically complements the basic patch in the reconfigured TgAMA1 groove (right).

Another feature of the TgAMA1-TgRON2sp co-structure is a substantial conformational change in the DII loop of AMA1 (Fig. 2C and fig. S3), an influential substructure in RON2 coordination (18, 19, 23, 24). In the apoliprotein (apo) TgAMA1 structure (18), the DII loop is intimately associated with the base of the hydrophobic groove through numerous interactions, including a pair of tryptophan residues that anchor the loop into hydrophobic pockets in the apical groove (Fig. 2C). By inducing the conformational changes leading to displacement of the DII loop, TgRON2sp is able to expose the fully functional binding surface on TgAMA1 that displays enhanced shape and charge complementarity (Fig. 2D). In its reconfigured form, the apical groove presents an extended basic patch that complements an acidic patch on the TgRON2sp N-terminal helix (Fig. 2D). These data suggest a stepwise binding mechanism where contact is initiated between the end of the AMA1 apical groove opposite the DII loop and the disulfide-tethered β-hairpin loop region of RON2 (Fig. 2B).

Guided by our high-resolution TgAMA1-TgRON2sp co-structure and sequence alignments (Fig. 3A), we identified key residues on TgRON2-2 and probed their functional relevance. Alanine substitutions in the N-terminal helix (D1297, H1301, D1304, and I1305) (25), the coil connecting the helix and cystine loop (P1309), the cystine loop (C1313, I1318, L1319, and C1323), and the C-terminal coil (T1332) all showed some degree of reduced binding to TgAMA1 as determined by SPR and ELISA (Fig. 3B and fig. S4) (21), consistent with an extensive, composite TgAMA1-TgRON2sp interface. Of particular note, the broadly conserved residues D1297, D1304, P1309, C1313, and C1323 were identified as critical for TgAMA1 binding, as was the less conserved L1319 located in the cystine loop. An important role for the cysteine-dependant conformation of RON2 in promoting complex formation is corroborated by the reduced ability of the TgRON2-2 mutants to bind TgAMA1 in the micronemes of T. gondii tachyzoites (fig. S5) and inhibit host cell invasion by T. gondii (Fig. 3C) (21).

Fig. 3

Hot spot residues in TgAMA1-TgRON2sp binding. (A) Alignment of apicomplexan RON2 sequences corresponding to TgRON2sp (25); secondary structure and numbering are based on TgRON2. (B) SPR-measured relative binding of TgRON2-2 mutants to TgAMA1. (C) Effect of TgRON2-2 mutants (6 μM) in inhibiting invasion by T. gondii, highlighting the crucial roles of C1313 and C1323. Student’s t test; *P < 0.05; **P < 0.01; ***P < 0.001; means  ±  SD for N = 3. (D) Binding consequences of TgAMA1 mutants and synergistic TgRON2-2 mutants in cell-based assays (top) and spatially represented on the surface of TgAMA1 (bottom). WT, wild type.

To probe the role of specific AMA1 residues in promoting complex formation, we selected a set of 11 TgAMA1 residues for mutagenesis studies (fig. S6 and table S2). Y230 was initially targeted as the analogous residue in PfAMA1 (Y251) was previously implicated in complex formation (24). The Y230→A230 (Y230A) substitution showed no detectable effect in cell-based (fig. S7B) and ELISA (fig. S7C) assays and only an order of magnitude weaker binding by SPR (Kd of 10−8 M versus Kd of 10−9 M) (fig. S7, D and E) (21). A minimal effect for Y230 in promoting or stabilizing the TgAMA1-TgRON2sp complex was confirmed by the TgAMA1 Y230A–TgRON2sp co-structure (fig. S8). The remaining TgAMA1 mutants were expressed on the surface of BHK-21 cells and tested for their ability to bind TgRON2-2 (Fig. 3D and fig. S9) (21). Substitutions of F197 (base of cystine loop) or Y213 (base of tyrosine cage surrounding P1309) to alanine impaired binding. In fact, Y213A showed no binding at any of the concentrations tested, suggesting that this highly conserved residue is a critical anchor point during complex formation. The F197A mutant showed a more moderate effect, with binding disrupted only at the lowest concentration. Thus, the F197A variant provided an opportunity to test the potential of synergistic TgAMA1-RON2 mutations. Total abrogation of complex formation was observed between TgAMA1 F197A and either TgRON2-2 I1318A or L1319A (tip of the cystine loop). These synergistic effects prompted us to generate three additional TgRON2-2 constructs comprising (i) the N-terminal helix to cystine loop, (ii) the coil connector to C terminus, and (iii) the N-terminal helix to coil connector (fig. S10). Analysis of these constructs by ELISA (21) demonstrated clear binding for the first two but no interaction with the third, effectively isolating the most critical region of TgRON2-2 to the cystine loop. These results enable the refinement of our proposed mechanism of binding, where the cystine loop in combination with the Y213-P1309 bridge act as a primary structural brace to anchor the peptide in the apical groove and promote displacement of the DII loop.

