Abstract
Apical membrane antigen-1 (AMA1) is a conserved malarial vaccine candidate essential for the formation of tight junctions with the rhoptry neck protein (RON) complex, enabling Plasmodium parasites to invade human erythrocytes, hepatocytes, and mosquito salivary glands. Despite its critical role, extensive surface polymorphisms in AMA1 have led to strain-specific protection, limiting the success of AMA1-based interventions beyond initial clinical trials. Here, we identify an i-body, a humanised single-domain antibody-like molecule that recognises a conserved pan-species conformational epitope in AMA1 with low nanomolar affinity and inhibits the binding of the RON2 ligand to AMA1. Structural characterisation indicates that the WD34 i-body epitope spans the centre of the conserved hydrophobic cleft in AMA1, where interacting residues are highly conserved among all Plasmodium species. Furthermore, we show that WD34 inhibits merozoite invasion of erythrocytes by multiple Plasmodium species and hepatocyte invasion by P. falciparum sporozoites. Despite a short half-life in mouse serum, we demonstrate that WD34 transiently suppressed P. berghei infections in female BALB/c mice. Our work describes the first pan-species AMA1 biologic with inhibitory activity against multiple life-cycle stages of Plasmodium. With improved pharmacokinetic characteristics, WD34 could be a potential immunotherapy against multiple species of Plasmodium.
Similar content being viewed by others
Introduction
Progress in reducing the global burden of malaria has stalled, and numerous challenges to established interventions have evolved, highlighting the need for new interventions to prevent malaria1. Plasmodium falciparum accounts for most of the malaria-induced morbidity and mortality worldwide, with P. vivax being the second major cause1,2,3. Invasion of erythrocytes by Plasmodium merozoites initiates all the clinical manifestations of malaria4. The invasion process is fast, sequential, and tightly regulated by multiple interactions between parasite ligands and receptors on the erythrocyte surface5,6. Following attachment to the erythrocyte membrane, the merozoite reorientates, allowing the apical end to form a tight junction as an entry point for invading host erythrocytes7. Parasite antigens involved in this attachment and invasion process have been explored as potential anti-malarial blood-stage vaccines and therapeutic targets8,9,10.
AMA1 is a micronemal protein released onto the parasite surface following the attachment of the merozoite to the erythrocyte membrane. AMA1 subsequently binds to a hairpin loop in the ectodomain of the RON2 protein on the erythrocyte surface to form a tight junction11,12,13. There is considerable experimental evidence that the AMA1-RON2 complex is also involved in sporozoite invasion of mosquito salivary glands and mammalian hepatocytes14,15,16. AMA1 is one of the few invasion proteins shared between all Plasmodium species causing human malaria, although the sequence varies between species. Previously, PfAMA1 was evaluated as a promising vaccine candidate due to its pivotal role in host cell invasion by multiple life-cycle stages of the parasite and being an immunogenic protein targeted by acquired immunity6,17,18. However, polymorphisms in AMA1 tend to induce strain-specific protection, and for this reason, clinical development of AMA1 has been challenging19,20,21,22,23. Even though a few cross-reactive monoclonal antibodies (mAbs) have been generated against AMA1, these antibodies lacked significant functional activity24,25,26,27,28.
i-bodies are human single immunoglobulin domains inspired by the structural features of the shark variable new antigen receptor (VNAR) and were designed to overcome some of the limitations of conventional mAbs29,30. The i-body consists of a human scaffold (first Ig domain of the human neural cell adhesion molecule) and two regions of hypervariable amino acids constituting complementarity determining regions (CDR) 1 & 3, where they would exist in the VNAR29,30. By randomising sequences of CDR1 (fixed length) and CDR3 (variable length), a library containing 1010 unique i-bodies was generated. Due to their smaller size compared to conventional antibodies and easy access to cryptic epitopes, i-bodies have been considered as potential therapeutics against diseases such as osteoporosis, idiopathic pulmonary fibrosis, and renal fibrosis29,31,32. In this study, we investigated the use of i-bodies to identify novel epitopes in AMA1, an important malaria vaccine candidate.
Phage display was used to identify a novel i-body, WD34, which binds to AMA1 from multiple Plasmodium species with low nanomolar affinity. By determining the crystal structures of WD34-AMA1 complexes, we have defined the binding footprint of WD34 in Pf and PvAMA1. Using several Plasmodium strains and species, we demonstrated the inhibitory role of WD34 in merozoite and sporozoite invasion. Furthermore, we showed that WD34 could transiently suppress P. berghei infections in BALB/c mice. Our findings identify WD34 as the first pan-species anti-malarial antibody-like molecule with high invasion inhibitory activity.
Results
Identification of strain transcending i-bodies to AMA1
The i-body library, described by Griffiths et al.29, was biopanned on recombinant AMA1 (recAMA1) of two P. falciparum strains, 3D7 and W2mef (termed haplotype-distance biopanning). These two isoforms of PfAMA1 were chosen because they differ at many of the polymorphic sites in the antigen and show substantial antigenic differences33,34,35. In addition to panning on individual PfAMA1 isoforms, a cross-panning strategy was employed in which phage selected after three rounds of panning on one PfAMA1 isoform were subjected to three further panning rounds on the other isoform (Fig. 1a).
Enrichment of phage expressing i-bodies binding to recAMA1 was observed after three rounds of panning on either the 3D7 or W2mef antigen (Fig. 1b, c). These phage pools contained i-bodies that bound to recAMA1 of other P. falciparum strains (7G8, FVO, HB3 and D10), indicating that the pools contained phage that expressed i-bodies with broad specificity (Supplementary Figs. 1, 2). At each round of cross-panning, there were binders to both the 3D7 and W2mef isoforms of recAMA1, which were used in the original panning campaign. Since these ELISAs were carried out using pooled phage, it was possible that the pools contained many clones that bound to one or other, but not both forms of recAMA1. Forty-eight clones from the last round of each panning campaign were expressed as i-bodies in the periplasm of E. coli. ELISA screening of these clones revealed a wide spectrum of binding to the different forms of AMA1. This presumably reflected the level of i-body expression in addition to the binding affinity for a particular form of AMA1 (Fig. 1d). Sanger sequencing of 117 of these positive clones from all panning campaigns identified 12 unique i-body sequences (Fig. 1e).
When 12 i-bodies representing these sequences were tested by ELISA for their binding ability to a range of PfAMA1 isoforms (3D7, W2mef, 7G8, FVO, D10, and HB3), different binding specificities were observed (Fig. 1f). The WD33 and WD34 i-bodies, which were identified exclusively in the cross-panning campaign, bound to all AMA1 isoforms examined, but not to reduced and alkylated AMA1, indicating that these two i-bodies bind to conserved conformational epitopes shared by all isoforms of PfAMA1 (Fig. 1g and Supplementary Fig. 3).
The WD33 and WD34 i-bodies bound to parasite PfAMA1 when used to immunoblot extracts of saponin-lysed schizonts (3D7, W2mef, FVO and 7G8) electrophoresed under non-reducing conditions, but not reducing conditions (Fig. 2a). Both WD33 and WD34 bound specifically to a band of the predicted size for full-length AMA1 and another band corresponding to a processed fragment of AMA1. The identity of both bands was confirmed using well-characterised anti-AMA1 mAbs (Supplementary Fig. 4). The ability of these i-bodies to bind parasite AMA1 was confirmed by indirect immunofluorescence microscopy, which showed both WD33 and WD34 binding to merozoites within segmented schizonts with the expected staining pattern for AMA1 and colocalising with mAb 5G8, an AMA1 pro-domain specific mAb (Fig. 2b). The specificity of WD34 binding to AMA1 was further validated by immunoprecipitation followed by mass spectrometry (Supplementary mass spectrometry data).
WD34 binds to the conserved hydrophobic cleft region of AMA1 with low nanomolar affinity
The ability of anti-AMA1 i-bodies to inhibit the interaction of PfRON2 with PfAMA1 was examined using a biotin-tagged synthetic peptide representing a hairpin loop of PfRON2 that has been shown to bind to AMA113,36. Initial screening revealed a dramatic reduction of RON2 peptide binding to 3D7 recAMA1 in the presence of an equivalent concentration of WD34 (Fig. 3a). The results of a competition ELISA confirmed that WD34 inhibited the binding of the RON2 peptide to AMA1 in a dose-dependent manner (Fig. 3b). None of the other 11 representative i-bodies tested, including WD33, inhibited the interaction of recPfAMA1 and the PfRON2 peptide, further confirming that the WD33 and WD34 epitopes differ. The binding of the WD33 and WD34 i-bodies to different isoforms of PfAMA1 was characterised by surface plasmon resonance (SPR). WD34 bound to all forms of AMA1 examined, consistent with the initial observations from ELISA, immunoblotting, and indirect IFA experiments, with all binding affinities (KD) in the low nanomolar range (Fig. 3c). The KD values of WD34 binding to the six isoforms of PfAMA1 ranged from approximately 1–20 nM with the lowest and highest KD values observed with 7G8 AMA1 (1.07 ± 0.78 nM) and HB3 AMA1 (19.7 ± 7.9 nM), respectively. In a control experiment, the RON2 peptide bound to PfAMA1 with a KD similar to previously reported values36.
