Hepatic HAX-1 Deficiency Prevents Metabolic Diseases in Mice

A dissertation submitted to the

Division of Graduate Studies of the University of Cincinnati

In partial fulfillment of the requirements for the degree of

Doctor of Philosophy in the Molecular, Cellular, and Biochemical Pharmacology Program Department of Pharmacology and Systems Physiology University of Cincinnati College of Medicine 2020

Fawzi Alogaili BS, Pharmaceutical Science, University of Baghdad MS, Biochemistry, University of Basra

Committee Chair: David Y. Hui, PhD

1 Abstract

Obesity is the leading cause of insulin resistance, Type 2 diabetes and nonalcoholic fatty liver disease. Metabolic regulation is largely dependent on mitochondria, which play an important role in energy homeostasis by metabolizing nutrient and production of ATP. Numerous recent studies have implicated alterations in mitochondrial function in the pathogenesis of insulin resistance and Type 2 diabetes. In addition, bile acid plays a critical role in regulation of glucose, lipid and energy metabolism through activation of farnesoid X receptor (FXR) and its downstream effects on other nuclear receptors.

Aberrations in bile acid metabolism and its recirculation are associated with several metabolic diseases. Recently, bile acid signaling pathways have become the source of promising novel drug targets to treat common metabolic and hepatic diseases. This study demonstrated that increasing hepatic mitochondrial activity through pyruvate dehydrogenase and elevating enterohepatic circulation of bile acids are promising new approaches for metabolic disease therapy, but neither approach alone can completely ameliorate disease phenotype in high fat diet-fed mice. Our study showed that diet-induced metabolic diseases including hepatic steatosis, hyperlipidemia, and insulin resistance can be completely ameliorated in mice with liver-specific HCLS1-associated protein X-1 (HAX-

1) inactivation. Mechanistically, the HAX-1 protein interacts with inositol 1,4,5-trisphosphate receptor-1 (IP3R1) in the liver, and its absence reduces IP3R1 levels thereby increasing calcium storage in the endoplasmic reticulum and limiting mitochondrial calcium to prevent excess nutrient-induced overload and mitochondrial dysfunction. As a result, the ablation of HAX-1 in the liver activates pyruvate dehydrogenase and increases mitochondria utilization of glucose and fatty acids to prevent hepatic steatosis, hyperlipidemia and insulin resistance. Hepatic HAX-1 also interacts with the bile salt exporter protein BSEP/ABCB11

2 and, in contrast to its regulation of IP3R1 levels, the absence of HAX-1 increases expression of the bile salt export protein BSEP/ABCB11 to promote enterohepatic bile acid recirculation, leading to the activation of bile acid responsive genes in the intestinal ileum to augment insulin sensitivity. Taken together, these data suggest that hepatic HAX-1 inactivation protects against metabolic diseases via dual mechanisms of increased mitochondrial respiration and enterohepatic bile acid recirculation. Hence, HAX-1 expression in the liver is a potential therapeutic target for metabolic disease intervention.

Acknowledgments

Foremost, I would like to express my deep and sincere gratitude to my research advisor and mentor Dr. David Y. Hui, professor and director of Metabolic Diseases

Research Center, for giving me opportunity to do research in his laboratory and providing invaluable guidance throughout this research. His dynamism, vision, and motivation have deeply inspired me. It was a great privilege and honor to work and study under his guidance. Besides my advisor, I would like to thank the rest of thesis committee: Dr.

Evangelia Kranias, Dr. Guo-Chang Fan, Dr. Terry Kirley and Dr. Michael Tranter for their encouragement, support, guidance and insightful comments. I also owe my deepest appreciation to all past and present members in the Hui laboratory: Dr. Anja Jaeschke,

James Cash, David Kuhel, Eddy Konaniah, Dr. Ravi Komaravolu, Dr. Siva Chinnarasu,

Patrick Wolfkiel, Emily Igel, Rachel Boody, Megan & Morgan Zeiser, April Haller, Melissa

Asman, Rob Janning and Nicolas Bedel for their assistance and hard-work. My special, and sincere thanks goes to Dr. Kranias and her lab members. Their insight has significantly improved my project and has developed my technical and critical thinking skills. I would like also to express my special thanks to Dr. Robert Rapoport and Ms.

Nancy Thyberg for their help on the paper-work and in all matters relating to graduate school. I would like also to thank my sponsor Higher Committee for Education

Development in Iraq (HCEDiraq) for giving me this opportunity to study abroad in the US.

Finally, I would like also to express my thanks to my wife and my children for their continued support, love and understanding throughout my graduate career.

5 Table of Contents Introduction ...... 10 Significance ...... 10 Normal glucose regulation ...... 11 Normal lipid metabolism ...... 12 Liver glucose metabolism ...... 13 Association of fatty liver in obesity and Type 2 diabetes ...... 14 Hepatic lipid metabolism and insulin resistance ...... 16 Fatty acid delivery to liver ...... 16 Hepatic uptake of triglyceride-rich lipoproteins ...... 18 De novo lipogenesis ...... 20 Hepatic lipid oxidation ...... 21 VLDL secretion in insulin resistance state ...... 22 ER stress and its contribution to metabolic diseases ...... 24 Mitochondrial dysfunction in metabolic diseases ...... 26 Role of ER-Mitochondria coupling in metabolic diseases ...... 27 Bile acid and metabolic diseases ...... 28 HS-1- associated protein X-1(HAX-1) ...... 32 Rationale ...... 34 Methods ...... 35 Mice ...... 35 Biochemical studies ...... 35 In vivo hepatic triglyceride secretion and clearance ...... 36 Lipid synthesis and secretion in hepatocytes ...... 36 Lipid accumulation in liver ...... 37 De Novo lipogenesis ...... 37 Weight and adiposity measurement ...... 37 Real-Time PCR analysis of gene expression ...... 37 Table 1: Primer sequences used for RT-PCR amplification of RNA ...... 38 Fecal Microbiota Analysis ...... 39 Western blot analysis ...... 39 Co-immunoprecipitation ...... 40 Cellular oxygen consumption and fatty acid oxidation ...... 40 Histology ...... 41 Calcium measurements ...... 41 Pyruvate dehydrogenase activity assays ...... 42

6 Bile acid analysis ...... 42 Fecal cholesterol measurements ...... 42 Fecal Microbiota Analysis ...... 42 Statistical Analysis ...... 43 Results ...... 43 HAX-1 is localized at mitochondria and ER in liver and its inactivation has no deleterious effects ...... 43 Hepatic HAX-1 deficiency reduces plasma and hepatic triglyceride levels in chow-fed mice ...... 44 L-Hax1-/- mice sustain euglycemia with lower insulin levels ...... 45 Hepatic HAX-1 inactivation protects against Western diet-induced hyperlipidemia and hyperinsulinemia ...... 46 Hepatic HAX-1 deficiency suppresses diet-induced hepatosteatosis by reducing glucose-induced lipogenesis ...... 47 Hepatic HAX-1 deficiency increases mitochondrial respiration and fatty acid oxidation ...... 48 Hepatic HAX-1 deficiency increases PDH activity ...... 50

Hepatic HAX-1 inactivation improves ER-mitochondria calcium homeostasis via IP3R1 ...... 50 Hepatic HAX-1 inactivation increases hepatic BSEP level and bile acid responsive gene expression in intestine and liver ...... 51 Hepatic HAX-1 inactivation affects gut microbiota in mice fed a Western diet ...... 53 Discussion ...... 53 Conclusion of dissertation ...... 76 Figures ...... 93 Figure 1. The role of fatty liver in the pathogenesis of hyperinsulinemia and hyperglycemia ...... 93 Figure 2. The role of fatty liver in the pathogenesis of hyperinsulinemia and hypertriglyceridemia ...... 93 Figure 3. Schematic representation of ER- Mitochondria interactions, including some of the key proteins that regulate Ca+2 transfer ...... 94 Figure 4. Bile acid metabolism in liver and intestine ...... 95 Figure 5. Distribution of HAX-1 in liver subcellular fractions and different mouse tissues ...... 96 Figure 6. Plasm alanine aminotransferase (ALT) levels and liver morphology after hepatic HAX-1 depletion ...... 97 Figure 7. Body weight (BW) and whole body composition after virus injection ...... 98 Figure 8. Plasma triglyceride levels in control and L-HAX-1-/- mice ...... 99 Figure 9. Plasma cholesterol levels in control and L-HAX-1-/- mice ...... 100

7 Figure 10. Plasma non-esterified fatty acid (NEFA)levels in control and L-HAX-1-/- mice ...... 101 Figure 11. Assessment of triglyceride synthesis or secretion and clearance after hepatic HAX-1 deletion ...... 102 Figure 12. Measurement of hepatic lipids after hepatic HAX-1 deletion in mice fed a chow diet...... 103 Figure 13. Impact of hepatic HAX-1 deletion on blood glucose and insulin levels in mice fed a chow diet ...... 104 Figure 14. Impact of hepatic HAX-1 deletion on glucose tolerance and insulin sensitivity in mice fed a chow diet ...... 105 Figure 15. Impact of hepatic HAX-1 deletion on gluconeogenesis in mice fed a chow diet ...... 106 Figure 16. Body composition of control and L-HAX-1-/- mice fed a western diet for 12 weeks...... 107 Figure 17. Plasma triglyceride levels in control and L-HAX-1-/- mice after 12 weeks on a western diet ...... 108 Figure 18. Plasma cholesterol levels in control and L-HAX-1-/- mice fed on western diet for 12 weeks ...... 109 Figure 19. Plasma non-esterified fatty acid (NEFA)levels in control and L-HAX-1-/- mice ...... 110 Figure 20. Effect of hepatic HAX-1 deletion on blood levels of glucose and insulin in mice fed a western diet for 12 weeks ...... 111 Figure 21. Effect of hepatic HAX-1 deletion on glucose tolerance in mice fed a western diet for 12 weeks ...... 112 Figure 22. Effect of hepatic HAX-1 deletion on insulin sensitivity in mice fed a western diet for 12 weeks ...... 113 Figure 23. Effect of hepatic HAX-1 deletion on gluconeogenesis in mice fed a western diet ...... 114 Figure 24. Lipid accumulation in samples of liver extracts from control and L-HAX-1-/- mice fed e western diet for 12 weeks ...... 115 Figure 25. Triglycerides accumulation in glucose treated primary hepatocytes isolated from control and L-HAX-1-/- mice fed a chow diet...... 116 Figure 26. De novo lipogenesis in vivo in control and L-HAX-1-/- mice fed a chow diet ...... 117 Figure 27. Mitochondrial respirometry in intact primary hepatocytes isolated from control and L-HAX-1-/- mice fed a chow diet ...... 118 Figure 28. Effect of hepatic HAX-1 deletion on the mitochondrial function parameters in hepatocytes from control and L-HAX-1-/- mice fed a chow diet...... 119 Figure 29. Effect of hepatic HAX-1 deletion on the spare respiratory capacity and coupling efficiency in hepatocytes from control and L-HAX-1-/- mice fed a chow diet. 120 Figure 30. Measurement of fatty acid oxidation in primary hepatocytes from control and L-HAX-1-/- mice fed a chow diet...... 121 Figure 31. Assaying of fatty acid oxidation in the presence of FCCP ...... 122

8 Figure 32. Impact of hepatic HAX-1 deletion on free fatty acid accumulation in liver in control and L-HAX-1-/- mice fed e western diet for 12 weeks ...... 123 Figure 33. Impact of hepatic HAX-1 deficiency on expression of fatty acid oxidation genes in liver of control and L-HAX-1-/- mice fed a chow diet ...... 124 Figure 34. Impact of hepatic HAX-1 deficiency on Pyruvate Dehydrogenase (PDH) enzymatic activity ...... 125 Figure 35. Impact of hepatic HAX-1 deficiency on Pyruvate Dehydrogenase Kinase (PDK) ...... 126 Figure 36. Impact of hepatic HAX-1 deficiency on Pyruvate Dehydrogenase (PDH) phosphorylation ...... 127 Figure 37. Impact of HAX-1 deletion on protein levels of LDL receptor ...... 128 Figure 38. Effect of hepatic HAX-1 deletion on ER calcium homeostasis in primary hepatocytes from control and L-HAX-1-/- mice fed a chow diet ...... 129 Figure 39. Effect of hepatic HAX-1 deletion on mitochondrial calcium homeostasis in primary hepatocytes from control and L-HAX-1-/- mice fed a chow diet ...... 130 Figure 40. Effect of hepatic HAX-1 deficiency on protein level of Inositol 1,4,5- trisphosphate (IP3) receptor 1 (IP3R1) in liver of control and L-HAX-1-/- mice fed a chow diet ...... 131 Figure 41. Impact of hepatic HAX-1 deletion on protein expression levels of BSEP in total liver homogenates and bile canalicular membrane in control and L-HAX-1-/- mice fed a chow diet...... 132 Figure 42. Measurement of bile acid pool size in control and L-HAX-1-/- mice fed a chow or western diet ...... 133 Figure 43. Fecal cholesterol contents from control and L-HAX-1-/- mice fed a chow or western diet ...... 134 Figure 44. Impact of hepatic HAX-1 deficiency on expression of bile acid-responsive genes in ileum of control and L-HAX-1-/- mice ...... 135 Figure 45. Impact of hepatic HAX-1 deficiency on expression of FXR-responsive genes in liver of control and L-HAX-1-/- mice ...... 136 Figure 46. Impact of hepatic HAX-1 deficiency on expression of bile-acid responsive genes in liver of control and L-HAX-1-/- mice ...... 137 Figure 47. Impact of HAX-1 deletion on protein levels of mitochondrial CYP27 A and StARD involved in bile acid synthesis ...... 138 Figure 48. Assessment of gut microbiota in control and L-HAX-1-/- mice fed a chow or western diet...... 139 Figure 49. Scheme depicting hepatic HAX-1 inactivation mediated protection against diet-induced metabolic complications ...... 140 Appendix: Publications, Abstracts and Awards ...... 141

9 Introduction Significance

The prevalence of metabolic diseases including obesity, diabetes, and fatty liver disease is approaching pandemic proportions in industrialized and developing countries where a majority of the population enjoys sumptuous meals that are rich in fat and carbohydrates. Non-alcoholic fatty liver disease (NAFLD), for a long time unnoted in the metabolic field, is becoming a condition mostly involved in pathogenesis of metabolic diseases. The magnitude of the problem is apparent when it is estimated that in the United

States more than one-third of adults aged 60 years or older (~ 57 million) have

“prediabetes” (defined as insulin resistance, impaired glucose tolerance, or a plasma A1c concentration between 5.7% and 6.4% [2]. In contrast to Type 1 diabetes, which is due to impaired insulin secretion, Type 2 diabetes, with its core defects of insulin resistance, accounts for 90-95% of patients with diabetes. Type 2 diabetes is sixth leading cause of death in the US and accounts for 17.2% of all deaths in those aged >25years [3].

Studies over recent years have shown that up to 70-75% patients with Type 2 diabetes have prevalence of NAFLD and that these patients are also at higher risk of developing non-alcoholic steatohepatitis (NASH) and have two- to four-fold higher risk of developing liver failure [4]. In addition, increasing epidemiological evidence suggests that there is a bidirectional relationship between NAFLD and Type 2 diabetes and that NAFLD may precede and /or promote the development of Type 2 diabetes [4-6]. In addition to

Type 2 diabetes, NAFLD is commonly associated with obesity and dyslipidemia, which are components of the metabolic syndrome, strongly supporting the notion that NAFLD is the hepatic manifestation of the syndrome [7, 8].

10 Normal glucose regulation

Glucose is one of the sources of energy for the body. It is an energy-rich monosaccharide sugar that is broken down in cells to produce ATP. Glucose and other monosaccharides derived from the diet are transported across the intestinal wall through the hepatic portal vein and then delivered to liver cells and other tissues. They are converted to fatty acids, amino acids, and glycogen, or are oxidized by the various catabolic pathways in cells. Despite periods of feeding and fasting, plasma glucose levels are carefully regulated to remain in a narrow range between 4 and 7mM in normal individuals. This process is tightly controlled between glucose absorption from the intestine, production by the liver and uptake and metabolism by peripheral tissues. To avoid postprandial hyperglycemia and fasting hypoglycemia, the body can adjust blood glucose levels by secreting two hormones, insulin and glucagon. These hormones work in opposing manner to maintain glucose homeostasis. During period of hyperglycemia, insulin is secreted from b-cells of pancreas to facilitate the transport of glucose into muscles, adipose tissue and other cells and suppress gluconeogenesis in the liver to decrease blood glucose levels. During hypoglycemia, glucagon is secreted from a-cells of pancreas, which in turn increases hepatic glucose production resulting in increasing blood glucose levels [9-11]. Insulin increases glucose uptake mainly in muscle and fat cells by stimulating the translocation of the glucose transporter Glut4 from intracellular sites to the cell surface. Insulin does not stimulate glucose uptake in liver but blocks glycogenolysis and gluconeogenesis, and stimulates glycogen synthesis [12, 13]. Other actions of insulin include the stimulation of fat synthesis in adipocytes, promotion of triglyceride storage in fat cells, promotion of protein synthesis in liver and muscle, and promote cell growth [14].

11 Normal lipid metabolism

A second energy source in the diet is fat, composed primarily of triglycerides but also includes cholesterol and phospholipids. After eating a meal, the ingested fat is emulsified with bile acids and phospholipids in the lumen of the small intestine prior to hydrolysis by pancreatic lipases to cholesterol, fatty acids (FFAs) and monoglycerides for absorption into enterocytes. Within enterocytes, fatty acids have several fates such as partitioning into cholesterol ester (CE) or phospholipid (PL), or oxidation for energy utilization, or re-esterification to form triglycerides (TG) for incorporation into chylomicrons lipoprotein particles for distribution to the body tissues, or stored within the enterocytes in a lipid droplet or triglycerides storage pool. Triglyceride in chylomicron metabolized at muscle and adipose tissues by lipoprotein lipase (LPL), which is located on the luminal surface of the capillary endothelial cells, to release fatty acid for tissue uptake. In these tissues, dietary-derived fatty acid can be stored as TG or oxidized depending on their energy needs. The depletion of TG results in the reduction of the size of the lipoprotein, forming a chylomicrons-remnant. This remnant particle is taken up by the liver and the

TG remaining in the particle can be repackaged into VLDL for secretion back into the plasma for recycling of the dietary fatty acids [15, 16]. The TG carried in VLDL are metabolized in muscle and adipose tissues by LPL, and IDL are formed. The IDL are further metabolized to LDL, which are taken up by LDL receptor in numerous tissues including the liver, which is the predominant site of uptake. In addition, reverse cholesterol transport begins with the formation of nascent HDL by the liver and intestine. These small

HDL can then acquire cholesterol and phospholipids that are effluxed from cells, a process mediated by ATP-binding cassette transporter (ABCA1) resulting in the formation of mature HDL. The HDL then transports the cholesterol from peripheral tissues to the

12 liver either directly by interacting with hepatic SR-B1 (The scavenger receptor class B type I) or indirectly by transferring the cholesterol to VLDL or LDL, a process facilitated by cholesteryl ester transfer protein (CETP)[17, 18].

Liver glucose metabolism

The liver plays a key role in the control of glucose and lipid homeostasis as it has the capacity to both consume and produce these energy-rich nutrient, depending on the body requirement. Seventy percent of cell number or 80% of the liver volume is composed of hepatocytes that fulfill the metabolic needs of the body. Blood glucose is taken by the hepatocytes through glucose transporter 2 (Glut2), which has a higher Km (17mM) than other glucose transporters of the same family. The low affinity and high capacity of this transporter allows efficient transport of glucose across the plasma membrane of hepatocytes and only when blood glucose is high (e.g. following feeding state) [19]. Once taken, glucose is phosphorylated to glucose-6-phosphate (G6P) by the liver glucokinase

(GCK), the rate limiting step for hepatic glucose utilization. Unlike other hexokinase, GCK has a significantly lower affinity and is only active when glycemia is high [20]. Therefore, both Glut2 and GCK act as a glucose sensor as they are activated only during glucose abundance to facilitate glucose uptake for glycolysis or glycogen synthesis. In the fed state, insulin is secreted from b-cells of pancreas in response to an increase glucose, amino acids, and fatty acids. Insulin stimulates glycogen synthase by activating protein kinas B, also known as (Akt), which phosphorylates and inactivates glycogen synthase kinase(GSK-3), thus, increasing glycogen synthesis. On the other hand, insulin enhances acetylation of glycogen phosphorylase, which promotes dephosphorylation and inhibition of glycogen phosphorylase by protein phosphatase-1, thus inhibiting glycogenolysis [21].

Moreover, insulin increases hepatocytes glucose uptake indirectly by stimulating the

13 expression of glucokinase to generate G6P [22]. G6P acts as either a precursor for glycogen synthesis, or metabolized to generate pyruvate through glycolysis. G6P is also metabolized by the pentose phosphate pathway to generate NADPH, which is required for lipogenesis and synthesis of other bioactive molecules [23]. In concert with this, pyruvate provides a carbon source for lipogenesis, linking glycolysis to lipogenesis. Thus, when dietary carbohydrates are abundant, the liver not only utilizes glucose as the main source of fuel but also converts glucose into fatty acid. In the fasted state, G6P is transported into endoplasmic reticulum (ER) and dephosphorylated by glucose-6- phosphatase (G6Pase) to release glucose [23]. cAMP-responsive element binding protein H (CREBH) is a ER membrane protein, its levels are higher in the fasted state.

CREBH promotes the expression of gluconeogenesis genes such as PEPCK-C and

G6Pase [24]. Furthermore, in the fasted state, small intestine secretes fibroblast growth factor 15/19 (FGF15/19) which stimulates glycogen synthase kinase (GSK-3), inhibiting glycogen synthase [25].

Association of fatty liver in obesity and Type 2 diabetes

Nutrient intake in excess of energy requirement to sustain cell viability and activity is stored in adipose tissues and other ectopic organs such as muscles, heart, and liver to cause dysfunction and increased risk of diseases. The presence of fatty liver in patients with obesity and Type 2 diabetes has long been reported [26, 27]. It is considered an incidental pathologic finding, with little or no clinical significance. However, fat accumulation in the liver due to NAFLD has also been reported to be an obesity- independent predictor of Type 2 diabetes in multiple prospective studies, suggesting that

NAFLD may represent the hepatic manifestation of metabolic syndrome [28]. NAFLD may progress from simple steatosis (i.e. fat accumulation ³5% of the hepatocytes) to NASH,

14 which is characterized by inflammation and fibrosis, and in some cases to cirrhosis and even to hepatocellular carcinoma [29].

Since the liver plays a fundamental role in coordinating system metabolic homeostasis and the adaptation to nutrient availability and deprivation, it seems reasonable to postulate that excess fat accumulation in liver alters glucose and lipid homeostasis, leading to development of insulin resistance. During the fasting state, the combination of high glucagon and low insulin levels increases hepatic glucose production

(HGP) to meet metabolic demands of peripheral tissues. The increase in glucose production in the liver originates from breakdown of stored glycogen by process of glycogenolysis, increased gluconeogenic substrate supply, and enhanced expression and activity of genes involved in gluconeogenic pathway enable the liver to generate glucose from 3-carbon precursors such as amino acids and lactate [30]. The process is opposite during response to absorptive state, in which insulin concentrations rise and glucagon levels decline, leading to increase liver glucose uptake, glycogen synthesis, inhibit HGP, and induce the synthesis of fatty acids for storage and subsequent utilization

[31]. In insulin-resistant disorders such as hepatic steatosis, obesity and Type 2 diabetes, insulin fails to suppress HGP in both absorptive and post absorptive state due to inability of insulin to inhibit glycogenolysis and gluconeogenesis (Fig. 1) [32] . In addition to the inability of insulin to regulate glucose homeostasis, insulin-resistant patients also have increased de novo lipogenesis and re-esterification, inducing fat accumulation in the liver.

Hence, insulin-resistant patients exhibit increased secretion and decreased clearance of triglycerides (TG) leading to elevation in blood lipid levels [33, 34]. Therefore, during insulin resistance, insulin loses its ability to restrain TG synthesis, secretion, and glucose production by liver leading to hyperglycemia and hypertriglyceridemia (Fig. 2).

15 Interestingly, liver fat content and serum insulin are closely correlated. The correlation between fasting insulin and liver fat is independent of obesity, although obesity is also independently associated with increases in liver fat content [35].

Hepatic lipid metabolism and insulin resistance

In the post absorptive state, insulin regulates the synthesis and storage of lipids by increasing de novo lipogenesis, inhibiting fatty acid oxidation and increasing triglyceride esterification and secretion [36]. Recent studies in animal models and humans have illustrated the significance of hepatic insulin action to the regulation of hepatic lipid metabolism and the development of steatosis during insulin resistance. Hepatic steatosis develops when the rate of fatty acid (FA) input (uptake and synthesis with increased esterification to TG is greater than the rate of FA output (oxidation and secretion). Thus,

TG accumulation in hepatocytes result from a complex interaction between: (a) hepatic fatty acid uptake derived from plasma free FA (FFA) released from hydrolysis of adipose tissue TG and FFA released from hydrolysis of circulating TG during post-absorptive phase; (b) hepatic uptake of triglyceride-rich lipoproteins (LDL and chylomicron remnants) mediated by LDL receptor and LDL receptor related-proteins during the absorptive phase;

Fatty acid delivery to liver

Hepatic lipid uptake is a function of substrate delivery and transport into hepatocytes. Studies in mice and humans have implicated adipose tissue lipolysis as an important source of fatty acid that promote NAFLD and hepatic insulin resistance. During fasting conditions, the major source of FFA delivered to liver is derived from FFA released from subcutaneous adipose tissue, which enter the systemic circulation and are then

16 transported to the liver by the hepatic artery and portal vein. Although lipolysis of visceral adipose tissue TG releases additional FFA directly into portal systems, but the relative contribution of portal vein FFA derived from visceral fat is small compared with FFA derived from subcutaneous fat; which is about 5% and 20% of portal vein FFA originate from visceral fat in lean and obese subjects, respectively [38]. However, the relationship between lipolysis and insulin sensitivity is largely independent of body mass. For example, insulin-resistant obese individuals have higher visceral fat content than their weight-matched, insulin sensitive counterparts [39]. Moreover, gene expression of hepatic lipase and hepatic lipoprotein lipase are higher in obese individuals with NAFLD than those without NAFLD, suggesting that FFA released from lipolysis of circulating TG contribute to hepatic FFA accumulation and steatosis [40, 41]. Therefore, this increase in hepatic lipase and hepatic lipoprotein lipase are responsible to increases in hepatic TG accumulation in obese individuals with Type 2 diabetes [42]. Membrane protein that direct the uptake of FFA from circulation into tissues are also involved in increased hepatic FFA uptake. Studies were reported the protein expression of FAT/CD36 is increased in liver and skeletal muscle but decreased in adipose tissue in obese individuals with NAFLD compared with obese individuals who have normal hepatic TG [43, 44] , suggesting that the membrane FA transport proteins redirect the uptake of plasma FFA from adipose tissue toward other tissues.

Increases in hepatic TG accumulation in NAFLD leads to increase in liver diacylglycerol (DAG). Many studies in rodents and humans have implicated that hepatic diacylglycerol content was the best predictor of hepatic insulin resistance in obese individuals [45, 46]. Increased hepatic diacylglycerol content leads to increased translocation of the primary novel protein kinase C (PKC) isoform in liver, protein kinase-

17 Ce (PKCe) to plasma membrane at which, it was reported to bind and inhibit the activity of the intracellular kinase domain of the insulin receptor. This would be leading to serine/threonine phosphorylation and concomitantly reduced insulin-stimulated tyrosine- phosphorylation of insulin receptor substrate 2 (IRS2) and Akt serine/threonine kinase 2 or protein kinase BB (Akt2) in liver. Ultimately, the ability of insulin to activate glycogen synthesis and inhibit gluconeogenesis would be impaired [46-48]. The diacylglycerol- induced protein kinase-Ce has emerged as a most common mechanism explaining the development of insulin resistance in liver and has convincingly demonstrated by knocking down expression of protein kinase-Ce in the liver of mice [48].

