Plasmodium falciparum CyRPA Glycan Binding Does Not Explain Adaptation to Humans

Abstract

The closest relatives of the human malaria parasite Plasmodium falciparum are a number of species, collectively known as the subgenus Laverania, that infect chimpanzees, bonobos, and western gorillas (Sharp et al. 2020). These parasites appear to be host specific: in extensive noninvasive surveys of wild African apes, the three species that infect gorillas have not been detected in samples from chimpanzees or bonobos, while the four parasite species that infect chimpanzees or bonobos have never been found in samples from gorillas (Liu et al. 2017; Loy et al. 2017). Similarly, none of these ape parasite species have been found to infect humans living sympatrically with, and exposed to the same mosquitoes as, wild apes (Loy et al. 2018). However, at some point in the recent past, P. falciparum arose via the zoonotic transmission of a parasite from gorillas (Liu et al. 2010; Sharp et al. 2020) and it is of obvious interest to understand how the ancestor of the human parasite overcame the host species barrier.

A potentially important element of the host species barrier is the pentameric PCRCR complex, that forms an essential component of the parasite's erythrocyte invasion pathway (Cowman et al. 2017; Farrell et al. 2024). Three proteins, the reticulocyte binding protein homolog 5 (RH5), the cysteine-rich protective antigen (CyRPA), and the RH5 interacting protein (RIPR), form the RCR complex which mediates binding to the erythrocyte membrane; RCR binds to a complex of two more proteins, cysteine-rich small secreted protein (CSS) and Plasmodium thrombospondin-related apical merozoite protein (PTRAMP), which is anchored to the parasite membrane. RH5 and RIPR are both thought to bind host proteins: RH5 binds basigin (Crosnier et al. 2011), while RIPR has been suggested to bind semaphorin 7A (Nagaoka et al. 2020), although this has been questioned (Williams et al. 2024). The RH5–basigin interaction appears to form part of the host species barrier: in vitro, RH5 proteins encoded by chimpanzee parasites do not bind gorilla basigin, while RH5 proteins from gorilla parasites do not bind chimpanzee basigin (Galaway et al. 2019). However, the RH5 proteins of two gorilla parasite species (the third was not tested) were found to bind human basigin, suggesting that, in this respect, the gorilla parasites were preadapted to infect humans (Galaway et al. 2019).

Until recently, CyRPA was thought to serve simply as a linker between RH5 and RIPR (Scally et al. 2022), but it has now been shown to have an essential function in the invasion process by directly binding to host glycans with sialic acids (Day et al. 2024). The major glycans found on the surface of red blood cells differ between humans and the other apes. The glycans on the surface of ape red blood cells end with the sialic acids N-glycolylneuraminic acid (Neu5Gc; about 75%) and N-acetylneuraminic acid (Neu5Ac; about 25%). Neu5Gc is produced by the enzyme cytidine monophosphate Neu5Ac hydroxylase (CMAH), which converts Neu5Ac to Neu5Gc. A deletion occurred within the CMAH gene in an ancestor of Homo sapiens about 2 to 3 million years ago, such that humans do not produce Neu5Gc and instead have only Neu5Ac on the glycans on their erythrocytes (Chou et al. 2002). Day et al. (2024) found that CyRPA from Plasmodium reichenowi (PrCyRPA), a chimpanzee parasite, binds both Neu5Gc and Neu5Ac to a similar extent, whereas P. falciparum CyRPA (PfCyRPA) has a 50-fold higher affinity for Neu5Ac than for Neu5Gc, suggesting that this protein is adapted to the sialoglycans of humans. They noted that PfCyRPA and PrCyRPA differ at two sites (residues 154 and 209) within one of two putative glycan binding sites. Furthermore, they found that replacing these two amino acids in PfCyRPA (Y154 and G209) with the residues found in PrCyRPA (H154 and R209) was sufficient to produce P. reichenowi-like glycan-binding properties, while substituting these two amino acids in PrCyRPA with the residues found in PfCyRPA had the opposite effect (Day et al. 2024). Thus, this CyRPA–glycan interaction appeared to be another host species barrier that had to be overcome, and the authors went on to conclude that CyRPA of P. falciparum has “changed sialic acid specificity during co-evolution with the human host.”

