Zoonotic origin of the human malaria parasite *Plasmodium malariae* from African apes

Plasmodium malariae is one of six Plasmodium species that commonly infect humans, yet it remains comparatively understudied due to typically low parasitemias and under-detection. While the origins of P. falciparum and P. vivax have been traced to African ape parasites, the evolutionary history of P. malariae has been unclear despite long-known close relatives in non-human primates. Historically, a morphologically indistinguishable quartan parasite in chimpanzees was named P. rodhaini, and cross-infection studies suggested limited host specificity between human and ape parasites. Additionally, P. brasilianum infects New World monkeys and is nearly identical at limited loci to P. malariae, leaving the direction of host transfer debated. This study aims to resolve how many P. malariae-related species exist, their host ranges, and the origin of human P. malariae by generating and analyzing mitochondrial and nuclear sequences from wild and captive African apes and mining existing genomic datasets.

Early 20th-century microscopy identified a P. malariae-like parasite in chimpanzees (P. rodhaini), with later experiments demonstrating reciprocal infectivity between humans and chimpanzees via blood inoculation and mosquito transmission, blurring species boundaries. Limited molecular studies, often restricted to mitochondrial cyt b and a few nuclear loci, detected P. malariae-related parasites in wild chimpanzees, gorillas, and bonobos, and in mosquitoes feeding on apes. Two chimpanzee-derived genome sequences (GA01 and GA02) closely related to P. malariae were previously labeled as a separate species (P. malariae-like), hypothesized to have diverged >3 million years ago. For New World monkeys, P. brasilianum sequences are near-identical to human P. malariae, and experimental transmissions have occurred both ways, raising debate over whether P. brasilianum is an anthroponosis from humans or the source of human P. malariae. These mixed findings motivated broader genomic and phylogenetic analyses.

- Sampling: Blood (including dried blood spots) from captive and wild chimpanzees and a gorilla across Central and West Africa; non-invasive fecal samples from habituated apes at multiple sites across equatorial Africa. Host mtDNA confirmed species/subspecies; permits and CITES regulations followed. - Diagnostic screening: Samples screened for P. malariae-related parasites using PCR targeting mitochondrial cyt b (~956 bp) and additional mitochondrial fragments, followed by phylogenetic assignment. - Sequence generation: Applied single-genome/limiting dilution PCR (SGA) to minimize PCR artifacts and chimeras, amplifying mitochondrial (cyt b) and nuclear loci (e.g., ldh, eef). Derived additional mtDNA genomes for GA01 and GA02. - Selective whole genome amplification (SWGA): Designed short primers (6–12 bp) frequent in P. malariae genome but rare in human DNA. Multiple rounds with alternate primer sets (e.g., m1_37, m1_28) amplified parasite DNA from ape-derived samples; products were purified and sequenced (MiSeq). - Data mining and assembly (M2 lineage): Mined public sequencing datasets (e.g., from mosquito blood meals and ape samples) by mapping reads to the P. malariae reference (Pm01) to extract P. malariae-related reads. De novo assembly (SPAdes), iterative scaffolding (ABACAS), and polishing (IMAGE), with careful curation to separate M1/M1-like from M2 contigs based on nucleotide identity thresholds. Annotation transfer (RATT) and manual curation; calculated genetic distances in R. - Read processing and variant calling: Filtered host reads; mapped to references; variant calling with GATK HaplotypeCaller; curated callable sites; removed problematic or heterozygous calls in single-clone data. - Phylogenetics: Alignments with Clustal Omega, MUSCLE, or TranslatorX; trees with RAxML (GTR+G for nucleotides, ProtGAMMAJIT for amino acids); phylogenetic networks with SplitsTree (Median-Joining) using SNP sets (e.g., 990 SNPs in 45 kb; ~2,050–2,500 SNPs across ~1.1 Mb) after masking non-callable regions. - Population genetics and diversity: Compared pairwise nucleotide diversity at zero- and fourfold degenerate sites; computed N/S ratios, neutrality index (NI), Direction of Selection (DoS) per gene; assessed transition/transversion ratios; site frequency spectra (SFS) at fourfold sites; sliding-window analyses across the core genome to detect regions of elevated diversity/divergence (e.g., ~45 kb region on chromosome 10).