Hepatic uptake of triglyceride-rich lipoproteins

As mentioned previously, ApoB-containing lipoproteins transport triglycerides through the plasma to liver, and adipose tissue, which stores fatty acids, and to muscle, which utilizes fatty acids for energy production. Within the intestinal epithelium, dietary triglycerides are ultimately packed into apoB48-containing chylomicrons that are secreted into the lymphatics [49, 50]. In the liver, endogenous triglycerides are packed into apoB100-containing VLDL particles that are secreted directly into the plasma. The cellular mechanism by which chylomicrons and VLDL are metabolized is quite similar but the

VLDL metabolism is mainly altered in the setting of insulin resistance and hepatic steatosis. Each VLDL particle is produced with a single embedded molecule of apoB100.

Within the liver, VLDL particles also includes other apolipoproteins such as ApoE, ApoCII, and ApoCIII. After VLDL secretion into plasma (described below), VLDL particles are directed to muscle and fat tissues where they acquire an additional complement of ApoCII molecules that circulate in association with HDL particles [51, 52]. In the capillaries, VLDL binds and activate lipoprotein lipase, which hydrolyzes triglycerides and releases FFA.

18 Under physiological conditions, lipoprotein lipase removes approximately 50% of triglycerides from the core of VLDL particles, at this point the conformational changes will reduce the binding affinity of apolipoproteins [52]. The rate of hydrolysis of chylomicron and VLDL triglycerides is controlled by ApoCIII, which inhibits lipoprotein lipase activity ant its upregulated in the setting of insulin resistance [51]. Upon dissociation of VLDL from Lipoprotein lipase, exchangeable apolipoprotein such as ApoCII and ApoCIII transfer spontaneously to HDL. In turn, ApoE transfers from HDL to VLDL [15]. The presence of ApoE on VLDL identifies it as a remnant particle and renders it a high affinity ligand for hepatic receptors that clear the particle from the circulation. However, only 50% of VLDL remnants are cleared by receptors and the other 50% are further metabolized by additional interaction with lipoprotein lipase. This further depletes triglycerides from the core and enriches the particle in cholesteryl ester resulting in smaller, dense remnant particles called IDL. VLDL remnant and a fraction of IDL particles (ApoE-containing particles) are cleared from plasma by hepatocytes when they become small enough to pass through the fenestrated endothelium of the liver and become trapped within the

Space of Disse by electrostatic interaction with large extracellular heparan sulfate proteoglycans (HSPG) [52, 53]. There, hepatic lipase remodels the particles, rendering them an optimal size to interact with cell surface receptors. These receptors include LDL receptor, LDL-receptor related protein (LRP-1) and HSPG. Additionally, the other fraction of IDL interacts with hepatic lipase [54], which hydrolyze most of the remaining triglycerides within the core leading to a conformational changes that result in ApoE dissociation, leaving only ApoB100. This lead to formation of a particle that is rich in cholesteryl ester defined as LDL [52]. Importantly, hepatic lipase activity is increased in insulin resistance leading to increase the proportion of LDL [55]. In contrast to ApoE-

19 containing remnant particles, LDL particles are only taken up by hepatic cells by the LDL receptor [56]. Approximately 70% of LDL receptors are localized to the liver, with remainder expressed on adrenocortical, gonadal and smooth muscle cells, as well as macrophages and lymphocytes. Upon binding to LDL receptor, LDL particles are removed from plasma by receptor-mediated endocytosis [56, 57]. In this pathway, both the lipoprotein particles and receptor are taken up into an endocytic vesicle that is delivered to lysosomes, where the particle dissociates from the receptor and is hydrolyzed.

Moreover, ApoB100 is degraded and cholesteryl esters are hydrolyzed to form free cholesterol, which is delivered to the ER via lysosomes by the activity of the Niemann-

Pick type C (NPC) proteins 1 and 2 [58]. The LDL receptor is separated and returned to the cell surface [56, 57].

De novo lipogenesis

Although hepatic de novo lipogenesis is thought to be a quantitatively minor pathway for hepatic TG accumulation relative to esterification [59], rates of postprandial de novo lipogenesis increases significantly in patients with NAFLD [33, 60, 61]. Compared with insulin-sensitive subjects, consumption of a high-carbohydrate meal was associated with a much lower rate of muscle glycogen synthesis and a diversion of most of the ingested glucose toward hepatic DNL in insulin-resistant individuals. Therefore, insulin resistance in skeletal muscle could promote TG accumulation by diverting ingested carbohydrate away from storage as muscle glycogen and toward de novo lipogenesis in liver [62] . An excess of glucose, insulin ultimately enhances de novo fatty acid synthesis through a complex cytosolic enzyme reactions in which acetyl CoA is converted to malonyl CoA by Acetyl-CoA carboxylase (ACC) to form one palmitate molecule. The rate of de novo lipogenesis is regulated by the fatty acid synthase (FAS), ACC1 and ACC2,

20 and stearoyl-CoA desaturase 1 (SCD1). The expression of these enzymes is regulated transcriptionally by several nuclear transcription factors such as sterol regulatory-element binding proteins (SREBPs), carbohydrates-responsive element binding proteins

(ChREBPs), liver X receptor a (LXRa), farnesoid X receptor (FXR), and peroxisome proliferator-activated receptors(PPARs) [63]. Interestingly, hepatic de novo lipogenesis is regulated independently by insulin and glucose through the activation of sterol regulatory- element binding protein-1c (SREBP-1c) [64] and carbohydrates-responsive element binding protein (ChREBP) [65], which activates the transcription of nearly all genes involved in de novo lipogenesis. Many studies provided evidence that hyperinsulinemia or overexpression of sterol regulatory-element binding protein-1c stimulates lipogenesis and causes hepatic steatosis [66, 67]. On the other hand, liver-specific carbohydrates- responsive element binding protein inhibition results in decreased hepatic de novo lipogenesis and ameliorated hepatic steatosis in ob/ob mice [68].

Hepatic lipid oxidation

The metabolic rate of liver tissue is nearly 20 times greater than the metabolic rate of resting skeletal muscle and 50 times greater than the metabolic rate of adipose tissue

[69]. It is estimated that liver consumes ~450 Kcal /d and accounts for ~20% of total resting energy expenditure. Since the liver used a mix of fuels, it is difficult to quantify accurately in vivo which substrate is dominant because of complicated exchange of metabolites between multiple biochemical pathways. It has been estimated that fatty acid and amino acid oxidation provide 90% of the fuel for basal hepatic energy requirements, and the use of FFA as a source of fuel decreases during the fed state [70].

The oxidation of fatty acid in liver occurs primarily within mitochondria. Transport of FA inside the mitochondrial matrix is regulated by a carnitine-dependent enzyme shuttle,

21 CPT1 in the outer mitochondrial membrane and CPT2 in the inner mitochondrial membrane. Mitochondrial b-oxidation shortens fatty acyl CoA by two carbon units at each cycle (released as acetyl CoA), through series of enzymes regulated transcriptionally by

PPAR-a [71]. Acetyl CoA derived from FAO can either enter tricarboxylic acid cycle for energy production in the liver or can be condensed to form ketone bodies (acetoacetate and beta- hydroxy butyrate) which are exported to provide fuel for other tissues [72]. Mice lacking PPAR-a are prone to NAFLD and fail to benefit from insulin sensitizer effects of omega-3 fatty acids [73]. In addition, any inhibition of ACC1 and ACC2 leads to both an increase in lipid oxidation and a decrease in lipid synthesis. ACC1 is a cytosolic enzyme, while ACC2 is located at the mitochondrial membrane and both catalyze the carboxylation of acetyl CoA to form malonyl CoA. The ACC1-generated malonyl CoA is utilized by FAS for the synthesis of FA in the cytosol. In contrast, the ACC2-generated malonyl CoA acts as inhibitor of CPT1 [74]. Animal studies have showed that mutation of serine residue in

ACC1 and ACC2 prevents inactivation by AMP-activated protein kinase (AMPK), resulting in increased hepatic DAG content, activation of PKCe and development of hepatic insulin resistance. On the other hand, increased AMPK activity leads to a reduction in ACC1 and

ACC2 activity, resulting in improved insulin resistance [75, 76]. In human, it has been reported that Type 2 diabetes is associated with a reduction in ATP production due to decreased FAO leading to impaired hepatic energy metabolism and insulin resistance

[77, 78].

VLDL secretion in insulin resistance state

Very low density lipoprotein (VLDL) is a large macromolecular aggregate of thousands of lipid molecules and several proteins, produced by liver and secreted into systemic circulation. VLDL formation provides an important mechanism for converting

22 water-insoluble TG into a water-soluble form that can export lipids from the liver and deliver them to peripheral tissues. There is general agreement that hepatic VLDL assembly involves the fusion of newly synthesized apolipoprotein B-100 (apoB-100) molecule with a TG droplet through the action of microsomal triglyceride transfer protein

(MTP) [79, 80]. Data from most studies have found that VLDL-TG secretion rate is greater in patients with NAFLD than in those with normal hepatic TG content, who were matched on BMI and percent body fat [81].

The role of insulin in the regulation of hepatic VLDL production is via regulation of

MTP expression and apoB-100 degradation, in addition to the role of insulin that already addressed in regulation of fatty acid flux to the liver, hepatic de novo lipogenesis and fatty acid oxidation in the liver, MTP gene expression is negatively regulated by insulin partly via FoxO1. Insulin normally phosphorylates and then exclude FoxO1 from nucleus resulting in the inhibition of hepatic MTP expression. However, in insulin resistance,

FoxO1 phosphorylation is reduced leading to increase MTP expression and then VLDL overproduction [82, 83]. In addition, insulin can directly affect the secretion of apoB-100 by targeting it for degradation. In cultured hepatocytes, acute exposure of cells to insulin inhibits the secretion of both TG and apoB-100 despite increased stimulation of TG synthesis. This inhibition of VLDL secretion from hepatocytes by insulin results from increased degradation of apoB-100 by PI3-kinase mediated mechanisms [84, 85].

Similarly, short-term hyperinsulinemia with euglycemia inhibited the secretion of VLDL to prevent hyperlipidemia in normal humans [86, 87]. In contrast, VLDL secretion was not inhibited in obese individuals [86] or people with Type 2 diabetes [88], two diseases characterized by insulin resistance, due to loss of responsiveness of insulin overtime and

23 increased stimulation of apoB-100 secretion by increased hepatic TG and increased fatty acid flux to the liver results in increased VLDL secretion and dyslipidemia.

ER stress and its contribution to metabolic diseases

The assembly and secretion of apoB100-containing triglyceride-rich lipoproteins is initiated in the ER where it is partially lipidated to form a primordial intermediate necessary for the appropriate folding of this highly hydrophobic 550kDa protein. The ER is an intracellular organelle that performs numerous functions related to the synthesis, folding and transport of proteins and plays an important role in lipid synthesis and Ca2+ homeostasis [89, 90]. It is well-reported that excess saturated fatty acids and cholesterol can induce ER stress and disrupt lipid metabolism in hepatocytes [91-94]. Particularly, long-term exposure to saturated fat and cholesterol can promote the formation of an ordered membrane domain in the ER that inhibits VLDL formation and SERCA activity causing loss of both lipid clearance and calcium homeostatic capabilities. Moreover, long- term exposure to lipids to formation of lipid droplets which require increased synthesis of phosphatidylcholine at the ER leading to perturbs the balance phosphatidylcholine/phosphatidylethanolamine concentrations in the ER resulting in

SERCA dysfunction [95, 96]. In the ER, millions of proteins are synthesized, but not all of them are able to be properly folded and processed. Once the misfolded protein load increases due to ER stress, ER initiates the unfolded protein response through the actions of canonical sensors PERK, IRE1 and ATF6 in an attempt to restore ER homeostasis [90, 97]. Collectively, these branches suppress protein synthesis, facilitate protein degradation, and increase the availability of chaperones and, if all fails, can drive cells to cell death [92]. There are currently a number of models to explain the mechanistic link among obesity, ER stress, insulin resistance, and Type 2 diabetes [92]. An important

24 discovery that led to the concept of metabolic regulation by ER has emerged from the recognition of the ability of ER to regulate insulin action and glucose metabolism [98, 99].

However, the ER stress has impact on metabolism beyond the boundaries of insulin, including direct actions on many metabolic pathways independent of this hormone.

ER stress contributes to fatty liver and hepatic insulin resistance by activating transcription factors that regulate the expression of lipogenic or gluconeogenic genes, depending on which branch of the ER stress response pathway is being activated. For example, ER stress activates the cAMP-response element-binding protein H (CREBH), which is a liver-specific ATF6 homolog, resulting in increasing the expression of gluconeogenic genes phosphoenolpyruvate carboxykinase (PEPCK) and glucose-6- phosphatase (G6Pase) [24]. In contrast, the activation of XBP-1 in IRE-1 pathway increases proteasomal degradation of FoxO1, resulting in reduced gluconeogenesis independently of insulin signaling pathways [100]. However, disruption of hepatic XBP-1 significantly reduced plasma TG, cholesterol and FA levels by down-regulating genes involved in de novo lipogenesis such as SCD1, DGAT2 and ACC2 [101]. In line with this, exposure of HepG2 cells to high concentration of glucose or palmitate induced ER stress, activating SREBP-1, and enhanced lipid accumulation through the activation of PERK- eIF2a pathway [102]. Furthermore, activation of this pathway leads also to increased expression of hepatic VLDL receptor, which is responsible for the uptake and intracellular accumulation of triglycerides that causes hepatic steatosis [103]. In addition, PERK-eIF2a activation leads to decreased global translation resulting in decreased ApoB expression, therapy reducing triglyceride secretion and the enhancement of hepatic steatosis.

25 Mitochondrial dysfunction in metabolic diseases

Another subcellular organelle that plays a key role in nutrient utilizing and sensing is the mitochondria. Dysfunctional hepatic mitochondrial energetics in the setting of hepatic insulin resistance, Type 2 diabetes and obesity plays a critical role in the development of steatosis [104]. Recent studies in rodents and human showed that there is inefficient mitochondrial adaptation to substrate overload in patients with obesity, Type

2 diabetes and hepatic steatosis [105, 106]. In line with this, hepatic insulin resistance and fatty liver reflect alterations in mitochondrial energetics in both rodent models and human. However, this alteration is dependent on the stage and severity of the disease, susceptibility of the metabolic pathway, and the ability of hepatocytes to buffer and store excess lipids [106, 107]. Additionally, dysfunction in mitochondrial energetics in NAFLD is concurrent to incomplete fatty acid oxidation, leading to accumulation of ceramide and

DAG, which further exacerbates the impairment of insulin signaling [104]. Along these lines, In vivo studies have shown that hepatic insulin resistance and steatosis are already established in mice fed a high fat diet before alteration in mitochondrial energetics become evident [104, 107].

It is worth mentioning that mitochondrial oxidative energetics are composed of multiple pathways including, fatty acid b-oxidation, ketogenesis, hepatic tricarboxylic acid

(TCA) cycle, respiratory chain activity, and ATP synthesis. b-oxidation, mitochondrial respiration, and TCA cycle flux have been reported to be induced in several animal models of nutritional overload, as well as in patients with obesity and/or hepatic steatosis

[104, 108-110]. Moreover, hepatic ketogenesis was also augmented in mice fed a high - fat diet, thus suggesting that increased b-oxidation leads to general high acetyl CoA which is followed by efficient ketone bodies (KB) production [111, 112]. Therefore, increased

26 mitochondrial energetics could reflect a hepatic compensatory mechanism to dispose excess acetyl CoA through complete oxidation. However, under chronic condition of nutritional overload in both mice and human, studies have shown that there were reductions in ketogenesis, mitochondrial respiratory chain activity [107], and rates of ATP synthesis [77, 78]. This means that mitochondrial respiration and oxidation may evolve from a hyperactive but inefficient state during moderate obesity and insulin resistance to a more generalized dysfunction during severe obesity and insulin resistance.

Role of ER-Mitochondria coupling in metabolic diseases

As discussed above, both mitochondria and ER have distinct roles in hepatocytes , but they physically and functionally interact at sites known as mitochondria-associated membranes (MAMs) to regulate lipid homeostasis, mitochondrial metabolism and Ca2+ homeostasis [113, 114]. The physical interactions between both organelles does not involve membrane fusion, but are mediated through protein bridges (Figure 3). Among the proteins involved in interaction between mitochondria and ER or in MAMs of mammalian cells are enzymes involved in lipid synthesis such as DGAT2, those involved

+2 in Ca handling such as IP3R1, VDAC, sigma-1 receptor (sigR1), those controlling mitochondrial dynamics such as dynamin -related protein (Mfn2), apoptosis like Bcl2 and protein sorting such as phosphofurin acidic cluster sorting protein 2 (PACS2) [115].

VDAC at the outer mitochondrial membrane interacts with IP3R1 at ER through the molecular chaperone glucose – regulated protein 75 (Grp75), allowing Ca2+ transfer from

ER to mitochondria [116]. It has been reported that MAM integrity is altered in palmitate- treated Huh-7 as well as in the liver of different models of obese and diabetic mice leading to hepatic insulin resistance. Loss of Ca2+ transfer from ER to mitochondria link MAM disruption to hepatic insulin resistance [117]. Ca2+ is important for modulating

27 mitochondrial enzyme activities and ATP synthesis, as well as the functionality of ER- based channels and buffering chaperone. On the other hand, ER homeostasis relies on the ATP levels and ROS that are generated by mitochondria. Therefore, any functional changes in one organelle could affect the function of other organelle and then impacting hepatic metabolic processes. Indeed, experimental mitochondrial dysfunction induced ER stress through an elevation of cytosolic free Ca2+ and led subsequently to insulin resistance, hepatic steatosis and increased hepatic glucose output [96, 118, 119].

Additionally, palmitate leads to increased ER Ca2+ efflux promoting downstream markers of mitochondrial lipotoxicity such as ROS accumulation and caspase activation [118].

2+ Importantly, IP3R1, an intracellular Ca release channel in hepatocytes, is one of the proteins that its expression is upregulated and enriched in MAMs of animal models of obesity. The reduction of IP3R1 levels results in improved mitochondrial function, reduced hepatic steatosis and enhanced glucose tolerance in obese mice [120, 121]. Conversely, loss of mitochondrial Mfn2 or Cyclophilin D induced ER stress, insulin resistance and impaired glucose tolerance in mouse liver [117, 122].

Bile acid and metabolic diseases

Bile acids are soluble products derived from catabolism of highly insoluble cholesterol. Bile acid synthesis is the primary pathway for cholesterol metabolism. An emerging evidence point towards the role of bile acids as important modulators of metabolic diseases such as obesity, Type 2 diabetes and NAFLD. The metabolic role of bile acids includes: 1) The formation of mixed micelles in the small intestine that facilitate digestion, solubilization and absorption of dietary lipid and fat soluble vitamins such as vitamin A,D,E and K; 2) Solubility of cholesterol to prevent cholesterol crystallization and gallstone formation in the gallbladder; 3) Induces bile flow from hepatocytes into the bile

28 canaliculi and then gallbladder; 4) Their formation from cholesterol and their subsequent excretion in the feces represent the major route for cholesterol excretion; and 5) Their bacteriostatic function that maintains sterility in the biliary tree. Thus, disruption of normal bile acid synthesis and metabolism is associated with cholestasis, gallstone, inflammation, malabsorption of lipids and fat soluble vitamins and atherosclerosis [123].

Bile acid biosynthesis involves modification of the ring structure of cholesterol, oxidation and shortening of the side chain, and finally conjugation of bile acid with an amino acid [124]. There are two pathways for bile acid synthesis: classical pathway(or neutral pathway) and alternative (or acidic) pathway. All enzymes and intermediates of these pathways are displayed in Figure 4. The classical bile acid pathway is regulated by cholesterol 7a-hydroxylase (CYP7A1), a cytochrome P450 enzyme converts cholesterol to 7a-hydroxycholesterol. This pathway generates cholic acid (CA) and chenodeoxycholic acid (CDCA) and contributes ~75% of total bile acid synthesis [125]. The alternative pathway generates CDCA and accounts for ~ 9% in human and 25% in mice of the total bile acid synthesis [126] (Fig. 4A). The acidic pathway is regulated by sterol 27- hydroxylase (CYP27A) located in the inner mitochondrial membrane and regulated by mitochondrial cholesterol transport steroidogenic acute regulatory protein (StARD) and the cholesterol transport by StARD is essential rate-limiting step of this pathway [127,

128]. The role of acidic pathway is to generate and control the levels of life sustaining regulatory oxysterols that help control cellular cholesterol and lipid homeostasis [129,

130]. The most abundant bile acid in human includes the primary bile acids cholic acid, and chenodeoxycholic acid and their respective secondary bile acids, deoxycholic acid

(DCA) and lithocholic acid (LCA). The secondary bile acids are formed by deconjugation and dihydroxylation by microbial enzymes in colon. In the mouse and rat,

29 chenodeoxycholic acid is mainly converted into muricholic acid [131]. Before bile acids are transported out of hepatocytes, most of them are conjugated to either taurine (in the mouse) or glycine (in humans) [132]. The conjugation helps to lower PKa and increase their solubility, therapy facilitating micelle formation in acidic environment of the duodenum.

The secretion of bile salts from hepatocytes into the canaliculi requires the bile salt export protein (BSEP/ABCB11), while transport of phospholipids requires ABCB4 (also known as MDR3 in humans or MDR2 in mice). On the other hand, cholesterol efflux into bile requires ABCG5/ABCG8 (Fig. 4). The bile salts together with phospholipids and cholesterol are passed into the gall bladder, where they are concentrated to form bile juice, which is composed of 85% water. The remaining solute is a complex mixture of bile salts (67%), phospholipids (22%) and cholesterol (4%), together with electrolytes, minerals and bilirubin and biliverdin pigments which give it a yellow green hue [133]. In addition, small amount of mucus and secretory IgA may contribute to the bacteriostatic functions of bile [134]. Hepatic secretion of bile is activated by the presence of food in the duodenum, which causes secretion of cholecystokinin (CCK) from the intestinal mucosa into circulation to promote contraction of gall bladder [135]. In the lumen, the bile salts contain mixed micelles to facilitate absorption of the fat soluble vitamins and the metabolism of dietary lipids by pancreatic enzymes before absorption. Following secretion, most of bile salts (95%) are reabsorbed in the distal ileum via the apical sodium- dependent transporter (ASBT) present on the enterocyte brush border (Fig. 4B). Then, bile salts efflux into blood by the heterodimeric transport (OSTa/OSTb) [136, 137].

However, a small amount of bile salts escape resorption and undergo deconjugation by bacterial flora before either being absorbed or converted into secondary bile acids. The

30 absorbed primary and secondary bile acids are transported back to the liver and actively transported into hepatocytes by sodium (Na+-taurocholate co-transporting polypeptide

(NTCP) and organic anion transporter (OATP) that mediate uptake of bile salts and bile acids, respectively. In the liver, bile acids are re-conjugated and then re-secreted together with newly synthesized bile salts to complete one cycle of enterohepatic circulation. In humans, bile acid pool contains ~2-4gm of bile acids. Recycling of bile acids between the liver and intestine occur six to ten times each day and transport 20-40gm bile acids.

However, 0.2-0.6gm of bile acid are excreted in the feces each day and this amount must be replenish by de novo synthesis from cholesterol [133, 137]. Importantly, the hepatic recovery of bile acids from portal vein is incomplete, therefore there is low levels of bile acids in the peripheral circulation of humans and mice [138, 139].

Bile acids have recently emerged as versatile signaling molecules having systemic endocrine functions. Bile acids affect glucose and lipid homeostasis and energy expenditure via activation of bile acid receptors in the liver, gut and peripheral tissues

[140]. The two major bile acid receptors that regulate host metabolism are the nuclear farnesoid X receptors (FXR) and membrane-bound Takeda G protein-coupled receptor

TGR5 [140]. Chenodeoxycholic acid and cholic acid are the main agonist, while secondary bile acids lithocholic acid and deoxycholic acid are the most potent TGR5 agonist. Bile acid activation of FXR regulates metabolic processes in different organs such as regulation of fat accumulation and inflammatory responses in liver during NAFLD development, lipid storage in white adipose tissue and insulin secretion in b- pancreatic cells. On the other hand, bile acids binding to TGR5 affects GLP-1 secretion in a cells in the pancreas, insulin sensitivity and energy expenditure in muscle and adipose tissue

[140]. Gut microbiota has profound effect on bile acid metabolism by promoting

31 deconjugation, dehydrogenation, and de-hydroxylation of primary bile acids in the distal small intestine and colon, thus increasing the formation of secondary bile acids and the chemical diversity of bile acids [141]. Conversely, bile acid can shape the gut microbiota community by promoting the growth of bile acid-metabolizing bacteria and inhibiting the growth of other bile acid-sensitive bacteria [142]. This indicate that the interaction between gut microbiota and bile acids is not unidirectional and synthesis, metabolism and distribution of bile acids in the body are regulated via an intriguing interaction between bile acids, their receptors FXR and TGR5 and the gut microbiota.

HS-1- associated protein X-1(HAX-1)

One protein that participates in the crosstalk between ER and mitochondria is the

HS1-associated protein x-1( HAX -1). This protein was first identified in a yeast two-hybrid assay on the basis of its binding to the hematopoietic cell-specific protein 1(HS-1) in lymphocytes to mediate lymphocyte differentiation [143]. Subsequent studies showed that HAX-1 is ubiquitously expressed in murine and human tissues and is localized in mitochondria and ER, but also present in nuclear envelope and plasma membrane [144].

Human mutation of HAX-1 causes increased neutrophil apoptosis resulting in autosomal recessive severe neutropenia and neurological impairments [145, 146]. Moreover, based on its apparent homology with the anti-apoptotic protein Bcl-2, HAX-1 was predicted to be involved in regulation of apoptosis [143]. An in vivo study support this notion and showed that the global HAX-1 deletion resulted in excessive apoptosis of neurons and postnatal lethality caused by loss of motor coordination and function, leading to failure to eat and drink [147]. HAX-1 has also been shown to interact with a number of cellular and viral proteins, thus indicating its potential involvement in multiple signaling pathway and cellular processes. In this regard, HAX-1 is reportedly associated with hypoxic tumor

32 progression [148], metastatic pancreatic cancer [149], and liver, lung, and breast cancers

[150]. Furthermore, HAX-1 has been shown to be involved in mammalian cell migration

[151].

In the heart, HAX-1 binds to phospholamban at ER/SR, where it increases phospholamban inhibition of sarco/endoplasmic Ca2+ ATPase (SERCA2a) to reduce

Ca2+ SR cycling and impact contraction-relaxation of cardiomyocytes [152, 153]. In addition, HAX-1 modulates mitochondria function through interacting with cyclophilin D to affect mitochondrial transition pore (MTP) integrity [154]. Interestingly, a recent study demonstrated that cardiac HAX-1ablation increases SERCA2a oxidation and reactive oxygen production at the ER/SR through direct interaction of HAX-1 with NOX4 [155].

Mechanistically, decreases of SERCA2 function result in elevated cytosolic Ca2+ and decreased SR Ca2+ load, which in turn causes ER stress and mitochondrial dysfunction

[96, 156]. Along these lines, previous studies have shown that HAX-1 may down-regulate

SERCA2 protein levels in non-muscle cells that may lead to ER stress and loss of mitochondrial function [157].

HAX-1 is also highly expressed in the liver and previous work showed that HAX-1 is a binding partner of the ATP-binding cassette-type protein (ABCb11), also known as bile salt export protein (BSEP), and is involved in the internalization of BSEP/ABCb11 via clarthrin-coated vesicles, suggesting that HAX-1 may have a role in receptor endocytosis and bile acid secretion [158, 159]. Moreover, HAX-1 has been shown to interact with Parl on the mitochondrial membrane of hepatocytes, maintaining cell survival through regulation of mitochondrial stability [147]. Taken together, these evidence suggest that

HAX-1 may play an important role in nutrient metabolism in hepatocytes.

33 Rationale As previously stated, obesity and the associated comorbidities are the most common and expensive causes of medical intervention in the developed world. A number of serious comorbid diseases including insulin resistance, metabolic syndrome, Type 2 diabetes and NAFLD are associated with obesity. The therapeutic options are limited due to critical scientific gaps in our knowledge of molecular mechanisms linking obesity with metabolic disturbances of insulin resistance, Type 2 diabetes and hepatic steatosis [160].

