CHARACTERIZATION STUDIES OF A PI(3)P BINDING LEGIONELLA EFFECTOR GLUCOSYLTRANSFERASE
The bacterial pathogen Legionella pneumophila is an intracellular pathogen that infects freshwater amoebae and human alveolar macrophages. During infection, L. pneumophila uses a Dot/Icm type IV secretion system to secrete over 300 effectors that manipulate cellular pathways to gain control over its host. The function and roles for many of these effectors have been identified, which has shed light on the intricate relationship between L. pneumophila and its hosts. Furthermore, the use of L. pneuomophila as a model organism has become a powerful tool for probing cellular pathways, finding novel proteins and discovering new enzymology. The work presented here highlights some of these examples. For Part I of my dissertation, I present work that focuses on one effector, SetA. SetA is an O-glucosyltransferase with broad substrate specificity and may serve several roles during infection. Using a high-throughput imaging screen, I found that SetA activates an autophagy transcription factor, TFEB, under normal nutrient conditions. Using mass spectrometry analysis, I identified several SetA glucosylation sites on TFEB. These glucosylation sites are located near two regulatory regions on TFEB: 1) a phosphosite that regulates a nuclear export signal and 2) a cluster of Ser and Thr adjacent to the regulatory 14-3-3 binding site. I found that glucosylation at either regulatory site is sufficient to dysregulate TFEB, demonstrating that SetA can robustly override signaling pathways to activate TFEB. In addition to its glucosyltransferase domain, SetA has a well-established C-terminal PI(3)P binding domain that has been implicated in regulating SetA activity. Using NMR, I solved the structure of this domain and identified several features that are important for PI(3)P recognition and membrane targeting. Part II of my dissertation focuses on characterizing a new effector, lpg2526. Until now, the function of this effector was not identified. I solved the structure of lpg2526 using X-ray crystallography and found that this effector resembles the macrodomain family of ADP-ribosyl hydrolases. Using biochemical approaches, I was able to show that lpg2526 has ADP-ribosyl hydrolase activity in vitro. Ongoing work in the lab is focused on determining the role of this effector during infection.