Changes in Gene Expression and Viral Titer in Varroa Jacobsoni Mites After a Host Shift Asian to European Honey Bees

Honey bees (Apis mellifera L.) are the most important insects for the pollination of crops and wildflowers. However, they have experienced increasing colony die-offs during the past two decades. Multiple species of parasitic mites have been described that affect honey bees. The most important species in beekeeping belong to the genus Varroa ( Varroa jacobsoniand Varroa destructor). Varroa mite parasitism of honey bees is thought to be the most significant cause of colony mortality worldwide, and mite resistance to active ingredients of acaricides has become common. V. destructor causes direct impacts on bee production as well as indirect effects on bee health by vectoring viruses and other pathogens. These large ectoparasitic mites are associated with a condition known as parasitic mite syndrome, or PMS. When colonies exhibit PMS pathogens, brood diseases and viruses are present at unusually high levels. The open wound caused during feeding can allow microorganisms to enter and weaken the host, and mites themselves are vectors for viruses and perhaps other bee pathogens. There are a number of studies that also suggest that the primary cause of colony mortality is the viruses associated with the mites. Until recently, Varroa jacobsoni was known to only live and reproduce in Asian honey bee (Apis cerana) colonies while V. destructor successfully reproduces in both A. cerana and A. mellifera colonies. However, we have sampled an island population of V. jacobsoni that is highly destructive to A. mellifera, the primary species used for pollination and honey production. These recently discovered populations of mites represent an enormous threat to apiculture. ^ For the first part of this project, we focused on investigating the differences in gene expression between populations of V. jacobsoni mites reproducing on A. cerana and those reproducing on A. mellifera(detailed description of the methods and results are discussed in chapter two). Briefly, we sequenced and assembled a de novo transcriptome of V. jacobsoni.We also performed a differential gene expression analysis contrasting biological replicates of V. jacobsoni populations that differed in their ability to parasitizeA. mellifera. Using the edgeR, EBSeq, and DESeq R packages for the differential gene expression analysis, we found 287 differentially expressed genes (FDR ≤ 0.05), of which 91% were up-regulated in mites paraistizing A. mellifera.Furthermore, we searched for orthologous genes in public databases and were able to associate 100 of these 287 differentially expressed genes with a functional description. The mites found parasitizing A. mellifera showed substantially more variation in expression among replicates. While a small set of genes including, putative transcription factors and digestive tract developmental genes showed reduced expression in the mites, the vast majority of differentially expressed genes were up-regulated. These up-regulated genes are associated with mitochondrial respiratory function and apoptosis, suggesting that mites on this host may have experienced higher stress levels and were less optimally adapted to parasitize them. Some genes involved in reproduction and oogenesis were also over-expressed, which should be further studied in regards to this host shift. ^ The second part of the study was carried out to survey, for the first time, the viruses associated with V. jacobsoni and to determine whether these viruses played a role in mite colonization (detailed description of methods and results are discussed in chapter three). Briefly, we assembled a virus transcriptome of V. jacobsoni to provide the first survey of pathogens in this species. Among the list of putative viruses found were Deformed Wing Virus (DWV), Dragonfly Cyclovirus 1, Farmington Virus, Formica Exsecta Virus 2, Halyomorpha Halys Virus, Heliconius Erato Iflavirus, Kakugo Virus, Kirsten Murine Sarcoma Virus, Sacbrood Virus, Spodoptera Exigua Iflavirus 1. Our findings suggest that overall all the mite samples had similar viruses, with slight variations in the abundance of certain sequences. A search against a honey bee associated microbe database revealed the likely presence of Macula-like virus (Tymoviridae), a microsporidian and a spiroplasma, all of which have been previously reported for A. mellifera. The results from the expression analysis suggested that there are four different viruses that are differentially expressed between mites on A. cerana and A. mellifera, three of which were found up-regulated on the A. mellifera hosts. Among those three sequences, we found a match to Dragonfly Cyclovirus PK5222. Although this exact sequence was not similarly expressed across all samples of the two hosts, this Cyclovirus appears to be one of the most abundant viruses that were common to all samples. This analysis also showed a clear geographical clustering of the samples according to the expression patterns. Samples collected in Solomon Islands (SI) clustered together and the samples collected in Papua New Guinea (PNG) also clustered together in their own group. The lack of clear expression differences between the two hosts suggests that the viruses are not critical for host acquisition or overcoming host defenses. We have solid evidence that DWV infects V. jacobsoni and it is surprising that it is most closely related to an isolate from North America. As far as we know this is the first time that DWV has been reported in V. jacobsoni and indeed the virus pathogens of this mite had not previously been determined