Several pathophysiological causes related to abnormal hepatic lipid metabolism have been associated with increased susceptibility to lipid accumulation in the liver and consequently lead to selective insulin resistance and hepatic steatosis. Specifically, previous studies have linked defects in the ER and mitochondrial function that follow nutrient overload. Obesity is associated with increased MAM formation which results in increased Ca2+ flux form ER to mitochondria in the liver leading mitochondrial dysfunction

[120]. On the other hand, bile acids have recently emerged as versatile signaling molecules that regulate glucose and lipid homeostasis. HAX-1 has already been identified as a key moderator of calcium homeostasis and localized to ER and mitochondria [157].

Furthermore, HAX-1 has a broad spectrum of binding partners such as bile salt export pump (BSEP/ABCb11) in the liver [159]. These unique features of HAX-1 distribution and interaction allow us to hypothesize that HAX-1 may play important role in nutrient metabolism in hepatocytes. The first specific aim of this study is to characterize the role of HAX-1 in increased Ca2+ transfer from ER to mitochondria, resulting in inhibition of normal mitochondrial respiration. The second aim is to define the role of HAX-1 in enterohepatic circulation of bile acids. To test our hypothesis, hepatocyte-specific HAX-1 deficiency mice were maintained on chow or Western diet for 12 weeks, which resulted in the development of insulin resistance and hepatic steatosis.

Methods Mice

The hepatic HAX-1 knock-out model was developed by injecting a floxed C57BL/6J HAX-

1 mouse [147] with adeno-associated virus (AAV). The Vector Biolabs generated the AAV vector (AAV8-TBG-eGFP catalog no. VB1743, and AAV8-TBG-iCre catalog no. VB

1724). We injected each AAV vector retro-orbitally at 5 X1011 genome copies per mouse and characterized the mice at 3 weeks after AAV-Tbg-iCre injection. We housed mice under a 12hr light and dark cycle (lights on at 7a.m and lights off at 7p.m). We use adult male mice at age of 10 weeks for virus injection. Mice were fed either a normal chow diet

(Teklad, Madison, WT) or a western diet containing 15.2 %Kcal protein, 42.7%Kcal carbohydrates and 42.7%Kcal fat (TD.8813, Teklad). All the animal care and use procedures followed the guidelines of the Institutional Animal Care and Use Committee of the University of Cincinnati in accordance with the guidelines of the US National

Institutes of Health.

Biochemical studies

Blood samples were collected by cardiac puncture or tail vein after an overnight fasting.

Blood glucose levels were measured using an Accu-Check Active Glucometer (Roche

Applied Science, Indianapolis, IN) by sampling from the tail. Plasma insulin levels were determined using a mouse ultrasensitive ELISA kit (Crystal Chem, Chicago, IL). Plasma

Cholesterol and triglycerides were determined using Infinity cholesterol and triglyceride kits (Thermo Fisher Scientific, Middletown, NJ). Plasma non-esterified fatty acid levels were measured using colorimetric assay kits (Wako chemicals, Richmond, VA, USA). For lipoprotein separation, 0.2ml plasma from 4-5 mice in each group, were subjected to fast performance liquid chromatography (FPLC) on a Superose 6 HR 10/30 column

35 (Amersham Bioscience, Piscataway, NJ, USA), and 0.5ml fractions were collected for cholesterol and triglycerides determination as described previously [161]. For glucose tolerance test, a glucose solution was administered orally by stomach gavage (2g per kg body weight) after an overnight fasting. For the insulin tolerance test, we intraperitoneally injected mice with insulin (Novolin R) into mice at 0.75U per kg body weight after 6-hr fast.

In vivo hepatic triglyceride secretion and clearance

After an overnight fasting, mice were injected intraperitoneally with Poloxamer 407 (1gm per kg body weight) diluted 10% in saline to assess triglyceride secretion. For postprandial lipid clearance, mice were given a 200µl bolus of olive oil (Sigma- Aldrich) by oral gavage. Blood samples (50µl) were collected from the tail vein prior intraperitoneal injection and at 1, 2, 3hr after injection to measure triglyceride levels.

Lipid synthesis and secretion in hepatocytes

Hepatocytes were isolated as previously described [162]. Cells were seeded on collagen

-coated 6 wells plate at 400,000 cells/well, allowed to attach for at least 4hrs, rinsed once with PBS and kept in 2 ml serum -free M199 media for 24hr either in non-stimulated media

(glucose 5mM plus dexamethasone 10-7 M) or stimulated media(glucose 25mM plus insulin 100nM and dexamethasone 10-7). Lipid extraction from the media and cell lysate with chloroform/methanol (2:1), the organic solvent was evaporated, and lipids were resuspended in 150µl of chloroform/ methanol. 50µl of the suspension was allowed to dry, then triglyceride concentration was quantified by calorimetric assay kit. 100 µl of cell lysate was used for protein determination by BCA kit (Pierce; Rockford, IL), with BSA as the standard. For assessment of lipid accumulation in primary hepatocytes; the cells were

36 stained with Oil Red Oil and visualized under a microscope at the end of the incubation period in high glucose M119 media

Lipid accumulation in liver

To quantify levels of triglyceride and cholesterol, 0.5gm of liver tissue was homogenized in 1ml of 50mM Tris, PH 7.4, 150mM NaCL, 5mM EDTA. We added 0.5µl of 1mCi/ml [

3H] cholesterol for extraction efficiency. The lipids were extracted twice with 1ml of petroleum ether, vortex thoroughly and centrifuge at 1000xg for 10min. 50µl of organic phase was collected into a fresh tube, allowed to dry and then we added 410µl of color reagent to quantify triglycerides, cholesterol and free fatty acid. The values were normalized to the liver weight.

De Novo lipogenesis

Lipogenesis in vivo was performed by feeding a [14C]glucose solution (2mg/g, 0.05μCi) by intragastric gavage. Tissues were harvested after 2hr for lipid extraction with chloroform/methanol (2:1). The rate of glucose incorporation into newly synthesized fatty acids were calculated as nmol of [14C]glucose incorporated into fatty acids per gm of tissue.

Weight and adiposity measurement

Age matched mice were housed with one to three mice per cage. Weights were obtained using a Denver 300K scale. Adiposity measurement were obtained using 1H magnetic resonance spectroscopy (EchoMRI-100, Echo-medical systems) as described previously

[163].

Real-Time PCR analysis of gene expression

RNA was extracted from tissue using Direct-Zol RNA kits according to manufacture instructions (Zymo Research, CA, USA). cDNA was generated using qScrip cDNA

37 synthesis kit (Quanta Bioscience, Gaithersburg, MD). PCR conditions were one cycle at

22°C for 5min; one cycle at 42°C for 30min; and one cycle at 85°C for 5min. Quantitative real-time PCR was performed on a StepOnePlus Fast Thermocycler using Fast SYBER

Green Master Mix(Applied Biosynthesis, Carlsbad, CA) with primers sequences as shown in table 1. All PCR reactions were performed in duplicate or triplicate.

Table 1: Primer sequences used for RT-PCR amplification of RNA

Gene Forward primer Reverse primer

Cyclophilin A TCATGTGCCAGGGTGGTGAC CCATTCAGTCTTGGCAGTGC

HAX-1 TCAATAGCATCTTCAGCGATATGG TCTCACCAGGTGTCTCTGACTCA

Echs1 CCACATCACCCGGGTCAA CAACCCCCACCAAGAGCATA

Hadha CGTCCCGGGCGATTG AAGCTGCGGCAAATGCA

ACC2 GGGCTCCCTGGATGACAAC TTCCGGGAGGAGTTCTGGA

CD36 GCCAAGCTATTGCGACATGA GATAGACCTGCAAATGTCAGAGGAA

CPT1a GGCAGAGCAGAGGTTCAAGCT GCCAGCGCCCGTCAT

CPT2 GAAGAAGCTGAGCCCTGATG GCCATGGTATTTGGAGCACT

FGF15 GGCCAAGCCCAGAGAACAG TTTCAAAGAAGGAGCGGTGAA

OSTa CTGAGCATAGTGGGCCTGTTC AGCTGCGCTCTTCTCAGAAATT

OSTb CAACAGCCAGGTCTTCCTAAGAG CTGGCAGAAAGACAAGTGATGAGT

Shp GCCTGGCCCGAATCCT GCGGAAGAAGAGATCTACCAGAAG

Total Bacteria CTGAACCAGCCAAGTAGCG CCGCAAACTTTCACAACTGACTTA

Bacteroidetes CATGTGGTTTAATTCGATGAT AGCTGACGACAACCATGCAG

Firmicutes ATGTGGTTTAATTCGAAGCA AGCTGACGACAACCATGCAC

CYP27A GCGGGCAGAGAGTGAATCAG ATGGCTTCCAAGGCAAGGT

StAR CGAGGGCAGCTGTAGAGTGTT TTCTCCACTGGCAGCCTGTT

ASBT TGGAAACTGGAATGCAGAACACT GAAGGTGAACACCAGGTTGAGAT

ABCG5 AACTTCACTTGTGGTGGATCCA TGGACCCCTTGGGTGATG

38 ABCG8 GAGCTGCCCGGGATGATA CCCGGAAGTCATTGGAAATCT

Fecal Microbiota Analysis

The bacteria in feces content were extracted using PowerSoil DNA Isolation Kit (MO Bio laboratory, Inc., Carlsbad, CA). After purification, o.5ng/ul DNA was used for qPCR reaction.

Western blot analysis

Livers, ileum and other tissues were excised and homogenized on ice-cold RIPA buffer

(ThermoFisher Scientific) containing protease and phosphatase inhibitor cocktail (Roche

Diagnostics). SDS samples were prepared by adding 4xSDS sample buffer [40%

(vol/vol) glycerol, 240mM Tris:HCL (PH6.8), 8% (wt/vol) SDS, 0.04% bromophenol blue ,

5% (vol/vol) b-mercaptoethanol) to four volumes lysate and boiling at 95 °C for 5 min.

Samples were run on Express Plus PAGE Gels(GenScript, Piscataway, NJ, USA), and gels were transferred to polyvinylidene difluoride (PVDF) membrane by semiwet transfer.

The membranes were blocked in Tris-buffered saline solution containing 5% milk and

0.1% Tween 20 for 1hr at 4°C and then incubated for 90 min with a 1:1000 dilution of primary antibodies. The membranes were washed and then incubated with horseradish peroxidase-conjugated secondary antibodies and visualized using Pierce enhanced chemiluminescence reagents. For assessment of BSEP expression, canalicular plasma membrane was isolated as described previously [164]. Primary antibodies used to determine protein expression were HAX-1 antibody (BD Transduction catalog no.

610825), Phospho-PDH Ser 300 (Millipore catalog no. ABS194), PDH (ThermoFisher catalog no. 459400), BSEP (ABCB11, Boster biological catalog no. PB9414) , IP3R1(

Santa Cruz catalog no. sc-271197), Cytochrome C antibody (Cell Signaling catalog no.

4272S), Calreticulin antibody (Cell Signaling catalog no. 12238S), LDLR antibody (Abcam

39 catalog no.ab30532), MRP2 antibody (Abcam catalog no. ab110740), b-actin (Cell

Signaling catalog no. 4967), CYP27A antibody (Abcam catalog no. ab126785), StARD antibody (Abcam catalog no. ab202060) and GAPDH ( Abcam catalog no. ab9482).

Co-immunoprecipitation

The HAX-1 specific antibodies were crosslinked to Protein A Dyna beads with BS3 crosslinker prior to incubation for 12hr at 4°C with liver lysates that had been pre-cleared with off-target antibody and Protein A-Sepharose beads. The samples were eluted with

2X NuPAGE lithium dodecyl sulfate sample buffer (Thermo Fisher Scientific, NP0008) for

5 min at 95°C and used for Western blot analysis.

Cellular oxygen consumption and fatty acid oxidation

OCR was measured in intact hepatocytes in real time using XF24 Extracellular Flux

Analyzer and the XF 24 v1.5.0.69 software ( Seahorse bioscience). Cells were seeded on collagen-coated XF24 plates (1hr at RT) at 10,000 cells/well. Cells were incubated overnight in William’s E media (Life Technologies) supplemented with 100U/ml

Penicillin/Streptomycin , 2mM L-glutamine , 5% FBS, 100nM insulin and 100nM dexamethasone. Before the study, medium was removed, and cells were rinsed twice by

PBS and kept in 500µl of DMEM base supplemented with 25mM glucose(Fisher), 1mM sodium pyruvate (Gibco) and 4mM Glutamax (Gibco) for 1hr in the seahorse CO2-free incubator at 37°C. After baseline measurement, additions were delivered in the following order through the instrument’s individual injection ports: Oligomycin 12 µg/ml (Sigma),

FCCP 1.5µM (Sigma) and Rotenone 1µM/Antimycin A 1.5µM (Sigma). For fatty acid oxidation, cells were incubated for 5hr in DMEM base without any substrate, cells then were washed with Krebs-Henseleit buffer (KHB; 11mM NaCl, 4.7 mM KCl, 2mM Mgso4,

1.2 mM Na2HPO4, 2.5mM glucose, 0.5mM carnitine and 0.1 µM insulin). Then 500µl KHB

40 was added to each well and plate was incubated in the seahorse CO2-free incubator at

37°C for 1hr. BSA conjugated palmitate (150 µl) was added to port A of the seahorse flux plate. Fatty acid stimulated oxygen consumption was measured after reading basal respiration rate. Etomoxir (Sigma) was added in port B for a final concentration of 50µM of the flux plate to inhibit respiration and to calculate the OCR and ATP produced specifically by fatty acid stimulation.

Histology

For the Hematoxylin and Eosin ( H & E) staining ( Sigma-Aldrich , catalog no. MHS32-

1L), we fixed tissues in 4% paraformaldehyde overnight, dehydrated the samples, paraffin embedded them and 5-micron sections were prepared for staining. For the Oil Red O staining, we prepared 5 micron frozen sections from snap-frozen liver tissues and fixed them in 37% formaldehyde for 1min. We then stained the section in 0.5% Oil Red O in propylene glycerol and then visualized under microscope.

Calcium measurements

Changes of intracellular calcium concentration in hepatocytes were determined using the

Fluo-4AM NW calcium assay kit (molecular probes, Eugene, OR). Following culture of cells with serum-free M119 media for 16hr in presence of 5mM or 25mM glucose, cells were loaded with Fluo-4AM in the presence of probenecid. Thapsigargin 450nM or

DMSO was added and fluorescence (490 excitation; 516nm emission) was measured over a 5min period. Data are expressed as fluorescence units in thapsigargin treated cells

– fluorescence units in DMSO treated cells. Mitochondrial calcium levels under basal and nutrient-stimulated conditions were assessed as described [118]. In brief, the hepatocytes were incubated with or without 150μM palmitate, or with 5 or 25mM glucose for 6hr prior to the addition of the mitochondria-specific calcium indicator Rhod-2AM. The cells were

41 washed 3 times and then replenished with fresh media to measure fluorescence intensity

(552 nm excitation; 581 nm emission).

Pyruvate dehydrogenase activity assays

Liver tissue lysates were prepared by homogenizing tissues in ice-cold PBS buffer containing protease inhibitor cocktail (Roche Diagnostics). Four hundred micrograms of tissue lysates was applied to measure PDH enzymatic activities using a commercially available kit (Abcam, ab109902). PDH enzymatic activity was measured at 450nm.

Bile acid analysis

Bile acid levels in plasma, liver, small intestine and feces were measured as described

[162]. Briefly, fasting mice were sacrificed and fresh tissues were collected, weighed and minced. Bile acids were extracted with 75% ethanol for 1hr at 50 °C. Bile acids were measured enzymatically using the mouse total bile acid assay kit (Cell Biolabs, STA-631).

Bile acid levels in each tissue were expressed as µmol/kg of body weight of each animal.

Fecal cholesterol measurements

Feces were collected from individually housed mice over 3 days. Dried feces (0.3-0.5 g) were then minced and extracted in 2.5ml saline and 2.5ml of chloroform/methanol (2:1 v:v). The organic phase that contains the extracted lipids was collected in a glass tube, air dried, and then re-dissolved in 400µl of 100% ethanol and cholesterol content was measured enzymatically using cholesterol assay kit (Thermo Fisher Scientific,

Middletown, NJ).

Fecal Microbiota Analysis

The bacteria in feces content were extracted using PowerSoil DNA Isolation Kit (MO Bio laboratory, Inc., Carlsbad, CA). After purification, 0.5ng/ul DNA was used for qPCR reaction.

42 Statistical Analysis

Statistical analysis was performed using GraphPad Prism version 8. Values are expressed as mean± SD. Multiple comparisons were tested by Student’s t test or analysis of variance (ANOVA). A value of P<0.05 was considered to be statistically significant.

Results HAX-1 is localized at mitochondria and ER in liver and its inactivation has no deleterious effects

Initial characterization of HAX-1 distribution in the liver revealed its predominant location in the mitochondria (Mito extract) and the ER (Fig. 5A). The detection of HAX-1 in liver

ER is not due to contamination of the ER fraction with mitochondria as indicated by the absence of Cytochrome C in this fraction. To investigate the function of hepatic HAX-1 in metabolic regulation, we ablated HAX-1 expression in adult liver by injection of 5 x1011 genome copies of recombinant AAV8 virus encoding Cre recombinase gene under the control of the strong liver-specific TBG promoter into Hax1flox/flox mice [155, 165]. We have chosen to use this approach instead of mating the Hax1flox/flox mice with liver-specific Cre transgenic mice because of the effectiveness and expediency of this procedure, and to minimize the uncertainty on the effect of HAX-1 deletion in liver development. Also, this acute deletion approach makes it possible to distinguish the direct effects of HAX-1 deletion from any secondary or compensatory effects. Results showed that AAV-TBG-

Cre administration effectively eliminated HAX-1 expression in the liver whereas injection of the control AAV-TBG-GFP has no effect on HAX1 expression (Fig. 5B). The specificity of liver HAX-1 deletion was verified by immunoblots showing AAV-TBG-Cre has no effect on HAX-1 levels in other tissues, except the liver of Hax1flox/flox mice (Fig. 5B). No

43 difference in serum ALT levels was observed between Hax1flox/flox mice with AAV-TBG-

GFP or AAV-TBG-Cre injection (Fig. 6A). Histological examination also revealed no abnormalities nor fibrosis in the liver tissues of these animals (Fig. 6B). These data indicate that HAX-1 inactivation does not cause liver injury or toxicity. Moreover, Mice with liver specific HAX-1 deletion weighed slightly less but not significantly different than control mice and displayed no differences in body composition when these mice fed a chow diet (Fig. 7A&B), suggesting the absence of a global energetic defect. All subsequent experiments comparing the impact of liver-specific HAX-1 inactivation were performed by using Hax1flox/flox mice injected with AAV-TBG-GFP as control mice and those injected with AAV-TBG-Cre as liver-specific HAX-1 knockout (L-Hax1-/-) mice.

Hepatic HAX-1 deficiency reduces plasma and hepatic triglyceride levels in chow- fed mice

The impact of liver HAX-1 deficiency on lipid metabolism was assessed by comparing plasma lipid levels between chow-fed control and L-Hax1-/- mice. Results revealed significantly lower fasting plasma triglyceride and VLDL levels in L-Hax1-/- mice compared to control mice (Fig. 8A,B). In contrast, fasting plasma cholesterol levels were not impacted by HAX-1 inactivation in liver in these mice fed a chow diet(Fig. 9A). Consistent with these results, cholesterol distribution among various lipoprotein fractions revealed similar IDL/LDL and HDL levels between control and L-Hax1-/- mice (Fig. 9B). Moreover, plasma non-esterified fatty acid (NEFA) reduced by (36%) in L-Hax1-/- mice but did not reach the statistical significance (Fig. 10). The difference in plasma triglyceride and VLDL levels in control and L-Hax1-/- mice was not due to differences in VLDL synthesis and secretion (Fig. 11A,B). In contrast, when control and L-Hax1-/- mice were administered a bolus lipid meal after an overnight fast, significantly less postprandial hypertriglyceridemia

44 was observed in L-Hax1-/- mice (Fig. 11C) although the area under the curve analysis of the data revealed that the difference did not reach statistical significance(Fig. 11D).

Further analysis revealed that triglyceride but not cholesterol levels in the liver of fasted

L-Hax1-/- mice were significantly lower (36% less) compared to control (Fig. 12A,B).

However, this difference was not observed in fed mice (Fig. 12C&D). Thus, the reduction in plasma lipids did not enhance hepatic lipid accumulation during the fasting state. These results indicate that hepatic HAX-1 deficiency does not influence VLDL synthesis and secretion under basal conditions but enhances triglyceride-rich lipoprotein clearance to lower plasma triglyceride levels and reduces triglyceride content in fasted mice

L-Hax1-/- mice sustain euglycemia with lower insulin levels

Fasting blood glucose levels were similar and within normal range in both age-matched

AAV-TBG-GFP injected control Hax1flox/flox mice (GFP) and AAV-TBG-Cre injected L-

Hax1-/- mice (Fig. 13A). Interestingly, while fasting insulin levels in the control mice were similar and within the normal range observed with wild type mice, fasting insulin levels were found to be significantly lower in L-Hax1-/- mice compared to the control mice (Fig.

13B), leading to a lower HOMA-insulin resistance index suggestive of elevated insulin sensitivity in mice with liver HAX-1 inactivation (Fig. 13C). Indeed, when a glucose solution (2 gm/kg ) was administered orally to these animals after an overnight fast, an enhanced glucose excursion rate was observed in L-Hax1-/- mice (Fig. 14A,B). While statistical differences were reached for glucose tolerance, no difference was observed for insulin tolerance at least in these mice fed a chow diet (Fig. 14C). Importantly, no difference was also observed for pyruvate tolerance (Fig. 15). Thus, these results indicate that pyruvate conversion into glucose was not different between control and L-Hax1-/- mice, which reflect no changes in gluconeogenetic pathways during hepatic HAX-1

45 inactivation. Taken together, these results indicate that the L-Hax1-/- mice are capable of maintaining euglycemia with lower insulin levels in a mechanism that is consistent with improved glucose tolerance and insulin sensitivity.

Hepatic HAX-1 inactivation protects against Western diet-induced hyperlipidemia and hyperinsulinemia

The low plasma triglyceride levels and their ability to maintain euglycemia with low basal fasting insulin levels in chow-fed L-Hax1-/- mice prompted us to investigate the influence of liver HAX-1 inactivation on metabolic disease manifestation in response to chronic feeding of a high fat-cholesterol type Western diet. The AAV-TBG-Cre injected mice displayed a slightly lower body weight and liver/body ratio compared to AAV-TBG-GFP injected control mice after 12 weeks of Western diet feeding (Fig. 16A,B), but the difference did not reach statistical significance. Measurement of body composition by nuclear magnetic resonance showed that also there is no difference in fat mass between

L-Hax1-/- and control mice (Fig. 16C). Importantly, plasma triglyceride and cholesterol levels throughout the 12-week period remained at normal levels of <100 mg/dL in the L-

Hax1-/- mice in contrast to the hypertriglyceridemia (Fig. 17A) and hypercholesterolemia observed in control mice (Fig. 18A). Differences in plasma triglyceride and cholesterol levels were reflected by the lower plasma VLDL and LDL levels in L-Hax1-/- mice (Fig.

17B and 18B). In addition to reduced plasma triglyceride and cholesterol, L-Hax1-/- mice had significantly lower non esterified free fatty acid (NEFA) levels compared to control mice (Fig. 19). Both the control and the L-Hax1-/- mice were capable of maintaining euglycemia throughout the 12-week Western diet feeding period and neither group developed overt hyperglycemia (Fig. 20A). However, the control mice developed hyperinsulinemia indicative of prediabetes and insulin resistance whereas insulin levels

46 in the L-Hax1-/- mice remained at low or normal levels indicative of resistance to diet- induced hyperinsulinemia (Fig. 20B). Estimation of insulin resistance by HOMA-IR index is consistent with the interpretation that the control mice became insulin resistant after 12 weeks of Western diet feeding, but the L-Hax1-/- mice remained insulin sensitive with a low HOMA-IR (Fig. 20C). Glucose tolerance test after oral administration of a glucose solution (Fig. 21A,B) as well as insulin sensitivity test administered by insulin injection

(Fig. 22A,B) confirmed that the Western diet-fed control mice were glucose intolerant and insulin resistant but the L-Hax1-/- mice remained glucose tolerant and were sensitive to insulin-induced glucose excursion from blood 12-weeks after feeding the Western type diet. Surprisingly, gluconeogenesis assessment by pyruvate tolerance test showed no differences between control and L-Hax1-/- mice after 12 weeks of western diet feeding, which consistent with that shown in mice fed a chow diet (Fig. 23)

Hepatic HAX-1 deficiency suppresses diet-induced hepatosteatosis by reducing glucose-induced lipogenesis

Examination of the livers from control and L-Hax1-/- mice after the 12-week Western diet feeding period revealed significantly more hepatic triglycerides and cholesterol in control mice compared to that observed in L-Hax1-/- mice (Fig. 20A,B,C). These observations indicated that hepatic HAX-1 deficiency also protects against diet-induced hepatosteatosis. To examine the mechanism underlying the differences in hepatosteatosis susceptibility between control and L-Hax1-/- mice, primary hepatocytes isolated from these animals were incubated in media containing 5 and 25mM glucose to mimic normal and hyperglycemia conditions and then assayed for triglyceride content in cell lysate and in the extracellular media. When the hepatocytes were incubated with

5mM glucose, no difference in intracellular triglyceride content between control and HAX-

47 1 deficient cells was observed (Fig. 25B,C). However, in the presence of 25 mM glucose, an ~2.5-fold increase in intracellular triglyceride content was observed in control hepatocytes but the L-Hax1-/- hepatocytes did not display glucose-induced steatosis and their intracellular triglyceride levels remained similar to those observed in control and L-

Hax1-/- hepatocytes cultured under normoglycemic conditions (Fig. 25A,B). The L-Hax1-

/- hepatocytes also secrete less triglyceride into the media under normoglycemic conditions (Fig. 25C). While hyperglycemic conditions increased the triglyceride levels in the media of L-Hax1-/- hepatocytes to levels similar to those observed when control hepatocytes were cultured under normoglycemic condition, significantly more triglycerides were found in the media of the control hepatocytes when cultured with 25 mM glucose (Fig. 25C). These in vitro data were corroborated with in vivo experiments showing reduced de novo lipogenesis in the liver of L-Hax1-/- mice (Fig. 26). Taken together, Taken together, these results indicate that hepatic HAX-1 inactivation reduces glucose-induced lipogenesis thereby lowers plasma triglyceride levels and protects against diet-induced hepatosteatosis in L-Hax1-/- mice.

Hepatic HAX-1 deficiency increases mitochondrial respiration and fatty acid oxidation

The lower plasma triglyceride levels as well as reduced glucose-induced lipogenesis and steatosis observed in L-Hax1-/- mice suggested that hepatic HAX-1 deficiency increases glucose utilization. To test this possibility, primary hepatocytes isolated from control and

L-Hax1-/- mice were incubated with normal media containing 25mM glucose and oxygen consumption rates were monitored over a 2-hour period. Oligomycin (12µg/mL) was added to inhibit ATP synthesis, followed by 1.5μM FCCP for mitochondria uncoupling and measurement of mitochondrial respiratory capacity, and then rotenone (1µM) and

48 antimycin A (1.5µM) were added to inhibit complex 1 for measurement of non- mitochondrial respiration (Fig. 27 A,B). Results revealed significant increase in basal and maximal respiration rates as well as ATP production (Fig. 28A-D) by hepatocytes from

AAV-TBG-Cre injected L-Hax1-/- mice compared to control hepatocytes isolated from

AAV-TBG-GFP injected mice. However, spare respiration capacity and coupling efficiency were not different between L-Hax1-/- and control hepatocytes (Fig. 29 A,B). The latter observations indicated that HAX-1 deficiency improves normal mitochondrial respiration that is unrelated to reserve respiratory capacity in response to stress or energy dissipation as heat. Furthermore, when oxygen consumption rates were measured in response to 150μM palmitate in the presence or absence of the fatty acid oxidation inhibitor etomoxir to assess fatty acid-specific oxidation, we found that the L-Hax1-/- hepatocytes also displayed significantly higher fatty acid oxidation (Fig. 30A,B) and fatty acid-specific ATP production compared to control hepatocytes (Fig. 31A-C). Consistent with the higher fatty acid oxidation rates of L-Hax1-/- hepatocytes, non-esterified fatty acid

(NEFA) levels were found to be lower in the livers of L-Hax1-/- mice (Fig. 32). Analysis of hepatic RNA expression revealed the altered expression of selected fatty acid oxidation genes including increased expression of carnitine palmitoyltransferase-1 (CPT1) that is responsible for fatty acid transport into the mitochondria and the reduced expression of acetyl CoA carboxylase-2 (ACC2), an enzyme that inhibits fatty acid oxidation and promotes fatty acid synthesis (Fig. 33). Taken together, these results support the hypothesis that improved hepatic glucose utilization and fatty oxidation rate is at least in part responsible for the improved glucose tolerance and resistance to diet-induced hyperinsulinemia and hyperlipidemia observed in L-Hax1-/- mice.