Having analyzed the AMA1-RON2 interaction in T. gondii, we set out to evaluate its molecular conservation across Apicomplexa. We recently showed that the AMA1-RON2 binding event is conserved in Plasmodium species (5). Sequence analysis revealed a two–amino acid insertion in the Pf/PvRON2 cystine loop (Fig. 3A), and, correspondingly, the PfAMA1 apo structure displayed an extended apical groove relative to TgAMA1 (18) that appeared ideally suited to complement the predicted PfRON2 cystine loop extension (Fig. 4A). Thus, although the cystine loop region is important for the initial interaction, it may also serve to regulate cross-species specificity.

Fig. 4

A model for Plasmodium invasion. (A) Homology model of the PfAMA1-PfRON2sp complex reveals an extended platform in the apical groove that accommodates the extended PfRON2sp cystine loop (black) relative to the overlaid TgRON2sp (green). (B) Apical surface views of PfAMA1 (gray) bound to invasion inhibitory molecules (green) with PfAMA1 residues implicated in binding shown in purple. PDB ID for IgNAR14l-1/mAb is 2Z8V, and for mAb 1F9, 2Q8A. Straight black lines denote the receptor binding groove.

The TgAMA1-TgRON2sp structure provides a basis for inhibitory mechanisms derived from recent antibody [monoclonal antibody 4G2 (mAb 4G2), mAb 1F9, immunoglobulin (Ig) NAR14l-1] and peptide (R1) studies (Fig. 4B). Mutagenesis studies suggested that the cross-reactive mAb 4G2 binds to the stem of the PfAMA1 DII loop (19, 24). It is now clear that mAb-mediated anchoring of the DII loop would impair the necessary conformational changes induced by RON2 binding. Opposite the DII loop, mAb 1F9 binds the PfAMA1 DI apical surface (23, 26), whereas IgNAR14l-1 forms an extensive interface with the same region (27), both effectively occluding the RON2 cystine loop binding surface. Nuclear magnetic resonance analysis showed that the R1 peptide contacts residues throughout the groove (28) and, interpreted in the context of our data, may displace the DII loop in a similar fashion to TgRON2sp. Collectively, these data provide a valuable resource for rational design of inhibitory small molecules, peptide mimetics, and engineered vaccines. Indeed, the ability of TgRON2sp to inhibit invasion at nanomolar concentrations is a promising step toward these goals.

Supporting Online Material

Materials and Methods

Figs. S1 to S10

Tables S1 to S3

References (2940)

References and Notes

  1. Materials and methods are available as supporting material on Science Online.
  2. Single-letter abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.
  3. Acknowledgments: This work was supported by Canadian Institutes of Health Research (CIHR) grant MOP82915 to M.J.B. and ANR-MIE-2009 grant to M.L. M.J.B. is a CIHR New Investigator and a Michael Smith Foundation for Health Research scholar. M.L.T. is supported by a Julie Payette Natural Sciences and Engineering Research Council Research Scholarship. M.H.L. is supported by ANR-MIE-2009. We are grateful to G. Labesse and R. Cerdan for helpful discussions and to J.-F. Dubremetz and M. Murphy for critical reading of the manuscript. Structure factors and coordinates have been deposited in the Protein Data Bank (PDB) under accession nos.: 2y8t, r2y8tsf (TgAMA1-TgRON2sp), 2y8r, r2y8rsf (TgAMA1 Y230A) and 2y8s, r2y8ssf (TgAMA1 Y230A–TgRON2sp). The CNRS and the University of Victoria Innovation and Development Corporation have filed a European Patent Application (application number EP 11305540.4) on the identification of a region of RON2 involved in AMA1 binding and its potential targeting by vaccines and drugs.

Stay Connected to Science

Navigate This Article