UCSF DIABETES, ENDOCRINOLOGY & METABOLISM TRAINING PROGRAM FACULTY RESEARCH SUMMARIES

Neural Networks of Obesity and Hypothalamic Development

A major focus of my laboratory is in the area of hypothalamic development. Despite the fact that this brain region regulates diverse behaviors, it is still poorly defined. Because steroidogenic factor 1 null mice exhibit profound defects in ventromedial hypothalamus development, we view SF-1 as a tool for studying neuroendocrine development of this region. In rodents, the ventromedial hypothalamus functions as a relay center controlling metabolic homeostasis and reproductive behavior. The genetic program that governs the development and physiological function of neurons within the VMH is unknown. Similarly, a precise map of the neuronal circuitry to and from the VMH is not established. We have shown that SF-1 is required for establishing the normal hypothalamic-limbic system circuitry, but is not required for survival or formation of the VMH. We have recently also shown that SF-1 heterozygous mice display obesity when fed a high fat diet and when group-housed.  We attribute this change in eating to a reduction in the levels of the neurotrophin (brain-derived neurotropic factor), which normally serves as a satiety factor. 

Ongoing projects in this area include: (1) defining new molecular markers for the VMH, (2) tracing projections from the VMH using the WGA trans-synaptic tag, (3) ablating the vesicular glutamate transporter (V-Glut2) in the VMH, (4) performing a mutation screen in zebrafish for genes controlling hypothalamic development.

Structure of Orphan Nuclear Receptors

We are actively determining how receptors belonging to NR5 subfamily are regulated.  To do this, we have turned to structural biology. It has remained controversial as to whether SF-1 is activated by a ligand-dependent mechanism. SF-1 contains a classic DNA binding domain (DBD), a large hinge domain, and a ligand binding domain (LBD) harboring a C-terminal activating (AF2) domain bearing the signature hexamer LxxLL motif. In collaboration with Robert Fletterick, we obtained the crystal structure of LRH-1 at 2.4 Å. There is no ligand, and this receptor is stabilized by a novel structural element exposed to solvent. In recent work, we have determined the structure of mouse SF-1 to 1.2 Å resolution.  In collaboration with R. Fletterick and K. Guy (UCSF) and T. Willson (GSK), we showed that hSF-1, mSF-1 and hLRH-1 are bound by phospholipids, leading us to propose that NR5 receptors are phosphatidyl inositol (PI) nuclear sensors.  This represents a novel, non-cytoplasmic mechanism by which PIs signal directly to the nucleus. This might be especially relevant to the proliferative roles of both SF-1 and LRH-1.   We also shown with J. Thornton that the rodent LRH-1 has recently evolved to diminish its affinity for PIs – this may explain why rodents handle cholesterol so differently from humans.

Current projects include: (1) determining the crystal structure of SF-1 and 3,4,5 PI, (2) establishing which biological PI is the best ligand for mSF-1/hLRH-1, (3) assessing how nuclear PI signaling affects SF-1/LRH-1 targets in mice/yeast, (4) humanizing the mouse for LRH-1 signaling.

Sumoylation in Nuclear Receptor Function

Our previous studies have shown that maximal SF-1-mediated transcription and recruitment of general cofactors depends on a single phospho-serine 203, located in the hinge region. Our studies show that phosphorylation stabilizes the LBD.  In a separate study, we also found that like other transcription factors, NR5A receptors are sumoylated.  Sumoylation is a newly described posttranslational modification that leads to protein conjugation and repression.  We have determined that Sumo-mediated repression of NR5A receptors may work through a DEAD-box protein.  We still do not know what triggers sumoylation, how repression is achieved, and how much receptor is actually sumoylated in vivo. 

Current projects include: (1) determining the in vivo significance of sumoylation in flies and mice, (2) determining the crystal structure of sumoylated/phosphorylated SF-1/LRH-1, and (3) generating a suitable antibody against sumoylated SF-1/LRH-1 receptor.

back to faculty list