49 Hepatic HAX-1 deficiency increases PDH activity

Additional experiments were performed to identify the mechanism by which HAX-1 deficiency increases glucose and fatty acid oxidation rates and protects against diet- induced metabolic diseases. In view of the importance of the PDH complex in glucose oxidation, we determined PDH activities in livers of L-Hax1+/+ and L-Hax1-/- mice and showed that HAX-1 inactivation increased PDH activity in the liver of mice fed a chow or

Western diet (Fig. 34A,B). While this increase was not due to changes in PDH kinase

(PDK)-2 and -4 expression (Fig. 35A,B), HAX-1 inactivation significantly reduced

PDHE1a phosphorylation under both chow- and Western diet-fed conditions (Fig.

36A,B). Increased mitochondrial respiration due to higher PDH activity has also been reported to elevate LDL receptor expression in hepatocytes [166]. Indeed, increased LDL receptor protein levels were observed in Western diet-fed L-Hax1-/- mice compared to those observed in control mice (Fig. 37). The increased level of LDL receptor may partly account for the lower plasma lipid levels observed in L-Hax1-/- mice.

Hepatic HAX-1 inactivation improves ER-mitochondria calcium homeostasis via

IP3R1

Increased PDH activity due to higher levels of non-phosphorylated PDH in the absence of differences in PDK2 and PDK4 suggested that HAX-1 inactivation may increase the dephosphorylation of PDH, a process that is mediated by calcium-dependent phosphatases in the mitochondria [167]. In view of emerging evidence showing that ER calcium homeostasis is an important determinant of mitochondrial respiration and metabolic diseases [120, 168, 169], and HAX-1 is present in both the ER and the mitochondria, we determined if HAX-1 inactivation influenced ER and mitochondria calcium homeostasis in the liver. Primary hepatocytes isolated from L-Hax1-/- mice were

50 found to display significantly higher ER calcium levels compared to control mice under both normoglycemic (5 mM glucose) and hyperglycemic (25 mM glucose) conditions (Fig.

38A,B). In contrast, mitochondrial calcium levels were found to be lower in L-Hax1-/- hepatocytes compared to control cells (Fig. 39). Importantly, HAX-1 inactivation also prevented palmitate- and hyperglycemia-induced mitochondrial calcium overload (Fig.

39).

Mitochondrial dysfunction due to over-nutrition is caused by excessive ER calcium transport to the mitochondria, a process that is mediated in part by IP3R1[118, 120].

-/- Interestingly, we found lower levels of IP3R1 protein in the liver of L-Hax1 mice compared to control mice (Fig. 40A,B). To determine how HAX-1 may regulate IP3R1 level in the liver, we performed co-immunoprecipitation experiments with HAX-1 specific antibodies and discovered that HAX-1 interacts with IP3R1 (Fig. 40A,B). The finding that

HAX-1 interacts with IP3R1 and its ablation reduces IP3R1 levels in hepatocytes is reminiscent to HAX-1 interaction with SERCA2 to suppress its degradation in the heart

-/- [155]. Thus, the reduced IP3R1 levels in L-Hax1 mice is likely to its enhanced degradation. Importantly, the reduced IP3R1 level due to HAX-1 deficiency limits ER calcium transport to the mitochondria, thereby preventing excessive nutrient-induced mitochondrial calcium overload and dysfunction.

Hepatic HAX-1 inactivation increases hepatic BSEP level and bile acid responsive gene expression in intestine and liver

Another liver protein that interacts with HAX-1 and its activity is sensitive to cellular calcium modulation is the bile salt export protein BSEP encoded by the ABCB11 gene

[159, 170]. Interestingly, we observed increased BSEP levels in liver lysates as well as bile canalicular membranes of L-Hax1-/- mice compared to control mice (Fig. 41A,B).

51 Moreover, and consistent with results reported by others [171], the elevated BSEP expression increased bile acid pool size as reflected by higher levels of bile acids in plasma and liver (Fig. 42A,B). However, fecal bile acid levels and bile acid levels in the intestine were not different between these animals (Fig. 42C,D) suggesting that hepatic

HAX-1 deficiency promotes the conservation of bile acids within the enterohepatic circulation similar to that observed in liver-specific ABCB11 transgenic mice [171, 172].

Interestingly, we also observed increased excretion of cholesterol in the feces of both chow- and Western diet-fed L-Hax1-/- mice (Fig. 43). Moreover, and similar to results reported for the ABCB11 transgenic mice [171], hepatic HAX-1 deficiency did not alter intestinal expression of the bile acid transporter ASBT but robustly increased the expression of bile acid-responsive genes including FGF15, Osta/b, and the short heterodimer partner Shp (Fig. 44A,B). Hepatic expression of bile acid-responsive genes such as Shp, Abcg5, and Abcg8 [173] were also higher in L-Hax1-/- mice (Fig. 45A,B).

Increased expression of Abcg5 and Abcg8 is consistent with higher fecal cholesterol content in L-Hax1-/- mice. Additionally, expression of the mitochondria cholesterol transport protein steroidogenic acute response protein (StAR) and the mitochondrial bile acid synthesis gene Cyp27a was also increased in the livers of Western diet-fed L-Hax1-

/- mice (Fig. 46,47). The increased expression of StAR that mediates cholesterol transport into the mitochondria [174], and Cyp27a that promotes the alternative acidic pathway of bile acid synthesis [127], may also partly account for the reduced hepatic cholesterol level and increased bile acid pool observed in L-Hax1-/- mice. Taken together, these data indicate that hepatic HAX-1 inactivation, via increased BSEP levels in bile canaliculus membrane and enterohepatic bile acid circulation [171], also protects against diet-

52 induced metabolic disease by activation of bile acid-responsive genes in the intestine and liver.

Hepatic HAX-1 inactivation affects gut microbiota in mice fed a Western diet

Previous studies have documented that alterations of gut microbiota are related to obesity, insulin resistance, Type 2 diabetes and NAFLD [175-177]. The human gut microbiota has emerged as an environmental factor, which contributes to host metabolism on the basis of the initial finding that germ-free(GF) mice have reduced adiposity, which can be reversed by colonization with a normal gut microbiota[178]. We examine changes in fecal microbiota in L-Hax1+/+ and L-Hax1-/- mice fed a chow or Western diet. We found that a feeding of Western diet is associated with increased population of Firmicutes and decreased population of Bacteroidetes (Fig. 48A), which consistent with studies showing that Firmicutes population are elevated in obese and diabetic subjects [179] and decrease after gastric bypass surgery [180]. Interestingly, hepatic HAX-1 inactivation significantly increased relative abundance of Bacteroidetes compared to control mice during a

Western diet feeding (Fig. 48A). This resulted in a significant increase in the ratio of

Firmicutes/ Bacteroidetes(F/B) (Fig. 48B). Consistent to our study, previous studies have shown that increased (F/B) ratio associated with improved insulin resistance and a decrease in adiposity [176].

Discussion Obesity and its associated comorbidity are among the most prevalent and challenging conditions affecting the medical profession in the 21st century. A major metabolic consequence of obesity is insulin resistance, which is strongly associated with deposition of triglycerides in the liver leading to hepatosteatosis. Thus, understanding the cellular and molecular mechanisms that contribute to the pathogenesis of these diseases

53 is vital for developing novel therapeutic strategy. Accumulative evidence suggests that both ER and mitochondria are key actors in energy homeostasis of liver and that hepatic insulin resistance is associated with mitochondrial dysfunction, ER stress, and altered lipid metabolism. In particular, emerging evidence suggests that disruption of Ca+2 homeostasis in the ER and mitochondria plays a central role in the development of abnormal insulin action and metabolic alteration in the liver [99, 181]. Hence, identification of proteins that regulate the proper function of both ER and mitochondria may provide significant insight for the development of therapeutic strategy to rectify dysfunction of both organelles. Along these lines, HAX-1 is one of these potential candidates.

HAX-1 was first identified to be a 35kDa protein that contains Bcl-2 homology domains 1 and 2 (BH1, BH2) at the N-terminal end [143]. Although the existence of predicted BH1 and BH2 suggests similarities between HAX-1 and other Bcl-2 family members, HAX-1 only shows a weak homology with a pro-apoptotic protein called Bcl-

2/adenovirus EIB 19KDa protein-interacting 3 (BNIP3) [143]. Moreover, the role and existence of purported BH1 and BH2 domains in HAX-1 have been disputed on the basis of data obtained by sequence analysis and structure prediction [182, 183]. Since crystal structure is currently unavailable, the existence of BH1 and BH2 domains remains largely unknown. HAX-1 contains a putative PEST( polypeptide sequences enriched in proline, glutamate, serine and threonine) from amino acids (104-117), which suggests rapid and regulated degradation of the protein by proteasome. HAX-1 also contains an acid box region in the N-terminus that contains primarily glutamic and aspartic acids, but the functional significance of this domain remains undefined. Furthermore, HAX-1 also shows a putative transmembrane domain near the C-terminus (amino acids 261-273), indicating a role of HAX-1 as an integral membrane protein as well as several identified protein-

54 binding regions that are also present mostly in the C-terminal part of the protein [143,

182]. Interestingly, sequences present in the N-terminus of HAX-1 are responsible for its mitochondrial targeting, whereas sequences in its C-terminus are essential for its localization at the ER membranes [184]. Many studies showed that HAX-1 is ubiquitously expressed in virtually every tissue [143, 144, 185]. HAX-1 has high protein and mRNA levels in liver, heart, and testis in healthy mouse, rat and human tissue samples [182,

186]. Discrepancies were found in the expression of HAX-1 in skeletal muscle, jejunum and brain. These differences in the expression may be due to the differences in posttranslational regulation such as mRNA stability, efficiency of translation or protein stability in individual organs [186].

HAX-1 expression is up-regulated in a broad variety of tumor tissues including lung cancer, leukemia, myeloma, breast cancer, hepatoma and, cervical tissue [187-190].

HAX-1 may be an upstream regulatory factor of pro-apoptotic protein P53, inhibiting the rise in its expression level [190]. Interestingly targeting HAX-1 in some types of cancer increases sensitivity of cancer to chemotherapy [191, 192]. Since HAX-1 contains the homology domain of Bcl-2 family and has weak homology with the apoptosis-related protein NIP3, it may function as an anti-apoptosis protein through this Bcl-2 homology domain. In this regard, HAX-1 is shown to play a critical role in fine-tuning the ER stress through interaction with inositol-requiring enzyme (IRE-1), a Ser/Thr kinase with an endonuclease activity [193]. Hyperactivation of IRE-1 promotes cell death through phosphorylation of JNK and activation of caspase-12 [194, 195].

In the heart, HAX-1 was identified as a regulator of mitochondrial permeability transition pore (mPTP) and mitochondrial membrane integrity through interaction with cyclophilin D. The presence of HAX-1 renders cyclophilin D more prone to ubiquitination

55 and degradation, resulting in inhibition of mPTP opening and cell death [154]. Amino acids

175-206 of HAX-1 are demonstrated to interact with procaspase- 9, which is capable of initiation of apoptosis via caspase -3 activation [196]. Moreover, the C-terminal of HAX-1 can directly interact with X-linked inhibitor of apoptotic protein (XIAP), preventing its ubiquitination. As ubiquitination targets protein for proteasomal degradation, the binding of HAX-1 preserves the XIAP protein and promotes cell survival [197]. Interestingly, HAX-

1 inhibits the apoptosis in glioblastoma cells by disruption of the interaction between Akt1 and Hsp90 [198].

The anti-apoptotic property of HAX-1 may be due to its regulation of ER calcium homeostasis. Previous studies showed that HAX-1 interacts with the ubiquitous calcium pump SERCA2 to regulate intracellular calcium homeostasis [157]. Amino acids 203-245 of HAX-1 binds to the SERCA2 nucleotide binding domain, where ATP hydrolysis normally catalyzes a SERCA2 domain rearrangement to allow SERCA2 to pump Ca+2 into the ER. HAX-1 overexpression in HEK 293 cells leads to a 37% reduction of Ca+2 released from the ER to the cytosol on treatment with thapsigargin, an irreversible

SERCA2 inhibitor [157]. The same group also demonstrated that this in vitro interaction of HAX-1 with SERCA2 decreases when Ca+2 levels are increased and interaction of

HAX-1 with SERCA2 promotes proteasomal degradation of calcium pump. Importantly, overexpressing SERCA2 abolished any protective effects of HAX-1 against apoptosis induced by hydrogen peroxide and thapsigargin [157].

In addition to its role in apoptosis and Ca+2 homeostasis, HAX-1 promotes cell migration, which may contributes to its tumorigenesis. It has been shown that HAX-1 in

Hela cells colocalizes with PKD2 in lamellipodia, which are important features during cell migration. In the same study, HAX-1 is also demonstrated to interacts with cortactin, an

56 F-actin associated protein, further suggesting that HAX-1 may play a role in cytoskeleton organization at lamellipodia [199, 200]. The potential role of HAX-1 in metastasis is further demonstrated in an oral squamous cancer cells study. The last 9 amino acids of HAX-1 are identified to be important for interaction with b6 subunit of integrin avb6 and the regulation of integrin endocytosis [151]. This study suggests that HAX-1 is involved in regulating clathrin-mediated endocytosis of integrin avb6 and down regulation of HAX-1 enhances the endocytosis of avb6 and suppresses cell migration.

In the heart, HAX-1 also regulates intracellular calcium homeostasis via interaction with an ER-resident protein phospholamban, the cardiac inhibitor of SERCA2 activity [152]. The HAX-1/phospholamban interaction has a physiological effect on cardiomyocytes: it reduces myocytes contraction and Ca+2 transient by stabilizing the phospholamban/SERCA2 complex, to decrease Ca+2 kinetics and SERCA Ca+2 affinity

[153]. Thus, HAX-1 and phospholamban may also work cooperatively to inhibit apoptosis through regulation of SERCA2 activity and ER Ca+2 homeostasis in the heart.

The association of HAX-1 with regulators of ER Ca+2 homeostasis is not limited to phospholamban and SERCA2. HAX-1 also interacts with polycystin-2, which is the product of second polycystic kidney disease gene, also known as PKD2/transient receptor potential cation channel, subfamily P, member2 (TRPP2) [200]. More

+2 importantly, polycystin-2 interacts with a MAM localized IP3R1, to inhibit IP3- induced Ca release from the ER. Therefore, cells expressing TRPP2 showed a diminished increase in the mitochondrial Ca+2 after stimulation with histamine, reflecting the reduced amount of released Ca+2 in the ER [201]. However, it remains to be elucidated if the HAX-1 is

+2 required for the TRPP2 and IP3R1. Therefore, regulation of ER-Ca homeostasis by

TRPP2 strongly suggests that the pathological TRPP2 mutation contributes to the

57 development of autosomal dominant polycystic kidney disease. The interaction of HAX-1 with calcium pump SERCA2 and polycystin-2, which associated with intracellular receptor

+2 IP3R1 implies a connection between HAX-1 and ER-mitochondrial Ca signaling that occur on the MAM [202, 203]. On this specialized sub-domain of ER, Ca+2 is transferred between two organelles to provide optimal bioenergetics by providing sufficient reducing equivalents to support oxidative phosphorylation [204]. These observations provided additional support for the hypothesis that HAX-1 plays essential role in ER-mitochondria communication through regulation of Ca+2 transfer. The importance of HAX-1 in intracellular organelle functions is highlighted by studies showing that human mutation in

HAX-1 have been linked to mitochondrial dysfunction [145, 182]. Interestingly, Hirasaka and coworkers suggested that HAX-1 has calcium binding capabilities [205]. They also showed that HAX-1 C-terminal domain required Ca+2 to bind with UCP3 in the mitochondria. These finding suggested that HAX-1 can directly bind Ca+2 and play important role in mitochondrial calcium sensing [205].

HAX-1 is also highly expressed in the liver. Since the liver represents a major metabolic organ that controls gluconeogenesis, glycogen storage, lipogenesis, and lipid metabolism, it is possible that HAX-1 may also regulate hepatocyte nutrient metabolism and sensitivity to excess nutrient-induced metabolic diseases through its interaction with a broad spectrum of binding partners in distinct subcellular organelles. Importantly, yet, whether HAX-1 functions in the liver to regulate mitochondrial fuel utilization and metabolic diseases is currently unknown. Therefore, the objective of this dissertation is to investigate the role of HAX-1 in hepatic glucose and lipid metabolism. We generated liver-specific knockout model (L-Hax1-/- ) mice to test the hypothesis that hepatic HAX-1

58 is a potential therapeutic target to prevent development of obesity, diabetes, steatosis, and metabolic syndrome.

In the current study, we discovered that HAX-1 serves as a crucial regulator of hepatocyte energy metabolism. In the liver, HAX-1 is localized in the mitochondria and

ER, and contributes to the development of diet-induced hepatic steatosis, hyperlipidemia and diabetes. Our study presents the first evidence that hepatic HAX-1 inactivation increases insulin sensitivity, resulting in protection against diet-induced hyperlipidemia, hyperinsulinemia, and hepatic steatosis. The impact of HAX-1 inactivation is highly potent as low insulin concentration is sufficient to maintain euglycemia in chow-fed L-Hax1-/- mice. The metabolic benefits of hepatic HAX-1 inactivation are multi-factorial due to: a) reduction of IP3R1 levels to maintain ER-mitochondria calcium homeostasis, thereby preventing excess nutrient-induced mitochondria calcium overload and dysfunction; b) decreased PDH phosphorylation, thereby improving mitochondrial respiration to increase nutrient utilization and LDL receptor expression to enhance postprandial lipid clearance; and c) increased BSEP levels and its presence in the bile canaliculus membrane, thereby improving enterohepatic bile acid circulation and the activation of intestinal bile acid- responsive gene expression to augment insulin sensitivity.

The ER is a multifunctional organelle that serves as the most important Ca+2 store in the cell. ER is able to accumulate Ca+2 at millimolar levels in both free and protein- buffered forms, whereas cytosolic Ca+2 levels are maintained at low levels (10-100nM)

[206, 207]. The ER calcium is essential for the regulation of protein posttranslational processing, folding, and export and can be consistently released to the cytosol and transported to the mitochondria to provide sustained and precise Ca+2 -mediated

+2 +2 +2 responses. In the liver, ER Ca is regulated by Ca release channels IP3Rs, Ca pump

59 SERCA2 and Ca+2 binding proteins [206-208]. The uptake of cytosolic Ca+2 into the ER occurs against its chemical gradient and requires energy released from ATP hydrolysis

[209]. This process is mediated by SERCA2. The Ca+2 stored in the ER is released to the

+2 cytosol through the IP3Rs channels to generate cytosolic signals. This released Ca leads to a depletion in ER luminal Ca+2. This depletion can be sensed by STIM1 and

STIM2 and respond by activating Ca+2 entry from extracellular milieu. STIMs are ER membrane proteins that possess two EF hand domains. Under basal conditions, Ca+2 is bound to these domains and the proteins remain in its monomeric form. When ER luminal

Ca+2 is depleted, Ca+2 dissociates from STIM protein, leading to oligomerization and trafficking of STIM through the ER membrane towards the plasma membrane contact sides. This localization allows STIM to couple with the plasma membrane Ca+2 channel-

Orai, to facilitate Ca+2 entry to the cells from the extracellular space. This process is called

Store-Operated Ca+2 Entry (SOCE). Upon the entrance of Ca+2 to the cytosol, SERCA2 pumps Ca+2 into ER to replenish the deficit [210, 211]. High levels of Ca+2 inside the ER are essential for protein folding and processing. These proteins include buffering protein such as calsequestrin, calreticulin and chaperones such as GFP78 (BiP, immunoglobulin binding protein), calnexin, GRP94 and GRP170. Another class of ER Ca+2 -regulatory proteins include ERp57, ERp72and ERO1L, which provide an electron transport pathway from thiol residue to molecular oxygen during disulfide bond formation [212]. Therefore,

ER Ca+2 depletion impairs the function of all of these proteins and if persistent, will lead to ER dysfunction and cell death.

Calcium level in the mitochondria also plays a vital role in the cellular homeostasis by regulating a wide range of processes from bioenergetics to cell death. The main force driving mitochondrial Ca+2 uptake is electrochemical gradient formed across

60 mitochondrial inner and outer membranes [213]. To get into the mitochondrial matrix, Ca+2 first crosses the outer mitochondrial membrane (OMM), which is permeable to Ca+2 through VDAC, which connects with IP3R1 in the ER through linkage with the chaperone

GRP75. This structure in the MAM allows the transfer of Ca+2 from ER to mitochondria

[116]. The inner mitochondrial membrane(IMM) is impermeable to ions, and the transport of Ca+2 occurs through the mitochondrial Ca+2 uniport (MCU), which has low affinity (Kd of 20-30uM) under physiological condition. Rapid Ca+2 uptake into mitochondria is counteracted by an extrusion mechanism mediated by the mitochondrial antiporter exchanging H+/ Ca+2 in the liver [214, 215]. Within mitochondria, low Ca+2 levels are required to increase the activity of matrix and citric acid cycle enzymes such as pyruvate dehydrogenase phosphatase (PDP), a-ketoglutarate dehydrogenase (KGDH), and isocitrate dehydrogenase (IDH). This culminates in increased NADH production that feeds the respiratory chain to produce ATP to meet the needs of the cell [216].

Conversely, excessive Ca+2 accumulation is associated with mitochondrial dysfunction due to increased mitochondrial ROS production and opening of the mitochondrial permeability transition pore (mPTP) that leads to collapse of membrane potential, inhibition of respiration and mitochondrial swelling with consequent loss of nucleotides and cytochrome C and is directly linked with apoptotic cell death [217, 218].

The importance of intracellular Ca+2 distribution among various organelles is well established in the cascade of events linking hormonal signaling to metabolic processes in the liver [219, 220]. Cytosolic Ca+2 in the liver is initiated by a hormonal or other agonist binding to G-protein-Coupled Receptor (GPCR). For example, glucagon (through glucagon receptor) and catecholamine (through b-adrenergic receptor) activates Gas protein, leading to stimulation of adenylate cyclase activity and the production of cyclic

61 AMP (cAMP). These signals activate protein kinase A (PKA), which is able to

2+ phosphorylate and activate the IP3R, leading to Ca release from the ER [221].

Catecholamine also signals through the a-adrenergic receptor to activate phospholipase

C (PLC), which catalyzes the hydrolysis of phosphatidylinositol 4,5 bisphosphate (PIP2) to produce the intracellular messengers IP3 and diacylglycerol (DAG) [222]. IP3 also

2+ binds to IP3R to stimulate Ca release from the ER. Opposing the effects of glucagon and catecholamine is insulin. Insulin influences cytosolic Ca2+ levels through the stimulation of AKT. It has been shown that AKT is able to directly phosphorylate IP3R, inhibiting its function, attenuating cytosolic Ca2+ signaling, and leading to increased ER

Ca2+ concentration [223]. Importantly, obesity is associated with increased phosphorylation status of IP3R (increase activity) as well as its protein expression are increased in the liver [224]. Higher cytosolic Ca2+ also stimulates the activity of calcineurin which ultimately mediates FoxO1 activation leading to increase gluconeogenic expression [224]. Another way that high cytosolic Ca2+ activate gluconeogenesis genes is by activating CaMKII, which phosphorylates and activates FoxO1, leading to FoxO1 translocation to the nucleus [225].

+2 The IP3Rs are only intracellular Ca release channels in hepatocytes [226]. There

+2 are three IP3Rs isoforms, with varying sensitivity to Ca , IP3 and distinct tissue and subcellular expression patterns [227]. In hepatocytes, the predominant isoforms are

IP3R1 and IP3R2. IP3R2 is mostly concentrated in the region of the ER along the apical

(canalicular) membrane and it is important for regulation of bile secretion [170, 226, 228].

In contrast, IP3R1 is diffusely distributed throughout the cytoplasm [226], but is also present in abundance in the MAM to catalyze Ca+2 release from the ER to mitochondria that is required for cellular bioenergetics. Interestingly, IP3R1 mRNA was increased in

62 NASH biopsy samples when compared with control liver biopsy in human. In addition, there is increased association between ER and mitochondria in human fatty liver and this association increases with worsening disease severity and is associated with elevated expression of IP3R1 in NASH [121]. Moreover, in animal model of obesity, the degree of

ER-mitochondrial connectivity and expression of IP3R are increased significantly in the

MAM isolated from the liver. This resulted in increased Ca+2 transfer from ER to mitochondria leading to enhanced oxidative stress and compromised oxidative phosphorylation [120]. Primary hepatocytes isolated from obese mice showed increased

Ca+2 transfer from ER to mitochondria through MAM connections leading to mitochondrial dysfunction [120]. Treatment of primary hepatocytes with palmitate decreased ER Ca+2 content levels, elevated mitochondrial Ca+2 content and elevated ROS. Remarkably, experimental treatment of these cells with a Ca+2 chelator abrogated the lipotoxic effect of palmitate [118]. This indicates that the excessive Ca+2 transfer to mitochondria is associated with a hepatic mitochondrial dysfunction and this dysfunction driven by Ca+2 overload also contributes to the development of hepatic steatosis [118]. On the other hand, total absence of ER Ca+2 transfer to the mitochondria also causes mitochondrial dysfunction due to impairment of Ca+2 -dependent activation of dehydrogenases

(pyruvate dehydrogenase, isocitrate and a-ketoglutarate dehydrogenase) that generate

NADH for ATP synthesis [204].

Our study showed that HAX-1 interacts with IP3R1 in hepatocytes as revealed by coimmunoprecipitation. To test whether this interaction was direct interaction, we will perform yeast two-hybrid screening test and GST pull-down assay in future experiments.

The inactivation of hepatic HAX-1 lowers but does not ablate IP3R1 expression, therapy preserving normal mitochondrial functions while preventing nutrient-induced

63 mitochondrial Ca+2 overload, reducing PDH phosphorylation and improving mitochondrial activity. These results are consistent with previous reports that silencing of IP3R1 (70% reduction in hepatic IP3R) improve glucose tolerance, insulin sensitivity and mitochondrial activity in obese mice [120]. Indeed, it was reported that basal Oxygen Consumption

Rates (OCR) were decreased by 60% in IP3R1 Knock-out (KO) cells compared with WT cells. In addition, maximal respiration was significantly reduced in KO cells. Surprisingly, the inhibited OCR was rescued by Ca+2 ionophore ionomycin, confirming that the availability of Ca+2 was specifically critical for normal oxidative phosphorylation [204]. In our study, we found that hepatic HAX-1 inactivation is associated with an increased expression of some of fatty acid oxidation genes as well as increased mitochondrial respiration in presence of high glucose or palmitate, resulting in decreased lipid accumulation in liver. Indeed, inhibition of ACC and activation of CPT1 protect mice against diet-induced insulin resistance [229]. Thus, our data support the notion that the

+2 constitutive IP3R1 Ca release is crucial for ongoing support of optimal mitochondrial function.

