Liver is a key regulator of systemic metabolism in physiology and pathology. It maintains cholesterol, triglyceride and glucose homeostasis in response to fasting and feeding. Transcription factors responsive to hormones and nutrients coordinate many of these responses. They regulate gene networks relating to lipid, glucose and amino acid metabolism during fasting and feeding, and factors such as sex, circadian rhythms and dietary content interact with these transcriptional programs, as reviewed in chapter 1. Understanding the different players in these processes and how they interact with each other will be key to develop strategies to combat metabolic diseases. Recent advances in probing the transcriptional landscape, coupled with high throughput sequencing approaches such as assay for transposase accessible chromatin (ATAC-Seq), have enabled the identification of transcriptional regulators of these processes and their interrelationships. In chapter 2, I describe results from a screen using ATAC-Seq and RNA-Seq to profile novel transcriptional regulators of the hepatic response to fasting and feeding. We determined GATA4 expression and activity to be upregulated by feeding through insulin. Knocking out Gata4 in adult liver impaired the transcriptional and metabolic response to feeding. Additionally, loss of hepatic GATA4 led to a reduction in HDL cholesterol and to the accumulation of liver triglycerides. These effects were accompanied by the downregulation of genes involved in cholesterol efflux and triglyceride hydrolysis, and the upregulation of genes involved in lipid uptake. Furthermore, hepatic GATA4 colocalized and collaborated with Liver X Receptor (LXR), a key regulator of cholesterol transport. In chapter 3, I investigated the dynamics of hepatic LXRs via ATAC-Seq and RNA-Seq in wild-type and LXRα/β knockout liver. Loss of LXRs in liver changed the transcriptional regulatory landscape by reducing accessibility at enhancers and increasing accessibility at promoters. A broad set of transcription factors that bind enhancers, including nuclear receptors, had reduced activity based on their motif accessibility, showing their reliance on LXRs. Moreover, this investigation demonstrated that LXR also functions as a transcriptional repressor of certain genes. Overall, these studies revealed important functions for GATA4 and LXR within hepatic transcriptional networks, extending our understanding of transcriptional regulation of hepatic lipid metabolism.

Multilocular adipocytes are a hallmark of thermogenic adipose tissue, but the factors that enforce this cellular phenotype are unknown. Here, we show that an adipocyte-specific product of the Clstn3 locus (CLSTN3beta) present only in placental mammals facilitates the rapid utilization of stored triglyceride by limiting lipid droplet (LD) size. CLSTN3beta is an integral ER membrane protein that localizes to ER-LD contact sites via a conserved hairpin-like domain. Mice lacking CLSTN3beta have abnormal LD morphology and altered substrate utilization in BAT, as well as increased susceptibility to cold-induced hypothermia, despite having no defect in adrenergic signaling. Conversely, forced expression of CLSTN3beta is sufficient to enforce a multilocular LD phenotype in cultured cells and BAT. Mechanistically, CLSTN3beta associates with CIDE proteins and impairs their ability to transfer lipid between LDs, thereby restricting LD expansion. These findings define a molecular mechanism that maximizes LD surface area to facilitate lipid utilization in thermogenic adipocytes.

The liver X receptors (LXRs) are members of the nuclear receptors superfamily of transcription factors. Multiple lines of evidence established LXRs as crucial regulators of metabolic homeostasis. Upon activation by cholesterol intermediates, LXRs induce the expression of a cluster of genes involved in cholesterol efflux, catabolism, and storage. In addition, activation of LXRs represses inflammatory signaling and enhances cellular survival against noxious stimuli. The diverse biologic roles of LXRs have made them attractive targets for pharmacologic manipulation.

In this work, we utilized unbiased high-throughput genome-wide screens to identify novel LXR target genes.