To identify whether WD34 binds directly to or at a site close to the conserved hydrophobic cleft involved in RON2 binding, competition ELISAs were performed with two mAbs (1F9 and 4G2) whose epitopes have been previously identified as proximal to the hydrophobic cleft21,28. Both mAbs inhibited the binding of WD34 in a dose-dependent manner, although 1F9 appeared to have a greater impact on WD34 binding than 4G2 (Fig. 3d). In contrast, the binding of WD33 was inhibited by mAb 4G2, but not mAb 1F9 (Supplementary Fig. 5). These data suggest that the WD34 epitope lies between the footprints of mAbs 1F9 and 4G2, which are at opposite ends of the conserved hydrophobic cleft in PfAMA1.
WD34 binds to AMA1 from other Plasmodium species
As their epitope(s) appeared to be highly conserved, the WD33 and WD34 i-bodies were tested for binding to recAMA1 from two other Plasmodium species: P. vivax (Pv-Palo alto strain), which is the second leading cause of human malaria, and the rodent parasite P. chabaudi adami (Pc-556KA strain). WD34 bound to both Pv and PcAMA1, indicating that this epitope is conserved across Pf strains and in at least some other Plasmodium species (Fig. 1g). In contrast to WD34, WD33 bound to neither Pv nor PcAMA1. The affinity of WD34 for PvAMA1 (119 ± 37 nM) and PcAMA1 (26.1 ± 4.1 nM) was approximately 30-fold and 8-fold lower, respectively, than the average KD value for PfAMA1 isoforms. These data indicate that WD34 binds strongly to many isoforms of PfAMA1 and AMA1 from other Plasmodium species despite extensive polymorphic changes in these molecules; it is estimated that approximately 10% of the ectodomain amino acid residues in Pf and PvAMA1 are polymorphic37. The slight differences in KD values between the different PfAMA1 isoforms and WD34 could be due to the involvement of some polymorphic residues in the WD34-AMA1 interaction. The affinity of WD33 for PfAMA1 isoforms ranged from 10–20 nM (Fig. 3c). These data indicate that the different abilities of WD34 and WD33 to inhibit RON2 binding were not due to differences in binding affinity of the two i-bodies.
AMA1 interacting residues of WD34 are conserved in Plasmodium
To understand the structural basis of WD34 binding to AMA1 and its pan-specificity, we determined the crystal structures of WD34 in complex with Pf (3D7) recAMA1 and Pv (Palo Alto) recAMA1 (both domain I + II) to resolutions of 2.4 Å and 3 Å, respectively (Fig. 4a–c, Supplementary Fig. 6 and Supplementary Table 2). Consistent with previous publications, both Pf and PvAMA1 constructs used here adopted the conserved two domain structure previously observed for AMA120,21,38,39,40. Similarly, WD34 has the typical compact globular structure described for an i-body29 featuring a series of beta-strands, a canonical disulphide bond in the scaffold, and two loop regions corresponding to CDR1 and CDR3. In both i-body-AMA1 complexes, the WD34 footprint spans the conserved hydrophobic cleft in AMA1, perpendicular to the RON2 peptide footprint (Fig. 4d–f). The AMA1 surface area buried by WD34 is 1645 Å and 1972 Å for Pf and PvAMA1, respectively. These buried surface areas are remarkably large, particularly considering that WD34 is a single Ig domain with only two CDRs. Indeed, the observed buried surface areas are larger than the footprints of previously identified invasion inhibitory anti-AMA1 antibodies, including 1F9 (1220 Å in PDB ID: 2Z8V) and IgNAR (14I1-M15) (1205 Å in PDB ID: 2Q8A)21,40. In contrast to 1F9 and IgNAR (14I1-M15), which interact with the highly polymorphic loop 1d of AMA1, WD34 binds mainly to relatively conserved residues in AMA1 domain I of both Plasmodium species, displaying only minor differences in the two binding interfaces (Fig. 4a, b and Supplementary Figs. 6, 7).
As expected, the CDR1 and CDR3 regions of the i-body are responsible for most of the interactions with AMA1 (Supplementary Table 2). In both complexes, an additional disulphide bond linking CDR1 (C22) and CDR3 (C31) constrains the flexibility of the CDR loops in WD34 (Fig. 4g). This disulphide bond is essential for the interaction of WD34 with AMA1, as mutating C22 and C31 to serine completely abolished the interaction (Fig. 4g). In both complexes, at least 12 of the 21 CDR residues in WD34 are involved in binding to AMA1 (Supplementary Table 2). Apart from interacting with the residues in the CDR3 and CDR1 loops of WD34, AMA1 also interacts with WD34 scaffold residues, forming several direct contacts. A detailed description of this interaction is provided in Supplementary information (Supplementary Fig. 8).
Although the WD34 footprint is nearly identical in size and location in Pf and PvAMA1, there are some differences in the i-body-antigen interactions in the two complexes that should be noted. One difference is that WD34 forms more hydrogen bonds with PfAMA1 than with PvAMA1 (Supplementary Table 2). In the WD34-AMA1Pf complex, G80WD34 and W88WD34 interact with AMA1 via hydrogen bonds, whereas these residues only have weak van der Waals interactions with PvAMA1. Another difference is the mode of engagement with Y234Pf (Y179Pv). In the WD34-AMA1Pf complex W88WD34 contacts Y234Pf via a π stacking interaction (Supplementary Fig. 9), which is absent in the WD34-AMA1Pv complex. FoldX position scan software showed that the WD34-AMA1Pf complex exhibits more free binding energy than the WD34-AMA1Pv complex, consistent with the higher binding affinity observed in the SPR analyses41 (Supplementary Fig. 10). This detailed structural information provides a molecular explanation for the observation that WD34 binds to AMA1 from these two major human Plasmodium species and does so with slightly different affinities.
WD34 inhibits erythrocyte invasion by merozoites of multiple Plasmodium species
The inhibitory activity of WD33 and WD34 on parasite growth was tested using in vitro parasite growth inhibition assays. WD34, but not WD33 inhibited all the P. falciparum strains tested (3D7, W2mef, 7G8 and Dd2) with the mean half-maximal inhibitory concentration (IC50) ranging from 13 μg/ml to 40 μg/ml (Fig. 5a–c). P. vivax cannot be maintained in continuous culture to study invasion. Because of this, we used a transgenic P. falciparum line engineered to express PvAMA1. In addition, P. knowlesi, which can be cultured in vitro, was also examined. WD34 inhibited the growth of both the Pf3D7 transgenic PvAMA1 parasite line and P. knowlesi parasites in a dose-dependent manner, with IC50S of 84 μg/ml and 128 μg/ml, respectively. The broadly similar potency of WD34 in inhibiting multiple parasite lines is consistent with the structural and functional data described above. Importantly, the IC50 values for WD34 reported here are lower than previously published values for AMA1 antibodies that inhibit merozoite invasion42,43.
WD34 inhibits hepatocyte invasion by sporozoites of P. falciparum
Previous reports have indicated that the AMA1-RON2 complex is involved in hepatocyte invasion by sporozoites14,15,44. To investigate the effect of WD34 on cell traversal, salivary gland sporozoites of the PfNF54 strain were incubated with HC-04 hepatocytes in the presence of fluorescein isothiocyanate (FITC)-dextran and increasing concentrations of WD34. WD34 did not significantly inhibit hepatocyte traversal when compared to the negative control i-body, 21H5, although there was some modest inhibition at the highest WD34 concentration tested (Fig. 5d). This finding, with P. falciparum sporozoites aligns with the recent observation made by Fernandes et al.15 that conditional loss of AMA1 and RON2 did not affect P. berghei sporozoite migration.
To test the effect of WD34 on parasite invasion of hepatocytes, sporozoites of PfNF54 were incubated with HC-04 cells in the presence of the i-body, and the number of cells containing CSP-positive intracellular parasites was quantified using flow cytometry 24 h after the treatment. Hepatocyte invasion by NF54 was moderately inhibited by WD34 in a dose-dependent manner (mean difference 62.9%; p = 0.0178), but NF54 was not significantly affected by the control i-body (mean difference 35.1%; p ≥ 0.4006) (Fig. 5e). These results are consistent with the AMA1-RON2 interaction being important for P. falciparum sporozoite invasion of hepatocytes14,15,44.
