To explore the diversity of species belonging to the Plasmodium
AGAH-lineage in Africa, we collected blood samples from 17 chimpanzees recently trapped from the wild and kept as pets in villages of Gabon by hunters and their families (see Figure S1
). Considering that only the subspecies Pan troglodytes troglodytes
has been found in Gabon, these 17 animals are likely to belong to this chimpanzee subspecies.
Among them, two were found to be infected with Plasmodium
by means of PCR assay or microscopy. The other 15 animals were found negative both by microscopy and PCR assay. For these two chimpanzees (named B and K), observed parasites under microscopy were falciparum
-like (ring stages with two chromatin dots and presence of multiply-infected red blood cells 
). Thick blood smears revealed low parasitemia in both individuals, approximately 300 parasites/µl for chimpanzee B and 2000 parasites/µl for chimpanzee K.
For both, we amplified and sequenced the parasite's Cytochrome b (Cyt b) gene. The Cyt b sequences obtained were similar between the two samples (identity of 99.8% based on 866 nucleotides (nt)), but different from all other Plasmodium Cyt b sequences known to date. The most similar sequences obtained using BLAST were Cyt b sequences from P. reichenowi and P. falciparum, which show 92% and 91% identity, respectively.
Because the Cyt b sequences were partial, we studied the whole mitochondrial DNA (mtDNA) of these two new isolates (named P. sp_K and P. sp_B). For isolate K, we amplified 5529 nt including three main genes: Cytochrome oxydase I (Cox I), Cytochrome oxydase III (Cox III) and Cytochrome b (Cyt b). Apart from short missing segments amounting to 420 nt, the mtDNA sequenced corresponds to the whole P. falciparum 3D7 mtDNA (5949 nt). For technical reasons (certainly due to the very low parasitemia and degraded DNA), we were unable to accomplish this sequencing for isolate B.
In order to determine the evolutionary relationships of this new Plasmodium
relative to other species, we compared its sequence to 17 known complete Plasmodium
mitochondrial genome sequences, with the bird apicomplexan parasite Leucocytozoon caulleryi
as an outgroup (see Table S1
). Maximum likelihood (ML) phylogenetic trees were reconstructed at both the nucleotide and amino acid levels on the whole mitochondrial genome sequence, considered as a single genetic unit 
DNA and protein analyses provided identical results: the parasite collected in chimpanzee K belongs to the AGAH-lineage but is more divergent from P. falciparum
than is P. reichenowi
(; see also Figure S3
for the tree reconstructed from the partial Cyt
b sequence (866 nt) and including both P. sp
_K and P. sp
_B). Over the entire mitochondrial genome, the genetic distance observed between the new taxon (P. sp_K
) and P. falciparum
0.213 substitutions per nucleotide site on the ML phylogram) or P. reichenowi
0.215) is almost four times higher than the distance observed between P. falciparum
and P. reichenowi
Phylogenetic relationships among Plasmodium species (including P. sp_K) and associated host groups.
To estimate the divergence time of the plasmodium AGAH-lineage, we used a calibration chosen within the hominid hosts. The age of the P. falciparum
split is generally considered to be similar to the one separating humans from chimpanzees 
, that is, between four and seven million years 
. Because of pervasive variations of mitochondrial substitution rates among malaria parasite lineages (), a Bayesian relaxed molecular clock was used, which revealed a divergence time of 21±9 Myrs between the new Plasmodium
species and the clade constituted by P. falciparum
and P. reichenowi
. Interestingly, this estimated time frame fits with the radiation of hominoids during the Miocene 
. Our results suggest therefore that the plasmodium AGAH-lineage may have been present in early hominoids 
and that this lineage may have also experienced a diversification during the early Miocene period as it occurred for their hosts 
. Obviously, this estimated time of divergence is dependent on the calibration used. Recently, Martin and colleagues 
suggested that the split between P. falciparum
might have occurred far earlier than previously considered. They propose that P. falciparum
originated from a recent transfer of P. reichenowi
to humans during the last 2.8 Myrs 
. Under this hypothesis, the new species would have diverged from P. falciparum/P. reichenowi
about 10 Myrs ago. As divergence data are lacking for those parasites from the fossil record, we are unable to distinguish between these two hypotheses. Further data on the diversity of Plasmodium
infecting great apes in Africa will certainly help resolve this particular aspect of the evolution of P. falciparum
In conclusion, we bring to light the existence of a new Plasmodium
species that infects chimpanzees in Gabon. We propose to name this new species Plasmodium gaboni
sp. nov. in reference to the country where we obtained it. Our discovery suggests that great apes and perhaps simian primates may host a far higher diversity of Plasmodium
species in Africa than previously recognised. Beyond the interest of this new species in the understanding of the evolution of this group of parasites, its position in the AGAH-lineage as the sister-group of P. falciparum
opens up the possibility of exploring lineage-specific evolution using comparative genomics, and hence, to look for the genes responsible for the adaptation of these parasites to their specific hosts. Comparison between genomes will advance understanding of the differences in pathology and the processes at work in the interaction with the vertebrate or the mosquito hosts 
. It is thus essential to complete the nuclear genome sequence of this new species of phylogenetic importance within the AGAH-lineage, in order to enhance our knowledge of the functional genomics of human malaria parasites.
Finally, this new species was discovered in two chimpanzees conserved as pets by villagers in Gabon. Given the recent history of primate to human shifts in several pathogens (e.g. HIV 
; Ebola 
; for Plasmodium
, the most recent involved P. knowlesi
and occurred from macaques to humans in Asia 
) and the close proximity between P. gaboni
and the most virulent agent of malaria, P. falciparum
, we think that the risk of transfer of this species to humans must be seriously considered.