+2 Previous studies implicated that increased IP3R1 Ca release was associated with increased hepatic glucose production and insulin-resistant state [224, 225]. Our results indicated that lowering of hepatic IP3R1 levels by hepatic HAX-1 inactivation in mice fed chow or western diet enhanced insulin sensitivity. These findings indicate that hepatic steatosis is closely associated with insulin resistance and hepatic IP3R1 is involved in the hepatoprotective role of HAX-1 under chow or western diet feeding conditions. We did not find differences in pyruvate tolerance test, which is currently utilized to assess gluconeogenesis. This is may be due to the fact that the protection of hepatic steatosis

64 could be disconnected from gluconeogenic phenotypes under the conditions of our experiments.

The importance of PDH complex activity in mitochondrial respiration is well established in the literature. PDH catalyzes irreversible oxidative decarboxylation of pyruvate to acetyl CoA for oxidation by the citric acid cycle or conversion to fat. PDH complex is an assembly of four proteins: pyruvate dehydrogenase (E1), dihydrolipoamide acyltransferase (E2), dihydrolipoyl dehydrogenase (E3), and one structural protein

(E2/E3 binding protein). PDH component proteins are arranged as a core of 60 E2 subunits around which are distributed 30 copies of E1, 12 copies of E3, and 12 copies of the E2/E3 binding protein. PDH activity is increased after meals for disposal of excess glucose but reduced during fasting to spare three carbon compounds (pyruvate, alanine and lactate) to be converted back to glucose. Also, its activity is regulated by feedback inhibition by its products of acetyl CoA and NADH [230]. Moreover, PDH complex activity is inactivated by phosphorylation with PDH kinases (PDK) and can be activated by PDH phosphatase (PDPs). There are four PKD isoforms and their expression in various tissues differs depending on the physiological condition among various tissues [230, 231]. For example PDK1 has a limited tissue distribution, but has been detected in heart and pancreatic islet [231, 232] , and it is responsible for the Warburg effect in cancer cells.

PDK2 is expressed ubiquitously [231]. PDK3 is found in testis, kidney and brain [231,

233]. PDK4 is highly expressed in heart, skeletal muscle, liver and kidney [231, 234, 235].

Numerous studies have shown that PDK4 is slightly increased in liver of animals while

PDK2 is primarily increased in the livers of starved and diabetic animals and humans

[230, 236], indicating that PDK2 is the main PDK in liver to regulate the PDH complex activity. As mentioned earlier, the activation of PDH complex is mediated by

65 dephosphorylation by PDH phosphatases (PDPs). There are two PDH phosphatase isoforms, which are mitochondrial Ser/Threo phosphatases that contain regulatory and catalytic subunits. The PDP1 is a dominant isoform in heart, brain and testis and also found in skeletal muscle and liver, although its basal level is low compared with that found in heart [237]. The PDP2 is found in liver and abundantly expressed in liver and adipose tissue, which are highly lipogenic [233] and is also abundant in kidney and heart. It has been observed that only PDP2 is down-regulated by starvation and diabetic state in those tissues [237]. In concert with this, the expression of PDK4 is greatly increased in skeletal muscle and diaphragm but not livers and kidney of wild-type mice fed a high at diet [238].

Interestingly, PDK4 knock-out mice fed high fat diet were modestly more glucose tolerant than wild-type fed the same diet and developed insulin resistance. Moreover, inactivation of PDK4 leads to increased lipid accumulation in the liver [238]. This indicates that PDK4 is not the important PDK that regulates PDH activity in the liver. In contrast, whole body or hepatic deficiency of PDK2 has been shown to be protective against high fat diet- induced glucose tolerance and hepatic steatosis [239]. These and other similar studies led to suggestion that the PDK enzymes may be therapeutic targets for diabetes management [240]. This possibility was supported by recent evidence showing that a liver-specific pan-PDK inhibitor was effective in improving insulin signaling and lowering hepatic lipid accumulation in high fat diet induce obese mice [241]

Pyruvate plays a critical role in both anabolic and catabolic metabolism depending upon the tissue and physiological condition of organism as it represents the link between glycolysis and TCA cycle [242]. In insulin-resistant condition, pyruvate is used for gluconeogenesis as well as de novo fat synthesis instead of ATP generation in the TCA cycle, resulting in induction of hyperglycemia and hepatic steatosis. Indeed increased

66 availability of pyruvate due to decreased PDH activity contributes to the development of hepatic steatosis in high fat-diet fed mice [238, 239]. The beneficial effect of active PDH in improving insulin resistance and reducing hepatic steatosis is due to different mechanisms. For example: increased activity of PDH leads to large decrease in TCA cycle intermediates, which is caused by decreased availability of oxaloacetate due to reduced anaplerotic influx from pyruvate, resulting in reduction of ATP production. To compensate for reduced ATP production in TCA cycle, fatty acid consumption via b- oxidation must be greatly increased, leading to less fat accumulation [239]. This is consistent with our finding that activation of PDH due to HAX-1 inactivation in the liver leads to increased fatty acid oxidation when hepatocytes were treated with palmitate.

Moreover, ATP production in HAX-1 deficient cells is higher than that in wild type cells.

Additionally, Wu and his colleagues [241]have claimed that increased TCA cycle which is caused by increased PDH activity results in dissipating of intermediates more rapidly leading to enhanced fatty acid oxidation to replenish TCA cycle. Mechanistically, liver-

PDK2 inhibition leads to reduced ChREBP activity, which promotes insulin-independent lipogenesis in liver. This reduction in ChREBP activity associated with essentially decreased expression of lipogenesis enzymes ( ACC1,FAS and SCD1). In consonance with our result, stimulation of PDH by dichloroacetate enhances oxidative phosphorylation

(OXPHOS) [166]. To fuel OXPHOS, cells could rely on fatty acid oxidation, increasing lipid catabolism [243, 244], which result in reduced hepatic fatty acid accumulation.

Importantly, increased PDH activity leads to increased LDL receptor expression which may augment cholesterol and triglycerides transport to the cells to increase fat availability.

Taken together, these results indicate that PDH activation via IP3R1 reduction is one

67 mechanism by which L-Hax1-/- mice were protected against Western diet-induced hyperinsulinemia, hyperlipidemia, and hepatosteatosis.

While obesity and metabolic disease continue to rise, the most successful approach for metabolic disease intervention is bariatric surgery. It is effective not only in reducing body weight but also in improving insulin related metabolic complication such as

Type 2 diabetes and NAFLD [245]. The most common bariatric surgery procedures are

Roux-en Y gastric bypass (RYGB) and vertical sleeve gastrectomy (VSG). Bariatric surgery basically aims to obtain malabsorptive and/or restrictive gut anatomy modifications. Significant metabolic improvement (improved insulin sensitivity) after bariatric surgery occurs as early as a few days after surgery prior to weight loss, which occurs later in follow-up [246]. The metabolic improvement of bariatric surgery has been attributed to numerous mechanisms, including altering nutrient transport to bypass the proximal intestine with rapid delivery to distal intestine, changes in gut microbiota, and increasing bile salt transport in the enterohepatic circulation [247-251]. Bile acids are commonly suggested as a mediator of the early metabolic beneficial effects of bariatric surgery [252]. This suggestion is based on the consistent finding that circulating bile acids are increased in both fasting and postprandial conditions after RYGB and VSG bariatric surgery [249, 253]. It has been proven that obesity is associated with a decreased circulating conjugated bile acids in human [247]. Interestingly, in obese insulin-resistant subjects, the administration of tauroursodeoxycholic acid (TUDCA), which is increased after RYGB improve hepatic and peripheral insulin sensitivity [254, 255]. Additionally, oral administration of cholic acid increases bile acid pool size and serum bile acid levels, and in parallel prevents obesity , insulin resistance, glucose intolerance during high fat feeding

[256].

68 As mentioned previously, in addition to their role as a detergent that facilitates digestion and absorption of dietary lipids, bile acids can act as hormones that activate

FXR and TGR5 receptors. FXR is mainly expressed in the liver, gut, kidney and adrenal cortex, while TGR5 is highly expressed in nonparenchymal cells of liver, which include the gallbladder epithelial cells, cholangiocytes (epithelial cells of bile duct), sinusoidal epithelial, as well as immune cells [257]. However, TGR5 is also expressed in brown adipose tissue, intestinal enteroendocrine cells and the brain [258-260].

Bile acids decrease their own synthesis from cholesterol through a delicate feedback inhibitory circuit that involves both the liver and the intestine. The first pathway starts when bile acids activate FXR in the liver, which in turn will induce hepatic expression of the short heterodimer partner (SHP). SHP binds to and interferes with activity of liver receptor homologue (LRH1) and liver X receptor to repress the transcriptional activation of CYP7A1, the rate limiting enzyme in the neutral pathway of bile acid biosynthesis from cholesterol [261-263]. The second pathway involves the FXR- mediated induction of FGF15 in the intestine. FGF15 is secreted into portal circulation to activate hepatic FGF receptor 4 (FGFR4)/b-klotho complex resulting in an alternative pathway that suppressed CYP7A1 expression in SHP-independent pathway [264, 265].

More importantly, the acidic pathway appears to lack the striking response seen with

CYP7A to bile acid in the neutral pathway. Interestingly, the expression of StARD is not regulated by bile acid feedback inhibition [266]. Moreover, overexpression of CYP27A in the liver slightly elevated bile acid synthesis compared to StARD overexpression due to low cholesterol content in inner mitochondrial membrane [127]. In our study, we have demonstrated that hepatic deletion of HAX-1 significantly increased both gene expression and protein levels of CYP27A and StARD resulting in significantly increases both serum

69 and hepatic bile acid levels in mice, which was consistent with previous studies [127,

128]. Furthermore our study is in agreement with a report that overexpression of StARD in liver reduces hepatic lipid accumulation, attenuates insulin resistance and regulates glucose metabolism [267].

The increased enterohepatic circulation of bile acids has a beneficial role in lipid and glucose metabolism. The original characterization of FXR-/- mice suggested that FXR controls lipid metabolism, since these mice had hepatic lipid accumulation, increased plasma triglycerides, cholesterol, FFA and lipoproteins (VLDL and LDL) [268].

Conversely, treatment of diabetic, obese or wild-type mice with bile acids or specific FXR agonist lowered plasma triglycerides, FFA, cholesterol and decreased hepatic steatosis

[269, 270]. The decreased in plasma lipoproteins levels appears to be due to FXR dependent induction of genes involved in lipoprotein clearance from the plasma. These include SR-B1 (HDL receptor), VLDL receptor, syndecan-1 and ApoCII, a cofactor for lipoprotein lipase that hydrolyze plasma triglyceride [271, 272]. Moreover, FXR activation showed to induce LDL expression and repress PCSK9, an LDL inhibitor [273].

Additionally, activation of FXR by bile acid induces hepatic expression of ABCG5 and

ABCG8, which promoted biliary free cholesterol secretion and fecal cholesterol loss [173].

Induction of FXR leads to repression of hepatic SREBP-1c in a mechanism involving

SHP, resulting in inhibition of lipogenesis genes such as ACC1, ACC2, FAS and G6Pase and malic enzyme [269, 274]. As discussed above, bile acid activates intestinal FXR leading to induction of FGF15/19 expression in intestine, which bind to FGR4 in the liver to suppress bile acid synthesis. Interestingly, the metabolic rate is increased and weight gain attenuated in transgenic mice overexpressing FGF15/19. This was attributed to increased brown adipose tissue mass and enhanced liver b-oxidation due to decreased

70 ACC2 expression [275]. These mice were protected from diet-induced obesity and had lower glucose, insulin, cholesterol and triglyceride levels. In addition to reducing plasma and hepatic lipids, FGF15 represses gluconeogenic genes by inhibiting activity of the transcription factor CREB, a key regulator of PGC-1a and other gluconeogenic genes

[276]. Furthermore, FGF15/19 stimulates protein and glycogen synthesis, making its action similar to insulin except FGF15/19 does not stimulate lipogenesis, which is the key advantage in considering FGF15/19 pathway as anti-diabetic therapy [277].

The current study showed that normal plasma glucose, insulin, and lipid levels can be achieved in Western diet-fed mice by liver-specific inactivation of HAX-1. Hepatic steatosis was also significantly reduced in Western diet-fed L-Hax1-/- mice compared to control mice. In addition to reducing PDH phosphorylation and improvement of mitochondrial activity, HAX-1 inactivation in the liver also increased BSEP/ABCB11 levels and its localization to the bile canaliculus. This latter result is consistent with a previous study reporting that HAX-1 interacts with BSEP/ABCB11 and the absence of HAX-1 increases BSEP/ABCB11 localization to the apical membranes of liver and kidney cells

[159]. Although the exact mechanism underlying the increased BSEP/ABCB11 remains unknown, it is tempting to speculate that the reduced expression of IP3R1 redirects IP3 to

IP3R-2 that has been shown previously to promote BSEP/ABCB11 translocation and bile salt export activity [170]. IP3R-2 is most concentrated in the region of ER beneath the apical membrane and it is important for the initiation and speed of polarized Ca+2 waves,

+2 as well as for bile secretion [226, 278]. Therefore, Ca released from IP3R-2 promotes

BSEP activity by enhancing exocytic insertion, as with Mrp2 [279]. BSEP constitutively recycles between a subapical vesicles and the canalicular membrane in hepatocytes.

Importantly, vesical fusion to apical membrane depends on highly increased in localized

71 Ca+2 levels. Thus, increased ER Ca+2 due to hepatic HAX-1 deletion may lead to increase

BSEP levels in the canalicular membrane and then increased bile secretion in liver from hepatic HAX-1 deficiency. More importantly, elevated BSEP/ABCB11 has also been shown to enhance enterohepatic cycling of bile salts without significant changes in fecal bile salt excretion [172]. Hence, hepatic HAX-1 inactivation mimics bariatric surgery with respect to enhancing bile salt recycling through the enterohepatic circulation. Also similar to bariatric surgery [280], hepatic HAX-1 inactivation increases expression of FGF15 and other bile acid responsive genes in intestine such as SHP, OSTa and OSTb. It has been proven that increase enterohepatic circulation of bile acids leads also to increased cholesterol excretion in feces due to increased bile flow [281]. This is consistent with our finding that mice with hepatic HAX-1 deletion had increased bile acid re-circulation, resulting in higher fecal cholesterol excretion compared to wild-type.

In addition to activation of FXR in the intestine and liver, increased bile acid circulation, also leads to activation of TGR5 on the intestinal entero-endocrine L cells, leading to secretion of GLP-1 into circulation. GLP-1 is known to promote insulin secretion and thus regulates glucose homeostasis [259, 282, 283]. As mentioned previously, TGR5 is also expressed in brown and white adipose tissues and low levels have also been detected in skeletal muscle. Activation of TGR5 receptor in these tissues by bile acids leads to increased energy expenditure in brown adipose tissue, preventing obesity and insulin resistance [256]. The metabolic effect of bile acids that bind TGR5 is highly dependent on the induction of cAMP-dependent thyroid hormone-activating enzyme type

2 iodothyronine deiodinase (D2), as this effect is lost in D2-/- mice. Bile acid treatment of brown adipocyte and human skeletal myocytes increases D2 activity and oxygen consumption [132].

72 In the present study, the observation that less insulin is required to maintain euglycemia in chow-fed liver HAX-1 knockout mice and their protection against Western diet-induced hyperinsulinemia is consistent with the interpretation that metabolic benefit of hepatic HAX-1 inactivation is at least in part due to increased hepatic BSEP/ABCB11 levels that mediate FGF15 expression in the intestine. However, despite these benefits of elevated BSEP/ABCB11 levels, its overexpression alone may be compromised due to increasing cholesterol absorption [284]. Thus, a strategy to promote postprandial lipid clearance is necessary to complement the BSEP/ABCB11 increase for optimal metabolic benefits. In this regard, the present study shows that hepatic HAX-1 inactivation also increases LDL receptor expression thereby suggesting that targeting hepatic HAX-1 may hold therapeutic promise along these lines.

Another key player in obesity and its related diseases is the gut microbiota. The human gut microbiota consist of at least 1014 bacteria, consist of approximately 1,100 prevalent species, with approximately 160 such species per individual. The gut microbiota is estimated to contain 150 fold more genes than our own host genomes [285].

Importantly, studies have shown that obesity is associated with profound changes in composition and metabolic function of the gut microbiota [178, 286, 287]. Mammalian gut microbiota belongs predominantly to 4 bacterial phyla: the Gram-negative Bacteroidetes and Proteobacteria and the Gram positive Actinobacteria and Firmicutes. Animal studies have demonstrated that ob/ob mice exhibit a major reduction in the abundance of

Bacteroidetes and a proportional increase in Firmicutes [286]. Similarly, feeding of a high fat/high polysaccharide diet to wild-type rodents led to similar microbial changes [288].

Consistent with animal studies, Ley et al found that the ratio of Firmicutes/ Bacteroidetes was increased in the distal gut microbiota in human obesity [289]. In contrast, Zhang et

73 al showed that Prevotellaceae, a subgroup of Bacteroidetes, are significantly enriched in obesity [290]. This is raising the potentially important issue of diet as a confounding factor, as the patients in the Ley study [289] were either on a fat-restricted or carbohydrate restricted diet whereas in the Zhang study [290], researchers did not limit dietary components. Interestingly, overweight pregnant patients also have reduced numbers of

Bifidobacteria and Bacteroidetes, whereas increased numbers of certain Firmicutes such as Staphylococus and Escherichia Coli [291]. Additionally, another study also described a decrease of Bacteroidetes in obesity and an increase in Firmicutes such as Lactobacilli

[292]. More Interestingly, studies in germ free-mice revealed that the gut microbiota enhances adiposity mainly by increased energy extraction from food by regulating fat storage [178, 293], and germ free-mice are protected from obesity and metabolic syndrome [178, 286, 294]. Whereas conventionalization (the restoration of conventional intestinal flora of germ-mice resulted in an increase in fat hepatic triglycerides, fasting blood glucose, and insulin resistance. Thus ob/ob mice have been reported to harvest energy from food more efficiently than lean wild-type mice [294]. Remarkably, this trait of obesity was transmissible through fecal transplants from obese (as compared to non- obese) to germ-free mice [288, 294].

While bile acids can influence the microbiota in the gut, gut microbiota have also profound effects on bile acid metabolism since the conversion of primary bile acids into secondary bile acids relies on the presence of intestinal microbiota. The gut microbiota deconjugate the natural occurring FXR antagonist in mice TbMCA and thus promote FXR signaling in mice [295] and is also required for the production of secondary bile acids acting as a ligands for TGR5 [296]. A study comparing bile acid composition in wild-type mice and mice harboring a disrupted FXR genes, raised under germ-free condition or in

74 a conventional manner, suggested the gut microbiota regulates FXR signaling not only by acting on conversion of primary bile acids to secondary bile acids, but also by regulating bile acid synthesis [295]. By comparing germ-free and FXR-deficient mice, it was demonstrated that the gut microbiota regulates FGF15 levels in the ileum through a

FXR-dependent pathway. By comparing the expression of direct FXR target genes in the liver (SHP and LRH1) and ileum (SHP and FGF15), it appears that gut microbiota primarily affects FXR target in the ileum and not the liver. Tissue-specific knockouts have shown that CYP7A1, but not CYP8B1 is predominantly regulated by FXR -FGF15 pathway in mice [295]. Thus, the gut microbiota modulates FXR signaling in the gut by repressing the activity of CYP7A1 in the liver. In mice, this leads to reduced synthesis of

TbMCA, which is an FXR antagonist, thus promoting FXR-dependent induction of FGF15 expression in the ileum. A study by Li et al [176] showed that the antioxidant tempol targets intestinal Lactobacillus, a major source of bile salt hydrolase (Bsh) activity in the murine gut. By knockout Bsh activity, production of secondary bile acids declined.

Additionally, increase in FXR antagonist TbMCA expanded the bile acid pool size. Indeed, tempol feeding results in a near conversion of the Firmicutes/Bacteroidetes ratio with concomitant increase in TbMCA levels in the intestine similar to the result obtained in mice with selective deletion of intestinal FXR that results in reduced obesity [176].

The wild-type mice displayed increased Firmicutes/Bacteroidetes ratio when they fed a Western diet [288]. In the present study, we showed that HAX-1 inactivation in mice fed Western diet for 12 weeks are associated with low Firmicutes/Bacteroidetes ratio.

Growing body of evidence suggests that the various bariatric surgery could exert their beneficial effects by reshaping the intestinal microbiota [297, 298]. For example: RYGB surgery is associated with a shift in the composition of the gut microbiota from

75 predominant Firmicutes and Bacteroidetes toward one where gamma proteobacteria

[297, 299]. Whereas another study showed that a decrease in Firmicutes/Bacteroidetes ratio after RYGB [300]. Thus, this is another evidence that hepatic HAX-1 inactivation may have a metabolic effect similar to those result from bariatric surgery.

As previously stated, microbiota can reshape bile acids pool. Conversely, bile acids shape the composition of intestinal microbiota. Bile acids have both direct antimicrobial effects on gut microbiota [142], and indirect effect through FXR-induced antimicrobial peptides or regulation of host immune response by action on both FXR and

TGR5 [301, 302]. Feeding of rodents diet containing cholic acid leading to complex signaling changes in gut microbiota with Firmicutes expanded from 54% to between 93% and 98% of the microbiome, at the class level, the Clostridia expanded from 39% in controls to approximately 70%. This indicates that increased bile acid levels in the gut appear to favor Gram-positive members of Firmicutes including bacteria that cause the

7a-dihydroxylation of host primary bile acids into toxic bile acids. By contrast, decreased levels of bile acids in the gut favor Gram-negative members of the microbiome, some of which produce potent lipopolysaccharide and potential pathogens [303, 304]. This indicates that changes in bile acid composition may have different impacts on gut microbiota. Although our study did not provide evidence for direct interaction of hepatic

HAX-1 and bile acid metabolism, these findings suggest that the impact of hepatic HAX-

1 inactivation on ameliorating diet-induced metabolic diseases may be also linked to modulation of intestinal microbiota. In-depth analysis of intestinal microbiota and bile acid composition following hepatic HAX-1 inactivation warrants in the future studies.

Conclusion of dissertation In conclusion, this dissertation for the first time elucidates a new role of HAX-1 in regulating glucose and lipid metabolism. Specifically, HAX-1 inactivation reduces ER Ca2+

76 transfer to mitochondria during over-nutrition status, therapy enhancing mitochondrial pyruvate dehydrogenase activity to increase nutrient oxidation, ameliorating hepatic steatosis, and hyperlipidemia by increasing insulin sensitivity. This study may provide a better understanding of the role of hepatic HAX-1 in metabolism and therefore provide new strategies for the treatment of complications associated with metabolic diseases such as obesity and Type 2 diabetes. Moreover, modulation of bile acid pool and enhancement of intestinal FGF15 levels by hepatic HAX-1 inactivation improve hepatic metabolism during diet-induced fatty liver (Fig. 49). Therefore, this dissertation underscores the potential role of HAX-1 to treat obesity, insulin resistance and Type 2 diabetes.

REFERENCE LIST

1. Yki-Jarvinen, H., Liver fat in the pathogenesis of insulin resistance and type 2 diabetes. Dig Dis, 2010. 28(1): p. 203-9. 2. Centers for Disease Control and Prevention. National diabetes fact sheet: national estimates and general information on diabetes and prediabetes in the United States,2011. Atlanta, GA: U.S. Department of Health and Human Services. 2011. 3. Tolman, K.G., et al., Spectrum of liver disease in type 2 diabetes and management of patients with diabetes and liver disease. Diabetes Care, 2007. 30(3): p. 734-43. 4. Bril, F. and K. Cusi, Management of Nonalcoholic Fatty Liver Disease in Patients With Type 2 Diabetes: A Call to Action. Diabetes Care, 2017. 40(3): p. 419-430. 5. Anstee, Q.M., G. Targher, and C.P. Day, Progression of NAFLD to diabetes mellitus, cardiovascular disease or cirrhosis. Nat Rev Gastroenterol Hepatol, 2013. 10(6): p. 330- 44. 6. Byrne, C.D. and G. Targher, NAFLD: a multisystem disease. J Hepatol, 2015. 62(1 Suppl): p. S47-64. 7. Marchesini, G., et al., Nonalcoholic fatty liver disease and the metabolic syndrome. Curr Opin Lipidol, 2005. 16(4): p. 421-7. 8. McCullough, A.J., Pathophysiology of nonalcoholic steatohepatitis. J Clin Gastroenterol, 2006. 40 Suppl 1: p. S17-29. 9. James, P. and R. McFadden, Understanding the processes behind the regulation of blood glucose. Nurs Times, 2004. 100(16): p. 56-8. 10. Saltiel, A.R. and C.R. Kahn, Insulin signalling and the regulation of glucose and lipid metabolism. Nature, 2001. 414(6865): p. 799-806. 11. Szablewski, L., Glucose Homeostasis and Insulin Resistance. 2011.

77 12. Abel, E.D., et al., Adipose-selective targeting of the GLUT4 gene impairs insulin action in muscle and liver. Nature, 2001. 409(6821): p. 729-33. 13. Bruning, J.C., et al., A muscle-specific insulin receptor knockout exhibits features of the metabolic syndrome of NIDDM without altering glucose tolerance. Mol Cell, 1998. 2(5): p. 559-69. 14. PE, C., Glucose homeostasis and hypoglyceia in William’s Textbook of Endocrinology., F.D. Wilson JD, Editor. 1992, Saunders: Philadelphia, Pa., W.B. p. 1223-1253. 15. Cooper, A.D., Hepatic uptake of chylomicron remnants. J Lipid Res, 1997. 38(11): p. 2173-92. 16. Redgrave, T.G., Formation of cholesteryl ester-rich particulate lipid during metabolism of chylomicrons. J Clin Invest, 1970. 49(3): p. 465-71. 17. Dallinga-Thie, G.M., et al., The metabolism of triglyceride-rich lipoproteins revisited: new players, new insight. Atherosclerosis, 2010. 211(1): p. 1-8. 18. Rosenson, R.S., et al., Cholesterol efflux and atheroprotection: advancing the concept of reverse cholesterol transport. Circulation, 2012. 125(15): p. 1905-19. 19. Thorens, B., GLUT2, glucose sensing and glucose homeostasis. Diabetologia, 2015. 58(2): p. 221-32. 20. Cardenas, M.L., A. Cornish-Bowden, and T. Ureta, Evolution and regulatory role of the hexokinases. Biochim Biophys Acta, 1998. 1401(3): p. 242-64. 21. Zhang, T., et al., Acetylation negatively regulates glycogen phosphorylase by recruiting protein phosphatase 1. Cell Metab, 2012. 15(1): p. 75-87. 22. Agius, L., Glucokinase and molecular aspects of liver glycogen metabolism. Biochem J, 2008. 414(1): p. 1-18. 23. Rui, L., Energy metabolism in the liver. Compr Physiol, 2014. 4(1): p. 177-97. 24. Lee, M.W., et al., Regulation of hepatic gluconeogenesis by an ER-bound transcription factor, CREBH. Cell Metab, 2010. 11(4): p. 331-9. 25. Kir, S., et al., FGF19 as a postprandial, insulin-independent activator of hepatic protein and glycogen synthesis. Science, 2011. 331(6024): p. 1621-4. 26. Silverman, J.F., W.J. Pories, and J.F. Caro, Liver pathology in diabetes mellitus and morbid obesity. Clinical, pathological, and biochemical considerations. Pathol Annu, 1989. 24 Pt 1: p. 275-302. 27. Stone, B.G. and D.H. Van Thiel, Diabetes mellitus and the liver. Semin Liver Dis, 1985. 5(1): p. 8-28. 28. Demir, M., S. Lang, and H.M. Steffen, Nonalcoholic fatty liver disease - current status and future directions. J Dig Dis, 2015. 16(10): p. 541-57. 29. Masarone, M., et al., Non alcoholic fatty liver: epidemiology and natural history. Rev Recent Clin Trials, 2014. 9(3): p. 126-33. 30. Ramnanan, C.J., et al., Physiologic action of glucagon on liver glucose metabolism. Diabetes Obes Metab, 2011. 13 Suppl 1: p. 118-25. 31. Lin, H.V. and D. Accili, Hormonal regulation of hepatic glucose production in health and disease. Cell Metab, 2011. 14(1): p. 9-19. 32. Rizza, R.A., Pathogenesis of fasting and postprandial hyperglycemia in type 2 diabetes: implications for therapy. Diabetes, 2010. 59(11): p. 2697-707. 33. Donnelly, K.L., et al., Sources of fatty acids stored in liver and secreted via lipoproteins in patients with nonalcoholic fatty liver disease. J Clin Invest, 2005. 115(5): p. 1343-51.