WD34 has a short suppressive effect on P. berghei infections in mice
To assess whether WD34 has an effect on P. berghei parasitaemia in mice, we performed a Peters' 4-day suppressive test45 using BALB/c mice. Treatment of mice once daily with WD34 (30 mg/kg, I.P.) for 4 days suppressed parasitaemia at day 4 by approximately 49% compared to untreated mice (p = 0.0294) (Fig. 6b, c). However, this suppressive effect was short-lived, persisting for no more than 2–3 days, after which the parasitaemia rapidly rose to the levels seen in untreated mice, with all animals succumbing to malaria by day 9 (p = 0.2124). It is possible that this transient effect was due to either poor serum stability or the short half-life of WD34. Therefore, we examined the thermal stability, serum stability and pharmacokinetic profile of WD34. In these experiments, WD34 was stable in both PBS and mouse serum (Supplementary Fig. 14) but had a very short half-life (Supplementary mass spectrometry information). Immunoblotting of P. berghei-infected erythrocytes demonstrated that WD34 specifically recognised P. berghei AMA1, with a single ~55 kDa band detected in infected cells, but not in uninfected cells. Under reducing conditions, the P. berghei-specific i-body signal was drastically reduced, suggesting WD34 recognised a disulphide bond-dependent epitope in PbAMA1 (Fig. 6f). Taken together, these data demonstrate that WD34 has a short suppressive effect on P. berghei growth in vivo.
Discussion
AMA1 has been studied extensively as a potential component of a vaccine against P. falciparum and P. vivax malaria. However, like most P. falciparum merozoite proteins that are natural immunogens, AMA1 is highly polymorphic, and this has prevented the generation of broadly protective immune responses in field trials with AMA1-containing vaccines23,46,47,48. Here, we used an i-body (WD34) isolated from a phage-display library to identify a previously unrecognised highly conserved epitope in AMA1. To isolate WD34 and other cross-reactive i-bodies, we used a cross-panning strategy on Pf3D7 and PfW2mef recAMA1, which were chosen because of their extensive sequence differences33. In contrast to the cross-reactive characteristics of the i-bodies selected using the cross-panning strategy, i-bodies selected by direct panning on one or other of the recAMA1 isoforms were more strain-specific.
WD34 reacted strongly with both isoforms of AMA1 used in the cross-panning strategy to select this i-body and with four other isoforms of PfAMA1 examined. Unexpectedly, WD34 cross-reacted with AMA1 of P. vivax (Palo Alto strain), the other major cause of human malaria, with AMA1 of P. knowlesi, which causes malaria in humans and other primates, and AMA1 of the rodent parasite, P. chabaudi adami (556KA strain).
Using recPfAMA1 and a synthetic peptide representing the extracellular loop of PfRON2, we showed that WD34 blocked the interaction between AMA1 and its ligand, which is inserted by the parasite into the erythrocyte cell membrane prior to merozoite invasion. As WD34 also blocked the binding of mAbs 1F9 and 4G2, whose epitopes lie at opposite ends of the RON2-binding cleft, we concluded that the WD34 epitope lies in the central region of this hydrophobic cleft. This was confirmed by solving the X-ray crystal structures of the WD34-AMA1Pf and WD34-AMA1Pv complexes to resolutions of 2.4 Å and 3 Å, respectively. The WD34 footprint consisted of relatively conserved residues in the central region of the RON2-binding cleft of both the Pf and Pv isoforms of AMA1, distant from the highly polymorphic loop 1d in P. falciparum, which is targeted by the “strain-specific” mAb, 1F921.
In both complexes, WD34 contacts a small number of residues in the hydrophobic cleft that are directly involved in RON2 binding, but most of the WD34 contacts are with surrounding residues on both sides of the cleft in loops 1a, 1b, 1c and 1e20 (Supplementary Fig. 11). Notably, of the six WD34 contact residues common to Pf and PvAMA1, five are strictly conserved in AMA1 of all isolates of multiple Plasmodium species collected in multiple countries where malaria is endemic. Only one residue (G172) is clearly polymorphic (Fig. 7). In ~32% of the PfAMA1 sequences surveyed and in the HB3 isoform studied here, G172, which makes contacts with G80WD34, is mutated to glutamic acid (Fig. 7 and Supplementary Fig. 12). This difference may be responsible for the WD34 binding affinity for HB3 PfAMA1 (KD = 19.7 nM) being slightly lower than for the other five PfAMA1 isoforms studied (KD = 1.07–7.16 nM). In <1% of Pf parasites surveyed, this residue was mutated to either valine or arginine, but the effect of these mutations on WD34 binding affinity was not examined (Supplementary Fig. 12). Furthermore, there is a high conservation of WD34-AMA1Pf contact residues in AMA1 of the rodent parasite P. chabaudi, which had a similar low nanomolar binding affinity in SPR analyses (Supplementary Fig. 12c). The finding that the binding affinity of WD34 for PvAMA1 was approximately 30-fold less than its binding affinity for PfAMA1 is not surprising given that WD34 was selected by panning the i-body-phage library on Pf3D7 AMA1. Although there are many conserved features to the WD34 epitope in these two forms of AMA1, there are also differences in the two interfaces, which could account for this difference in binding affinity. Notably, WD34 forms more polar interactions with PfAMA1 than with PvAMA1 (Supplementary Table 2). For example, G80 and W88 of WD34 form hydrogen bonds to PfAMA1 G172 (3.2 Å) and 2.82 Å Y236 (2.82 Å), respectively, but equivalent interactions are lacking in the WD34-AMA1Pv complex.
A characteristic property of camelid nanobodies or shark single-domain antibodies (VNARs) is their ability to interact via their extended CDR3 loops with concave epitopes not accessible to conventional antibodies49. The WD34 i-body studied here also has an extended CDR3 loop, but rather than penetrating the hydrophobic RON2-binding cleft of AMA1 like the IgNAR (14I1-M15), the CDR3 and CDR1 loops, with some framework residues, form a paratope that spans the RON2-binding cleft40. Despite the small size of the i-body compared to conventional antibodies, the WD34 footprint is considerably larger than that of previously characterised anti-AMA1 antibodies21,40. This may, at least partly, be due to the disulphide bond between CDR1 and CDR3 facilitating a spatial arrangement of the CDR loops that provides a flattened paratope surface. An important consequence of the large interface between WD34 and the surface of AMA1 may be a tolerance of polymorphisms, contributing to the broad cross-reactivity described here.
The binding of PfAMA1 to a hairpin loop of PfRON2 is important for the invasion of host cells by merozoites and sporozoites11,12,13,14,15,16,50,51. Although there is evidence suggesting that PfAMA1 and PvAMA1 may have another binding interaction in addition to the RON2-loop interaction52, it was notable that the WD34 i-body inhibited the invasion of erythrocytes and hepatocytes by P. falciparum merozoites and sporozoites, respectively53. The inhibitory potency of WD34 in merozoite growth inhibition assays was broadly similar for multiple strains of P. falciparum and for P. knowlesi, with the IC50 values reported here being some of the lowest for an “antibody” recognising multiple parasite lines. To our knowledge, this is the first time a cross-species anti-AMA1 antibody-like molecule with potent invasion inhibitory properties has been reported. Previously, Collins et al.28 and Igonet et al.27 described two AMA1 mAbs with broad strain and species specificity, however, both these mAbs showed rather poor in vitro inhibitory activity27,43,54. Furthermore, antibodies to other Plasmodium antigens having similar potent inhibitory activity and broad strain and species specificity have not been reported.
In a preliminary in vivo study, WD34 suppressed parasitaemias in mice by 50% 4 days after challenge with P. berghei. However, this effect was short-lived, and parasitaemias returned to control levels shortly after the last WD34 treatment. Pharmacokinetic and serum stability studies indicated that the transient effect observed is primarily attributed to the short half-life resulting from rapid renal clearance, which aligns with the smaller molecular weight of WD34. These results supported the previously published pharmacokinetic data for another i-body (AD-114), where the serum half-life was estimated to be 11 min55. That study showed that PEGylation dramatically extended the half-life of the i-body. In view of these results, in vivo half-life extension for WD34 will be explored in future.
