Discussion
Here we report the genomic analysis of the largest collection (n=1474) of high-quality P. vivax genomes to date, originating from 31 countries across Asia, Africa, Oceania and Americas. The global P. vivax population displays high genetic diversity and separates into three main continental populations, i.e. Eastern Asia and Oceania, Africa and West Asia, and Latin America, similar to other reports . The geographical clustering largely matches the previously defined zoogeographical zones with similar ecological and epidemiological characteristics and P. vivax relapse rates . Most of the clusters identified in the admixture analysis are relatively isolated within one region, with little admixture between regions, with the exception of Africa and West Asia, as reported previously . In contrast, a high degree of admixture is observed within regions, especially in South-East Asia and in LAM. In this study we increased the number of genomes compared to previous reports, in particular from West Africa, Vietnam, Brazil and Peru, with the largest genomic analysis of Latin American isolates (n= 399) to date. With additional samples from South-East Asia compared to earlier studies , we do not find additional subpopulations in admixture analysis in this region. However, with the increased sample size in LAM, we do find an additional subpopulation in Latin America compared to earlier studies . This increased resolution is confirmed in another very recent study with additional samples from Brazil .
P. vivax isolates from Latin America form a distinct group within the global P. vivax population, characterized by structured subpopulations, sporadic clonal clusters at specific sites, and lower genetic diversity compared to other regions in the world. This analysis confirms the pattern of low-local but relatively high-regional genetic diversity in LAM measured previously using microsatellites .
The lower genetic diversity of P. vivax in LAM might be explained by a founder effect of relatively recent colonization events of this parasite species in this region, potentially through multiple waves of human migrations . In addition, a lower transmission intensity in LAM compared to other regions in the world can contribute to the lower diversity . In turn, Brazil, which is the country with the highest incidence of malaria -among those included in the LAM region in this study- exhibited the greatest level of genetic diversity. Furthermore, in comparison to genetic diversity of P. falciparum populations in high transmission African regions P. vivax diversity in LAM is high . This is typically the effect of random genetic drift in small populations that remain relatively isolated from each other; rare alleles disappear (decreased diversity) and increased divergence between sites. Gene flow and admixture were detected across populations from Mexico, Panama, and the North Coast of Colombia, and across countries in the Amazon region, however connectivity between these two separate regions is highly limited. This marked population differentiation between these two regions might be attributed to factors such as differentiated ecological environments, distinct vector populations, different human occupation and mobility patterns, and history of antimalarial interventions . These differentiation patterns are very similar to those observed in P. falciparum populations in South America . While the Amazon region and coastline regions of South and Central America are suitable habitats for various malaria vectors, there are ecological and human barriers between the Central and Northern South American coastlines and the Amazon region. These regions are separated by mountain ranges and have clear differences in climate and vegetation where the north coast region has a tropical climate with a dry season, while the Amazon rainforest has a tropical rainforest climate with year-round rainfall. The observed limited connectivity of parasite populations also suggests limited human mobility and thereby limited transmission between these regions.
Indeed, the Amazon basin is a vast territory of tropical rainforest encompassing several countries and including large regions of indigenous territories intersected by numerous rivers and scattered cities and villages that offer natural barriers contributing to the observed subpopulation structure . On the other hand, connectivity and admixtures within this region in all likelihood is due to patterns of human mobility, which are closely related to occupational patterns like logging and illegal gold mining . In addition, transmission of malaria (and gene flow) across large geographic distances is facilitated by the large human reservoir of asymptomatic and low-density P. vivaxinfections characteristic for this region as well as the dormant reservoir of hypnozoites characteristic for P. vivax in general .
Conversely, further North, in Colombia, Panama and Mexico we observe less admixture than in the Amazon region and distinct populations by country. We identified a specific subpopulation that is predominant in Colombia, but also observed (in a lower proportion) in Panama and even into Mexico likely reflecting human migration facilitating spread of this population from the North coast of Colombia into countries further to the North. According to a recent study, the majority of P. vivax genomes in Panama are part of a highly clonal population that has been present in the country for at least ten years . In addition, the same study reported likely imported parasites, which we identified to belong to Colombian population.