Adipocyte differentiation is accompanied by a robust program of cell-type specific gene expression. Although transcriptional regulators of this process are well defined, the contribution of posttranscriptional factors to the establishment of the adipocyte phenotype is poorly understood. This thesis presents the RNA-binding protein paraspeckle protein 1 (Pspc1), a component of the paraspeckle complex, that promotes adipogenesis in vitro and is important for mature adipocyte function in vivo. Cross-linking and immunoprecipitation followed by RNA sequencing revealed that Pspc1 binds to intronic and 3′ untranslated regions of a battery of adipocyte RNAs, including that encoding the transcriptional regulator Ebf1. Purification of the paraspeckle complex from adipocytes further showed that Pspc1 associates with the RNA export factor Ddx3x in a differentiation-dependent manner. Remarkably, Pspc1 relocates from the nucleus to the cytoplasm during differentiation, coinciding with enhanced export of adipogenic RNAs. Mice lacking Pspc1 in fat show reduced lipid storage and adipose tissue mass and are resistant to diet-induced obesity and insulin resistance due to a compensatory increase in energy expenditure. These findings highlight a role for Pspc1-dependent RNA maturation in the posttranscriptional control of adipose development and function.

Individual cells maintain tight control over their cholesterol content and its intracellular distribution. Abnormal cholesterol accumulation is cytotoxic and promotes coronary artery disease. How cholesterol is synthesized and exchanged between cells was elucidated in the past century. However, cholesterol transport between intracellular membranes is poorly understood. Most cellular cholesterol is found at the plasma membrane (PM), yet the endoplasmic reticulum (ER) is key for cholesterol sensing, biosynthesis, and storage. The pathway between cholesterol deposition at the plasma membrane and its subsequent transport to the ER represents a major gap in our understanding. We sought to address this by using genome-wide screens to identify novel target genes of the cholesterol-responsive transcription factor LXR. We discovered Gramd1b as a�previously uncharacterized LXR target gene. Through bioinformatic analyses we determined that Gramd1b has two paralogs, Gramd1a and Gramd1c, that together encode a family of GRAM domain containing proteins. Secondary structural analyses of this family revealed remarkable similarity to StAR-family proteins despite little primary sequence conservation. We named the Gramd1a-c protein products as the Aster (Greek, aster for “star”) proteins. Through binding experiments and x-ray crystallography we determine the Aster proteins contain a unique cholesterol-binding ASTER domain. Next, we used live cell super-resolution microscopy and determined that the Aster proteins are tethered to the ER by a single pass transmembrane domain. Remarkably, cholesterol loading of the plasma membrane results in the dynamic recruitment of Aster-A, -B, and –C to ER-PM contact sites. The N-terminal GRAM domain showed high affinity for phosphatidylserine in binding experiments and mediates cholesterol-dependent Aster recruitment to the PM. To determine the biological role of Aster proteins we knocked down Aster-A in cells and generated Aster-B knockout mice. Cells lacking Aster-A have slowed PM to ER cholesterol transport. Mice lacking Aster-B are deficient in adrenal cholesterol ester storage and steroidogenesis due to defective cholesterol transport from SR-BI to the ER. Our findings identify a nonvesicular pathway for plasma membrane to ER cholesterol transport that is critical in cellular uptake of HDL-derived cholesterol.

Phospholipids (PLs) are important structural components of biological membranes and precursors of numerous signaling molecules. The fatty acyl composition of PLs determines the biophysical characteristics of membranes. Multiple lines of evidence demonstrated that changes in fatty acyl composition could potentially affect the properties of proteins associated with membranes and influence the biological processes that occur on membranes. However, there is little understanding of how regulatory pathways control PL fatty acyl composition in vivo or how such regulation dictates physiological responses. In this work, we investigated the regulation of membrane fatty acyl composition by the Liver X Receptor (LXR)-Lysophosphatidylcholine Acyltranferase 3 (Lpcat3) pathway and its physiological or pathological relevance in lipid homeostasis and metabolic diseases.

In chapter 2, we define a nuclear receptor pathway for the dynamic modulation of membrane composition in response to changes in cellular lipid metabolism. Ligand activation of LXR preferentially drives the incorporation of polyunsaturated fatty acids into PLs through induction of the remodeling enzyme Lpcat3. Promotion of Lpcat3 activity decreases endoplasmic reticulum (ER) stress induced by saturated free fatty acids in vitro or by obesity and hepatic lipid accumulation in vivo. Conversely, Lpcat3 knockdown in liver exacerbates ER stress and inflammation. Mechanistically, Lpcat3 modulates inflammation both by regulating c-Src (Proto-oncogene tyrosine-protein kinase Src) and JNK (c-Jun N-terminal kinases) activation through changes in membrane composition and by affecting substrate availability for inflammatory mediator production. These results outline an endogenous mechanism for the preservation of membrane homeostasis during lipid stress and identify Lpcat3 as an important mediator of LXRs effects on metabolism.