However, the chimpanzee parasite P. reichenowi did not give rise to P. falciparum. Instead, the human parasite originated from a recent cross-species transmission of a parasite species, termed Plasmodium praefalciparum, that infects gorillas (Rayner et al. 2011). Furthermore, while P. praefalciparum and P. reichenowi are generally quite closely related, the gene encoding CyRPA in P. falciparum is unusually divergent from that of P. reichenowi (Fig. 1), because most of it lies, along with the gene encoding RH5, in a short region of chromosome 4 that was transferred into an ancestor of P. praefalciparum from an ancestor of a much more distantly related gorilla parasite, Plasmodium adleri (Sundararaman et al. 2016). Part of the first exon of the CyRPA gene, encoding residues 1 to 41 (residues 1 to 30 form a signal peptide), exhibits the evolutionary relationships typical of the whole genome. However, the remainder of the gene, encoding residues 42 to 362, has the unusual phylogeny resulting from the introgression event (Fig. 1). The two amino acid replacements tested by Day et al. (2024) could have occurred at any time since the common ancestor of P. falciparum and P. reichenowi, which in the case of the CyRPA gene was the last common ancestor of the entire Laverania clade (Fig. 1). We have compared the CyRPA sequences of seven Laverania species (Hall et al. 2002; Otto et al. 2018) to identify when these two changes took place, and more generally to investigate the pathways of host adaptation of CyRPA.

Fig. 1.

Evolutionary relationships among members of the subgenus Laverania, which includes Plasmodium species infecting African apes, as well as P. falciparum. Species in red (P. reichenowi, P. billcollinsi, P. gaboni) infect chimpanzees, in blue (P. praefalciparum, P.blacklocki, P. adleri) infect western gorillas, and in purple (P. lomamiensis) infect bonobos. a) The phylogeny relevant to most of the genome (Sharp et al. 2020). b) The phylogeny found for a region of about 8 kb on chromosome 4, which includes the genes encoding RH5 and CyRPA (Sundararaman et al. 2016; Sharp et al. 2020).

Open in new tabDownload slide

As a consequence of the introgression event, the CyRPA sequences of P. falciparum and P. reichenowi are unexpectedly divergent, differing at 63 residues. In contrast, the CyRPA sequences of P. falciparum and P. praefalciparum differ at only seven sites. Remarkably, we found that the residues at the two sites (154 and 209), whose replacement was found to switch the glycan binding affinities of CyRPA (Day et al. 2024), are identical in P. falciparum and the gorilla parasite P. praefalciparum (Fig. 2). Gorillas have a functional CMAH gene, and thus the same Neu5Gc-rich glycans as chimpanzees. Since these two sites did not change during the origin of the human parasite, obviously neither was involved in the adaptation of P. falciparum to using Neu5Ac after transmission from gorillas. Both sites differ between the CyRPAs of P. praefalciparum and P. adleri, which has N154 and R209 (Fig. 2). From the alignment of the Laverania as a whole (supplementary fig. S1, Supplementary material online), both changes appear to have occurred in the ancestor of P. praefalciparum, but these changes are unlikely to reflect host adaptation since both P. praefalciparum and P. adleri infect gorillas.

Fig. 2.

Alignment of the region of CyRPA containing the two sialic-acid-recognizing sites identified by Day et al. (2024); residues 93 to 212 are shown. Sequences are from P. falciparum (from human), P. praefalciparum (gorilla), P. adleri (gorilla), and P. reichenowi (chimpanzee). Residues in site 1 (148, 149,152, 155, and 209) and in site 2 (98, 100, 132, 134, and 148) are highlighted in lighter font (blue). Residues at two sites (154 and 209) that were experimentally switched between the P. falciparum and P. reichenowi proteins (Day et al. 2024) are boxed in red; note that P. praefalciparum shares the same amino acid as P. falciparum at both sites. The one site within this region that differs between P. falciparum and P. praefalciparum is highlighted in red. See supplementary fig. S1, Supplementary Material online for the full sequence alignment from seven Plasmodium (subgenus Laverania) species.