Malaria vaccine development has made important advances in the last decade with two vaccines targeting P. falciparum sporozoites now approved for use56,57,58,59. As discussed earlier, the polymorphisms in AMA1 have proven to be a major barrier to efforts aimed at developing a vaccine incorporating this asexual blood-stage antigen. One approach to overcoming this problem has been to immunise with an AMA1-RON2 complex. This approach has shown promise in preclinical studies and has demonstrated that there are conserved or less antigenically diverse epitopes in AMA1 that can be targeted by protective antibodies60,61,62. However, it is not yet clear whether the highly conserved WD34 epitope described here can be targeted by a natural antibody response or antibodies induced by vaccination. If a form of AMA1 can be engineered that diverts the induced antibody response away from highly polymorphic regions, such as loop 1d, towards relatively conserved regions, such as the WD34 footprint defined here, AMA1 may have renewed potential as a vaccine candidate, especially given the possibility of using this approach for both P. falciparum and P. vivax malaria25. However, it is possible that the WD34 epitope is poorly immunogenic, and its identification here may have resulted from the use of the large naive i-body library in which the range of CDR sequences was not restricted by immunisation or infection, which usually proceeds the creation of nanobody or mAb libraries. Whether the WD34 epitope is immunogenic needs to be investigated in future studies to further inform vaccine strategies.
In recent years, evidence has accumulated showing that the control of malaria with drugs and a vaccine may be complemented by using human mAbs to prevent malaria63,64,65. The human mAb, CIS43LS, which targets PfCSP, protected human volunteers against P. falciparum infection in both experimental challenge and natural exposure settings64,66. Although the major focus of current immunoprophylaxis research has been the P. falciparum CSP antigen, antibodies targeting asexual blood-stage antigens should be considered as a complementary strategy67. AMA1 is an attractive candidate for this approach as it is expressed in both pre-erythrocytic and asexual blood stages of the parasite life cycle and present in all Plasmodium species infecting humans. The WD34 i-body in its current small, single-domain form is not a candidate for assessment as an immunotherapeutic because of its short half-life. However, it did provide some suppression of P. berghei parasitaemia in mice and extending its half-life, for example, with an Fc tag could provide an attractive format for development as an immunoprophylactic agent against multiple species of malaria. A similar strategy has been effective in other fields. AD-214, an i-body with a human IgG Fc against idiopathic pulmonary fibrosis has been shown to be safe and well tolerated in a human Phase 1 safety and tolerability study in healthy volunteers (ClincalTrials.gov: NCT04415671). This suggests a potential future pathway for a WD34-derived antibody-based therapeutic against malaria.
Methods
Mammalian cells, parasite cultures, peptides, and antibodies
Expi293 and ExpiCHO (ThermoFisher) cells were maintained in Expi293 and ExpiCHO expression mediums (ThermoFisher) at 37 °C, 8% CO2, 125 rpm. Plasmodium falciparum 3D7, DD2, 7G8, FVO and W2mef and P. knowlesi parasites were maintained in in vitro cultures at 37 °C according to previously published protocols68,69. Biotin-tagged and untagged RON2 peptide (DITQQAKDIGAGPVASCFTTRMSPPQQICLNSVVNTALSTSTQSAMK) with cyclised cysteines were synthesised by Mimotopes, Australia.
mAbs 4G2, 1F9 and 5G8 were obtained from Robin Anders. Anti-c-myc mAb 9E10 and anti-i-body scaffold mAb 7G4 were made at the WEHI Antibody Facility, Bundoora, Australia. Fluorophore and HRP-conjugated Abs were used as follows: Anti-M13 HRP mAb (Catalogue no: 11973-MM05T-H, Sino Biological), anti-His HRP mAb (Catalogue no: A7058, Sigma-Aldrich), goat anti-human IgG (Fc-specific) HRP (Catalogue no: A0170, Sigma-Aldrich), goat anti-mouse IgG (Catalogue no: AP127P, Sigma-Aldrich), Streptavidin HRP (Catalogue no: 21130, ThermoFisher), goat anti-human Alexa fluor 488 (Catalogue no: H10120, Invitrogen), goat anti-rabbit Alexa Fluor 568 (Catalogue no: A11036, Molecular Probes), goat anti-rabbit 546 (Catalogue no: A11081, Invitrogen), goat anti-mouse 568 (Catalogue no: A11031, Invitrogen), goat anti-mouse FITC (Catalogue no: F0257, Sigma-Aldrich).
Expression and purification of recombinant AMA1
Codon-optimised full ectodomain (domain I, II and III) and domain I + II (for X-ray structure determination experiments) sequences of Pf3D7, W2mef, 7G8, FVO, HB3, D10 and PcAMA1 were cloned into pET-28a (+)-Tev vector (Genscript, Singapore). AMA1 protein was expressed in inclusion bodies, purified and refolded according to standard protocols20,70. The codon-optimised PvAMA1 (Palo Alto) sequence in PcDNA3.1 was kindly provided by James Beeson, Burnet Institute. PvAMA1 was expressed in Expi293 cells and purified from the Expi293 expression medium using Ni-NTA chromatography50.
Biopanning using the phage-displayed i-body library
Phage panning was performed as described previously with some modifications29,40. Phage-displayed i-body library (~1010 random i-body sequences cloned into pADL-23c phagemid vector) was amplified in E. coli TG1 cells. At OD600 = 0.4–0.5, E. coli TG1 cells were infected with M13K07 helper phage (Antibody Design Laboratories, USA) to display the i-bodies on phage particles. The amplified i-body phage library was biopanned independently on immobilised recombinant Pf3D7 and PfW2mef AMA1 (2.5 µg/ml). Six biopanning rounds were carried out in each biopanning campaign. For cross-biopanning, a portion of the round three phage from each panning campaign was incubated with immobilised recombinant AMA1 of the other isoform (from Pf3D7 enriched phage to PfW2mef AMA1 and vice versa). These then underwent a further three rounds of biopanning.
Phage enrichment ELISA, single clone screening and sequencing
Phage ELISAs were performed as described previously29. Briefly, 96-well plates (Nunc, Maxisorp) were coated with the antigens and blocked with 5% skim milk. Amplified phage pools of biopanning rounds or single clone phage were incubated at a 1:10 dilution against immobilised antigens. The wells were washed three times with 0.1% PBST, and bound phage were detected with anti-M13 HRP mAb (Sino-biologicals) followed by 3,3′,5,5′-tetramethylbenzidine (ThermoFisher) as enzyme substrate. The enzyme reaction was stopped with 0.1 M HCl, and absorbance was measured at 450 nm.
Phage-infected TG1 single colonies were grown as 2 ml cultures in 48 deep-well plates (Corning Axygen) and induced with 1 mM IPTG for 16 h. The level of i-body expression and AMA1 binding of periplasmic extracts for each clone were assessed by ELISA. The positive clones for both i-body expression and AMA1 binding were sequenced by Sanger sequencing at the Australian Genome Research Facility, Melbourne, Australia.
Expression and purification of i-bodies
AMA1-specific i-body sequences in pADL-23c vector plasmids (Antibody Design Laboratories, USA) were transformed into E. coli BL21 cells. Cultures grown overnight at 37 °C in 2YT media supplemented with 0.1% (w/v) glucose and ampicillin (100 μg/ml) were diluted to OD600 of 0.1 and incubated at 37 °C. At OD600 = 0.7, cells were induced with 1 mM IPTG and subjected to further growth overnight at 28 °C. Periplasmic extraction of the i-bodies was conducted as described previously71. Briefly, the centrifuged cell pellets were resuspended in full-strength spheroplast buffer with lysozyme (40 µg/ml) and incubated on ice for 10 min followed by incubation in half-strength spheroplast buffer. The periplasmic fraction was isolated by centrifugation, and i-bodies were purified using immobilised metal affinity chromatography (IMAC) followed by anion exchange (Cytiva-CaptoQ) and size-exclusion chromatography.
Sequences of Human IgG1 Fc region-tagged WD33 and WD34 were designed by cloning into a PcDNA3.1 vector (Genscript, Singapore) to express i-body Fc versions in Expi293 and ExpiCHO (Invitrogen, USA) expression systems. Fc-tagged i-bodies were purified using affinity purification (Mabselect PrismA-Cytiva) and size-exclusion chromatography.
Standard and competition ELISAs
Indirect i-body binding ELISAs were used to determine the specificity of AMA-1-specific i-bodies. All incubations were performed at 100 μl per well at room temperature for 1 h or overnight at 4 °C. Washing steps included at least three washes with 0.1% PBS-Tween. The screen of i-bodies with the ability to block the RON2 peptide was performed by incubating the same concentrations (50 nM) of biotin-tagged RON2 peptide and i-body against immobilised Pf3D7 AMA1 (2.5 µg/ml) and measuring the bound RON2 peptide by using streptavidin HRP.
In competition ELISAs, 150 nM of the biotin-tagged RON2 peptide and increasing concentrations of WD34 and WD33 were incubated in wells coated with 2.5 µg/ml Pf3D7 AMA1. Bound RON2 peptide was measured using streptavidin HRP.
Competition ELISAs of WD34 with mAbs 1F9 and 4G2 were performed with 100 nM of human Fc-tagged i-body (WD33 or WD34) and increasing concentrations of mAbs. This mixture of mAb and i-body was added to wells coated with Pf3D7 AMA1. Bound WD34 was measured using goat anti-human Fc-specific HRP antibody.
