A Feast of Malaria Parasite Genomes

Abstract

The Plasmodium genus has evolved over time and across hosts, complexifying our understanding of malaria. In a recent Nature paper, Rutledge et al. (2017) describe the genome sequences of three major human malaria parasite species, providing insight into Plasmodium evolution and raising the question of how many species there are.

A budding malariologist in the 1990s was steeped in the dogma that there were four species of malaria parasite responsible for millions of cases of malaria in tropical and sub-tropical regions of the world. Plasmodium falciparum was the deadliest, causing two to three million deaths each year due to the development of cerebral malaria in children under the age of 5 in sub-Saharan countries. Plasmodium vivax, on the other hand, was the comparatively “benign” species found in South American and Central and Southeast Asian countries, little studied because of the greater threat posed by P. falciparum and the lack of an in vitro culture system. Plasmodium malariae and Plasmodium ovale were species on the periphery, perceived as causing so few infections as to be almost negligible in impact. A clade of monkey malaria species, primarily studied because of their potential as in vivo models for P. falciparum and P. vivax, were also known to infect Old World primates in Southeast Asia and included Plasmodium knowlesi, Plasmodium cynomolgi, and Plasmodium inui.

Fast forward 20 years and how things have changed! The number of global human malaria infections has dramatically declined, and there now exists a greater understanding of the importance and threat that P. vivax presents and acknowledgment that this species requires its own research and elimination agenda (WHO, 2015). Compelling evidence also exists that P. ovale is in fact two species, P. ovale curtisi and P. ovale wallikeri (Sutherland et al., 2010), morphologically similar and inhabiting sympatric ranges but genetically distinct. And most astonishing of all has been an evaluation of the zoonotic (host-switching) potential of Plasmodium species, due to pioneering work by Balbir Singh and colleagues identifying P. knowlesi as a species causing a heavy burden of clinical infections in Malaysia (Singh et al., 2004) and several groups providing evidence for the existence of additional Plasmodium species naturally infecting wild-living chimpanzees and gorillas (reviewed in Loy et al., 2017). Indeed, among the latter findings include evidence that P. vivax found in African apes has the potential to infect humans, suggesting that human and ape P. vivax parasites represent a single species. P. malariae-like and P. ovale-like sequences too have been found in African great apes, although their zoonotic potential remains unclear.

While reference genome sequences of P. falciparum, P. vivax, and P. knowlesi have been published, the community has been waiting anxiously for high-quality genomes of P. malariae and the two species of P. ovale. Excitingly, the paper by Rutledge and colleagues in the January 2017 issue of Nature (Rutledge et al., 2017) does not disappoint and presents reference/high-quality drafts, as well as shotgun sequences, for five P. malariae, two P. malariae-like, three P. ovale curtisi, and two P. ovale wallikeri isolates (Table 1). For several of these, a means to identify pre-existing sequence data (“co-infection mining”) involved screening the sequencing reads of ~2,500 samples from the MalariaGen Pf3K global collection generated at the same genome-sequencing center, the Wellcome Trust Sanger Institute, where many of the authors are based. Fortuitously (although not for the patients), four samples of P. malariae were obtained from travelers returning to Australia from Uganda, Papua Indonesia, Malaysia, and Guinea, and one of them contained sufficiently high parasite DNA with low human host contamination for it to be used to generate a high-quality, manually curated P. malariae reference genome.

Table 1.

The Current Satus of References, Draft Assemblies, and Shotgun Sequences for Six Human-Infective Species of Malaria Parasite

	Pf	Pv	Pk	Pm	Poc	Pow
No. genomes sequenced	1 reference*, 2 draft assemblies*, 1,000 s shotgun sequences	1 reference*, 5 draft assemblies, 100 s shotgun sequences	1 reference*, ~50 shotgun sequences	1 reference, 7 draft assemblies*/shotgun sequences	1 reference, 3 draft assemblies*/shotgun sequences	1 reference, 2 draft/shotgun sequences
Geographical location of sequenced genomes	Global	Global	Malaysia	Uganda, Papua Indonesia, Mali, Malaysia, Guinea	Ghana, Cameroon	Ghana, Cameroon
Zoonotic	No	Possibly	Yes	Yes	Not known	Not known
Notable multi-gene families	pir, var, tra, ETRAMP, PHIST, EBP	pir, STP1, tra, ETRAMP, PHIST, RBP, DBP	pir, SICAvar, tra, ETRAMP, PHIST, RBP, DBP	pir, STP1, tra, ETRAMP, PHIST, RBP, DBP, fam-l, fam-m	pir, STP1, tra, ETRAMP, PHIST, RBP, DBP	pir, STP1, tra, ETRAMP, PHIST, RBP, DBP

Pf, P. falciparum; Pv, P. vivax; Pk, P. knowlesi; Pm, P. malariae; Poc, P. ovale curtisi; Pow, P. ovale wallikeri. Data compiled from Ansari et al. (2016) and Rutledge et al. (2017).

