Title of Document : Fructose intake and circulating triglycerides : an examination of the role of APOC 3

Title of Document: Fructose intake and circulating triglycerides: an examination of the role of APOC 3. Eric Spencer Campbell, Nutrition, 2014 Directed By: Professor, Dr. Thomas Castonguay, Nutrition and Food Science Department of Nutrition and Food Science, University of Maryland, College Park MD USA Fructose consumption can lead to marked increases in plasma triglycerides in both humans and laboratory animals. We have observed that overnight access to a 16% fructose solution can promote hypertriglyceridemia in rats. Several investigators have suggested that APOC 3 may be implicated in promoting fructose-induced hypertriglyceridemia. Wehave examined the role of APOC 3 in liver and blood taken from rats that had been given access to a fructose solution overnight as a supplement to standard laboratory chow. Hepatic APOC3 mRNA expression from fructose alone resulted in a 14 % reduction compared to control. Interestingly, hepatic APOC3 expression was increased by about 250% in sucrose, high fructose corn syrup and glucose groups. The serum protein levels of APOC3 did not differ across groups. Contrary to our hypothesis, these results indicate that glucose containing sugars increased hepatic APOC3 mRNA expression but no sugar was capable of increasing the serum protein level. FRUCTOSE INTAKE AND CIRCULATING TRIGLYCERIDES: AN EXAMINATION OF THE ROLE OF APOC 3

I would like to thank all friends and family for such great support during my thesis. I would like to give a special thanks to Dr. Thomas Castonguay for helping me "stay the course".

I. Introduction
Nikkila and Ojala 1 were among the first to report that fructose consumption leads to a dramatic increase in circulating triglycerides. Since that time fructoseinduced hypertriglyceridemia (FIH) has been reported by dozens of investigators 2-4 , many of whom have proposed mechanisms to explain how fructose uniquely and quickly is converted from sugar to circulating fat. This review attempts to summarize the major factors involved in FIH. It is by no means comprehensive, and we apologize if we have inadvertently omitted someone's contribution to the literature.
On the other hand, we hope that this introduction to the phenomenon contributes to a deeper understanding of how fructose can be added to the long list of factors promoting obesity, particularly abdominal obesity.
The obesity epidemic that we are experiencing in the US is a national concern that affects all Americans. It does not discriminate among age, race, class, or gender.
In 1985 the rate of obesity across our country was under 14%; but by 2010 the rate exceeded 30% 5 . Furthermore, obesity is quickly becoming a leading cause of major health problems and death 6 . The research community, along with policy makers and health care officials are grappling to find the causes of this relatively sudden increase in American's girth. Several researchers have targeted high-fructose corn syrup as a potential accomplice supporting the obesity epidemic [7][8][9][10] . They and their supporters have pointed out that fructose has become more prevalent in our diets over the past century. In 1900 the average fructose intake was 15g/day. Fructose was consumed mainly through eating fruits and vegetables, which have the added benefit of fiber 11 .
However, as of 2010, fructose consumption had risen to 73g/day and was being consumed in highly processed forms 11 . One particularly attractive hypothesis linking obesity and fructose consumption is that increased fructose intake can disrupt normal hepatic metabolism leading to increased hepatic lipogenesis 12 . The purpose of this review is to summarize the evidence supporting this hypothesis all the while focusing on one manifestation of this disruption, that of fructose-induced hypertriglyceridemia.

Fructose and Appetite
It is currently believed that normal hunger and satiety signals operate to alert the body when to initiate or end eating. One of these protein signals, leptin, decreases food intake when in excess and is therefore said to have anorexigenic effects 12,13 .
Leptin is produced within adipose cells in relation to the amount of adipose in the body and travels to the arcuate nucleus. When excess adipose tissue exists, there is excess leptin produced signaling the body to stop eating. Another signal, ghrelin, is produced in the absence of food in the stomach and stimulates food intake 14,15 .
Ghrelin is said to have orexigenic effects and travels through the blood system to the hypothalamus to signal hunger 15 . Recently, we have reported that overnight access to sucrose or fructose solutions can dramatically increase liver 11 β hydroxysteroid dehydrogenase -1, an enzyme that is linked to omental obesity 16,17 . Further, we have demonstrated that both fructose and sucrose can have significant effects on hypothalamic leptin, ghrelin and several other factors that help to control appetite 18,19 .
We believe that dietary obesity arises when normal hunger and satiety signals are disrupted causing food intake disregulation in the form of overeating. While fructose has become a target of many research topics surrounding obesity, it is important to remember that sucrose (table sugar) is comprised of two molecules, one of glucose and one of fructose. Not surprisingly, sucrose and fructose have both been shown to have similar effects on weight gain.
Fructose can increase the palatability of some foods and as a result can cause normal signals for fullness to be ignored, thereby allowing for increased intake and subsequent weight gain 20,21 . Others have confirmed this effect with other sugars, showing sugary foods motivate intake despite satiety 22 . Rats fed high fructose diets had increased body adiposity after 3-6 weeks [23][24][25] . Furthermore, rats continually fed fructose develop leptin resistance, which in turn accelerates high-fat induced obesity 26 . Additionally, diet-induced obesity causes leptin resistance in both leptin transport across the blood-brain barrier as well as central leptin signaling in mice 27 .
Thus the long-term consequences of fructose consumption and weight gain may override normal satiety signals.