Immunoblotting and immunofluorescence of recombinant proteins and parasites
Late-stage schizonts of P. falciparum were purified by using MACS magnetic columns (Miltenyi Biotec) and saponin lysis. Isolated schizonts were lysed with sample buffer and subjected to SDS-PAGE on 4–12% acrylamide gels (BOLT, Invitrogen). Plasmodium berghei-infected and uninfected murine erythrocytes were collected from BALB/c mice, and saponin pellets were solubilised in reducing or non-reducing sample buffer prior to equal sample loading and SDS-PAGE through 10% Bis-Tris gels. Proteins were transferred to nitrocellulose membrane for 7 min (iBlot2, Invitrogen) or for 60 min in SDS-PAGE transfer buffer (P. berghei blots), blocked with 10% skim milk and probed with 25 μg/ml of i-bodies (in 5% skim milk). Membranes were then washed three times with 0.05% PBST and incubated with anti-myc or HRP (1:5000) secondary antibodies for 1 h before washing and development of membranes. Western blots were developed with West Pico PLUS chemiluminescent reagents (ThermoFisher, USA) and visualised using the Biorad GelDoc or ChemiDoc system.
For fixed immunofluorescence imaging, smears of synchronised late-stage schizonts on glass microscope slides were fixed with 100% methanol for 2 min and incubated in a blocking buffer (4% BSA in PBS) for at least 1 h. i-bodies (10 μg/ml), and fluorophore-tagged secondary antibodies (1:1000) were used. Labelled parasites were imaged with a Zeiss LSM 880 inverted microscope with ×63/1.4 oil objective, appropriate excitations, and an Airyscan detector. Captured images were analysed with ImageJ software.
Immunoprecipitation and mass spectrometry analysis
Late-stage Pf (3D7) schizonts were purified by saponin lysis and washed with PBS (pH 7.4) three times. Schizont pellets were lysed by RIPA buffer (Thermo Scientific) according to the manufacturer’s protocol. Parasite lysates (100 µl) were incubated with 50 µg of WD34 and 21H5 i-bodies at room temperature for 2 h. Parasite lysate and i-body mixtures were incubated with anti-c-myc Magnetic Beads (Thermo Scientific) overnight, which were then washed with 0.001% PBS-Tween (pH 7.4) five times and PBS (pH 7.4) five times. On-bead tryptic digestion was performed for LC-MS/MS analysis of bound proteins as previously described with minor modifications72. Briefly, beads were washed with 20 mM NH4HCO3 (pH 8.0) three times and denatured in 8 M urea. Proteins were reduced and alkylated with 2 mM tris-2-carboxyethyl-phosphine and iodoacetamide. Each bead sample was diluted with 800 µl of 50 mM Tris HCl (pH 8.3) and subjected to tryptic digestion (1:100 ratio with antibody load) overnight at 37 °C. Tryptic peptides were recovered, desalted and concentrated according to the published protocol72.
The tryptic peptides were separated on Ultimate 3000 RSLCnano UHPLC (Thermo Scientific) and analysed on a Q-Exactive HF Orbitrap LC-MS/MS system (Thermo Scientific). The mass-spectral database search was conducted using Proteome Discoverer 2.4 and the Sequest HT search engine, with label-free quantitation. Each replicate was searched separately, and precursor ion area and abundances were not normalised as part of the label-free quantitation process.
Surface plasmon resonance studies
Kinetic analysis and equilibrium dissociation constants (KD) were measured using a Biacore T200 system (Cytiva). The surfaces of flow cells one, two and four were activated for 14 min with a 1:1 mixture of 0.1 M NHS (N-hydroxysuccinimide) and 0.4 M EDC (N-(3-dimethylaminopropyl)-N’-ethylcarbodiimide hydrochloride) at a 5 μl/min flow rate. The ligands (AMA1) at a concentration of 80 μg/ml in 10 mM sodium acetate, pH 4.5, were immobilised at a density of 500 RU on flow cells two, three and four. Flow cell one was immobilised with bovine serum albumin to serve as a reference surface. All the surfaces were blocked with a 7-min injection of 1 M ethanolamine, pH 8.0. i-bodies (14 kDa, 90% purity based on SDS-PAGE) in 10 mM Tris, 150 mM NaCl,0.005% Tween-20,1% BSA pH 8, were injected over the flow cells at a concentration range at a flow rate of 45 μl/min and a temperature of 25 °C. Both multi-cycle kinetic and single-cycle kinetic analyses were used to describe the AMA1-i-body interactions.
Structure determination of the WD34-AMA1 complexes
WD34-bound PfAMA1 and PvAMA1 protein samples were prepared for crystallisation as described previously29. The protein-protein complexes were in 50 mM Tris (pH 8.0) and initial sparse matrix crystallisation trials were performed in 96-well sitting drop trays (Swissci, Neuheim, Switzerland) at a molar ratio of 1:1 using an in-house Gryphon LCP (Art Robbins Instruments). Initial screenings were set up with 0.2 μl protein sample and 0.2 μl reservoir per drop. Subsequent crystal optimisation trials were conducted in 24-well Limbro plates (Hampton Research). All crystallisation trials were performed at 20 °C. WD34-AMA1Pf complex crystals were obtained in 0.1 M HEPES, 70% MPD, pH 7.5 at 5 mg/ml concentration and WD34-AMA1Pv complex crystals were obtained in 0.1 M MES, 12% w/v PEG20000, pH 6.5 at 5 mg/ml concentration.
X-ray diffraction data was collected on the MXII beamline of Australian synchrotron using an Eiger detector (Dectris, Baden-Dättwil, Switzerland). Data were processed to 2.7 Å for WD34-AMA1Pf and 3 Å for WD34-AMA1Pv using the XDS package73. Structures were solved by molecular replacement in Phaser using the structural coordinates of PDB IDs—1Z40 (PfAMA1), 5NQF (PvAMA1) and 5AEA (i-body scaffold)74,75. Structure building and refinement were performed using Coot and PHENIX76,77. The atomic coordinates and structure files were deposited in the Protein Data Bank under PDB ID 8QU7 (WD34-AMA1Pf) and 8QUS (WD34-AMA1Pv). All the figures of the structures were prepared using PyMOL78.
Blood-stage invasion inhibition assays
Two-cycle growth inhibition assays were performed as follows: synchronised trophozoite stage parasites were diluted to 0.5% parasitaemia (P. knowlesi) or 0.2% parasitaemia (P. falciparum) and 1% (v/v) haematocrit in 25 µl RPMI 1640, 25 mM HEPES media with i-bodies and added into 96-well round-bottomed culture plate. Parasites were incubated for 32 h (P. knowlesi) or 48 h (P. falciparum) at 37 °C. Parasites were then washed with PBS, stained with 10 μg/ml ethidium bromide (BioRad) and 1 × 105 erythrocytes were counted using an Accuri flow cytometer (BD Biosciences). i-body activity was assessed across three independent biological replicates and IC50 values were calculated using GraphPad Prism 9.2 (GraphPad, USA).
Blood-stage invasion inhibition assays for transgenic parasites
Invasion assays were performed with WD34 and 21H5 as previously described, with minor changes79. Parasites were synchronised with sorbitol, late trophozoite and schizont stages were then inoculated at 0.5% parasitaemia in 4% haematocrit in 45 µl RPMI media, with 5 µl i-bodies in PBS. The ring stages were harvested by fixing with 0.25% glutaraldehyde. Parasites were stored in PBS at 4 °C. Cultures were later stained with 1× SYBR Green I (S7563, Invitrogen), and parasitaemia was assayed by flow cytometry in a BD FACSCanto™ II system, with further analysis in FlowJo™ v10.9.0 (BD Life Sciences). Parasites were gated from uninfected erythrocytes using the FITC-A and PE-A channels. Net growth rates were calculated by subtracting the parasitaemia of parasites inoculated with 1 mg/ml heparin, then dividing by the parasitaemia of parasites inoculated with 5 µl PBS. Growth percentages were subtracted from 100% to give invasion inhibition.
IC50 calculations were performed in R using the packages drc and ggplot279,80,81,82. Assays were performed in biological triplicate, and IC50s calculated independently for each replicate. The parasite lines used were based on W2mef, with the AMA1 locus replaced with either the P. vivax or the P. falciparum 3D7 locus17,33.
