Structural and mechanistic characterization of the core mammalian Hippo tumor suppressor pathway

The Hippo Tumor Suppressor Pathway regulates eukaryotic cell growth by sensing and integrating diverse signals of cell contact, cell shape, cell density, and tissue organization. The canonical core of the Hippo pathway consists of a kinase cascade, in which upstream MST1/2 kinases phosphorylate and activate downstream LATS1/2 kinases. Activated LATS1/2 then phosphorylate and inhibit transcription coactivators, YAP and TAZ, thus silencing growth, proliferation, and anti-apoptosis. Dysregulation of the Hippo pathway leads to tissue overgrowth, tumors, and cancers. In my Ph.D., I focused on investigating the molecular basis of how the core Hippo pathway is regulated. One of the major questions that I have addressed in this thesis is the molecular mechanism of how MOB1 protein coordinates the interactions between the upstream MST1/2 kinases and the downstream LATS1/2 kinases. A second theme that I have investigated is the structural basis for the auto-inhibitory mechanism in NDR1, a paralog of LATS1/2. To accomplish these aims, I employed a combination of biochemistry, biophysics, proteomics, and structural biology approaches. In the first chapter, I systematically examine the phosphorylation dependent interactions between MST1/2 kinases and MOB1, and determine the phosphopeptide binding specificity of MOB1. I demonstrate that autophosphorylation of MST1/2 kinase provides multiple MOB1 binding sites, which in turn contributes to a redundancy of MST1/2-MOB1 interactions in cells. In the second chapter, I quantitatively assess how phosphorylation of MOB1 regulates its interactions with both the upstream MST1/2 kinases and the downstream LATS/NDR family kinases in vitro. Using interaction proteomics, I validate the significance of the in vitro studies and also discover that the MOB1 phosphorylation dependent recruitment of phosphatase PP6 and Rho guanine exchange factors (DOCK6-8) differ in key respects from that elucidated for MST1/2 kinases. In the third chapter, I describe the crystal structure of NDR1 kinase domain (paralog of LATS1/2), revealing its auto-inhibitory atypically long activation segment. I present a novel NDR1 activation mechanism, in which MOB1 binding and mutagenesis by disrupting the auto-inhibition operate independently yet additively to relieve NDR1 auto-inhibition. Together, the results of these studies illustrate and support a sophisticated communication network in the core Hippo pathway mediated by MOB1.Ph.D