Open in new tabDownload slide

Three of the differences between the CyRPA sequences of P. praefalciparum and P. falciparum (Y5F, I292V, and H320Y) reflect changes that likely occurred in P. praefalciparum, because the sequence of the human parasite is the same as that of P. adleri. The other four differences (S45N, Y91N, L187F, and K218N) reflect changes that likely occurred in the particular lineage of P. praefalciparum that gave rise to the human parasite, or after transmission from gorillas (supplementary fig. S1, Supplementary material online). To evaluate the possible significance of these changes, we examined their position in the RCR structure (Farrell et al. 2024). Notably, none of these sites lie in, or close to, the putative sialic acid binding sites (Fig. 3), and so it seems highly unlikely that these four changes would influence the glycan binding preferences of PfCyRPA. Given its sequence identity at all residues in or near the glycan binding sites, it seems likely that the P. praefalciparum CyRPA, like that of P. falciparum, prefers binding NeuAc over NeuGc.

Fig. 3.

Structure of P. falciparum CyRPA (aqua, filled) in complex with RH5 (pink, ribbon) and RIPR (gold, ribbon) (Farrell et al. 2024; PDB ID: 8CDD), visualized using ChimeraX (Meng et al. 2023); two views rotated by 180° are shown. Sites in the P. falciparum CyRPA (D98, T100, T132, H134, E148, I149, S152, I155, G209) involved in glycan binding (Day et al. 2024) are shown in purple. Sites that likely changed after transmission from gorillas (N45, N91, F187, and N218) are shown in red. Note that the carboxy-terminal loop of RH5 extends to near CyRPA N45, which is predicted to form a hydrogen bond with RH5 Q516.

Open in new tabDownload slide

Two of the sites that have changed in the recent evolution of P. falciparum CyRPA, N45 and F187, are close to the regions interacting with RH5 (Fig. 3). There are six residues in RH5 that appear to have changed in the specific lineage of P. praefalciparum that gave rise to P. falciparum, or after the transmission from gorillas to humans (Plenderleith et al. 2018; Galaway et al. 2019); however, none of these lie close to CyRPA in the RH5 structure (Farrell et al. 2024).

In summary, our data suggest that the gorilla-infecting precursor of P. falciparum was preadapted to infect humans not only because of changes in its RH5 protein (Galaway et al. 2019), but also because of the sialoglycan binding affinities of its CyRPA protein. Given that the two genes encoding these proteins were both introduced into an ancestor of P. praefalciparum by a horizontal gene transfer (Sundararaman et al. 2016), it seems likely that this event was a prerequisite for the zoonotic transfer that generated P. falciparum. However, how this transfer mediated changes to the glycan-binding properties of the CyRPA protein is not immediately obvious. Thus, it will be important to examine the CyRPA protein of P. praefalciparum (and also that of other gorilla parasites), in the context of the entire invasion complex, to determine whether it was already capable of efficiently binding human sialoglycoproteins; if this were the case, then no further adaptation of the P. falciparum CyRPA would have been needed. However, such a finding would raise the question of why the gorilla parasite acquired the preference for NeuAc, despite infecting a host with an intact CMAH enzyme. In the unlikely event that the P. praefalciparum CyRPA was found to preferentially bind Neu5Gc, then the means by which the P. falciparum CyRPA switched its preference to Neu5Ac would remain a mystery.

Supplementary Material

Supplementary material is available at Genome Biology and Evolution online.

Funding

This work was supported by the National Institutes of Health (grant numbers R01 AI120810, R37 AI150590, and P30 AI045008 to B.H.H.).

Data Availability

All sequence data are publicly available through the GenBank database at NCBI. Accession numbers and coordinates of the sequences are given in the Supplementary material online.

Literature Cited