78 34. Schwarz, J.M., et al., Hepatic de novo lipogenesis in normoinsulinemic and hyperinsulinemic subjects consuming high-fat, low-carbohydrate and low-fat, high- carbohydrate isoenergetic diets. Am J Clin Nutr, 2003. 77(1): p. 43-50. 35. Kotronen, A., et al., Liver fat in the metabolic syndrome. J Clin Endocrinol Metab, 2007. 92(9): p. 3490-7. 36. Leavens, K.F. and M.J. Birnbaum, Insulin signaling to hepatic lipid metabolism in health and disease. Crit Rev Biochem Mol Biol, 2011. 46(3): p. 200-15. 37. Fabbrini, E., S. Sullivan, and S. Klein, Obesity and nonalcoholic fatty liver disease: biochemical, metabolic, and clinical implications. Hepatology, 2010. 51(2): p. 679-89. 38. Nielsen, S., et al., Splanchnic lipolysis in human obesity. J Clin Invest, 2004. 113(11): p. 1582-8. 39. Weiss, R., et al., The "obese insulin-sensitive" adolescent: importance of adiponectin and lipid partitioning. J Clin Endocrinol Metab, 2005. 90(6): p. 3731-7. 40. Pardina, E., et al., Increased expression and activity of hepatic lipase in the liver of morbidly obese adult patients in relation to lipid content. Obes Surg, 2009. 19(7): p. 894- 904. 41. Westerbacka, J., et al., Genes involved in fatty acid partitioning and binding, lipolysis, monocyte/macrophage recruitment, and inflammation are overexpressed in the human fatty liver of insulin-resistant subjects. Diabetes, 2007. 56(11): p. 2759-65. 42. Ravikumar, B., et al., Real-time assessment of postprandial fat storage in liver and skeletal muscle in health and type 2 diabetes. Am J Physiol Endocrinol Metab, 2005. 288(4): p. E789-97. 43. Fabbrini, E., et al., Intrahepatic fat, not visceral fat, is linked with metabolic complications of obesity. Proc Natl Acad Sci U S A, 2009. 106(36): p. 15430-5. 44. Greco, D., et al., Gene expression in human NAFLD. Am J Physiol Gastrointest Liver Physiol, 2008. 294(5): p. G1281-7. 45. Magkos, F., et al., Intrahepatic diacylglycerol content is associated with hepatic insulin resistance in obese subjects. Gastroenterology, 2012. 142(7): p. 1444-6.e2. 46. Samuel, V.T., et al., Mechanism of hepatic insulin resistance in non-alcoholic fatty liver disease. J Biol Chem, 2004. 279(31): p. 32345-53. 47. Dries, D.R., L.L. Gallegos, and A.C. Newton, A single residue in the C1 domain sensitizes novel protein kinase C isoforms to cellular diacylglycerol production. J Biol Chem, 2007. 282(2): p. 826-30. 48. Samuel, V.T., et al., Inhibition of protein kinase Cepsilon prevents hepatic insulin resistance in nonalcoholic fatty liver disease. J Clin Invest, 2007. 117(3): p. 739-45. 49. Arias, I.M., et al., The liver: biology and pathobiology. 2011: John Wiley & Sons. 50. Iqbal, J. and M.M. Hussain, Intestinal lipid absorption. Am J Physiol Endocrinol Metab, 2009. 296(6): p. E1183-94. 51. Shachter, N.S., Apolipoproteins C-I and C-III as important modulators of lipoprotein metabolism. Curr Opin Lipidol, 2001. 12(3): p. 297-304. 52. Vance, J.E. and D.E. Vance, Biochemistry of lipids, lipoproteins and membranes. 2008: Elsevier. 53. Mahley, R.W. and Z.S. Ji, Remnant lipoprotein metabolism: key pathways involving cell- surface heparan sulfate proteoglycans and apolipoprotein E. J Lipid Res, 1999. 40(1): p. 1-16.

79 54. Young, S.G. and R. Zechner, Biochemistry and pathophysiology of intravascular and intracellular lipolysis. Genes Dev, 2013. 27(5): p. 459-84. 55. Lewis, G.F., et al., Hepatic lipase mRNA, protein, and plasma enzyme activity is increased in the insulin-resistant, fructose-fed Syrian golden hamster and is partially normalized by the insulin sensitizer rosiglitazone. Diabetes, 2004. 53(11): p. 2893-900. 56. Brown, M.S. and J.L. Goldstein, A receptor-mediated pathway for cholesterol homeostasis. Science, 1986. 232(4746): p. 34-47. 57. Goldstein, J.L. and M.S. Brown, The LDL receptor. Arterioscler Thromb Vasc Biol, 2009. 29(4): p. 431-8. 58. Kwon, H.J., et al., Structure of N-terminal domain of NPC1 reveals distinct subdomains for binding and transfer of cholesterol. Cell, 2009. 137(7): p. 1213-24. 59. Diraison, F., P. Moulin, and M. Beylot, Contribution of hepatic de novo lipogenesis and reesterification of plasma non esterified fatty acids to plasma triglyceride synthesis during non-alcoholic fatty liver disease. Diabetes Metab, 2003. 29(5): p. 478-85. 60. Flannery, C., et al., Skeletal muscle insulin resistance promotes increased hepatic de novo lipogenesis, hyperlipidemia, and hepatic steatosis in the elderly. Diabetes, 2012. 61(11): p. 2711-7. 61. Petersen, K.F., et al., Reversal of muscle insulin resistance by weight reduction in young, lean, insulin-resistant offspring of parents with type 2 diabetes. Proc Natl Acad Sci U S A, 2012. 109(21): p. 8236-40. 62. Petersen, K.F., et al., The role of skeletal muscle insulin resistance in the pathogenesis of the metabolic syndrome. Proc Natl Acad Sci U S A, 2007. 104(31): p. 12587-94. 63. Musso, G., R. Gambino, and M. Cassader, Recent insights into hepatic lipid metabolism in non-alcoholic fatty liver disease (NAFLD). Prog Lipid Res, 2009. 48(1): p. 1-26. 64. Shimomura, I., et al., Insulin selectively increases SREBP-1c mRNA in the livers of rats with streptozotocin-induced diabetes. Proc Natl Acad Sci U S A, 1999. 96(24): p. 13656- 61. 65. Yamashita, H., et al., A glucose-responsive transcription factor that regulates carbohydrate metabolism in the liver. Proc Natl Acad Sci U S A, 2001. 98(16): p. 9116-21. 66. Shimano, H., et al., Isoform 1c of sterol regulatory element binding protein is less active than isoform 1a in livers of transgenic mice and in cultured cells. J Clin Invest, 1997. 99(5): p. 846-54. 67. Shimomura, I., Y. Bashmakov, and J.D. Horton, Increased levels of nuclear SREBP-1c associated with fatty livers in two mouse models of diabetes mellitus. J Biol Chem, 1999. 274(42): p. 30028-32. 68. Uyeda, K. and J.J. Repa, Carbohydrate response element binding protein, ChREBP, a transcription factor coupling hepatic glucose utilization and lipid synthesis. Cell Metab, 2006. 4(2): p. 107-10. 69. Klein, S. and K. Jeejeebhoy, The malnourished patient: nutritional assessment and management in Sleisenger and Fordtran’s Gastrointestinal and Liver Disease M. Feldman, L. Freidman, and M. Sleisenger, Editors. 2002, W.B. Saunders Co: Philadelphia p. 265-285. 70. Muller, M.J., Hepatic fuel selection. Proc Nutr Soc, 1995. 54(1): p. 139-50. 71. Desvergne, B. and W. Wahli, Peroxisome proliferator-activated receptors: nuclear control of metabolism. Endocr Rev, 1999. 20(5): p. 649-88.

80 72. McGarry, J.D. and D.W. Foster, Regulation of hepatic fatty acid oxidation and ketone body production. Annu Rev Biochem, 1980. 49: p. 395-420. 73. Neschen, S., et al., n-3 Fatty acids preserve insulin sensitivity in vivo in a peroxisome proliferator-activated receptor-alpha-dependent manner. Diabetes, 2007. 56(4): p. 1034- 41. 74. Wakil, S.J. and L.A. Abu-Elheiga, Fatty acid metabolism: target for metabolic syndrome. J Lipid Res, 2009. 50 Suppl: p. S138-43. 75. Birkenfeld, A.L., et al., Deletion of the mammalian INDY homolog mimics aspects of dietary restriction and protects against adiposity and insulin resistance in mice. Cell Metab, 2011. 14(2): p. 184-95. 76. Fullerton, M.D., et al., Single phosphorylation sites in Acc1 and Acc2 regulate lipid homeostasis and the insulin-sensitizing effects of metformin. Nat Med, 2013. 19(12): p. 1649-54. 77. Schmid, A.I., et al., Liver ATP synthesis is lower and relates to insulin sensitivity in patients with type 2 diabetes. Diabetes Care, 2011. 34(2): p. 448-53. 78. Szendroedi, J., et al., Abnormal hepatic energy homeostasis in type 2 diabetes. Hepatology, 2009. 50(4): p. 1079-86. 79. Fisher, E.A., et al., The degradation of apolipoprotein B100 is mediated by the ubiquitin- proteasome pathway and involves heat shock protein 70. J Biol Chem, 1997. 272(33): p. 20427-34. 80. Zhou, M., E.A. Fisher, and H.N. Ginsberg, Regulated Co-translational ubiquitination of apolipoprotein B100. A new paradigm for proteasomal degradation of a secretory protein. J Biol Chem, 1998. 273(38): p. 24649-53. 81. Fabbrini, E., et al., Alterations in adipose tissue and hepatic lipid kinetics in obese men and women with nonalcoholic fatty liver disease. Gastroenterology, 2008. 134(2): p. 424-31. 82. Kamagate, A. and H.H. Dong, FoxO1 integrates insulin signaling to VLDL production. Cell Cycle, 2008. 7(20): p. 3162-70. 83. Lin, M.C., D. Gordon, and J.R. Wetterau, Microsomal triglyceride transfer protein (MTP) regulation in HepG2 cells: insulin negatively regulates MTP gene expression. J Lipid Res, 1995. 36(5): p. 1073-81. 84. Dashti, N., D.L. Williams, and P. Alaupovic, Effects of oleate and insulin on the production rates and cellular mRNA concentrations of apolipoproteins in HepG2 cells. J Lipid Res, 1989. 30(9): p. 1365-73. 85. Phung, T.L., et al., Phosphoinositide 3-kinase activity is necessary for insulin-dependent inhibition of apolipoprotein B secretion by rat hepatocytes and localizes to the endoplasmic reticulum. J Biol Chem, 1997. 272(49): p. 30693-702. 86. Lewis, G.F., et al., Effects of acute hyperinsulinemia on VLDL triglyceride and VLDL apoB production in normal weight and obese individuals. Diabetes, 1993. 42(6): p. 833-42. 87. Malmstrom, R., et al., Effects of insulin and acipimox on VLDL1 and VLDL2 apolipoprotein B production in normal subjects. Diabetes, 1998. 47(5): p. 779-87. 88. Malmstrom, R., et al., Defective regulation of triglyceride metabolism by insulin in the liver in NIDDM. Diabetologia, 1997. 40(4): p. 454-62. 89. Park, S.W., et al., Sarco(endo)plasmic reticulum Ca2+-ATPase 2b is a major regulator of endoplasmic reticulum stress and glucose homeostasis in obesity. Proc Natl Acad Sci U S A, 2010. 107(45): p. 19320-5.

81 90. Walter, P. and D. Ron, The unfolded protein response: from stress pathway to homeostatic regulation. Science, 2011. 334(6059): p. 1081-6. 91. Colgan, S.M., A.A. Hashimi, and R.C. Austin, Endoplasmic reticulum stress and lipid dysregulation. Expert Rev Mol Med, 2011. 13: p. e4. 92. Fu, S., S.M. Watkins, and G.S. Hotamisligil, The role of endoplasmic reticulum in hepatic lipid homeostasis and stress signaling. Cell Metab, 2012. 15(5): p. 623-34. 93. Wang, S., et al., IRE1alpha-XBP1s induces PDI expression to increase MTP activity for hepatic VLDL assembly and lipid homeostasis. Cell Metab, 2012. 16(4): p. 473-86. 94. Zhang, K., et al., The unfolded protein response transducer IRE1alpha prevents ER stress- induced hepatic steatosis. Embo j, 2011. 30(7): p. 1357-75. 95. Krahmer, N., et al., Phosphatidylcholine synthesis for lipid droplet expansion is mediated by localized activation of CTP:phosphocholine cytidylyltransferase. Cell Metab, 2011. 14(4): p. 504-15. 96. Fu, S., et al., Aberrant lipid metabolism disrupts calcium homeostasis causing liver endoplasmic reticulum stress in obesity. Nature, 2011. 473(7348): p. 528-31. 97. Malhotra, J.D. and R.J. Kaufman, The endoplasmic reticulum and the unfolded protein response. Semin Cell Dev Biol, 2007. 18(6): p. 716-31. 98. Hotamisligil, G.S., Endoplasmic reticulum stress and the inflammatory basis of metabolic disease. Cell, 2010. 140(6): p. 900-17. 99. Ozcan, U., et al., Endoplasmic reticulum stress links obesity, insulin action, and type 2 diabetes. Science, 2004. 306(5695): p. 457-61. 100. Zhou, Y., et al., Regulation of glucose homeostasis through a XBP-1-FoxO1 interaction. Nat Med, 2011. 17(3): p. 356-65. 101. Lee, A.H., et al., Regulation of hepatic lipogenesis by the transcription factor XBP1. Science, 2008. 320(5882): p. 1492-6. 102. Li, H., et al., AMPK activation prevents excess nutrient-induced hepatic lipid accumulation by inhibiting mTORC1 signaling and endoplasmic reticulum stress response. Biochim Biophys Acta, 2014. 1842(9): p. 1844-54. 103. Jo, H., et al., Endoplasmic reticulum stress induces hepatic steatosis via increased expression of the hepatic very low-density lipoprotein receptor. Hepatology, 2013. 57(4): p. 1366-77. 104. Patterson, R.E., et al., Lipotoxicity in steatohepatitis occurs despite an increase in tricarboxylic acid cycle activity. Am J Physiol Endocrinol Metab, 2016. 310(7): p. E484-94. 105. Cusi, K., Treatment of patients with type 2 diabetes and non-alcoholic fatty liver disease: current approaches and future directions. Diabetologia, 2016. 59(6): p. 1112-20. 106. Koliaki, C., et al., Adaptation of hepatic mitochondrial function in humans with nonalcoholic fatty liver is lost in steatohepatitis. Cell Metab, 2015. 21(5): p. 739-46. 107. Satapati, S., et al., Elevated TCA cycle function in the pathology of diet-induced hepatic insulin resistance and fatty liver. J Lipid Res, 2012. 53(6): p. 1080-92. 108. Iozzo, P., et al., Fatty acid metabolism in the liver, measured by positron emission tomography, is increased in obese individuals. Gastroenterology, 2010. 139(3): p. 846- 56, 856.e1-6. 109. Miele, L., et al., Hepatic mitochondrial beta-oxidation in patients with nonalcoholic steatohepatitis assessed by 13C-octanoate breath test. Am J Gastroenterol, 2003. 98(10): p. 2335-6.

82 110. Sunny, N.E., et al., Excessive hepatic mitochondrial TCA cycle and gluconeogenesis in humans with nonalcoholic fatty liver disease. Cell Metab, 2011. 14(6): p. 804-10. 111. Cotter, D.G., et al., Ketogenesis prevents diet-induced fatty liver injury and hyperglycemia. J Clin Invest, 2014. 124(12): p. 5175-90. 112. Sunny, N.E., et al., Progressive adaptation of hepatic ketogenesis in mice fed a high-fat diet. Am J Physiol Endocrinol Metab, 2010. 298(6): p. E1226-35. 113. Rizzuto, R., et al., Close contacts with the endoplasmic reticulum as determinants of mitochondrial Ca2+ responses. Science, 1998. 280(5370): p. 1763-6. 114. Voelker, D.R., Bridging gaps in phospholipid transport. Trends Biochem Sci, 2005. 30(7): p. 396-404. 115. Rieusset, J., Contribution of mitochondria and endoplasmic reticulum dysfunction in insulin resistance: Distinct or interrelated roles? Diabetes Metab, 2015. 41(5): p. 358-68. 116. Szabadkai, G., et al., Chaperone-mediated coupling of endoplasmic reticulum and mitochondrial Ca2+ channels. J Cell Biol, 2006. 175(6): p. 901-11. 117. Rieusset, J., et al., Disruption of calcium transfer from ER to mitochondria links alterations of mitochondria-associated ER membrane integrity to hepatic insulin resistance. Diabetologia, 2016. 59(3): p. 614-23. 118. Egnatchik, R.A., et al., ER calcium release promotes mitochondrial dysfunction and hepatic cell lipotoxicity in response to palmitate overload. Mol Metab, 2014. 3(5): p. 544- 53. 119. Lim, J.H., et al., Coupling mitochondrial dysfunction to endoplasmic reticulum stress response: a molecular mechanism leading to hepatic insulin resistance. Cell Signal, 2009. 21(1): p. 169-77. 120. Arruda, A.P., et al., Chronic enrichment of hepatic endoplasmic reticulum-mitochondria contact leads to mitochondrial dysfunction in obesity. Nat Med, 2014. 20(12): p. 1427- 35. 121. Feriod, C.N., et al., Hepatic Inositol 1,4,5 Trisphosphate Receptor Type 1 Mediates Fatty Liver. Hepatol Commun, 2017. 1(1): p. 23-35. 122. Sebastian, D., et al., Mitofusin 2 (Mfn2) links mitochondrial and endoplasmic reticulum function with insulin signaling and is essential for normal glucose homeostasis. Proc Natl Acad Sci U S A, 2012. 109(14): p. 5523-8. 123. de Aguiar Vallim, T.Q., E.J. Tarling, and P.A. Edwards, Pleiotropic roles of bile acids in metabolism. Cell Metab, 2013. 17(5): p. 657-69. 124. Russell, D.W., The enzymes, regulation, and genetics of bile acid synthesis. Annu Rev Biochem, 2003. 72: p. 137-74. 125. Schwarz, M., et al., Disruption of cholesterol 7alpha-hydroxylase gene in mice. II. Bile acid deficiency is overcome by induction of oxysterol 7alpha-hydroxylase. J Biol Chem, 1996. 271(30): p. 18024-31. 126. Schwarz, M., et al., Alternate pathways of bile acid synthesis in the cholesterol 7alpha- hydroxylase knockout mouse are not upregulated by either cholesterol or cholestyramine feeding. J Lipid Res, 2001. 42(10): p. 1594-603. 127. Pandak, W.M., et al., Transport of cholesterol into mitochondria is rate-limiting for bile acid synthesis via the alternative pathway in primary rat hepatocytes. J Biol Chem, 2002. 277(50): p. 48158-64. 128. Ren, S., et al., Overexpression of cholesterol transporter StAR increases in vivo rates of bile acid synthesis in the rat and mouse. Hepatology, 2004. 40(4): p. 910-7.

83 129. Arnon, R., et al., Cholesterol 7-hydroxylase knockout mouse: a model for monohydroxy bile acid-related neonatal cholestasis. Gastroenterology, 1998. 115(5): p. 1223-8. 130. Javitt, N.B., 25R,26-Hydroxycholesterol revisited: synthesis, metabolism, and biologic roles. J Lipid Res, 2002. 43(5): p. 665-70. 131. Botham, K.M. and G.S. Boyd, The metabolism of chenodeoxycholic acid to beta- muricholic acid in rat liver. Eur J Biochem, 1983. 134(1): p. 191-6. 132. Thomas, C., et al., Targeting bile-acid signalling for metabolic diseases. Nat Rev Drug Discov, 2008. 7(8): p. 678-93. 133. Dawson, P.A., M.L. Hubbert, and A. Rao, Getting the mOST from OST: Role of organic solute transporter, OSTalpha-OSTbeta, in bile acid and steroid metabolism. Biochim Biophys Acta, 2010. 1801(9): p. 994-1004. 134. Sung, J.Y., J.W. Costerton, and E.A. Shaffer, Defense system in the biliary tract against bacterial infection. Dig Dis Sci, 1992. 37(5): p. 689-96. 135. Chandra, R. and R.A. Liddle, Cholecystokinin. Curr Opin Endocrinol Diabetes Obes, 2007. 14(1): p. 63-7. 136. Chiang, J.Y., Bile acids: regulation of synthesis. J Lipid Res, 2009. 50(10): p. 1955-66. 137. Di Ciaula, A., et al., Bile Acid Physiology. Ann Hepatol, 2017. 16(Suppl. 1: s3-105.): p. s4- s14. 138. Angelin, B., et al., Hepatic uptake of bile acids in man. Fasting and postprandial concentrations of individual bile acids in portal venous and systemic blood serum. J Clin Invest, 1982. 70(4): p. 724-31. 139. Zhang, Y.K., G.L. Guo, and C.D. Klaassen, Diurnal variations of mouse plasma and hepatic bile acid concentrations as well as expression of biosynthetic enzymes and transporters. PLoS One, 2011. 6(2): p. e16683. 140. Molinaro, A., A. Wahlstrom, and H.U. Marschall, Role of Bile Acids in Metabolic Control. Trends Endocrinol Metab, 2018. 29(1): p. 31-41. 141. Ridlon, J.M., D.J. Kang, and P.B. Hylemon, Bile salt biotransformations by human intestinal bacteria. J Lipid Res, 2006. 47(2): p. 241-59. 142. Inagaki, T., et al., Regulation of antibacterial defense in the small intestine by the nuclear bile acid receptor. Proc Natl Acad Sci U S A, 2006. 103(10): p. 3920-5. 143. Suzuki, Y., et al., HAX-1, a novel intracellular protein, localized on mitochondria, directly associates with HS1, a substrate of Src family tyrosine kinases. J Immunol, 1997. 158(6): p. 2736-44. 144. Lees, D.M., I.R. Hart, and J.F. Marshall, Existence of multiple isoforms of HS1-associated protein X-1 in murine and human tissues. J Mol Biol, 2008. 379(4): p. 645-55. 145. Klein, C., et al., HAX1 deficiency causes autosomal recessive severe congenital neutropenia (Kostmann disease). Nat Genet, 2007. 39(1): p. 86-92. 146. Matsubara, K., et al., Severe developmental delay and epilepsy in a Japanese patient with severe congenital neutropenia due to HAX1 deficiency. Haematologica, 2007. 92(12): p. e123-5. 147. Chao, J.R., et al., Hax1-mediated processing of HtrA2 by Parl allows survival of lymphocytes and neurons. Nature, 2008. 452(7183): p. 98-102. 148. Jiang, Y., et al., Gene expression profiling in a renal cell carcinoma cell line: dissecting VHL and hypoxia-dependent pathways. Mol Cancer Res, 2003. 1(6): p. 453-62. 149. Velculescu, V.E., et al., Serial analysis of gene expression. Science, 1995. 270(5235): p. 484-7.

84 150. Yamaguchi, H., J. Wyckoff, and J. Condeelis, Cell migration in tumors. Curr Opin Cell Biol, 2005. 17(5): p. 559-64. 151. Ramsay, A.G., et al., HS1-associated protein X-1 regulates carcinoma cell migration and invasion via clathrin-mediated endocytosis of integrin alphavbeta6. Cancer Res, 2007. 67(11): p. 5275-84. 152. Vafiadaki, E., et al., Phospholamban interacts with HAX-1, a mitochondrial protein with anti-apoptotic function. J Mol Biol, 2007. 367(1): p. 65-79. 153. Zhao, W., et al., The anti-apoptotic protein HAX-1 is a regulator of cardiac function. Proc Natl Acad Sci U S A, 2009. 106(49): p. 20776-81. 154. Lam, C.K., et al., HAX-1 regulates cyclophilin-D levels and mitochondria permeability transition pore in the heart. Proc Natl Acad Sci U S A, 2015. 112(47): p. E6466-75. 155. Bidwell, P.A., et al., HAX-1 regulates SERCA2a oxidation and degradation. J Mol Cell Cardiol, 2018. 114: p. 220-233. 156. Janssen, K., et al., Inhibition of the ER Ca2+ pump forces multidrug-resistant cells deficient in Bak and Bax into necrosis. J Cell Sci, 2009. 122(Pt 24): p. 4481-91. 157. Vafiadaki, E., et al., The anti-apoptotic protein HAX-1 interacts with SERCA2 and regulates its protein levels to promote cell survival. Mol Biol Cell, 2009. 20(1): p. 306-18. 158. Kong, J., et al., Enhancement of interaction of BSEP and HAX-1 on the canalicular membrane of hepatocytes in a mouse model of cholesterol cholelithiasis. Int J Clin Exp Pathol, 2014. 7(4): p. 1644-50. 159. Ortiz, D.F., et al., Identification of HAX-1 as a protein that binds bile salt export protein and regulates its abundance in the apical membrane of Madin-Darby canine kidney cells. J Biol Chem, 2004. 279(31): p. 32761-70. 160. Samuel, V.T. and G.I. Shulman, Mechanisms for insulin resistance: common threads and missing links. Cell, 2012. 148(5): p. 852-71. 161. Hofmann, S.M., et al., Defective lipid delivery modulates glucose tolerance and metabolic response to diet in apolipoprotein E-deficient mice. Diabetes, 2008. 57(1): p. 5- 12. 162. Hamlin, A.N., et al., LRP1 Protein Deficiency Exacerbates Palmitate-induced Steatosis and Toxicity in Hepatocytes. J Biol Chem, 2016. 291(32): p. 16610-9. 163. Hofmann, S.M., et al., Adipocyte LDL receptor-related protein-1 expression modulates postprandial lipid transport and glucose homeostasis in mice. J Clin Invest, 2007. 117(11): p. 3271-82. 164. Meier, P.J. and J.L. Boyer, Preparation of basolateral (sinusoidal) and canalicular plasma membrane vesicles for the study of hepatic transport processes. Methods Enzymol, 1990. 192: p. 534-45. 165. Chen, S.J., et al., Biodistribution of AAV8 vectors expressing human low-density lipoprotein receptor in a mouse model of homozygous familial hypercholesterolemia. Hum Gene Ther Clin Dev, 2013. 24(4): p. 154-60. 166. Khan, A.U.H., et al., The PDK1 Inhibitor Dichloroacetate Controls Cholesterol Homeostasis Through the ERK5/MEF2 Pathway. Sci Rep, 2017. 7(1): p. 10654. 167. Denton, R.M., Regulation of mitochondrial dehydrogenases by calcium ions. Biochim Biophys Acta, 2009. 1787(11): p. 1309-16. 168. Arruda, A.P. and G.S. Hotamisligil, Calcium Homeostasis and Organelle Function in the Pathogenesis of Obesity and Diabetes. Cell Metab, 2015. 22(3): p. 381-97.

