Effects of population size on virus evolution: a baculovirus perspective

This thesis explores the population genetics of the baculovirus infection process and the consequences for virus evolution. Using Autographa californica multicapsid nucleopolyhedrovirus (AcMNPV) and lepidopteran insect larvae as a model system, we attempt to characterize (1) elemental virus-host and virus-virus interactions pertinent to virus invasion and disease of the host, focussing largely on the number of virus individuals invading a host insect and including the first strong experimental evidence that a single virion can cause disease in a host animal, and (2) the interactions that can occur between virus genotypes during the process of diseasing the host, specifically competition.
The development of two technologies was necessary to address these questions. First, a method for generating clonal baculovirus populations containing molecular tag sequences was devised: insect larvae were transfected with full-length baculovirus genomic DNA modified to allow replication in Escherichia coli (bacmids). Second, a quantitative real-time PCR (qPCR) based assay was developed to measure the frequency of a baculovirus genotype in a mixed population.
In order to estimate the number of pathogen individuals invading the host and causing death, a basic probabilistic model that links this number to host survival was developed and experimentally tested. This model is based on the independent action hypothesis (IAH): the notion that (1) every pathogen individual has a fixed probability of invading the host, (2) if a pathogen individual invades the host this will irrevocably lead to host death, and (3) that pathogen individuals act independently. Here the model was specifically developed to predict how often hosts are invaded by two pathogen genotypes, when being challenged by a pathogen population consisting of two genotypes. Model predictions were tested using two near-identical bacmid-derived virus genotypes, larvae of three host species, and the qPCR assay. IAH model predictions were confirmed in early instar (L3) larvae of permissive host species (Spodoptera exigua and Trichoplusia ni). This strongly suggests that in these instances it is the action of one invading virion which leads to host death. Model predictions did not hold for late instar larvae (L5) of permissive host species and both early and late larvae of a semi-permissive species (Mamestra brassicae), which show greater resistance to AcMNPV. In these instances there was a too high frequency of invasions with both genotypes, suggesting that more virions had invaded these larvae to cause disease than predicted by IAH.