Fructose and Reward
In addition to the effect of fructose on satiety signals, sucrose can induce addiction-like behaviors such as bingeing and dependence [28][29][30] . Hoebel and his associates have shown that underweight rats release more dopamine in the nucleus accumbens during a sucrose binge, indicating a potential mechanism for how sucrose may induce addiction-like behaviors 31 . Additionally, intermittent sucrose bingeing decreased D2 receptors and µ-opioid receptors in rats, which is an effect similar to that seen with morphine, indicating a role for sucrose in reward via its dopamine pathway 32 . These data make it clear that sugar (either sucrose or fructose) can have a major impact on reward and possibly addiction, as both phenomena are mediated by dopaminergic pathways. Additionally this research may shed light on how sugary foods override satiety and result in overeating through mechanisms other than directly affecting satiety signals of the hypothalamus.

Fructose and Liver
Sugar also has effects on peripheral metabolic mechanisms that influence intake. For example, fructose may increase lipogenesis in the liver beginning at the molecular level. Miyazaki 33 reported that sterol regulatory element-binding protein-1 (SREBP-1), an integral part of hepatic lipogenesis, was increased in the livers of rats receiving high-fructose diets. This enzyme is a critical factor that transforms excess energy into triglycerides. Thus, when in excess SREBP-1 will increase triglyceride formation. Additionally, elevated levels of glucocorticoids have been linked to visceral obesity and metabolic complications 34 . In a recent study 11-β hydroxysteroid dehydrogenase -1 (11-βHSD-1, a regulator of intracellular glucocorticoids) was increased in the livers of rats after 24-hours of fructose access 17 . Additionally, sucrose increases 11β-HSD1 in adipose tissue, which is a characteristic of human and animal models of obesity 15 . In a study that assessed the effect of fructose and glucose in humans, overweight and obese subjects consuming 25% of their energy (10 weeks) as either glucose or fructose sweetened beverages gained weight regardless of the type of sugar used 34 . Interestingly, in subjects consuming fructose the visceral adipose volume was significantly higher than both control and glucose groups 34 . Rats fed high fructose diets (6 weeks) displayed hypertriglyceridemia, hyperinsulinemia, and became glucose intolerant 35 . Liver X receptors (β and α) are nuclear receptors that help regulate carbohydrate and lipid metabolism, which decrease and increase respectively in response to chronic fructose consumption 35 . Taken together, these results indicate a potential connection between liver metabolism dysfunction, fructose, and increased visceral obesity.
Based upon all of the above, it seems to be clear that sugars, and particularly fructose, have the potential of being major contributors in supporting at least part of the obesity epidemic. However, not all authorities agree that sugar consumption is solely responsible for it. For example, the American Heart Association concluded that CVD and sugar consumption is not linked and that the effect on triglycerides seen with a high-sugar, high-carb diet can be reduced through exercise. Fried has proposed that long term, randomized clinical studies are needed to assess if there is a threshold at which fructose and sucrose content in diets can cause abnormalities in triacylglycerol metabolism.
The work summarized above reviews how sugars enter our diets and our bodies, and how they bypass some of the controls of intake and metabolism. The following section of this review attempts to outline how fructose is metabolized, and how it uniquely contributes to adiposity via hypertriglyceridemia.

II. Fructose metabolism -An Overview
Before attempting to explain how fructose causes hypertriglyceridemia, a short review of how it is absorbed and metabolized may be of help.