85 169. Rieusset, J., Endoplasmic reticulum-mitochondria calcium signaling in hepatic metabolic diseases. Biochim Biophys Acta Mol Cell Res, 2017. 1864(6): p. 865-876. 170. Kruglov, E.A., et al., Type 2 inositol 1,4,5-trisphosphate receptor modulates bile salt export pump activity in rat hepatocytes. Hepatology, 2011. 54(5): p. 1790-9. 171. Henkel, A.S., et al., Hepatic overexpression of Abcb11 in mice promotes the conservation of bile acids within the enterohepatic circulation. Am J Physiol Gastrointest Liver Physiol, 2013. 304(2): p. G221-6. 172. Figge, A., et al., Hepatic overexpression of murine Abcb11 increases hepatobiliary lipid secretion and reduces hepatic steatosis. J Biol Chem, 2004. 279(4): p. 2790-9. 173. Li, T., et al., Overexpression of cholesterol 7alpha-hydroxylase promotes hepatic bile acid synthesis and secretion and maintains cholesterol homeostasis. Hepatology, 2011. 53(3): p. 996-1006. 174. Miller, W.L., Steroidogenic acute regulatory protein (StAR), a novel mitochondrial cholesterol transporter. Biochim Biophys Acta, 2007. 1771(6): p. 663-76. 175. Le Roy, T., et al., Intestinal microbiota determines development of non-alcoholic fatty liver disease in mice. Gut, 2013. 62(12): p. 1787-94. 176. Li, F., et al., Microbiome remodelling leads to inhibition of intestinal farnesoid X receptor signalling and decreased obesity. Nat Commun, 2013. 4: p. 2384. 177. Qin, J., et al., A metagenome-wide association study of gut microbiota in type 2 diabetes. Nature, 2012. 490(7418): p. 55-60. 178. Backhed, F., et al., The gut microbiota as an environmental factor that regulates fat storage. Proc Natl Acad Sci U S A, 2004. 101(44): p. 15718-23. 179. Larsen, N., et al., Gut microbiota in human adults with type 2 diabetes differs from non- diabetic adults. PLoS One, 2010. 5(2): p. e9085. 180. Liou, A.P., et al., Conserved shifts in the gut microbiota due to gastric bypass reduce host weight and adiposity. Sci Transl Med, 2013. 5(178): p. 178ra41. 181. Mantena, S.K., et al., Mitochondrial dysfunction and oxidative stress in the pathogenesis of alcohol- and obesity-induced fatty liver diseases. Free Radic Biol Med, 2008. 44(7): p. 1259-72. 182. Fadeel, B. and E. Grzybowska, HAX-1: a multifunctional protein with emerging roles in human disease. Biochim Biophys Acta, 2009. 1790(10): p. 1139-48. 183. Jeyaraju, D.V., et al., Hax1 lacks BH modules and is peripherally associated to heavy membranes: implications for Omi/HtrA2 and PARL activity in the regulation of mitochondrial stress and apoptosis. Cell Death Differ, 2009. 16(12): p. 1622-9. 184. Yap, S.V., et al., HAX-1: a multifaceted antiapoptotic protein localizing in the mitochondria and the sarcoplasmic reticulum of striated muscle cells. J Mol Cell Cardiol, 2010. 48(6): p. 1266-79. 185. Carlsson, G., et al., Compound heterozygous HAX1 mutations in a Swedish patient with severe congenital neutropenia and no neurodevelopmental abnormalities. Pediatr Blood Cancer, 2009. 53(6): p. 1143-6. 186. Hippe, A., et al., Expression and tissue distribution of mouse Hax1. Gene, 2006. 379: p. 116-26. 187. Banerjee, A., et al., Hepatitis C virus core protein and cellular protein HAX-1 promote 5- fluorouracil-mediated hepatocyte growth inhibition. J Virol, 2009. 83(19): p. 9663-71.

86 188. Sun, X., et al., MicroRNA-223 Increases the Sensitivity of Triple-Negative Breast Cancer Stem Cells to TRAIL-Induced Apoptosis by Targeting HAX-1. PLoS One, 2016. 11(9): p. e0162754. 189. Trebinska, A., et al., HAX-1 overexpression, splicing and cellular localization in tumors. BMC Cancer, 2010. 10: p. 76. 190. Qian, B., et al., Role of the tumour protein P53 gene in human cervical squamous carcinoma cells: Discussing haematopoietic cell-specific protein 1-associated protein X-1- induced survival, migration and proliferation. Oncol Lett, 2018. 16(2): p. 2629-2637. 191. Hu, G., et al., miR-125b regulates the drug-resistance of breast cancer cells to doxorubicin by targeting HAX-1. Oncol Lett, 2018. 15(2): p. 1621-1629. 192. Wu, G., et al., miR-100 Reverses Cisplatin Resistance in Breast Cancer by Suppressing HAX-1. Cell Physiol Biochem, 2018. 47(5): p. 2077-2087. 193. Lam, C.K., et al., Novel role of HAX-1 in ischemic injury protection involvement of heat shock protein 90. Circ Res, 2013. 112(1): p. 79-89. 194. Fonseca, S.G., J. Gromada, and F. Urano, Endoplasmic reticulum stress and pancreatic beta-cell death. Trends Endocrinol Metab, 2011. 22(7): p. 266-74. 195. Lipson, K.L., et al., Regulation of insulin biosynthesis in pancreatic beta cells by an endoplasmic reticulum-resident protein kinase IRE1. Cell Metab, 2006. 4(3): p. 245-54. 196. Han, Y., et al., Overexpression of HAX-1 protects cardiac myocytes from apoptosis through caspase-9 inhibition. Circ Res, 2006. 99(4): p. 415-23. 197. Kang, Y.J., et al., Molecular interaction between HAX-1 and XIAP inhibits apoptosis. Biochem Biophys Res Commun, 2010. 393(4): p. 794-9. 198. Deng, X., et al., HAX-1 Protects Glioblastoma Cells from Apoptosis through the Akt1 Pathway. Front Cell Neurosci, 2017. 11: p. 420. 199. Radhika, V., et al., Galpha13 stimulates cell migration through cortactin-interacting protein Hax-1. J Biol Chem, 2004. 279(47): p. 49406-13. 200. Gallagher, A.R., et al., The polycystic kidney disease protein PKD2 interacts with Hax-1, a protein associated with the actin cytoskeleton. Proc Natl Acad Sci U S A, 2000. 97(8): p. 4017-22. 201. Sammels, E., et al., Polycystin-2 activation by inositol 1,4,5-trisphosphate-induced Ca2+ release requires its direct association with the inositol 1,4,5-trisphosphate receptor in a signaling microdomain. J Biol Chem, 2010. 285(24): p. 18794-805. 202. de Brito, O.M. and L. Scorrano, An intimate liaison: spatial organization of the endoplasmic reticulum-mitochondria relationship. Embo j, 2010. 29(16): p. 2715-23. 203. Simmen, T., et al., Oxidative protein folding in the endoplasmic reticulum: tight links to the mitochondria-associated membrane (MAM). Biochim Biophys Acta, 2010. 1798(8): p. 1465-73. 204. Cardenas, C., et al., Essential regulation of cell bioenergetics by constitutive InsP3 receptor Ca2+ transfer to mitochondria. Cell, 2010. 142(2): p. 270-83. 205. Hirasaka, K., et al., UCP3 is associated with Hax-1 in mitochondria in the presence of calcium ion. Biochem Biophys Res Commun, 2016. 472(1): p. 108-13. 206. Mekahli, D., et al., Endoplasmic-reticulum calcium depletion and disease. Cold Spring Harb Perspect Biol, 2011. 3(6). 207. Prins, D. and M. Michalak, Organellar calcium buffers. Cold Spring Harb Perspect Biol, 2011. 3(3). 208. Clapham, D.E., Calcium signaling. Cell, 2007. 131(6): p. 1047-58.

87 209. Toyoshima, C. and G. Inesi, Structural basis of ion pumping by Ca2+-ATPase of the sarcoplasmic reticulum. Annu Rev Biochem, 2004. 73: p. 269-92. 210. Carrasco, S. and T. Meyer, STIM proteins and the endoplasmic reticulum-plasma membrane junctions. Annu Rev Biochem, 2011. 80: p. 973-1000. 211. Soboloff, J., et al., STIM proteins: dynamic calcium signal transducers. Nat Rev Mol Cell Biol, 2012. 13(9): p. 549-65. 212. Gorlach, A., P. Klappa, and T. Kietzmann, The endoplasmic reticulum: folding, calcium homeostasis, signaling, and redox control. Antioxid Redox Signal, 2006. 8(9-10): p. 1391- 418. 213. Rizzuto, R., et al., Pathways for Ca2+ efflux in heart and liver mitochondria. Biochem J, 1987. 246(2): p. 271-7. 214. Chaudhuri, D., et al., MCU encodes the pore conducting mitochondrial calcium currents. Elife, 2013. 2: p. e00704. 215. Palty, R., M. Hershfinkel, and I. Sekler, Molecular identity and functional properties of the mitochondrial Na+/Ca2+ exchanger. J Biol Chem, 2012. 287(38): p. 31650-7. 216. Rizzuto, R., et al., Mitochondria as sensors and regulators of calcium signalling. Nat Rev Mol Cell Biol, 2012. 13(9): p. 566-78. 217. Kowaltowski, A.J., et al., Mitochondria and reactive oxygen species. Free Radic Biol Med, 2009. 47(4): p. 333-43. 218. Peng, T.I. and M.J. Jou, Oxidative stress caused by mitochondrial calcium overload. Ann N Y Acad Sci, 2010. 1201: p. 183-8. 219. Assimacopoulos-Jeannet, F.D., P.F. Blackmore, and J.H. Exton, Studies on alpha- adrenergic activation of hepatic glucose output. Studies on role of calcium in alpha- adrenergic activation of phosphorylase. J Biol Chem, 1977. 252(8): p. 2662-9. 220. Kraus-Friedmann, N. and L. Feng, The role of intracellular Ca2+ in the regulation of gluconeogenesis. Metabolism, 1996. 45(3): p. 389-403. 221. Bugrim, A.E., Regulation of Ca2+ release by cAMP-dependent protein kinase. A mechanism for agonist-specific calcium signaling? Cell Calcium, 1999. 25(3): p. 219-26. 222. Foskett, J.K., et al., Inositol trisphosphate receptor Ca2+ release channels. Physiol Rev, 2007. 87(2): p. 593-658. 223. Khan, M.T., et al., Akt kinase phosphorylation of inositol 1,4,5-trisphosphate receptors. J Biol Chem, 2006. 281(6): p. 3731-7. 224. Wang, Y., et al., Inositol-1,4,5-trisphosphate receptor regulates hepatic gluconeogenesis in fasting and diabetes. Nature, 2012. 485(7396): p. 128-32. 225. Ozcan, L., et al., Calcium signaling through CaMKII regulates hepatic glucose production in fasting and obesity. Cell Metab, 2012. 15(5): p. 739-51. 226. Hirata, K., et al., The type II inositol 1,4,5-trisphosphate receptor can trigger Ca2+ waves in rat hepatocytes. Gastroenterology, 2002. 122(4): p. 1088-100. 227. Wojcikiewicz, R.J., Type I, II, and III inositol 1,4,5-trisphosphate receptors are unequally susceptible to down-regulation and are expressed in markedly different proportions in different cell types. J Biol Chem, 1995. 270(19): p. 11678-83. 228. Cruz, L.N., et al., Regulation of multidrug resistance-associated protein 2 by calcium signaling in mouse liver. Hepatology, 2010. 52(1): p. 327-37. 229. Koves, T.R., et al., Mitochondrial overload and incomplete fatty acid oxidation contribute to skeletal muscle insulin resistance. Cell Metab, 2008. 7(1): p. 45-56.

88 230. Wu, P., et al., Starvation increases the amount of pyruvate dehydrogenase kinase in several mammalian tissues. Arch Biochem Biophys, 2000. 381(1): p. 1-7. 231. Bowker-Kinley, M.M., et al., Evidence for existence of tissue-specific regulation of the mammalian pyruvate dehydrogenase complex. Biochem J, 1998. 329 ( Pt 1): p. 191-6. 232. Sugden, M.C., et al., Expression and regulation of pyruvate dehydrogenase kinase isoforms in the developing rat heart and in adulthood: role of thyroid hormone status and lipid supply. Biochem J, 2000. 352 Pt 3: p. 731-8. 233. Huang, B., et al., Isoenzymes of pyruvate dehydrogenase phosphatase. DNA-derived amino acid sequences, expression, and regulation. J Biol Chem, 1998. 273(28): p. 17680- 8. 234. Peters, S.J., et al., Muscle fiber type comparison of PDH kinase activity and isoform expression in fed and fasted rats. Am J Physiol Regul Integr Comp Physiol, 2001. 280(3): p. R661-8. 235. Sugden, M.C., K. Bulmer, and M.J. Holness, Fuel-sensing mechanisms integrating lipid and carbohydrate utilization. Biochem Soc Trans, 2001. 29(Pt 2): p. 272-8. 236. Jeoung, N.H., et al., Role of pyruvate dehydrogenase kinase isoenzyme 4 (PDHK4) in glucose homoeostasis during starvation. Biochem J, 2006. 397(3): p. 417-25. 237. Huang, B., et al., Starvation and diabetes reduce the amount of pyruvate dehydrogenase phosphatase in rat heart and kidney. Diabetes, 2003. 52(6): p. 1371-6. 238. Jeoung, N.H. and R.A. Harris, Pyruvate dehydrogenase kinase-4 deficiency lowers blood glucose and improves glucose tolerance in diet-induced obese mice. Am J Physiol Endocrinol Metab, 2008. 295(1): p. E46-54. 239. Go, Y., et al., Inhibition of Pyruvate Dehydrogenase Kinase 2 Protects Against Hepatic Steatosis Through Modulation of Tricarboxylic Acid Cycle Anaplerosis and Ketogenesis. Diabetes, 2016. 65(10): p. 2876-87. 240. Jeoung, N.H., Pyruvate Dehydrogenase Kinases: Therapeutic Targets for Diabetes and Cancers. Diabetes Metab J, 2015. 39(3): p. 188-97. 241. Wu, C.Y., et al., Targeting hepatic pyruvate dehydrogenase kinases restores insulin signaling and mitigates ChREBP-mediated lipogenesis in diet-induced obese mice. Mol Metab, 2018. 12: p. 12-24. 242. Divakaruni, A.S. and A.N. Murphy, Cell biology. A mitochondrial mystery, solved. Science, 2012. 337(6090): p. 41-3. 243. Michelakis, E.D., L. Webster, and J.R. Mackey, Dichloroacetate (DCA) as a potential metabolic-targeting therapy for cancer. Br J Cancer, 2008. 99(7): p. 989-94. 244. Stacpoole, P.W., The pharmacology of dichloroacetate. Metabolism, 1989. 38(11): p. 1124-44. 245. Sjostrom, L., et al., Lifestyle, diabetes, and cardiovascular risk factors 10 years after bariatric surgery. N Engl J Med, 2004. 351(26): p. 2683-93. 246. Isbell, J.M., et al., The importance of caloric restriction in the early improvements in insulin sensitivity after Roux-en-Y gastric bypass surgery. Diabetes Care, 2010. 33(7): p. 1438-42. 247. Ahmad, N.N., A. Pfalzer, and L.M. Kaplan, Roux-en-Y gastric bypass normalizes the blunted postprandial bile acid excursion associated with obesity. Int J Obes (Lond), 2013. 37(12): p. 1553-9.

89 248. Kohli, R., et al., Weight loss induced by Roux-en-Y gastric bypass but not laparoscopic adjustable gastric banding increases circulating bile acids. J Clin Endocrinol Metab, 2013. 98(4): p. E708-12. 249. Myronovych, A., et al., Vertical sleeve gastrectomy reduces hepatic steatosis while increasing serum bile acids in a weight-loss-independent manner. Obesity (Silver Spring), 2014. 22(2): p. 390-400. 250. Nakatani, H., et al., Serum bile acid along with plasma incretins and serum high- molecular weight adiponectin levels are increased after bariatric surgery. Metabolism, 2009. 58(10): p. 1400-7. 251. Patti, M.E., et al., Serum bile acids are higher in humans with prior gastric bypass: potential contribution to improved glucose and lipid metabolism. Obesity (Silver Spring), 2009. 17(9): p. 1671-7. 252. Evers, S.S., D.A. Sandoval, and R.J. Seeley, The Physiology and Molecular Underpinnings of the Effects of Bariatric Surgery on Obesity and Diabetes. Annu Rev Physiol, 2017. 79: p. 313-334. 253. Albaugh, V.L., et al., Bile acids and bariatric surgery. Mol Aspects Med, 2017. 56: p. 75- 89. 254. Albaugh, V.L., et al., Early Increases in Bile Acids Post Roux-en-Y Gastric Bypass Are Driven by Insulin-Sensitizing, Secondary Bile Acids. J Clin Endocrinol Metab, 2015. 100(9): p. E1225-33. 255. Kars, M., et al., Tauroursodeoxycholic Acid may improve liver and muscle but not adipose tissue insulin sensitivity in obese men and women. Diabetes, 2010. 59(8): p. 1899-905. 256. Watanabe, M., et al., Bile acids induce energy expenditure by promoting intracellular thyroid hormone activation. Nature, 2006. 439(7075): p. 484-9. 257. Keitel, V. and D. Haussinger, TGR5 in the biliary tree. Dig Dis, 2011. 29(1): p. 45-7. 258. Keitel, V., et al., The bile acid receptor TGR5 (Gpbar-1) acts as a neurosteroid receptor in brain. Glia, 2010. 58(15): p. 1794-805. 259. Keitel, V. and D. Haussinger, Perspective: TGR5 (Gpbar-1) in liver physiology and disease. Clin Res Hepatol Gastroenterol, 2012. 36(5): p. 412-9. 260. Vassileva, G., et al., Targeted deletion of Gpbar1 protects mice from cholesterol gallstone formation. Biochem J, 2006. 398(3): p. 423-30. 261. Brendel, C., et al., The small heterodimer partner interacts with the liver X receptor alpha and represses its transcriptional activity. Mol Endocrinol, 2002. 16(9): p. 2065-76. 262. Goodwin, B., et al., A regulatory cascade of the nuclear receptors FXR, SHP-1, and LRH-1 represses bile acid biosynthesis. Mol Cell, 2000. 6(3): p. 517-26. 263. Lu, T.T., et al., Molecular basis for feedback regulation of bile acid synthesis by nuclear receptors. Mol Cell, 2000. 6(3): p. 507-15. 264. Holt, J.A., et al., Definition of a novel growth factor-dependent signal cascade for the suppression of bile acid biosynthesis. Genes Dev, 2003. 17(13): p. 1581-91. 265. Inagaki, T., et al., Fibroblast growth factor 15 functions as an enterohepatic signal to regulate bile acid homeostasis. Cell Metab, 2005. 2(4): p. 217-25. 266. Schwarz, M., et al., Identification and characterization of a mouse oxysterol 7alpha- hydroxylase cDNA. J Biol Chem, 1997. 272(38): p. 23995-4001. 267. Qiu, Y., et al., Steroidogenic acute regulatory protein (StAR) overexpression attenuates HFD-induced hepatic steatosis and insulin resistance. Biochim Biophys Acta Mol Basis Dis, 2017. 1863(4): p. 978-990.

90 268. Sinal, C.J., et al., Targeted disruption of the nuclear receptor FXR/BAR impairs bile acid and lipid homeostasis. Cell, 2000. 102(6): p. 731-44. 269. Watanabe, M., et al., Bile acids lower triglyceride levels via a pathway involving FXR, SHP, and SREBP-1c. J Clin Invest, 2004. 113(10): p. 1408-18. 270. Zhang, Y., et al., Activation of the nuclear receptor FXR improves hyperglycemia and hyperlipidemia in diabetic mice. Proc Natl Acad Sci U S A, 2006. 103(4): p. 1006-11. 271. Calkin, A.C. and P. Tontonoz, Transcriptional integration of metabolism by the nuclear sterol-activated receptors LXR and FXR. Nat Rev Mol Cell Biol, 2012. 13(4): p. 213-24. 272. Zhang, Y. and P.A. Edwards, FXR signaling in metabolic disease. FEBS Lett, 2008. 582(1): p. 10-8. 273. Langhi, C., et al., Activation of the farnesoid X receptor represses PCSK9 expression in human hepatocytes. FEBS Lett, 2008. 582(6): p. 949-55. 274. Horton, J.D., J.L. Goldstein, and M.S. Brown, SREBPs: activators of the complete program of cholesterol and fatty acid synthesis in the liver. J Clin Invest, 2002. 109(9): p. 1125-31. 275. Tomlinson, E., et al., Transgenic mice expressing human fibroblast growth factor-19 display increased metabolic rate and decreased adiposity. Endocrinology, 2002. 143(5): p. 1741-7. 276. Potthoff, M.J., et al., FGF15/19 regulates hepatic glucose metabolism by inhibiting the CREB-PGC-1alpha pathway. Cell Metab, 2011. 13(6): p. 729-38. 277. Owen, B.M., D.J. Mangelsdorf, and S.A. Kliewer, Tissue-specific actions of the metabolic hormones FGF15/19 and FGF21. Trends Endocrinol Metab, 2015. 26(1): p. 22-9. 278. Nathanson, M.H., A.D. Burgstahler, and M.B. Fallon, Multistep mechanism of polarized Ca2+ wave patterns in hepatocytes. Am J Physiol, 1994. 267(3 Pt 1): p. G338-49. 279. Augustine, G.J., F. Santamaria, and K. Tanaka, Local calcium signaling in neurons. Neuron, 2003. 40(2): p. 331-46. 280. Bozadjieva, N., K.M. Heppner, and R.J. Seeley, Targeting FXR and FGF19 to Treat Metabolic Diseases-Lessons Learned From Bariatric Surgery. Diabetes, 2018. 67(9): p. 1720-1728. 281. Zhang, J., et al., Chronic Over-expression of Fibroblast Growth Factor 21 Increases Bile Acid Biosynthesis by Opposing FGF15/19 Action. EBioMedicine, 2017. 15: p. 173-183. 282. Katsuma, S., A. Hirasawa, and G. Tsujimoto, Bile acids promote glucagon-like peptide-1 secretion through TGR5 in a murine enteroendocrine cell line STC-1. Biochem Biophys Res Commun, 2005. 329(1): p. 386-90. 283. Thomas, C., et al., TGR5-mediated bile acid sensing controls glucose homeostasis. Cell Metab, 2009. 10(3): p. 167-77. 284. Wang, H.H., et al., Transgenic overexpression of Abcb11 enhances biliary bile salt outputs, but does not affect cholesterol cholelithogenesis in mice. Eur J Clin Invest, 2010. 40(6): p. 541-51. 285. Qin, J., et al., A human gut microbial gene catalogue established by metagenomic sequencing. Nature, 2010. 464(7285): p. 59-65. 286. Ley, R.E., et al., Obesity alters gut microbial ecology. Proc Natl Acad Sci U S A, 2005. 102(31): p. 11070-5. 287. Turnbaugh, P.J., et al., The human microbiome project. Nature, 2007. 449(7164): p. 804- 10. 288. Turnbaugh, P.J., et al., Diet-induced obesity is linked to marked but reversible alterations in the mouse distal gut microbiome. Cell Host Microbe, 2008. 3(4): p. 213-23.

91 289. Ley, R.E., et al., Microbial ecology: human gut microbes associated with obesity. Nature, 2006. 444(7122): p. 1022-3. 290. Zhang, H., et al., Human gut microbiota in obesity and after gastric bypass. Proc Natl Acad Sci U S A, 2009. 106(7): p. 2365-70. 291. Santacruz, A., et al., Gut microbiota composition is associated with body weight, weight gain and biochemical parameters in pregnant women. Br J Nutr, 2010. 104(1): p. 83-92. 292. Armougom, F., et al., Monitoring bacterial community of human gut microbiota reveals an increase in Lactobacillus in obese patients and Methanogens in anorexic patients. PLoS One, 2009. 4(9): p. e7125. 293. Backhed, F., et al., Mechanisms underlying the resistance to diet-induced obesity in germ-free mice. Proc Natl Acad Sci U S A, 2007. 104(3): p. 979-84. 294. Turnbaugh, P.J., et al., A core gut microbiome in obese and lean twins. Nature, 2009. 457(7228): p. 480-4. 295. Sayin, S.I., et al., Gut microbiota regulates bile acid metabolism by reducing the levels of tauro-beta-muricholic acid, a naturally occurring FXR antagonist. Cell Metab, 2013. 17(2): p. 225-35. 296. Kuipers, F., V.W. Bloks, and A.K. Groen, Beyond intestinal soap--bile acids in metabolic control. Nat Rev Endocrinol, 2014. 10(8): p. 488-98. 297. Furet, J.P., et al., Differential adaptation of human gut microbiota to bariatric surgery- induced weight loss: links with metabolic and low-grade inflammation markers. Diabetes, 2010. 59(12): p. 3049-57. 298. Risstad, H., et al., Five-year outcomes after laparoscopic gastric bypass and laparoscopic duodenal switch in patients with body mass index of 50 to 60: a randomized clinical trial. JAMA Surg, 2015. 150(4): p. 352-61. 299. Graessler, J., et al., Metagenomic sequencing of the human gut microbiome before and after bariatric surgery in obese patients with type 2 diabetes: correlation with inflammatory and metabolic parameters. Pharmacogenomics J, 2013. 13(6): p. 514-22. 300. Kong, L.C., et al., Gut microbiota after gastric bypass in human obesity: increased richness and associations of bacterial genera with adipose tissue genes. Am J Clin Nutr, 2013. 98(1): p. 16-24. 301. Cipriani, S., et al., The bile acid receptor GPBAR-1 (TGR5) modulates integrity of intestinal barrier and immune response to experimental colitis. PLoS One, 2011. 6(10): p. e25637. 302. Vavassori, P., et al., The bile acid receptor FXR is a modulator of intestinal innate immunity. J Immunol, 2009. 183(10): p. 6251-61. 303. Islam, K.B., et al., Bile acid is a host factor that regulates the composition of the cecal microbiota in rats. Gastroenterology, 2011. 141(5): p. 1773-81. 304. Ridlon, J.M., et al., Cirrhosis, bile acids and gut microbiota: unraveling a complex relationship. Gut Microbes, 2013. 4(5): p. 382-7.

Figures

Figure 1. The role of fatty liver in the pathogenesis of hyperinsulinemia and hyperglycemia [1]

Figure 2. The role of fatty liver in the pathogenesis of hyperinsulinemia and hypertriglyceridemia [1]

Figure 3. Schematic representation of ER- Mitochondria interactions, including some of the key proteins that regulate Ca+2 transfer. PACS-2: phosphofurin acidic cluster sorting protein 2; IP3R1: inositol 1,4,5-trisphosphate receptor-1; SigR1: sigma- 1 receptor; Grp75: glucose-regulated protein 75; Mfn-2: mitofusin-2; VDAC: voltage- dependent anion channel; CypD: Cyclophilin D.

Figure 4. Bile acid metabolism in liver and intestine [123] A, A representative cartoon of hepatocytes showing the classic and alternative bile acid biosynthesis that generates glycine or taurine-conjugated bile acids. The location of the lipid transporter BSEP, MDR2/3, and ABCG5/ABCG8 on the apical membrane, and the location of NTCP and OATP transporters. FGF15 is shown bound to FGFR4/b-Klotho on the hepatocytes. FGF15 is secreted from intestine to portal blood leading to suppression of bile acid synthesis. The target genes of FXR in the liver are shown in the nucleus. B, A representative cartoon of enterocyte showing bile acid absorption from the lumen occurring via ASBT and bile acid efflux of the cell via OSTa/OSTb. FXR target genes in the intestine are shown to be induced by bile acids.

Figure 5. Distribution of HAX-1 in liver subcellular fractions and different mouse tissues. A, liver homogenate was centrifuged twice at 740 g for 5 min, then at 90,000 g for 10 min to pellet crude mitochondria. The resultant supernatant contains lysosomes and microsomes which was further centrifuged to obtain a supernatant consisting of ER and cytosolic fractions. Equal amounts of protein were immunoblotted (40µg) from each subcellular fractions. B, immunoblot analysis of HAX-1 flox/flox mouse tissues at 3 weeks after AAV injection. GFP, treatment with AAV-TBG-GFP; Cre treatment with AAV-TBG- Cre. GAPDH is used as a reference protein.

Figure 6. Plasm alanine aminotransferase (ALT) levels and liver morphology after hepatic HAX-1 depletion. Age - matched male AAV-TBG-GFP and AAV-TBG- Cre injected mice maintained on chow diet. A, Plasma ALT levels in AAV-TBG-GFP (white bars) and AAV-TBG-Cre (red bars) injected mice at 2 and 10 weeks after injection with indicated AAV. n= 4-5/group. Data represent the mean ± SD. B, Hematoxylin and eosin, and Sirius red staining of livers from the mice described in A. The scale bars represent 100 µM.

