Some pathogens have evolved to manipulate the behavior of their host. This behavioral manipulation may be beneficial for the pathogen and act as an adaptive manipulation strategy to increase the pathogen’s dispersal. The entomopathogenic fungus, Entomophthora muscae, infects house flies
(Musca domestica) and causes the fly host to exhibit an abnormal behavior, called summiting, where the fly will seek an elevated position and assume a stance beneficial for spore dispersal. These are the last acts of the fly before it dies, and shortly after death E. muscae extrudes from the cadaver and initiate sporulation. While the fungus-infected cadaver is discharging its infectious conidia, male flies has been observed to attempt copulation with this fungus-ridden cadaver, inevitably leading to infection of the male fly. This has led to hypotheses that the sporulating cadavers elicited cues which increased attraction from males as part of an adaptive manipulation strategy. In Chapter II of this
thesis, we investigated similar instances of entomopathogenic fungus-induced changes in sexual behavior through revision of already-published literature. Here, we found that only few instances exist, but we hypothesize that this understudied area likely is larger than previously thought and that increased awareness will lead to more discovered instances in the future.
In Chapter III, we investigated the hypothesis of E. muscae induced sexual behavior manipulation of un-infected conspecifics by using behavioral experiments of healthy flies towards conspecifics killed and sporulating with E. muscae, and coupled these with chemical- and transcriptomic analysis of the
infected fly cadavers. By quantifying the male sexual behavior towards female cadavers, we found that they exhibit a significantly increased sexual behavior towards E. muscae sporulating females, but only when the cadaver was in a late stage of sporulation. We also noticed that both healthy males and females exhibited feeding behavior to the E. muscae infectious conidia discharged from cadavers, which led us to investigate attraction to traps covered in these conidia. Both males and females were significantly attracted to conidia alone, indicating that these contain semiochemicals attracting to flies. We then investigated the antennal responses to headspace from sporulating cadavers and conidia and found significantly higher antennal response to all E. muscae headspace samples. Next, we utilized Gas chromatography-Mass Spectrometry (GC-MS) to investigate a known house fly pheromone fraction, cuticular hydrocarbons, in sporulating flies and found distinct changes in the profile of E. muscae. We also analyzed the collected headspace and found several novel compounds not otherwise associated with house flies, including ethyl octanoate and several unidentified sesquiterpenes. Lastly, we utilized transcriptomic analysis to determine if fly- or fungal genes were implicated in the production of these compounds. While expression of selected house fly genes were downregulated, we found several fungal genes related to ethyl octanoate and sesquiterpene synthesis, which were constitutively expressed and in some instances significantly upregulated throughout sporulation.
In Chapter IV, we assessed the potential of E. muscae as a future biological control agent by isolating three strains of E. muscae from Drosophila sp. and one strain from house flies, and experimentally infected selected fly lines. Here, we used three fly lines of the agricultural pest, Drosophila suzukii, two lines of its relative Drosophila melanogaster, and one line of house flies (Musca domestica), and assessed the fungal virulence of each strain on each fly line. We found that the fungal strains and fly lines exhibited a genotype-genotype infection compatibility and not one fungal strain was the most virulent in all fly lines. We furthermore investigated the fungal strains’ phylogenetic relationship by sequencing of the Internal Transcribed Spacer (ITS) with previously found E. muscae strains and found that E. muscae isolates from Drosophila sp. are distinct from E. muscae isolated from house flies. Additionally we found that all strains contains sequences of a previously described iflavirus, Entomophthovirus, which has previously been found to infect the fungus itself.
As summiting behavioral manipulation is a hallmark of E. muscae infection and it was induced in our experimental infections of house flies and Drosophila flies alike, we hypothesized that the fungus utilizes changes in house fly internal chemistry, such as secretion of specialized metabolites with a potential neuroactive effect. In Chapter V, we investigated this, by utilizing Liquid chromatography-Quadrupole-Time of Flight-Mass Spectrometry (LC-Q-TOF-MS) to assess the utilization of small metabolites during infection. We chose time points during the vegetative stages and two vital processes in the E. muscae infection, summiting behavior and sporulation, to obtain an overview of metabolome changes specifically associated with these processes. While we did not identify the compounds, a Factor Analysis of Mixed Data (FAMD)-analysis of the chemical features revealed metabolic shifts specifically related to summiting and sporulation, indicating that many specialized metabolites are implicated in these processes.
Collectively, these findings elucidate several parts of E. muscae host manipulation and contributes to the understanding of a host-manipulating EPFs lifecycle. In addition, we report one of the few known instances of pathogen-induced adaptive manipulation of host sexual behavior and perhaps the first
instance of a pathogen that not only manipulates the focally infected host, but also uninfected conspecifics.