A genetic, transgenic, and transcriptomic analysis of larval salivary gland physiology in Drosophila melanogaster

Cholesterol is the precursor to a unique class of signaling molecules called steroid hormones that initiate the development of sexual characteristics, reactions to stress, and maintenance of metabolism, among many other functions. Although much progress has been made in understanding the function of these signaling hormones, we do not fully understand how a single steroid can cause many distinct, tissue-specific responses. Drosophila melanogaster is an effective model for understanding steroid hormone action because of its simplicity. The steroid molting hormone 20-hydroxyecdysone (hereafter, 20E) is the primary active steroid in Drosophila and mediates not only larval molts and the emergence of a sexually mature adult, but also is an effector of many other tissue-specific actions from the embryonic to larval stages and into adulthood. An effective approach for studying hormone signaling is to use molecular genetics in which genes that have roles in 20E-signaling pathways are overexpressed or silenced in a specific tissue at a defined time during development. Even more, Drosophila is a tractable model organism with many tools that are widely available, and large-scale genetic screens are easily executed. I chose to focus on the salivary gland of larval Drosophila as a primary model for 20E signaling in that several specific, measurable responses to the steroid occur at different times during development. Using this model system, my dissertation research was centered on two major aims: 1) identify genes in the 20E-regulated process of glue secretion in the larval salivary gland; and 2) test the hypothesis that an ATP-Binding Cassette (ABC) transporter, E23, participates in a negative feedback loop that acts directly on the hormone 20E. Thus, I mapped four homozygous recessive, loss-of-function mutations that block the expulsion of the glue protein from the salivary glands of mutant larvae. One of the mutations maps to a gene encoding the enzyme Choline Acetyltransferase. In addition, I characterized a 20E-activated gene named E23, which encodes an ABC transporter. In humans, cholesterol secretion is mediated by a subset of ABC transporters that are similar to E23. Of the more than 50 ABC transporter genes in Drosophila melanogaster, E23 is the only known gene to be activated by 20E (mediated by the ecdysone receptor). We tested the hypothesis that E23 is part of a negative feedback loop that acts directly on the hormone 20E. We overexpressed E23 in specific tissues (including the larval salivary gland) that respond to 20E, at a defined time during development. We found that E23 overexpression phenocopies the loss of the ecdysone receptor, but this can be overcome by exposing tissues to high levels of 20E in vivo. Using Illumina RNA-seq on the larval salivary gland, we also found that E23 overexpression has widespread consequences on gene expression, including the expression of temporally-specific 20E targets. These results support our hypothesis that E23 has a negative regulatory role on 20E signaling through modulating the levels of 20E. This work is novel in that it suggests a previously unknown mechanism for effectively controlling the hormone exposure of target tissues, which may contribute to the diverse responses to a single hormone in flies, and furthermore, it serves as a springboard for future experiments to test whether E23 is a pump specific for the expulsion of the steroid hormone 20E.