Sporozoite cell traversal and hepatocyte invasion assays
HC-04 hepatocytes were seeded at near-confluence (100,000 cells per well) in 96-well plates83. Plasmodium falciparum NF54 sporozoites were generated by feeding stage V gametocytes at 0.3% parasitaemia to 3-day-old female Anopheles stephensi mosquitoes (kindly provided by M. Jacobs-Lorena, Johns Hopkins University) via water-jacketed glass membrane feeders as described84. On day 17 post bloodmeal, sporozoites were dissected from mosquito salivary glands in Scheider’s medium, enumerated by a haemocytometer, and resuspended in IMDM + 5% FBS + 0.5 mg/ml FITC-dextran at a final concentration of 450 sporozoites/μl (multiplicity of infection [MOI]: 0.45). WD34 and 21H5 i-bodies were added at the indicated concentrations and incubated with sporozoites and FITC-dextran for 15 min on ice. mAb 2A10 against P. falciparum CSP (a kind gift from F. Zavala, Johns Hopkins University) was used as a positive control at 10 μg/ml to validate the assays. After incubation, sporozoites were added to cells and centrifuged at 500 × g for 3 min. After a 4 h incubation at 37 °C, 5% CO2, cells were washed, trypsinised, and half of the cells processed to quantify cell traversal (percentage of FITC-Dextran positive HC-04 cells) by flow cytometry and the remaining half were replated in 96-well plates and incubated overnight at 37 °C, 5% CO2 for quantification of invasion at 24 h post-infection (percentage of cells positive for anti-CSP conjugated to ALEXA 647) by flow cytometry, as previously described85.
P. berghei infections in mice and Peters’ 4-day suppressive test
Six to eight weeks old female BALB/c acceptor mice were injected intraperitoneally (I.P.) with 1 × 107 erythrocytes infected with P. berghei expressing mCherry-Luciferase (a kind gift from Volker Heussler) withdrawn from donor mice86. Mice were simultaneously treated with WD34 (30 mg/kg in a once-daily ‘Q.D regimen’, I.P.) for four consecutive days or chloroquine (50 mg/kg, I.P., once), or untreated control mice received PBS (vehicle). Twelve hours after administration of the last i-body dose, parasitaemias were measured by flow cytometry of mCherry-positive infected cells, as previously described87. Mean parasitaemia values in each group of 3–4 mice are expressed as percent parasitaemia. Giemsa-stained blood smears were used to confirm parasitaemia. All mice eventually developed patent infections and succumbed to malaria or were euthanised according to WEHI Animal Ethics Committee-approved procedures. Data were analysed by one-way ANOVA with multiple comparisons using GraphPad Prism 9.
Differential scanning fluorometry analysis
Melting temperatures and thermal stability of i-bodies were determined by the ViiA 7 real-time PCR system (ThermoFisher); 0.5 mg/ml and 1 mg/ml WD33 and WD34 i-bodies in PBS (pH 7.4) mixed with 10 × Sypro Orange Dye (Invitrogen). Samples were tested in duplicate starting at 25 °C and increasing in 0.5 °C steps, at a rate of 1 °C/min, to 95 °C. TAMRA reporter was set for excitation (~550 nm) and emission (~587 nm) channels. Data was normalised and a sigmoidal non-linear regression was performed using GraphPad Prism 9.2 (GraphPad, USA).
In vitro serum stability assay
WD34 was mixed with BALB/c mouse serum to a concentration of 100 µg/ml and incubated at 37 °C. Time-point samples (0, 6, 18, 24, 36, 48, 72 and 96 h) were collected and evaluated by ELISA for binding to AMA1 (Coating concentration, 1 µg/ml). mAb 5G8 was used as a reference for AMA1 binding.
Animal ethics statement
BALB/c mice were obtained from Jackson Laboratory and housed under Physical Containment level 2 conditions at La Trobe Animal Research and Training Facility (LARTF) and Walter and Elisa Hall Institute of Medical Research (WEHI) Animal Research facility. All procedures in the pharmacokinetic experiments were approved by the La Trobe University Animal Ethics Committee and performed in LARTF. All procedures in the P. berghei infection study were approved by the WEHI Animal Ethics Committee and performed at WEHI.
Pharmacokinetic study of WD34 by LC/MS-MS
Eight to ten weeks old male BALB/c mice (n = 3) were injected with a single dose of WD34 (30 mg/kg, I.P.) and blood samples were collected at 0.5, 6, 12, 24, and 48 h post-dosing. WD34 in the plasma was quantified by monitoring a signature peptide LTWSLGLPEATVNVK contained both CDR3 and framework residues of WD34, using liquid chromatography-tandem mass spectrometry, as described previously55,88. Briefly, serum samples were adjusted to 2% SDS in ammonium bicarbonate (20 mM), vortexed, and heated to 95 °C for 15 min, followed by sonication for 10 min. Lysate containing 40 µg of protein was reduced and alkylated with 5 mM tris (2-carboxyethyl) phosphine and 20 mM iodoacetamide. The single pot, solid phase, sample preparation strategy was used to clean up samples for mass spectrometry from solubilised lysates89. Proteins were captured onto carboxylate-modified magnetic SpeedBeads (Cytiva, Sweden), and beads were washed with 80% ethanol. Tryptic digestion was carried out in 20 mM ammonium bicarbonate overnight at 37 °C. The tryptic peptides were then desalted using the StageTip method and dried in a speedvac72. Tryptic peptides were separated on a Thermo Ultimate 3000 RSLCnano UHPLC system and analysed on a Thermo Q-Exactive HF Orbitrap mass-spectrometer (ThermoFisher Scientific, Waltham, MA, USA). Peptides (800 ng) were loaded onto an Acclaim PepMap C18 3 µm 0.075 mm × 20 mm trap column (ThermoFisher Scientific, Waltham, MA, USA) and washed at 5 µl/min for 6 min (Buffer C: 0.1% (v/v) trifluoracetic acid, 2% (v/v) ACN) before switching the pre-column in line with the analytical column held at 55 °C (nanoEase M/Z Peptide BEH C18 Column, 1.7 µm, 130 Å and 75 µm ID × 25 cm, Waters). The separation of peptides was performed at 250 nL/min using a linear Acetonitrile gradient. Blank runs were performed between sample injections.
Three injections per sample were performed in a 90-min run. CID MS/MS spectra were collected using a 90-min method with Orbitrap HCD parameters using Orbitrap HCD parameters. Dynamic exclusion parameters were set as follows: exclude isotope on, exclude after n = 1 times, duration 25 s, default charge state of 2, and use the peptide monoisotopic peak determination mode. Other instrument parameters were: MS1 scan at 60,000 resolution, m/z 350–1500, AGC target 3e6, injection time max 30 ms. The mass-spectral database search and quantitation were conducted using Skyline 23.1.0.380 and the MS Amanda search engine with label-free quantitation matched against the Mouse reference proteome (Downloaded March 2023) and the antibody sequence for WD3490. The peptide fixed mod was carbamidomethyl of C; no variable mods were set. Precursor charges set to 2, 3, and 0.05 m/z ion match tolerance. Precursor tolerance was set to 20 ppm and fragment tolerance to 0.05 Da. Digest was set to Trypsin, and one missed cleavage was permitted. The included static modification was carbamidomethyl of C. Peptide quantification settings used MS1 precursor ion area and converted to ng protein using a matrix-matched calibration curve of WD34Fc in mouse serum (800 ng, 400 ng, 200 ng, 200 ng, 100 ng, 50 ng, 25 ng, 25 ng), with linear in log-space regression fit. WD34 abundance was normalised to mouse serum peptide MFASFPTTK (515.25733 m/z) to control technical variance.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
The crystal structures reported in this manuscript are deposited in the Protein Data Bank, www.rcsb.org (PDB IDs 8QU7 and 8QUS). Source data are provided with this paper as a Source Data file.
References
World Health Organization. World Malaria Report 2022 (WHO, 2022).
Phillips, M. A. et al. Malaria. Nat. Rev. Dis. Prim. 3, 17050 (2017).
Phyo, A. P., Dahal, P., Mayxay, M. & Ashley, E. A. Clinical impact of vivax malaria: a collection review. PLoS Med. 19, e1003890 (2022).
Miller, L. H., Ackerman, H. C., Su, X.-Z. & Wellems, T. E. Malaria biology and disease pathogenesis: insights for new treatments. Nat. Med. 19, 156–167 (2013).
Gaur, D., Mayer, D. C. G. & Miller, L. H. Parasite ligand–host receptor interactions during invasion of erythrocytes by Plasmodium merozoites. Int. J. Parasitol. 34, 1413–1429 (2004).
Baum, J., Maier, A. G., Good, R. T., Simpson, K. M. & Cowman, A. F. Invasion by P. falciparum merozoites suggests a hierarchy of molecular interactions. PLoS Pathog. 1, e37 (2005).
Aikawa, M., Miller, L., Johnson, J. & Rabbege, J. Erythrocyte entry by malarial parasites. A moving junction between erythrocyte and parasite. J. Cell Biol. 77, 72–82 (1978).