Absorption
Glucose is readily absorbed through the intestinal lumen via a diffusive GLUT-2 transporter protein and an active sodium glucose co-transporter pump (SGLT) in humans [36][37][38][39] . GLUT-2 is found on the basolateral membrane of epithelial cells to transport glucose into circulation 37 . GLUT-2 exhibits a low affinity for fructose and is responsible for a fraction of fructose absorption 37 . In contrast, many investigators reported that intestinal fructose absorption is incomplete in humans 40-42 .
Fructose is predominantly taken into the intestinal epithelial cells from the intestinal lumen passively via GLUT-5 protein transporter in humans 42 . GLUT-5 is also expressed on the basolateral side of the intestinal epithelial cells to allow for fructose to exit the epithelial cell and enter into circulation 43 . Some evidence suggests that fructose can exit epithelial cells via GLUT-2 as well as GLUT-5 44 .
The liver can extract 40-70% of dietary fructose in both rats and humans 45-48 .
The high efficiency of fructose extraction by the liver is due to the fact that the liver is the main source of GLUT-5 42 . Anatomically, the liver is the first organ to receive absorbed nutrients via the portal vein and, in that way contributes to its role in fructose extraction. By contrast, the liver accounts for a smaller percentage of glucose extraction for glycogen restoration 49 .

Fructose Metabolism
Fructose metabolism is best understood by considering three enzymes: fructokinase, aldolase B, and triokinase. All three of these are only found in the liver and kidneys of rats and humans. In the liver, fructose is rapidly converted to fructose-1-phosphate via fructokinase. Fructose -1-phosphate is then converted into the trioses dihydroxyacetone-phosphate and glyceraldehyde via the enzyme aldolase B.
Aldolase B also functions in the liver for normal glycolysis (glucose metabolism).
Glyceraldehyde is then converted to glyceraldehyde-3-phosphate via triokinase. Up until this point in fructose metabolism, there have been no rate limiting steps and as a result there is an increased amount of substrate leading to metabolic pathways from triose phosphate (ie, glycolysis. glycogenesis, glyconeogenesis, and lipogenesis, fatty-acid esterification). Glyceraldehyde-3-phosphate is next metabolized to pyruvate via the rate-limiting enzyme pyruvate kinase. Fructose enhances activation of pyruvate kinase and thus leads to an increased flux of pyruvate into the Kreb's cycle.
It should be noted that fructose metabolism does not promote an insulin response.

Glucose Metabolism
Glucose typically enters liver cells and is phosphorylated via glucokinase to form glucose-6-phosphate. Glucose-6-phosphate has two fates: conversion to fructose-6-phosphate via phosphoglucose isomerase, or storage in the form of glycogen. Fructose-6-phosphate is further phosphorylated by phosphofructokinase to produce fructose-1, 6-bisphosphate. Phosphofructokinase is the main rate-limiting step in glycolysis. Aldolase, as seen in fructose metabolism, next converts fructose-1, 6-bisphosphate into dihydroxyacetone-phosphate and glyceraldehyde-3-phosphate.
Dihydroxyacetone-phosphate is at the crossroads between glucose and fructose metabolism 51,12 . The metabolic fate of glucose relies heavily on the body's energy needs at the time. The integrated pathways of glucose metabolism include glycogenesis, glycogenolysis, glycolysis, pentose phosphate pathway, lipogenesis and the tricarboxylic acid cycle. …………………………………..