Figure 7. Body weight (BW) and whole body composition after virus injection. Age- matched male mice injected with AAV-TBG-GFP or AAV-GFP-Cre and fed chow diet. A, Body weight of injected mice with indicated virus 10 weeks after injection. n=8/group. Data are presented as Mean± SD. B, Analysis of the fat, lean, and fluid (water) content of these mice by using EchoMRI. The total body weight was measured for each mouse before the analysis. n=9/group. Data are presented as Mean ± SD.

Figure 8. Plasma triglyceride levels in control and L-HAX-1-/- mice. Age-matched male control and L-HAX1-/- mice fed on normal chow. A, Fasting plasma levels of triglyceride in mice at 4 weeks after the indicated AAV- injection. n= 11-12/ group B, Lipoprotein triglyceride profiles in plasma from pooled fasted mice 10 weeks after injection with the indicated AAV- injection. n=4-5/pool. All data represent the mean ± SD. ****P＜0.0001 for the comparison of the AAV-TBG-GFP and AAV-TBG-Cre by student’s two-tailed t test.

99 A

Figure 9. Plasma cholesterol levels in control and L-HAX-1-/- mice. Age-matched male control and L-HAX1-/- mice fed on normal chow. A, Fasting plasma levels of cholesterol in mice at 4 weeks after the indicated AAV- injection. n= 11-12/ group B, Lipoprotein cholesterol profiles in plasma from pooled fasted mice 10 weeks after injection with the indicated AAV- injection. n=4-5/pool. All data represent the mean ± SD

100

Figure 10. Plasma non-esterified fatty acid (NEFA)levels in control and L-HAX-1-/- mice. Age-matched male control and L-HAX1-/- mice fed on normal chow. Fasting plasma levels of NEFA in mice were measured at 4 weeks after the indicated AAV- injection. n= 6-10/ group.

101 A B

C D

Figure 11. Assessment of triglyceride synthesis or secretion and clearance after hepatic HAX-1 deletion. Age-matched male control and L-HAX1-/- mice fed on normal chow and used at 4 weeks after virus injection. A, Hepatic triglyceride secretion rate evaluated by measuring the time-dependent increase of plasma triglyceride levels after blocking triglyceride turnover with poloxamer 407. AAV-GFP and AAV-Cre treated mice (7-9 mice / group) were fasted overnight , and Poloxamer 407 was injected into mice intravenously. After injection, blood was collected at 0 min, 1hr, 2hr, and 3hr, and triglyceride levels were measured in plasma. B, Area Under the Curve (AUC) was measured. C, Postprandial triglyceride clearance in fasted AAV-GFP and AAV-Cre treated mice (6-9 mice / group) after an oral gavage of olive oil (200 μl). Changes in triglyceride levels were measured after gavage. D, Area Under the Curve (AUC) was calculated from changes in triglyceride levels in Figure B. Data are presented as means ± SD. *P < 0.05.

102 A B

C D

Figure 12. Measurement of hepatic lipids after hepatic HAX-1 deletion in mice fed a chow diet. Age-matched male control (white bars) and L-HAX-1-/- (red bars) mice maintained on chow diet and used 4 weeks after the indicated AAV- injection. A, Liver triglyceride in overnight- fasted mice. B, Liver cholesterol in overnight-fasted mice. C, liver triglyceride in fed mice. D, Liver cholesterol in fed mice. Data are presented as Mean ± SD. *P < 0.05.

103

A B

Figure 13. Impact of hepatic HAX-1 deletion on blood glucose and insulin levels in mice fed a chow diet. Age-matched male control (white bars) and L-HAX-1-/- (red bars) mice used at 4 weeks after the indicated AAV- injection. A, Blood glucose levels in mice fed normal chow after overnight fasting. B, Plasma insulin levels in mice fed normal chow after overnight fasting. C, Calculation of Homeostasis Model Assessment of Insulin Resistance (HOMA-IR) from glucose and insulin levels. HOMA-IR= [glucose] x [insulin]/22.5. All data represented as Mean ± SD. n= 8-9/ group. ***P< 0.001.

104

A B

Figure 14. Impact of hepatic HAX-1 deletion on glucose tolerance and insulin sensitivity in mice fed a chow diet. A, Oral glucose tolerance test (OGTT) was performed 4 weeks after virus injection. Glucose level was measured at times shown after 2g/kg of oral glucose in mice fasted overnight. B, AUC analyses of data shown in (A). C, Insulin tolerance test after intraperitoneal injection of 0.75 U/kg of insulin in mice after 6hr fasting. The glucose tolerance and insulin tolerance tests data were from two separate experiments each performed with 3-4 mice in each group. All data represented as Mean ± SD. n= 6-7/ group. * P < 0.05, **P <0.01.

105

Figure 15. Impact of hepatic HAX-1 deletion on gluconeogenesis in mice fed a chow diet. Pyruvate tolerance test was performed 4 weeks after the indicated AAV- injection. Glucose level was measured at times shown after 1g/kg of intraperitoneal pyruvate injection in mice fasted overnight. All data represented as Mean ± SD. n= 6-8/ group.

106

A B

z z

z z C z z

z z

z Figure 16. Body composition of control and L-HAX-1-/- mice fed a western diet for 12 weeks. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on western diet for 12 weeks after 3 weeks of viral injection. A, Total body weight determined for control (white bars) and L-HAX-1-/- (red bars) mice. n= 10-11/group. B, Liver weight normalized to total body weight. n= 6-8/group. C, Analysis of the fat, lean, and fluid(water) content of these mice by using EcoMRI, and normalized to total body weight. n=10-11mice/group. Data are presented as Mean±SD. * P < 0.05.

107

Figure 17. Plasma triglyceride levels in control and L-HAX-1-/- mice after 12 weeks on a western diet. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on western diet for 12 weeks after 3 weeks of viral injection. A, Plasma triglycerides levels in control (white bars) and L-HAX-1-/- (red bars) mice after overnight fasting. n=8-9 mice/group. B, Lipoprotein triglyceride profiles in plasma from pooled fasted mice n=4-5/pool. Data are presented as Mean ± SD. ***P< 0.001

108

Figure 18. Plasma cholesterol levels in control and L-HAX-1-/- mice fed on western diet for 12 weeks. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on western diet for 12 weeks after 3 weeks of viral injection. A, Plasma cholesterol levels in control (white bars) and L-HAX-1-/- (red bars) mice after overnight fasting. n=8-9 mice/group. B, Lipoprotein cholesterol profiles in plasma from pooled fasted mice. n=8/pool. All data represent the mean ± SD. * P < 0.05.

109

Figure 19. Plasma non-esterified fatty acid (NEFA)levels in control and L-HAX-1-/- mice. Age-matched male control and L-HAX1-/- mice fed western diet 3 weeks after the indicated AAV-injection. Fasting plasma levels of NEFA in mice were measured after 12 weeks of western diet feeding. n= 8/ group

110

A B

Figure 20. Effect of hepatic HAX-1 deletion on blood levels of glucose and insulin in mice fed a western diet for 12 weeks. Age-matched male AAV-TBG-GFP and AAV- GFP-Cre injected mice maintained on western diet for 12 weeks after 3 weeks of viral injection. A, Blood glucose levels in control (white bars) and L-HAX-1-/- (red bars) mice after overnight fasting. n=8 mice/group. B, Plasma insulin levels in control (white bars) and L-HAX-1-/- (red bars) mice after overnight fasting. n=7 mice/group. C, Calculation of Homeostasis Model Assessment of Insulin Resistance (HOMA-IR) from glucose and insulin levels in these mice after western diet feeding. HOMA-IR= [glucose] x [insulin]/22.5. Data are presented as Mean ± SD. * P < 0.05, **P< 0.01

111

Figure 21. Effect of hepatic HAX-1 deletion on glucose tolerance in mice fed a western diet for 12 weeks. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on western diet for 12 weeks after 3 weeks of viral injection. A, Oral glucose tolerance test (OGTT) was performed in mice fed western diet for 12 weeks. Glucose level was measured at times shown after 2g/kg of oral glucose in mice fasted overnight. B, AUC analyses for OGTT data shown in (A). n=7-8 mice/group. Data are presented as Mean ± SD. * P < 0.05, **P< 0.01

112

Figure 22. Effect of hepatic HAX-1 deletion on insulin sensitivity in mice fed a western diet for 12 weeks. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on western diet for 12 weeks after 3 weeks of viral injection. A, Insulin tolerance test (ITT) was performed in mice fed western diet for 12 weeks. Glucose level was measured at times shown after intraperitoneal injection of 0.75 U/kg insulin in mice fasted 6hr. B, AUC analyses for ITT data shown in (A). n= 6-8 mice/group. Data are presented as Mean ± SD. * P < 0.05, **P< 0.01, ***P< 0.001.

113

Figure 23. Effect of hepatic HAX-1 deletion on gluconeogenesis in mice fed a western diet. Pyruvate tolerance test was performed in age-matched male AAV-TBG- GFP and AAV-GFP-Cre injected mice maintained on western diet for 12 weeks after 3 weeks of viral injection. Glucose level was measured at times shown after 1g/kg of intraperitoneal pyruvate injection in mice fasted overnight. All data represented the Mean ± SD. n= 6 mice/ group.

114 A

GFP Cre

B C

Figure 24. Lipid accumulation in samples of liver extracts from control and L-HAX- 1-/- mice fed e western diet for 12 weeks. Age-matched male AAV-TBG-GFP and AAV- GFP-Cre injected mice maintained on western diet for 12 weeks after 3 weeks of viral injection. Liver extracts collected from mice fasted overnight. A, Representative photomicrographs of Oil Red-O staining on frozen liver sections from control and L-HAX- 1-/- following western diet feeding. The scale bars indicate 100µM. B, Quantification of the extracted triglyceride content of liver tissue (500 mg) from control (white bars) and L- HAX-1-/- (red bars) mice by petroleum ether extraction. C, Quantification of the extracted cholesterol content of liver tissue(500 mg) from control (white bars) and L-HAX-1-/- (red bars) mice by petroleum ether extraction. n=8 mice/group. Data are presented as Mean ± SD. * P < 0.05.

115

GFP Cre

Lysate B C Medium

Figure 25. Triglycerides accumulation in glucose treated primary hepatocytes isolated from control and L-HAX-1-/- mice fed a chow diet. Hepatocytes isolated from age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice after 3 weeks of viral. A, Representative photomicrographs of Oil Red-O stained isolated primary hepatocytes following treatment with 25µM glucose. Nuclei counterstained with hematoxylin. The scale bars indicate 100µM. B, Quantification of the extracted triglycerides content in lysate of treated hepatocytes with 5µM or 25µM glucose for 24 hr. C, Quantification of the extracted triglycerides content in medium of treated hepatocytes with 5µM or 25µM glucose for 24 hr. n=4 mice/group. Data are presented as Mean ± SD. * P < 0.05, ***P< 0.001

116

Figure 26. De novo lipogenesis in vivo in control and L-HAX-1-/- mice fed a chow diet. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice after 3 weeks of viral injection. In vivo de novo lipogenesis was performed by feeding fasted mice with 2 mg/g of [14C]glucose. The incorporation of the radiolabel into lipids was determined after 2 hr. n=5-6/group. Data are presented as Mean ± SD. **P< 0.01

117

Figure 27. Mitochondrial respirometry in intact primary hepatocytes isolated from control and L-HAX-1-/- mice fed a chow diet. Hepatocytes isolated from age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice 3 weeks after viral injection. The assay medium was supplemented with 25mM glucose, 1mM sodium pyruvate and 4mM glutamax. A, Oxygen consumption rates (OCR) were measured in primary hepatocytes from AAV-GFP and AAV-Cre treated mice by a seahorse XF24 extracellular flux analyzer. Following the measurement of basal respiration, oligomycin was added to measure ATP production. The subsequent addition of FCCP allowed for quantification of maximum and reserved respiration. Addition of rotenone(R) and antimycin A (A), which block both complex I and complex II, allowed for determination of non-mitochondrial respiration. Representative data from four independent experiments are shown. n= 5 mice/group. B, D OCR was calculated by subtracting basal OCR from the maximum OCR induced by FCCP in presence of 25mM glucose. n=3 mice/group. All data are expressed as Mean± SD. * P < 0.05.

118

(A) Basal Respiration (B) Max. Respiration

Figure 28. Effect of hepatic HAX-1 deletion on the mitochondrial function parameters in hepatocytes from control and L-HAX-1-/- mice fed a chow diet. Oxygen consumption rate data from Figure 22 A presented by indicated key mitochondrial function parameters. A, Basal respiration= (Last rate measurement before first injection)- (non-mitochondrial respiration). B, Maximal respiration= ( maximum rate measurement after FCCP addition)- (non-mitochondrial respiration). C, ATP production=( Last rate measurement before oligomycin injection)- (minimum rate measurement after oligomycin injection). D, non-mitochondrial respiration= minimum rate measurement after rotenone and antimycin A injection. n=4-5 mice/group. All data are expressed as Mean± SD. * P < 0.05.

119

(A) Spare Resp. Capacity

(B) Coupling Efficiency

Figure 29. Effect of hepatic HAX-1 deletion on the spare respiratory capacity and coupling efficiency in hepatocytes from control and L-HAX-1-/- mice fed a chow diet. Oxygen consumption rate data of spare respiratory capacity and coupling efficiency were calculated from Figure 22A. A, Spare respiratory capacity = (Maximal respiration) - ( Basal Respiration). B, Coupling efficiency = (ATP production rate) / (basal respiration rate) X 100. n= 4-5 mice/group.

120

Figure 30. Measurement of fatty acid oxidation in primary hepatocytes from control and L-HAX-1-/- mice fed a chow diet. Fatty acid-stimulated OCR was measured by the seahorse XF24 extracellular flux analyzer. Hepatocytes were isolated from age- matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice 3 weeks after AAV- injection. A, OCRs were measured from basal rate and after addition of palmitate (150uM) and then the carnitine palmitoyl transferase inhibitor etomoxir (50uM). The assay medium was the substrate-free base medium supplemented with 2.5mM glucose and 0.5mM carnitine. B, Fatty acid oxidation was expressed as % OCR plotted using measurement 3 as the baseline. Results are representative of three independent experiments. n=3 mice/group. All data are expressed as Mean± SD. * P < 0.05, **P< 0.01.

121 .

(B) FA-stimulated OCR (C) ATP production by FAO

Figure 31. Assaying of fatty acid oxidation in the presence of FCCP. Fatty acid- stimulated OCR was measured by the seahorse XF24 extracellular flux analyzer. Hepatocytes were isolated from age- matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice 3 weeks after AAV- injection. The assay medium was the substrate-free base medium supplemented with 2.5mM glucose and 0.5mM carnitine. A, OCRs were measured from basal rate and after addition of palmitate (150uM) and then FCCP added to allow the quantification of maximal respiration. The subsequent addition of etomoxir to inhibit respiration and to confirm that the rates measured were specific to fatty acid stimulated, and then oligomycin was added to measure ATP produced in presence of palmitate. B, The effect of hepatic HAX-1 deletion on fatty acid oxidation was determined by measuring fatty acid stimulated oxygen consumption by subtracting measurement 12 from measurement 6. C, The effect of hepatic HAX-1 deletion on ATP produced from fatty acid oxidation was determined by subtracting measurement 15 from measurement 9. Results are representative of three independent experiments. n=3-4 mice/group. All data are expressed as Mean± SD. * P < 0.05, **P< 0.01, ***P< 0.001.

122

Figure 32. Impact of hepatic HAX-1 deletion on free fatty acid accumulation in liver in control and L-HAX-1-/- mice fed e western diet for 12 weeks. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on western diet for 12 weeks after 3 weeks of viral injection. Liver extracts collected from mice fasted overnight. Free fatty acid levels in the liver (N=7-8) were determined enzymatically after lipid extraction. n=7-8 mice/group. All data are expressed as Mean± SD. **P< 0.01.

123

Figure 33. Impact of hepatic HAX-1 deficiency on expression of fatty acid oxidation genes in liver of control and L-HAX-1-/- mice fed a chow diet. Age- matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on chow diet and used 3 weeks after AAV-injection. Liver extracts obtained from mice fasted overnight and analyzed for gene expression. Quantitative real time PCR analysis of genes involved in fatty acid oxidation in liver was performed. The relative levels of mRNA transcripts detected was normalized by housekeeping gene cyclophilin. n=3-5 mice/group. All data are expressed as Mean± SD. * P < 0.05, **P< 0.01.

124 A

Figure 34. Impact of hepatic HAX-1 deficiency on Pyruvate Dehydrogenase (PDH) enzymatic activity. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on chow or western diet and used 3 weeks after AAV- injection. 400 µg of liver tissue lysate was applied to measure PDH enzymatic activities .A, PDH activity measurement in 30 min time course in mice fed chow diet. PDH activity represents the rate between two time points where the increase in absorbance is the most linear. B, PDH activity measurement in 30 min time course in mice fed western diet. PDH activity represents the rate between two time points where the increase in absorbance is the most linear. All data are expressed as Mean± SD. * P < 0.05.

125

B A

Figure 35. Impact of hepatic HAX-1 deficiency on Pyruvate Dehydrogenase Kinase (PDK). Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on chow or western diet and used 3 weeks after AAV- injection. Liver extracts collected from mice fasted overnight. A, Immunoblot analysis of PDKs in liver samples derived from control and L-HAX-1-/- mice fed chow or Western diet B, Quantification of relative PDKs expression in liver samples derived from control and L- HAX-1-/- mice fed a western diet for 12 weeks. Protein loading was verified with b- actin. n=3 mice/group. All data are expressed as Mean± SD. * P < 0.05.

126

A Chow diet

B Western diet

Figure 36. Impact of hepatic HAX-1 deficiency on Pyruvate Dehydrogenase (PDH) phosphorylation. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on chow or western diet and used 3 weeks after AAV- injection. Liver extracts collected from mice fasted overnight. A, Immunoblot analysis and quantification of relative PDH phosphorylation at serine 300 (pS -PDH) in liver samples derived from control and L-HAX-1-/- mice fed a chow diet. B, Immunoblot analysis and quantification of relative PDH phosphorylation at serine 300 (pS -PDH) in liver samples derived from control and L-HAX-1-/- mice fed a western diet for 12 weeks. Protein loading was verified with total PDH and b-actin. n=3 mice/group. All data are expressed as Mean± SD. * P < 0.05.

A 127

Figure 37. Impact of HAX-1 deletion on protein levels of LDL receptor. Age- matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on western diet for 12 weeks after AAV-injection. A, Livers were collected and homogenized on ice-cold RIPA buffer. Samples were separated by gradient SDS-PAGE and transferred to PVDF membranes. The membrane was probed with antibodies against LDLR and b-actin. Each lane represents a different mouse (n=4/group). B, Densitometry analysis of the Western blots presented in A. All data are expressed as Mean± SD. * P < 0.05.

128

5mM glucose A

25mM glucose B

Figure 38. Effect of hepatic HAX-1 deletion on ER calcium homeostasis in primary hepatocytes from control and L-HAX-1-/- mice fed a chow diet. Age- matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice used 3 weeks after AAV- injection. A, Primary hepatocytes were incubated for 16hr in 5mM glucose media. B, Primary hepatocytes were incubated for 16hr in 25mM glucose media. Following incorporation of Fluo-4 AM, cells were provided DMSO or thapsigargin (450nM) and fluorescence was measure over 5min period. Data are expressed as fluorescence units in thapsigargin treated cells – fluorescence units in DMSO treated cells. Thapsigargin was used to inhibit Ca 2+ uptake by SERCA2b. n=4 mice/group. All data are expressed as Mean ± SD. **P< 0.01, ***< 0.001, ****P< 0.0001

129

Figure 39. Effect of hepatic HAX-1 deletion on mitochondrial calcium homeostasis in primary hepatocytes from control and L-HAX-1-/- mice fed a chow diet. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice used 3 weeks after AAV- injection. Mitochondrial calcium levels assessed by Rhod-2 fluorescence in primary hepatocytes treated with or without 150µM palmitate, or with 5mM or 25mM glucose for 6hr. All data are expressed as Mean± SD. * P < 0.05, **P< 0.01.

130

Figure 40. Effect of hepatic HAX-1 deficiency on protein level of Inositol 1,4,5- trisphosphate (IP3) receptor 1 (IP3R1) in liver of control and L-HAX-1-/- mice fed a chow diet. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on chow diet and used 3 weeks after AAV- injection. A, Immunoblot analysis of immunoprecipitants indicated that HAX-1 co-immunoprecipitated with IP3R1 in liver homogenates. B, Western blot analyses and densitometry quantification of IP3R1 in liver homogenates. n=4 mice/group. All data are expressed as Mean ± SD. * P < 0.05.

131

Figure 41. Impact of hepatic HAX-1 deletion on protein expression levels of BSEP in total liver homogenates and bile canalicular membrane in control and L-HAX-1-/- mice fed a chow diet. Age-matched male AAV-TBG- GFP and AAV-GFP-Cre injected mice used 3 weeks after AAV- injection. A, Immunoblot of liver homogenates of control and L-HAX-1-/- mice showing the relative protein level of total BSEP normalized to b- actin. B, Immunoblot of bile canalicular membrane extracts from liver of control and L-HAX-1-/- showing the relative protein level of BSEP normalized to canaliculus resident protein MRP2. n=-3-4 mice/group. All data are expressed as Mean ± SD. *P< 0.05, **P< 0.01.

132

A B

C D

Figure 42. Measurement of bile acid pool size in control and L-HAX-1-/- mice fed a chow or western diet. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice and fed chow and western diet. Mice were fasted overnight, tissue were collected and bile, and bile acid concentrations were determined by enzymatic reactions. A, Bile acid content in plasma. B, Bile acid content in liver. Bile acid levels were normalized to liver weight of each mouse. C, Bile acid content in small intestine. An estimation of total amount of bile acids in small intestine was obtained by adding bile acid level values obtained from duodenum, jejunum and ileum. Values were normalized to body weight of each mouse. D, Bile acid content in feces. Bile acid levels were normalized to body weight of each mouse. All data are expressed as Mean ± SD. *P< 0.05.

133

Figure 43. Fecal cholesterol contents from control and L-HAX-1-/- mice fed a chow or western diet. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice and fed chow and western diet for 12 weeks. Feces collected from individually housed mice fed a chow diet (A) or western diet (B) and then lipid extracted to measure fecal cholesterol levels. n= 5-7 mice/group All data are expressed as Mean ± SD.**P< 0.01.

134

(A) Chow diet

(B) Western diet

Figure 44. Impact of hepatic HAX-1 deficiency on expression of bile acid- responsive genes in ileum of control and L-HAX-1-/- mice. Age-matched male AAV- TBG-GFP and AAV-GFP-Cre injected mice maintained on chow diet (A) or western diet (B) for 12 weeks. Ileal extracts obtained from mice fasted overnight and analyzed for gene expression. Quantitative real time PCR analysis of genes involved in bile acid recycling in small intestine was performed. The relative levels of mRNA transcripts detected was normalized by housekeeping gene cyclophilin. n=5-10 mice/group. All data are expressed as Mean± SD. * P < 0.05.

135

(A) Chow diet

(B) Western diet

Figure 45. Impact of hepatic HAX-1 deficiency on expression of FXR-responsive genes in liver of control and L-HAX-1-/- mice. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on chow diet (A) or western diet (B) for 12 weeks. Liver tissues obtained from mice fasted overnight and analyzed for gene expression. Quantitative real time PCR analysis of hepatic FXR-responsive genes was performed. The relative levels of mRNA transcripts detected was normalized by housekeeping gene cyclophilin. n=8-10 mice/group. All data are expressed as Mean± SD. * P < 0.05.

136

Figure 46. Impact of hepatic HAX-1 deficiency on expression of bile-acid responsive genes in liver of control and L-HAX-1-/- mice. Age-matched male AAV- TBG-GFP and AAV-GFP-Cre injected mice maintained on chow diet or western diet for 12 weeks. Liver tissues obtained from mice fasted overnight and analyzed for gene expression. Quantitative real time PCR analysis of hepatic expression of bile acid synthesis genes was performed. The relative levels of mRNA transcripts detected was normalized by housekeeping gene cyclophilin. n=7-10 mice/group. All data are expressed as Mean± SD. , **P< 0.01, ***P< 0.001.

137 A Chow diet Western diet

Figure 47. Impact of HAX-1 deletion on protein levels of mitochondrial CYP27 A and StARD involved in bile acid synthesis. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice maintained on chow or western diet for 12 weeks after AAV-injection. A, Livers were collected and homogenized on ice-cold RIPA buffer. Samples were separated by gradient SDS-PAGE and transferred to PVDF membranes. The membrane was probed with antibodies against CYP27A, StARD and b-actin. Each lane represents a different mouse (n=4/group). B, Densitometry analysis of the Western blots presented in A. All data are expressed as Mean± SD. * P < 0.05, **P< 0.01, ****P< 0.0001.

138

Figure 48. Assessment of gut microbiota in control and L-HAX-1-/- mice fed a chow or western diet. Age-matched male AAV-TBG-GFP and AAV-GFP-Cre injected mice used 3 weeks after AAV-injection. A, Relative phylum abundance of gut bacteria in mice fed a chow or western diet for 12 weeks. B, Ratio of Firmicutes to Bacteroidetes in mice fed a chow or western diet for 12 weeks. All data are expressed as Mean± SD. * P < 0.05, **P< 0.01, ****P< 0.0001, ## P<0.01 compared to AAV-GFP injected mice maintained on chow diet.

139

Figure 49. Scheme depicting hepatic HAX-1 inactivation mediated protection against diet -induced metabolic complications.

140

Appendix: Publications, Abstracts and Awards

Peer-reviewed papers

1. Liu GS, Gardner G, Adly G, Jiang M, Cai WF, Lam CK, Alogaili F, Robbins N, .

Rubinstein J, Kranias EG. A novel human S10F-Hsp20 mutation induces lethal

peri-partum cardiomyopathy. J Cell Mol Med 2018;doi: 10.1111/jcmm.

2. Fawzi Alogaili, Sivaprakasam Chinnarasu, Anja Jaeschke, Evangelia G.

Kranias, and David Y. Hui. J. Biol. Chem. (2020) 295(14) 4631–4646

3. Conference Abstracts

1- Fawzi Alogaili, Sivaprakasam Chinnarasu, Anja Jaeschke, Evangelia G. Kranias,

and David Y. Hepatic HAX-1 inactivation improves insulin sensitivity and

mitochondrial energetics in mice. Presented at 33rd annual meeting of Ohio

Physiological Society (OPS), University of Cincinnati College of Medicine,

September 28-29, 2018.

2- Sivaprakasam Chinnarasu, Fawzi Alogaili, Anja Jaeschke and David Y. Hui.

Importance of hepatic LDL receptor-related protein 1 (LRP1) in maintenance of

mitochondrial integrity and function. Presented at South Eastern Lipid Research

Conference (SELRC in Cashiers, NC, Nov 6-8, 2018).

3- Fawzi Alogaili, Sivaprakasam Chinnarasu, Anja Jaeschke, Evangelia G. Kranias,

and David Y Hui . Hepatic HAX-1 inactivation improves insulin sensitivity and

mitochondrial energetics in mice. Presented at the 39th Annual University of

Cincinnati College of Medicine Graduate Student Research Forum, Cincinnati,

OH, October 25, 2018.

4- Fawzi Alogaili, Sivaprakasam Chinnarasu, Anja Jaeschke, Evangelia G. Kranias,

and David Y Hui. Hepatic HAX-1 inactivation in mice prevents metabolic diseases

141 by dual mechanisms of enhanced mitochondrial activity and bile salt export.

Presented at South Eastern Lipid Research Conference (SELRC) in Cincinnati,

OH, September 11-13, 2019.

5- Fawzi Alogaili, Sivaprakasam Chinnarasu, Anja Jaeschke, Evangelia G. Kranias,

and David Y Hui. Hepatic HAX-1 inactivation in mice prevents metabolic diseases

by dual mechanisms of enhanced mitochondrial activity and bile salt export.

Presented at the 40th Annual University of Cincinnati College of Medicine

Graduate Student Research Forum, Cincinnati, OH, November 15, 2019.

Awards

1- 2nd Scientific place at Graduate Student Research Forum, University of Cincinnati,

2019.

142