Barry, A. E. & Arnott, A. Strategies for designing and monitoring malaria vaccines targeting diverse antigens. Front. Immunol. 5, 359 (2014).
Mitran, C. J. & Yanow, S. K. The case for exploiting cross-species epitopes in malaria vaccine design. Front. Immunol. 11, 335 (2020).
Burns, A. L. et al. Targeting malaria parasite invasion of red blood cells as an antimalarial strategy. FEMS Microbiol. Rev. 43, 223–238 (2019).
Besteiro, S., Michelin, A., Poncet, J., Dubremetz, J.-F. & Lebrun, M. 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 (2009).
Riglar, D. T. et al. Super-resolution dissection of coordinated events during malaria parasite invasion of the human erythrocyte. Cell Host Microbe 9, 9–20 (2011).
Srinivasan, P. et al. Binding of Plasmodium merozoite proteins RON2 and AMA1 triggers commitment to invasion. Proc. Natl Acad. Sci. USA 108, 13275–13280 (2011).
Yang, A. S. P. et al. AMA1 and MAEBL are important for Plasmodium falciparum sporozoite infection of the liver. Cell. Microbiol. 19, e12745 (2017).
Fernandes, P. et al. The AMA1-RON complex drives Plasmodium sporozoite invasion in the mosquito and mammalian hosts. PLoS Pathog. 18, e1010643 (2022).
Loubens, M. et al. Plasmodium sporozoites on the move: switching from cell traversal to productive invasion of hepatocytes. Mol. Microbiol. 115, 870–881 (2021).
Drew, D. R. et al. Functional conservation of the AMA1 host-cell invasion ligand between P. falciparum and P. vivax: a novel platform to accelerate vaccine and drug development. J. Infect. Dis. 217, 498–507 (2017).
Beeson, J. G. et al. Merozoite surface proteins in red blood cell invasion, immunity and vaccines against malaria. FEMS Microbiol. Rev. 40, 343–372 (2016).
Kennedy, M. C. et al. In vitro studies with recombinant Plasmodium falciparum apical membrane antigen 1 (AMA1): production and activity of an AMA1 vaccine and generation of a multiallelic response. Infect. Immun. 70, 6948–6960 (2002).
Bai, T. et al. Structure of AMA1 from Plasmodium falciparum reveals a clustering of polymorphisms that surround a conserved hydrophobic pocket. Proc. Natl Acad. Sci. USA 102, 12736–12741 (2005).
Coley, A. M. et al. Structure of the malaria antigen AMA1 in complex with a growth-inhibitory antibody. PLoS Pathog. 3, 1308–1319 (2007).
Hodder, A. N., Crewther, P. E. & Anders, R. F. Specificity of the protective antibody response to apical membrane antigen 1. Infect. Immun. 69, 3286–3294 (2001).
Thera, M. A. et al. A field trial to assess a blood-stage malaria vaccine. N. Engl. J. Med. 365, 1004–1013 (2011).
Remarque, E. J., Faber, B. W., Kocken, C. H. M. & Thomas, A. W. 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 (2008).
Dutta, S. et al. Overcoming antigenic diversity by enhancing the immunogenicity of conserved epitopes on the malaria vaccine candidate apical membrane antigen-1. PLoS Pathog. 9, e1003840 (2013).
Kusi, K. A. et al. Safety and immunogenicity of multi-antigen AMA1-based vaccines formulated with CoVaccine HT™ and Montanide ISA 51 in rhesus macaques. Malar. J. 10, 182 (2011).
Igonet, S. et al. Cross-reactivity studies of an anti-Plasmodium vivax apical membrane antigen 1 monoclonal antibody: binding and structural characterisation. J. Mol. Biol. 366, 1523–1537 (2007).
Collins, C. R., Withers-Martinez, C., Hackett, F. & Blackman, M. J. An inhibitory antibody blocks interactions between components of the malarial invasion machinery. PLoS Pathog. 5, e1000273 (2009).
Griffiths, K. et al. i-bodies, human single domain antibodies that antagonize chemokine receptor CXCR4. J. Biol. Chem. 291, 12641–12657 (2016).
Streltsov, V. A. et al. Structural evidence for evolution of shark Ig new antigen receptor variable domain antibodies from a cell-surface receptor. Proc. Natl Acad. Sci. USA 101, 12444–12449 (2004).
Cao, Q. et al. A single-domain i-body, AD-114, attenuates renal fibrosis through blockade of CXCR4. JCI Insight 7, e143018 (2022).
Qiu, H. et al. ADR3, a next generation i-body to human RANKL, inhibits osteoclast formation and bone resorption. J. Biol. Chem. 299, 102889 (2023).
Drew, D. R. et al. Defining the antigenic diversity of Plasmodium falciparum apical membrane antigen 1 and the requirements for a multi-allele vaccine against malaria. PLoS ONE 7, e51023 (2012).
Terheggen, U. et al. Limited antigenic diversity of Plasmodium falciparum apical membrane antigen 1 supports the development of effective multi-allele vaccines. BMC Med. 12, 183 (2014).
Coley, A. M. et al. The most polymorphic residue on Plasmodium falciparum apical membrane antigen 1 determines binding of an invasion-inhibitory antibody. Infect. Immun. 74, 2628–2636 (2006).
Vulliez-Le Normand, B. et al. Structural and functional insights into the malaria parasite moving junction complex. PLoS Pathog. 8, e1002755 (2012).
Elliott, S. R. et al. Research priorities for the development and implementation of serological tools for malaria surveillance. F1000Prime Rep. 6, 100 (2014).
Lim, S. S. et al. Structure and dynamics of apical membrane antigen 1 from Plasmodium falciparum FVO. Biochemistry 53, 7310–7320 (2014).
Vulliez-Le Normand, B., Saul, F. A., Hoos, S., Faber, B. W. & Bentley, G. A. Cross-reactivity between apical membrane antgen 1 and rhoptry neck protein 2 in P. vivax and P. falciparum: a structural and binding study. PLoS ONE 12, e0183198 (2017).
Henderson, K. A. et al. Structure of an IgNAR-AMA1 complex: targeting a conserved hydrophobic cleft broadens malarial strain recognition. Structure 15, 1452–1466 (2007).
Schymkowitz, J. et al. The FoldX web server: an online force field. Nucleic Acids Res. 33, W382–W388 (2005).
Seidel-Greven, M. et al. Isolation and light chain shuffling of a Plasmodium falciparum AMA1-specific human monoclonal antibody with growth inhibitory activity. Malar. J. 20, 37 (2021).
Maskus, D. J. et al. Characterization of a novel inhibitory human monoclonal antibody directed against Plasmodium falciparum apical membrane antigen 1. Sci. Rep. 6, 39462 (2016).
Silvie, O. et al. A role for apical membrane antigen 1 during invasion of hepatocytes by Plasmodium falciparum sporozoites. J. Biol. Chem. 279, 9490–9496 (2004).
Peters, W. The chemotherapy of rodent malaria, XXII. The value of drug-resistant strains of P. berghei in screening for blood schizontocidal activity. Ann. Trop. Med. Parasitol. 69, 155–171 (1975).
Crewther, P. E., Matthew, M. L., Flegg, R. H. & Anders, R. F. Protective immune responses to apical membrane antigen 1 of Plasmodium chabaudi involve recognition of strain-specific epitopes. Infect. Immun. 64, 3310–3317 (1996).
Dicko, A. et al. Impact of a Plasmodium falciparum AMA1 vaccine on antibody responses in adult Malians. PLoS ONE 2, e1045 (2007).
Ouattara, A. et al. Lack of allele-specific efficacy of a bivalent AMA1 malaria vaccine. Malar. J. 9, 175 (2010).
Henry, K. A. & MacKenzie, C. R. Antigen recognition by single-domain antibodies: structural latitudes and constraints. mAbs 10, 815–826 (2018).
Drew, D. R. et al. Defining species-specific and conserved interactions of apical membrane protein 1 during erythrocyte invasion in malaria to inform multi-species vaccines. Cell. Mol. Life Sci. 80, 74 (2023).
Scally, S. W. et al. PCRCR complex is essential for invasion of human erythrocytes by Plasmodium falciparum. Nat. Microbiol. 7, 2039–2053 (2022).
Lyke, K. E. et al. Low-dose intravenous and subcutaneous CIS43LS monoclonal antibody for protection against malaria (VRC 612 Part C): a phase 1, adaptive trial. Lancet Infect. Dis. 23, 578–588 (2023).
Lee, S.-K. et al. The direct binding of Plasmodium vivax AMA1 to erythrocytes defines a RON2-independent invasion pathway. Proc. Natl Acad. Sci. USA 120, e2215003120 (2023).