Theories on Fructose-Induced Hypertriglyceridemia
A diet high in fructose can produce adverse metabolic changes such as increased plasma triglycerides (TG), hepatic insulin resistance, and hepatic steatosis 52 . These changes usually occur in people consuming large amounts of fructose and under hypercaloric conditions 52 . Tappy has suggested that sugar calories should be limited to 140kcal/day for men and 100kcal/day for women 52 . Fructose can alter normal lipid metabolism in the liver by generating unregulated surges of pyruvate. Pyruvate enters the mitochondria via pyruvate dehydrogenase and forms acetyl-CoA. Acetyl-CoA acts as a carbon source for three different pathways: the citric acid cycle, lipogenesis, and the formation of ketone bodies. In the lipogenic pathways Acetyl-CoA is shuttled across the mitochondrial membrane as citrate and is restored back to Acetyl-CoA in the mitochondrial cytosol via ATP-citrate lyase. Here Acetyl-CoA provides substrate to the production of long-chain fatty acids. Fatty acid synthase is a multi-enzyme complex that helps facilitate the conversion of Acetyl-CoA to long-chain fatty acids. Rats fed 63% fructose for 24-48 hours and then fasted developed liver steatosis because fructose is highly lipogenic 53,54 .
Most of what has been presented above focuses on the potential effects of fructose consumption. We should point out that not only can these shifts in metabolism take place, but there is growing evidence that they are taking place in the population. For example, the amount of added sugars in the average American diet is currently 477kcal/day 10 . This is in contrast to the recommendation from the WHO that only 10% energy of total daily intake come from added sugars 55 . Stanhope and her associates have shown that leptin production drops by 20-30% when subjects consumed fructose-sweetened beverages (30% caloric intake) compared to glucosesweetened beverages 56,57 . They hypothesize that the leptin reduction observed during prolonged consumption of diets high in fructose may lead to increased energy intake and/or decreased energy expenditure and weight gain. In fact, in subjects consuming fructose there is a decrease in energy expenditure and fat oxidation when compared to subjects consuming glucose. In rhesus monkeys, the consumption of fructose over a 3 and 6 month period resulted in significant weight gain and decreased energy expenditure 10 . In a steady state fructose feeding study, postprandial levels of VLDL-TG increased in parallel with total TG, and it is likely that delayed TG clearance also contributes to this rise 10 . Rebollo 55 has reported that sucrose sweetened beverages significantly increases TG deposition in the liver, skeletal muscle, and visceral adipose tissue. Additionally, fructose consumption can increase TG over a 24-hour period, as well as apoB concentration 56 . In a study comparing the effects of consuming glucose, fructose or HFCS (25% energy) TG increased during HFCS and fructose consumption over a 24-hour period 57 . Fasting LDL and apoB concentrations were also increased as a result of fructose and HFCS consumption 57 . In an earlier separate study assessing the effect of fructose and glucose in humans, Stanhope 34 reported that overweight and obese subjects consuming 25% of their energy over 10 weeks as either glucose or fructose sweetened beverages all gained weight regardless of the type of sugar used. Interestingly, in subjects consuming fructose the visceral adipose volume was significantly higher than in both control and glucose groups.

Additional Factors Contributing to Fructose Induced Hypertriglyceridemia
Recently, it has been discovered that hepatic lipogenesis is regulated by several transcription factors, most notably the sterol regulatory element-binding protein (SREBP-1c) and the carbohydrate responsive element binding protein (CHREBP) 58,59 . SREBP-1c can induce expression of the fatty acid synthase complex, which leads to an increase in triglyceride production and secretion 60  Additional discoveries of peroxisome proliferator-activated receptors alpha and beta provide additional insights into the regulation of fatty acids in the liver 62 .
Peroxisome proliferator-activated receptor alpha (PPAR-α) is a liver transcription factor that enhances fatty acid β-oxidation. Fructose directly suppresses PPAR-α, which leads to less fatty acid oxidation and in that way promotes hypertriglyceridemia. In rats fed 10% fructose solutions PPAR-α was reduced while these changes were not observed in glucose fed rats 64 . Additionally, fructose reduced levels of hepatic fatty oxidation and increased levels of the pro-inflammatory transcription factor, nuclear factor Κ-β (NFK-β) 64 . Many compounds, such as fibrates and some fatty acids successfully enhance PPAR-α activity and remedy fructoseinduced hypertriglyceridemia 61 .
No list of factors promoting fructose-induced hypertriglyceridemia would be complete without including the forkhead transcription factor (FOXO1), which is involved in carbohydrate and lipid metabolism. FOXO1 is required for the activation of gluconeogenic gene expression in hepatic cells 65 . Fructose can stimulate FOXO1 expression, which favors a gluconeogenic state 66 . Additionally, it has been suggested that FOXO1 increases hepatic expression of a potent lipoprotein lipase (LPL) inhibitor, Apolipoprotein C-III (APOC3) 67 . APOC-3 is produced in the liver and released into circulation attached to very low-density lipoproteins 68 . In the presence of insulin LPL plays a pivotal role in the clearance of plasma triglycerides 69 . Thus, it seems that fructose is a dual threat to plasma triglyceride clearance; by its ability to activate APOC3 and its inability to produce significant insulin release.
Lastly, protein tyrosine phosphatase-1B (PTP-1B) is a receptor localized on the cytoplasmic side of endoplasmic reticulum. It negatively regulates insulin signaling in vivo 70,71 . Normally, PTP-1B will regulate insulin signaling to match that of the body's demand. The greatest expression of PTP-1B occurs in the absence of insulin and in insulin resistant states. Overexpression of PTP-1B leads to a desphosphorylation of insulin receptors and a down regulation of insulin signaling molecules 72 . Chronic fructose feeding causes PTP-1B overexpression in hamsters 73 and mice 74 . PTP-1B overexpression shifts hepatic lipid metabolism to favor an increase in lipoprotein secretion by increasing apoB100, a required substrate for lipoprotein production 74 . In that way, fructose promotes increased VLDL synthesis and release, and contributes to FIH.