We identified genomic regions shared across Latin American parasite populations that exhibit signatures of positive selection. Genes in these shared regions are predominantly involved in DNA replication binding and repair, RNA transcription and processing, parasite invasion, as well as microtubule-related motility, suggesting that genes involved in these biological processes are key for P. vivax evolution and survival. Positive selection on DNA replication, binding, and repair genes suggests adaptive evolution to environmental challenges such us host immune responses, while positive selection on genes associated with RNA transcription and processing hints at the potential role of post-transcriptional gene regulation in parasite adaptation and survival to varying conditions , such as parasite transition between hosts or life cycle stages (such as the male and female development through MD1 and FD1). Moreover, microtubules are essential components of the cytoskeleton and play a critical role in cell division, motility, and intracellular transport.
The selection signals in LAM are dissimilar to P. vivaxpopulations in other regions, as previous investigations with globalP. vivax isolate collections detected selective sweeps at drug resistance-associated loci (e.g., dhfr, dhps , andmdr1 ), . However, these studies also found evidence of local adaptation within distinct P. vivax populations and differential selection on surface antigens such as msp genes. Results from this study are in agreement with previous studies of P. vivaxpopulations from South America , supporting the validity of our IBD-approach including polyclonal infections for selection analysis. In contrast to previous studies, we included sub-telomeric regions in the selection analysis. We identified genes under selection putatively involved in antigenicity and host–parasite interactions, including parasite invasion, which is poorly understood in P. vivax . Positive selection in invasion genes has the potential to identify new candidate genes with a role in reticulocyte invasion and hence, can inform vaccine development, for example against antigens such aslsa3 and msp3 found under selection here. In P. falciparum , lsa3 is a candidate vaccine target and has been used to elicit sterile immunity in animal models .
Sub-telomeric regions are often excluded in genome analysis due to challenges in aligning their highly variable and polymorphic short reads, particularly with the incomplete PvSalI reference genome, but newer genomes like PvP01 now provide improved assembly for these regions . Our results of positive selection in these sub-telomeric regions match well with a proposed model of P. vivax adaptation from a genomic study of P. vivax strains compared to P. cynomolgi andP. knowlesi . That study proposed that positive and negative selection might be less effective in low recombination areas (i.e. in the core genome). They suggested that in P. vivax, genome structure might be an adaptive mechanism to deal with changing environments, like the host’s immune system, in addition to adaptations through genetic variation. In contrast, in P. falciparum populations, which need to survive in the host for a longer period before they can be transmitted, strong selection signals and high IBD are found in the core genome, for example surrounding drug resistance associated genes .
This study used a convenience sampling approach, limiting the ability to generalize findings. The majority of genomes included were from Brazil, Peru, and the Colombian coastal region, while genomes from other highP. vivax burden areas such as Venezuela, Nicaragua, and the Amazon region of Colombia are lacking. To complete the picture of genetic diversity in Latin America genomes from these regions, often remote or challenging to access, need to be generated.
This regional analysis of P. vivax populations in Latin America highlights the significant genetic diversity within the continent, and regional adaptation of the parasites to their hosts and different environmental challenges, which may contribute to the resilience ofP. vivax to current malaria control strategies. Connectivity patterns between parasite populations are highly relevant for control and elimination programs in the region where human mobility is a major driver of malaria transmission. IBD-analysis confirmed connectivity between different ancestral clusters within and between countries, whilst showing a lack of connectivity between parasites in northern regions of Central and South America and the Amazon parasite population. While temporal differences in relatedness can serve as evidence for changes in transmission over time , the populations identified were very stable over time, indicating ongoing and stable transmission.
Genomic surveillance of parasitic pathogens, coupled with epidemiological data, offers valuable insights for control and elimination programs. This study highlights the use of genetic information to uncover patterns to address key epidemiological questions related to P. vivax in Latin America, and demonstrates how parasite genetic information can reveal transmission patterns and connections among populations. Furthermore, understanding population-level diversity is important for controlling malaria effectively, as it can provide insights into drug resistance development, immune evasion strategies, and transmission dynamics. For example, measuring genetic diversity and within-host infection complexity can indicate the intensity of P. vivax transmission in Latin America, similar to approaches used for P. falciparumpopulations . In contrast, these indicators have been less accurate forP. vivax in areas with higher genetic diversity like Southeast Asia and Oceania .
Finally, as malaria risk is increasing in spatial and temporal variability worldwide, the diversity of available malaria control tools continues to grow, and funding constraints on control programs increase, there is an increasing need for malaria-endemic countries to adopt intervention policies that move away from a one-size-fits-all approach to one that is specifically tailored to their subnational context. However, the high level of connectivity between regions and countries observed in this study rather shows that a wider view and regional approach remains very important in this continent and should remain incorporated in regional malaria control and elimination strategies.