Intra-specific variation in avian malaria : Linking infection dynamics to haplotypes

Abstract: Avian Malaria infects thousands of species of birds across the Aves class. The most widespread (geographically and phylogenetically)morphological species is Plasmodium relictum, and of that, the mitochondrial lineage of SGS1 is the most common and invasive. It infects146 bird species, with large differences in infection outcome, from low intensity chronic disease to shorter, highly virulent infections. InPaper 1, I studied the gene expression of SGS1 in experimentally infected birds pre and post maximum parasitemia. The parasitemiavaried widely between birds and especially over the course of the infection. These differences were significantly linked to genes relatedto cell replication and cellular movement in high parasitemia infections and cellular metabolism in low parasitemia infections. We foundthat over time, variation in gene expression increased between samples, possibly illustrating individual responses of the parasites to theirhosts, and a desynchronisation in their lifecycles. Paper 2 explored the phylogeography of SGS1, and its related lineage GRW11, in thepalearctic region. Because the lineage system is defined by a highly conserved single mitochondrial gene, the nuclear polymorphic cellinvasion gene, merozoite surface protein 1 (msp1), was selected to study the genetic variation present in infected resident and migrant hostpopulations. We found extremely little variation, suggesting SGS1and GRW11 in Europe have an epidemic population structure, or thereis strong purifying selection pressure on the msp1 gene despite the wide host range. Paper 3 developed a genomic sequence capture methodusing 1035 probes designed for SGS1, and tested it on a range of SGS1, GRW11, and GRW4 samples. The probes effectively isolatedDNA from all three lineages, but sequencing success was low for samples with less than 1% parasitemia. We selected 25 genes to describethe higher-than-expected variation within SGS1 and with GRW11 and GRW4 samples. In Paper 4, two different host sources of SGS1infected blood were used to infect two groups of canaries. The groups differed in parasitemia and mortality, and from each group the threebirds with largest differences in infection outcome were selected for RNA sequencing to survey the underlying genomic variation. Thesource of the infection reliably separated the samples phylogenetically, with relatively less variation observed within the groups. Thissuggests that an infection is made up of a population of genetically diverse parasites. Paper 5 expanded on this idea by using the genomicsequence capture method from Paper 3 and refined bioinformatic methods from Paper 4 on some of the same samples from Paper 1.Samples collected at the same time points (8 and 20 days post infection) were sequenced. This allowed analyses of how the predominanthaplotypes change during an infection, and then link those haplotypes to the disease severity. We found that the least suppressed/mostvirulent haplotypes had genetic variants in genes related to cell invasion and immune evasion. The combined results of my thesis havefar-reaching implications that extend beyond the particular organism under investigation. The notion of genetic diversity within a singleinfections and the resulting parasite population dynamics offers exciting prospects for future research.