Human migration and the spread of malaria parasites to the New World

Abstract

We examined the mitogenomes of a large global collection of human malaria parasites to explore how and when Plasmodium falciparum and P. vivax entered the Americas. We found evidence of a significant contribution of African and South Asian lineages to present-day New World malaria parasites with additional P. vivax lineages appearing to originate from Melanesia that were putatively carried by the Australasian peoples who contributed genes to Native Americans. Importantly, mitochondrial lineages of the P. vivax-like species P. simium are shared by platyrrhine monkeys and humans in the Atlantic Forest ecosystem, but not across the Amazon, which most likely resulted from one or a few recent human-to-monkey transfers. While enslaved Africans were likely the main carriers of P. falciparum mitochondrial lineages into the Americas after the conquest, additional parasites carried by Australasian peoples in pre-Columbian times may have contributed to the extensive diversity of extant local populations of P. vivax.

Figure 1

Bayesian phylogenetic tree (A) and median-joining network (B) of the global sample of Plasmodium falciparum mitochondrial lineages (n = 1795). Circle sizes in B are proportional to haplotype frequencies, and pairs of haplotypes connected by a straight line differ by a single mutational step. The following color code was used to identify the geographic origin of parasites: red = Africa, dark blue = South America, light blue = Central America, orange = South Asia, green = Southeast Asia, and pink = Melanesia. Branches with posterior probabilities >0.70 are indicated by black circles in the phylogenetic tree; selected well-supported clades indicated in the figure (Ame1, Ame2, Global1, and Global2) are further discussed in the main text.

Figure 2

Bayesian phylogenetic tree (A) and median-joining network (B) of the global sample of Plasmodium vivax mitochondrial lineages (n = 941). Circle sizes in B are proportional to haplotype frequencies, and pairs of haplotypes connected by a straight line differ by a single mutational step. The following color code was used to identify the geographic origin of parasites: red = Africa, dark blue = South America, light blue = Central America and Mexico, yellow = Atlantic Forest from southeast and South Brazil, brown = Middle East and Central Asia, orange = South Asia, green = Southeast Asia, dark purple = East Asia, and pink = Melanesia. Branches with posterior probabilities >0.70 are indicated with black circles in the phylogenetic tree; selected well-supported clades indicated in the figure (Ame1, Ame2, Atl, Afr1, Mel1, Sea1, and Sea2) are further discussed in the main text.

Figure 3

Median-joining network of Plasmodium vivax/P. simium mitochondrial lineages from the Atlantic Forest of Southeast and South Brazil. Circle sizes are proportional to haplotype frequencies, and pairs of haplotypes connected by a straight line differ by a single mutational step. Yellow indicates samples from monkeys, and blue indicates samples from humans. Haplotypes Atl1, Atl2, and Ame1 are indicated. Sample collection sites and dates, as well as their respective hosts, are listed in Supplementary Table 7; sample collection sites are plotted on a map in Fig. S2.

Figure 4

Magnitude and directionality of historical gene flow between regional populations of Plasmodium falciparum (A) and P. vivax (B and C). Estimates of median mutation-scaled pairwise migration rates obtained with the best-supported migration model for each species are shown next to the arrows. Migration models tested are described in Supplementary Fig. 13 and compared in Supplementary Tables 14 and 15; the major difference between the models shown in B and C is that the former assumes an out-of-Africa spread of P. vivax, whereas the latter assumes a Southeast Asian origin of this parasite. The geographic origins of mitochondrial lineages are indicated on the map at the country level using the same color code as those of Fig. 1 (for P. falciparum) and Fig. 2 (for P. vivax) to represent geographic regions. Maps were built using the open-access R software library rworldmap: mapping global data combined with the ggplot2 library, which are both available at http://www.R-project.org/ (R Core Team, R: A Language and Environment for Statistical Computing.Vienna, Austria: R Foundation for Statistical Computing, 2017).