Woehlbier, U., Epp, C., Hackett, F., Blackman, M. J. & Bujard, H. Antibodies against multiple merozoite surface antigens of the human malaria parasite Plasmodium falciparum inhibit parasite maturation and red blood cell invasion. Malar. J. 9, 77 (2010).
Griffiths, K. et al. Half-life extension and non-human primate pharmacokinetic safety studies of i-body AD-114 targeting human CXCR4. mAbs 11, 1331–1340 (2019).
Stanisic, D. I. & Good, M. F. Malaria vaccines: progress to date. BioDrugs 37, 737–756 (2023).
Devi, S. 12 countries to get first doses of malaria vaccine. Lancet 402, 172 (2023).
Datoo, M. S. et al. Efficacy and immunogenicity of R21/Matrix-M vaccine against clinical malaria after 2 years’ follow-up in children in Burkina Faso: a phase 1/2b randomised controlled trial. Lancet Infect. Dis. 22, 1728–1736 (2022).
Adepoju, P. RTS,S malaria vaccine pilots in three African countries. Lancet 393, 1685 (2019).
Srinivasan, P. et al. Immunization with a functional protein complex required for erythrocyte invasion protects against lethal malaria. Proc. Natl Acad. Sci. USA 111, 10311–10316 (2014).
Patel, P. N. et al. Structure-based design of a strain transcending AMA1-RON2L malaria vaccine. Nat. Commun. 14, 5345 (2023).
Yanik, S. et al. Structure guided mimicry of an essential P. falciparum receptor-ligand complex enhances cross neutralizing antibodies. Nat. Commun. 14, 5879 (2023).
Kisalu, N. K. et al. A human monoclonal antibody prevents malaria infection by targeting a new site of vulnerability on the parasite. Nat. Med. 24, 408–416 (2018).
Kayentao, K. et al. Safety and efficacy of a monoclonal antibody against malaria in Mali. N. Engl. J. Med. 387, 1833–1842 (2022).
Gaudinski, M. R. et al. A monoclonal antibody for malaria prevention. N. Engl. J. Med. 385, 803–814 (2021).
Wu, R. L. et al. Low-dose subcutaneous or intravenous monoclonal antibody to prevent malaria. N. Engl. J. Med. 387, 397–407 (2022).
Chandley, P., Ranjan, R., Kumar, S. & Rohatgi, S. Host-parasite interactions during Plasmodium infection: implications for immunotherapies. Front. Immunol. 13, 1091961 (2023).
Trager, W. & Jensen, J. B. Continuous culture of Plasmodium falciparum: its impact on malaria research. Int. J. Parasitol. 27, 989–1006 (1997).
Moon, R. W. et al. Adaptation of the genetically tractable malaria pathogen Plasmodium knowlesi to continuous culture in human erythrocytes. Proc. Natl Acad. Sci. USA 110, 531–536 (2013).
Chatzileontiadou, D. S. M., Szeto, C., Jayasinghe, D. & Gras, S. Protein purification and crystallization of HLA-A∗02:01 in complex with SARS-CoV-2 peptides. STAR Protoc. 2, 100635 (2021).
Minsky, A., Summers, R. G. & Knowles, J. R. Secretion of beta-lactamase into the periplasm of Escherichia coli: evidence for a distinct release step associated with a conformational change. Proc. Natl Acad. Sci. USA 83, 4180–4184 (1986).
Rappsilber, J., Mann, M. & Ishihama, Y. Protocol for micro-purification, enrichment, pre-fractionation and storage of peptides for proteomics using StageTips. Nat. Protoc. 2, 1896–1906 (2007).
Kabsch, W. XDS. Acta Crystallogr. D Biol. Crystallogr. 66, 125–132 (2010).
McCoy, A. J. et al. Phaser crystallographic software. J. Appl. Crystallogr. 40, 658–674 (2007).
McCoy, A. J., Grosse-Kunstleve, R. W., Storoni, L. C. & Read, R. J. Likelihood-enhanced fast translation functions. Acta Crystallogr. D Biol. Crystallogr. 61, 458–464 (2005).
Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and development of Coot. Acta Crystallogr. D Biol. Crystallogr. 66, 486–501 (2010).
Adams, P. D. et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr. D Biol. Crystallogr. 66, 213–221 (2010).
The PyMOL Molecular Graphics System (Schrodinger LLC, 2015).
Persson, K. E. M., Lee, C. T., Marsh, K. & Beeson, J. G. Development and optimization of high-throughput methods to measure Plasmodium falciparum-specific growth inhibitory antibodies. J. Clin. Microbiol. 44, 1665–1673 (2006).
Wickham, H. ggplot2: Elegant Graphics for Data Analysis (Springer, 2016).
R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2023).
Ritz, C., Baty, F., Streibig, J. C. & Gerhard, D. Dose-response analysis using R. PLoS ONE 10, e0146021 (2016).
Sattabongkot, J. et al. Establishment of a human hepatocyte line that supports in vitro development of the exo-erythrocytic stages of the malaria parasites Plasmodium falciparum and P. vivax. Am. J. Trop. Med. Hyg. 74, 708–715 (2006).
Jennison, C. et al. Inhibition of plasmepsin V activity blocks Plasmodium falciparum gametocytogenesis and transmission to mosquitoes. Cell Rep. 29, 3796–3806.e3794 (2019).
Lopaticki, S. et al. Tryptophan C-mannosylation is critical for Plasmodium falciparum transmission. Nat. Commun. 13, 4400 (2022).
Prado, M. et al. Long-term live imaging reveals cytosolic immune responses of host hepatocytes against Plasmodium infection and parasite escape mechanisms. Autophagy 11, 1561–1579 (2015).
Favuzza, P. et al. Dual plasmepsin-targeting antimalarial agents disrupt multiple stages of the malaria parasite life cycle. Cell Host Microbe 27, 642–658.e612 (2020).
Marin, C. et al. Determination of plasma concentration of Belimumab by LC-MS/MS: method development, validation, and clinical application. J. Pharm. Biomed. Anal. 236, 115730 (2023).
Hughes, C. S. et al. Single-pot, solid-phase-enhanced sample preparation for proteomics experiments. Nat. Protoc. 14, 68–85 (2019).
Pino, L. K. et al. The Skyline ecosystem: informatics for quantitative mass spectrometry proteomics. Mass Spectrom. Rev. 39, 229–244 (2020).
Acknowledgements
We thank Dr Kaye Truscott and Deepti Verghese Jose for facilitating the protein purification in MBB laboratory (La Trobe University), Dr Chris Szeto and Professor Stephanie Gras for training on SPR, Liana Theoridis and Dr Teresa Carvalho for supplying parasites for initial experiments, Ornella Romeo at Adelaide University for preparation of parasite samples, La Trobe Proteomics and Metabolomics Platform and Bioimaging Facility for facilitating SPR, confocal imaging and mass spectrometry services, AdAlta for providing reagents and consumables to conduct the study, Australian Red Cross Blood Service for blood. X-ray crystallographic data were acquired using the MX2 beamline at Australian Synchrotron, The Australian Nuclear Science and Technology Organization. D.A. was supported by La Trobe University Graduate Research Scholarship (LTGRS), and La Trobe University Full Fee Research Scholarship (LTUFFRS). J.C. and K.H.L. were supported by ARC-RTP Scholarships. D.W.W. acknowledges Hospital Research Foundation Fellowship and collaborative research grant. J.A.B. and J.G.B. acknowledge NHMRC Investigator Grants 1176955 and 1173046, respectively. The Burnet Institute is supported by NHMRC and Victorian State Government infrastructure grants.
Author information
Authors and Affiliations
Contributions
M.F., R.F.A., and D.A. conceived the study. D.A., J.C., J.C.M., E.H., S.C., K.H.L., L.M.Y., J.M., D.H.O., C.C., and D.W.W. planned and performed the experiments. M.F., R.F.A., H.P., J.B., M.K., J.A.B., and D.W.W. supervised experiments. D.A., J.C., J.C.M., E.H., S.C., L.M.Y., J.A.B., and D.W.W. analysed the data. D.A., R.F.A., and M.F. wrote the original manuscript. All authors reviewed and approved the final manuscript.
Corresponding author
Ethics declarations
Competing interests
This study was a collaboration between AdAlta Limited and La Trobe University. M.F. is the founding chief scientist and a shareholder in AdAlta Ltd., and R.F.A. is also a shareholder in AdAlta. The other authors have declared no competing interests.
Peer review
Peer review information
Nature Communications thanks Lea Barfod and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Source data
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Angage, D., Chmielewski, J., Maddumage, J.C. et al. A broadly cross-reactive i-body to AMA1 potently inhibits blood and liver stages of Plasmodium parasites. Nat Commun 15, 7206 (2024). https://doi.org/10.1038/s41467-024-50770-7
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-024-50770-7
- Springer Nature Limited