Exploring Plasmodium vivax transmission dynamics and population genetics through genetics and genomics

© 2019 Abebe Alemu FolaGlobally, malaria remains one of the most life-threatening and highest burden infectious diseases. Amongst the six Plasmodium species that cause malaria in humans, Plasmodium vivax and Plasmodium falciparum are the predominant cause of morbidity and mortality. P. falciparum is the most virulent species; however, P. vivax has a wider geographic range and is more difficult to control and eliminate. The burden of malaria in Papua New Guinea (PNG) is among the highest in the Asia Pacific region with a hyperendemic transmission of P. falciparum and P. vivax. Since 2004, PNG has intensified malaria control efforts to alleviate this malaria burden and with the ultimate goal to eliminate malaria from the country by 2030. P. vivax is recognized as a major obstacle to achieve this elimination goal as evidenced by a smaller reduction in P. vivax prevalence relative to P. falciparum after intensified control efforts throughout the country. Indeed, in many areas where P. falciparum and P. vivax are co-endemic, P. vivax is now becoming the dominant malaria parasite. The main reason for this shift in malaria epidemiology may result from the use of malaria control measures primarily targeted at P. falciparum, and the fact that P. vivax is more difficult to treat and diagnose because of its distinct biology, in particular, its ability to form dormant liver-stage infections, known as hypnozoites. Therefore, approaches to eliminate P. vivax malaria need to address this unique parasite biology through the development and adoption of novel tools and strategies. To facilitate the discovery of novel strategies to contain Vivax malaria, this PhD project aimed to investigate P. vivax biology, molecular epidemiology and transmission dynamics using parasite population genetics, and through the development of a novel barcoding tool for high resolution genotyping. To address the existing knowledge gaps around the impact of the unique biology of P. vivax on malaria epidemiology, I compared the prevalence, multiplicity of infection (MOI) and genetic diversity of sympatric P. vivax and P. falciparum parasite populations throughout PNG (Chapter 3). A total of 892 P. vivax and 758 P. falciparum isolates from 16 Provinces (including 49 villages) were obtained from a 2008/2009 PNG national malaria survey completed after the nationwide intensification of control. All P. vivax and P. falciparum isolates were genotyped using the size polymorphic markers Pvms16/Pvmsp1F3, and Pfmsp2, respectively. The comparative analyses of within host multiplicity of infection (MOI) and population level genetic diversity showed striking differences between these sympatric Plasmodium species. Approximately 70% of P. vivax infected individuals carry more than one clone (i.e. MOI>1, known as polyclonal infections) compared to 20% for P. falciparum. There was a strong association between MOI and infection prevalence for P. falciparum but not for P. vivax, consistent with the maintenance of high infection complexity for P. vivax despite the reduction in new infections resulting from control interventions, and therefore suggesting a higher effective transmission potential for this species. The results showed that both species have high genetic diversity in populations throughout PNG. However, while immune selection on the antigen loci used (Pfmsp2 and Pvmsp1F3) and unusually high diversity of Pvms16, facilitates the detection of unique clones, it may limit the resolution of the comparative genetic diversity analyses. To further explore P. vivax genetic diversity and population structure throughout PNG, I then conducted a detailed population genetic analyses of P. vivax using a panel of 10 well-validated and putatively neutral microsatellite markers (Chapter 4). I genotyped a total of 230 P. vivax isolates with low complexity infections (maximum MOI=2) from eight geographically and ecologically distinct regions (“geographic populations”) throughout PNG. Population genetic analyses of microsatellite haplotypes/genotypes revealed a spectrum of genetic diversity, which was associated with prevalence. Multilocus linkage disequilibrium (mLD) was identified in six of the eight parasite populations indicating significant inbreeding. Strong geographic population structure (clustering of isolates according to their geographic origins) was identified among regions, dividing the Mainland (lowlands), Highlands and Islands, and the genetic divergence was significantly correlated with geographic distance. Sources and sinks of vivax transmission and parasite population connectivity between regions were defined. Patterns of P. vivax gene flow appear to follow major human migration routes, implying that the human host could be the main driver for the spread of infections and potentially, drug resistance. This find suggests assessing the contribution of the P. vivax asymptomatic reservoir and imported infections for sustainable local transmission are necessary to design targeted control interventions. Small panels (~10) of microsatellite markers used have been the “gold standard” genotyping tool for population genetic analyses of P. vivax. However, because of their small number, extremely high diversity and propensity for genotyping error they may have limited resolution for identifying local population structure in high transmission areas (such as the north coast of PNG). In addition, several other technical challenges make them unsuitable for large scale malaria surveillance. Therefore, I developed a novel genotyping tool known as a ‘single nucleotide polymorphism (SNP) barcode’ for high resolution monitoring of P. vivax parasite populations in PNG (Chapter 5). The barcode was specifically designed to capture the genetic diversity of PNG parasite populations by using SNPs identified amongst 20 high quality published PNG genomes. A total of 24283 SNPs with minor allele frequencies (MAF) of more than 10% were identified in putatively neutral regions of the P. vivax parasite genome. A filtered subset of 178 evenly spaced informative SNPs were selected for development of a high throughput parallel targeted amplicon sequencing (PTAS) assay using a multiplex PCR and Illumina MiSeq platform. To validate the developed barcode, I tested the genotyping success of 20 SNPs using small number of mono- and polyclonal infections, followed by genotyping of 96 P. vivax isolates from four catchment areas on the north coast of PNG collected in 2012-2014. Compared to the available microsatellite data for the same samples (E. Kattenberg, unpublished), the SNP barcode revealed more variable genetic diversity and stronger population structure with greater resolution to assign genotypes according to their geographic origins than the microsatellite panel. The developed SNP barcode has low genotyping error, is relatively cheap (~$-26-30/isolate) and transfer of the technology to field settings like PNG is technically feasible. As transmission declines, measuring changes in parasite population structure is important to assess the impact of control on malaria transmission, and to guide future control and elimination of malaria. To explore P. vivax population genetic signals with changing transmission due to intensified control, I applied the developed SNP barcode assay to a total of 376 P. vivax isolates obtained from serial cross-sectional surveys in two north coast Provinces of PNG (East Sepik and Madang) (Chapter 6). Samples were collected between 2005 and 2016, a period of initially declining transmission and later rebound. This investigation revealed interesting spatio-temporal patterns with declining transmission including a reduction in polyclonal infections, genetic diversity and population connectedness as indicated by an increase in population structure. Significant multilocus linkage disequilibrium (mLD) and a large cluster of parasites with high genomic relatedness (identity by descent, IBD) was also detected at low transmission. These population genetic signatures signify a strong population bottleneck, highly clonal transmission and inbreeding resulted from significant reductions in prevalence. For Madang Province, I also observed some genetic signatures related to the sources of malaria rebound providing insight into malaria resurgence through increasing migration of parasites from other endemic areas in addition to expansion of the residual parasite population. Containment of P. vivax infections and drug resistance may be facilitated by the existence of distinct parasite populations at low transmission yet will be undermined by increased parasite migration due to frequent human movement between endemic areas. Overall, this thesis demonstrates how genotyping a relatively small number of parasite isolates from representative geographic areas can provide surrogate signals of malaria transmission dynamics and infection spread. Moreover, it reveals the power of molecular epidemiology and population genetics and genomics for malaria surveillance that in combination with epidemiological data will allow the prioritisation and optimisation of malaria control and elimination strategies. I also discuss potential approaches to integrate population genetics and genomics into current malaria surveillance systems and how the control efforts can be benefited from this approach