^*indicates one or more assemblies generated from culture-adapted or chimpanzee/monkey-adapted strains.

What are the major findings of this work? First, the P. malariae genome appears to be indistinguishable from sequences of a species, Plasmodium brasilianum, found in New World primates, confirming a recent report describing them as a single anthrozoonotic species circulating freely between monkeys and humans in the Venezuelan Amazon (Lalremruata et al., 2015). Second, using 1,000 manually curated orthlogs in 12 Plasmodium species, Rutledge et al. (2017) constructed maximum likelihood trees that appear to show rodent malaria parasites as an outgroup to P. ovale. This intriguing finding suggests that there was an ancestral host switch from primates to the ancestor of African thicket rats that host the rodent malaria species. If this finding holds true, it may well be that rodent malaria parasites could be a good model system for the study of P. ovale compared to other human-infective species.

Finally, one goal of “malariomics” is to relate the preference of different Plasmodium species for specific hosts and developmental stages of red blood cells (RBCs) to genomic differences. P. ovale and P. vivax, for instance, infect immature RBCs of humans, while P. malariae prefers mature RBCs in humans and monkeys, and P. knowlesi infects both RBC stages and both hosts. Understanding the genomic bases of these distinctions would doubtless shed light on the mechanisms of zoonoses as well. As Plasmodium genomes have become available, it has become clear that what mainly distinguishes them are their multigene families. Such families can comprise a few or hundreds of genes, typically encoding proteins expressed on the parasite surface or exported to the cytoplasm and surface of infected RBCs, with putative functions that include cell adhesion and invasion, and evasion of the host’s immune system. A species or strain’s unique genomic repertoire of such proteins might determine its host and host cell range. But multigene families tend to cluster in the repetitive, difficult-to-assemble subtelomeric regions of chromosomes, making it hard to tally gene numbers accurately. Rutledge et al. (2017)’s use of long reads using the Pacific Biosciences next-generation sequencing platform, in combination with extensive manual curation, allows them to overcome that common obstacle and upgrade earlier analyses that were based on much lower-quality assemblies (Ansari et al., 2016). They quantitate large differences between P. ovale and P. malariae in gene complements of the pir, STP1, and surfin families of surface/exported proteins and, most notably, discover two large, previously undistinguished gene families in P. malariae that they relate to a known P. falciparum RBC invasion protein using 3D protein modeling. They also identify, from surveying the rbp family of RBC adhesion proteins across Plasmodium species, one member that may be required specifically for invading mature RBCs—an intriguing finding given that the same protein (RBP3, also called NBPXa) in P. knowlesi has been shown to be necessary for invading human RBCs (Moon et al., 2016).

While the study by Rutledge et al. (2017) provides missing pieces in the puzzle of the complex evolution and phylogeography of the Plasmodium genus, it also illustrates the benefits and drawbacks of sequencing laboratory-adapted strains versus extant clinical or field isolates. Whereas previous human malaria reference strains were chosen for sequencing because of their availability to researchers as pure frozen stocks from the Malaria Research and Reference Reagent Resource (https://www.beiresources.org/MR4Home.aspx) and the body of research describing them, this came at a cost. For example, many were isolated decades ago and may not be representative of today’s genetic diversity, and others were adapted to growth in a novel system, such as in vitro culture, a procedure now known to induce loss-of-function mutants that are not normally seen in natural isolates (Claessens et al., 2017). The alternative—to generate high-quality genome sequences from clinical isolates that have not been previously studied and for which there may be limited biological material for others to use—also has constraints. It’s wise to remember, then, that there really is no such thing as the reference genome, a one-size-fits-all, for a Plasmodium species, especially in the light of extensive genetic diversity and differences in gene copy number (Hupalo et al., 2016) that can exist even within a single species of this fascinating lineage.