Characterizing the honey bee genome for parent-of-origin effects on gene expression

Genomic imprinting has been studied extensively in higher plants and mammals where it plays a crucial role in development and has been linked to cancer and other diseases in mammals. The honey bee is one of the few invertebrate species in which genomic imprinting has been studied and could give new insights into parent-of-origin effects on behavior. The hypothesis tested was that parent-of-origin effects do exist in the honey bee and was done by using RNA-seq on two life stages of reciprocal crossed honey bees. Two different RNA-seq systems were used. The project was first done using the SOLiD 4 system, but validation of the results was poor due to a large number of false positives created from systematic and alignment biases that were not taken into account by the statistical methods used. The second system was the Illumina HiSeq 2 and the study was done using replicates and generalized linear mixed model statistics that could model and correct for systematic biases introduced during sequencing and alignment. The Illumina study showed that parent-of-origin effects do take place in the honey bee and the results were validated by sequencing PCR products of parent-of-origin effect transcripts using the Illumina MiSeq. The Illumina study showed that parent-of-origin effects are more evident in adult honey bees than in early larval stages, contrary to mammals and plants where the majority genomic imprinting takes place in early development. Regardless of life stage parent-of-origin effects in the honey bee are primarily biased towards maternal expression with the strongest maternal expression found in nuclear copies of mitochondrial genes or numts. The numts are not complete genes but gene fragments or pseudo-genes that are actively transcribed by the nuclear genome, which suggests a biological function and evolutionary selection for conserved maternal expression