Files
Abstract
Viruses are the most abundant, most diverse, and least understood organic structures on earth. While the best-studied viruses are pathogenic, many viruses engage in commensal or even mutualistic relationships with eukaryotes. In the pea aphid, Acyrthosiphon pisum, a facultative bacterial mutualist, Hamiltonella defensa, provides anti-hymenopteran parasitoid defense only when it is itself infected by virus APSE (Acyrthosiphon pisum Secondary Endosymbiont). While this is the only known instance of mutualism-by-proxy, the prevalence of mobile genetic elements in symbionts makes it unlikely that it will be the last. At present, however, the diversity, prevalence, and effects of viruses infecting insect-associated bacteria are largely unknown. I conducted four studies to elucidate the biology of the tripartite aphid/H. defensa/APSE interaction. Firstly, within-host bacterial abundance can affect conferred phenotypes and rates of transmission. All stable beneficial heritable symbiont infections must have sufficient titer to produce the beneficial phenotype and ensure vertical transmission to progeny, without engaging in over-replication that might affect host fitness. Using quantitative-PCR and standard fitness tests, I determined that in the absence of APSE H. defensa replicates at rates detrimental to its hosts development and reproduction. Secondly, previous sampling of North American APSEs in A. pisum was limited to a handful of samples from two New York counties and one in Utah. I tested more than 1100 aphids from six states for H. defensa presence and phage strain identity: in the process I identified a new strain of APSE, called APSE-8, and I present its associated defensive phenotype here. Thirdly, most studies examine effects of single heritable symbiont infection, but ~25% of individual aphids carry multiple facultative symbiont lineages despite previous reports of lower fidelity in the transmission of double infections and additional costs to hosts of superinfection I created artificially differentially-infected clonal lineages using H. defensa and its sister species, the anti-fungal defender Regiella insecticola, and determined that, under most conditions, coinfection is actually beneficial. Finally, I present the first pass at parsing aphid transcriptional responses to wasp parasitism in the presence and absence of aphid genotypic resistance and H. defensa-induced resistance, finding little-to-no evidence for the mobilization of innate immunity.