In chapter 3, we show that Lpcat3 is a critical determinant of triglyceride secretion due to its unique ability to catalyze the incorporation of arachidonate into membranes. Mice lacking Lpcat3 in the intestine fail to thrive during weaning and exhibit massive enterocyte lipid accumulation and reduced plasma triglycerides. Mice lacking Lpcat3 in the liver show reduced plasma triglycerides, hepatosteatosis, and secrete lipid-poor very low density lipoprotein (VLDL) lacking arachidonoyl PLs. Mechanistic studies indicate that Lpcat3 activity controls membrane lipid mobility in living cells, suggesting a biophysical basis for the requirement of arachidonoyl PLs in lipidating lipoprotein particles. These data identify Lpcat3 as a key factor in lipoprotein production and illustrate how manipulation of membrane composition can be used as a regulatory mechanism to control metabolic pathways.

Ketogenesis requires fatty acid influx from stored lipid droplets or extrahepatic tissues to hepatic mitochondria. However, whether inter-organelle contact sites can regulate this process is unknown. Recent studies have revealed a role for Calsyntenin-3b (Clstn3b), an endoplasmic reticulum-lipid droplet contact site protein, in control of lipid utilization in adipose tissue. Here we show that Clstn3b transcription is induced in liver by PPARa in settings of high lipid utilization, including fasting and ketogenic diet feeding. Hepatocyte-specific loss of Clstn3b in mice inhibits ketogenesis independent of changes in PPARa activation. Conversely, hepatic overexpression of Clstn3b promotes ketogenesis in mice. Furthermore, the absence of Clstn3b alters LD- mitochondrial crosstalk, as revealed by impaired fatty acid oxidation and mitochondrial respiration, and a decreased mitochondrial integrated stress response. These findings define an important function for Clstn3b in ketogenesis and lipid utilization in hepatocytes.

The study of fat tissue or adipose over the last half century has dramatically altered the perception of the fat cell, also known as the adipocyte. Once only thought to store fat, the adipocyte is currently recognized to be a highly critical and important metabolically active cell. Adipogenesis, the process of adipocyte formation, has now become an intense subject of basic scientific research in large part due to the up-surge in obesity and metabolic disease in modern day society. PPARgamma is the master regulator of adipogenesis and is also a validated target of anti-diabetic drug therapy. However, a renewed sense of urgency in discovering novel factors and mechanisms that regulate the expression and/or activity of PPARgamma now exists due to recent scrutiny in the clinical use of molecules that directly activate PPARgamma. Therefore, this thesis attempts to address this need by investigating and uncovering entirely new modes of regulation in the differentiating adipocyte.

Part one of this thesis presents the construction of a transgenic mouse line that carries a reporter vector containing a cloned set of highly conserved non-coding genomic sequences endogenously found within and adjacent to the PPARgamma locus in the mouse genome. This examination represents an effort to identify a genomic regulatory sequence that confers tissue selective expression of adipose-specific genes such as PPARgamma. Identification of such an enhancer would undoubtedly shed new light on the regulation of gene expression in adipocytes and usher in a new paradigm in the study of adipose tissue.

Part two of this thesis presents the identification of Vestigial-like 3 (Vgll3) as an inhibitor of adipocyte differentiation. Vgll3 is a conserved transcriptional co-activator that is down-regulated during adipogenesis. This gene was initially observed to be differentially regulated between pre-adipocyte cell lines that exhibit contradictory potential to become lipid-laden adipocytes. When overexpressed in differentiating adipocytes in vitro, Vgll3 induces a potent block in PPARgamma expression and adipocyte formation. Furthermore, ectopic expression of Vgll3 was observed to up-regulate the expression of genes previously determined to be inhibitors of adipogenesis and genes associated with other mesenchymal-derived cellular differentiation programs. These results point to Vgll3 as a gene whose expression must be carefully modulated during the formation of fully differentiated, mature adipose tissue.