- Identified three distinct lineages within the P. malariae-related clade: 1) M1: Human-infective P. malariae, including P. brasilianum and some ape-derived samples. 2) M1-like: An ape-infective lineage (from chimpanzees, bonobos, gorillas) very closely related to M1. 3) M2: A previously unknown, highly divergent ape parasite species infecting chimpanzees, bonobos, and gorillas across Central/West Africa; all P. malariae-related sequences from ape-fed mosquitoes clustered with M2. - Phylogenies from mitochondrial cyt b and nuclear loci robustly separated M1 from M1-like and both from M2 with strong bootstrap support. - Minimal evidence of recombination/genetic exchange between M1 (human P. malariae) and M1-like (ape lineage), consistent with long-term reproductive isolation. - Human P. malariae (M1) shows markedly reduced genetic diversity relative to M1-like: - Transition/transversion ratio similar or slightly higher in M1 (1.31) than M1-like (1.25), arguing against sequencing error as the diversity driver. - Pairwise nucleotide diversity difference between lineages is ~32-fold at zerofold sites and ~56-fold at fourfold sites, indicating a relative excess of nonsynonymous polymorphism in M1. - N/S ratio for polymorphisms is higher in M1 than M1-like; genome-wide DoS distribution in M1 is shifted negative, consistent with recent population expansion after a bottleneck. - Fixed divergence between M1 and M1-like is ~0.0069 substitutions per site. - A single short region (~45 kb on chromosome 10; ~0.2% of the core genome) exhibits elevated diversity among M1 strains; otherwise, little sign of introgression from ape parasites into human P. malariae. - P. brasilianum nests within the diversity of human P. malariae and clusters with African M1 strains in SNP networks, supporting a recent anthroponosis (human-to-New World monkey transfer), likely following human movements from Africa to the Americas. - Collectively, patterns of diversity and phylogenetic placement indicate human P. malariae arose via recent zoonotic transmission from African apes, analogous to the origin of P. falciparum from gorilla Laverania.

The study clarifies the evolutionary relationships among P. malariae-related parasites. Discovery of two ape-infective species (M1-like and M2) alongside human P. malariae (M1) shows that apes harbor multiple cryptic species previously conflated under P. rodhaini. The close relationship yet strong genomic separation and low diversity of M1 support a recent zoonotic origin of human P. malariae from an ape parasite, followed by a severe bottleneck and rapid expansion. The near absence of introgression from ape parasites suggests current reproductive isolation and effective speciation of human P. malariae. This mirrors the pattern seen in P. falciparum’s origin from gorilla parasites in Laverania, though host specificity differs across clades. The placement of P. brasilianum within African-lineage human P. malariae indicates a recent human-to-New World monkey transfer, resolving a long-standing debate. These findings refine our understanding of malaria parasite zoonoses, host switching, and their implications for surveillance and control.

This work demonstrates that human P. malariae originated via zoonotic transmission from African apes and that apes host two distinct P. malariae-related species: a closely related M1-like lineage and a highly divergent M2 lineage. The human lineage exhibits signature reductions in genetic diversity and an excess of nonsynonymous polymorphisms consistent with a bottleneck and expansion, with minimal subsequent introgression from ape parasites. P. brasilianum represents a recent anthroponosis nested within African human P. malariae diversity. These results place the origin of the P. malariae-related clade in Africa and clarify host switch directions. Future research should expand genomic sampling of ape parasites (especially M2), assess vector competence and transmission barriers, investigate the functional basis of host specificity, and conduct genomic surveillance to detect rare cross-species transmission events.

- Limited numbers of high-coverage genomes for ape lineages, particularly M1-like (only a few genomes) and M2 (assembled from mixed and low-identity contigs), may reduce resolution of divergence timing and gene flow. - Some analyses rely on selective whole genome amplification (SWGA) and mined reads from heterogeneous datasets, which can introduce coverage biases and assembly artifacts despite careful curation. - Phylogenetic network resolution near the base of African M1 is low due to dominance of differences between M1 and M1-like, limiting fine-scale inference of within-lineage ancestry. - Geographic and host sampling remains uneven, and noninvasive fecal sampling yields lower parasite DNA quality, potentially limiting detection of mixed infections and recombination. - Historical transmission experiments inform host range but may not reflect current natural transmission dynamics or vector compatibility.