III. Therapeutics
Recently, some researchers have focused on finding therapeutic drugs to ameliorate fructose-induced hypertriglyceridemia. Several exogenous compounds have been tested for their ability to interfere at different steps in the fructose metabolism pathway. Some of these findings are discussed below.
Insulin requires nitric oxide to stimulate glucose uptake. Fructose raises uric acid levels and uric acid inhibits nitric oxide bioavailability. Allopurinol is a zanthene oxidase inhibitor, and in that way can interfere with uric acid synthesis.
Nakagawa reported that Allopurinol effectively reduced uric acid even in combination with a HFCS diet 75 . Additionally, it inhibited hyperinsulinemia, hypertriglyceridemia, and hyperuricemia seen in rats fed only HFCS diets without the drug. These data provide evidence that uric acid may be important in the connection between fructose consumption and metabolic syndrome development.
Vila has reported that atorvastatin can inhibit triglycerides in plasma, and was effective in preventing increased plasma and liver TG accumulation in rats fed a high fructose diet 76 . Atorvastatin also reduced fatty acid beta oxidation. It also increased several genes that control fatty acid catabolism, prevented activation of CREBP transcriptional activity, blunted liver incorporation of fructokinase and prevented histological damage by fructose-induced liver inflammation. Rasineni 77 has shown that C. roseus (a subshrub also known as Lochnera rosea) can prevent fructoseinduced hypertriglyceridemia and lower hepatic fructokinase activity. Another therapeutic, Enicostemma littorale Blume extract (EL), successfully lowered TG levels by ameliorating glucose tolerance 78 . EL has HMG-CoA reductase activity that allows it to have a hypolipidemic effect as well as beneficial effects on platelet aggregation and blood coagulation 78 .
Finally, the antisense oligonucleotide ChREBP ASO (a therapeutic aimed at reducing TGs) successfully decreased plasma TG concentrations compared with control ASO treatments in rats fed high fat and high fructose diets 79 . Additionally fructose fed rats treated with ChREBP ASO had increased plasma uric acid, alanine transaminase and aspartate aminotransferase concentrations due to decreased fructose aldose and fructokinase expression 79 .
Fibrates are a class of plasma triglyceride lowering drugs that are used routinely in clinical practice 80 . The fibrates act as ligands for PPAR-α and induce expression of their associated fatty acid beta-oxidation genes. Additionally, fibrates also act to increase the sensitivity of LPL by reducing the expression of the LPL inhibitor, APOC3 81 . These three actions by fibrates can lower plasma triglycerides by increasing fatty acid beta-oxidation, by interfering with lipolysis, and by increasing LPL sensitivity, respectively. Several investigators have reported that fibrates can ameliorate fructose-induced hypertriglyceridemia in animal models 61,66,82,83 .
Thiazolidinediones (TZD) are a class of anti-diabetic drugs that target peroxisome proliferator activated receptor-gamma (PPAR-γ) in adipocytes to improve insulin sensitivity . TZD's can reverse and/or prevent fructose-induced hypertriglyceridemia is animal models 84-87 .

IV. Summary
Long-term exposure to dietary fructose (as opposed to glucose) results in marked increases in several enzymes that lead to metabolic dysfunction in rats and humans. The increases in gene expression such as those noted above, as well as a lack of key regulatory steps in initial fructose metabolism favors de novo lipogenesis (synthesis of triglycerides) in the liver. As a consequence, there is an overproduction of very low-density lipoprotein and triglycerides that are released into circulation, leading to hypertriglyceridemia 10,55,56 . In contrast, glucose does not increase gene expression that facilitates de novo lipogenesis, and thus spares a massive unregulated surge of metabolites favoring lipid production. One of the questions for future research deals with the safety of the country's continued use of fructose as a principal sweetening agent in our food supply. As attributed to Paracelsus "The dose makes the poison". Is there a safe level of exposure? If so, how does it vary across the life-span? Also, there are financial implications to altering a major component of the country's diet that need to be factored in to making such a decision. We look forward to several years of scientific debate.
The review of fructose intake and metabolism presented above has been written to facilitate a thorough examination of FIH. Presented below is a summary of the work that has been reviewed. The list is by no means exhaustive, but is offered to the reader as an initial guide into an important literature as we continue to grapple with developing a thorough assessment of the effects of our use of sugars in our diets.
These summaries are divided into three types of FIH papers: reviews (like this one), papers that focus on some component of the pathway (from fructose intake to circulating triglycerides) and papers that focus on assessing the efficacy of exogenous compounds that suppress circulating triglycerides.  [1][2]. This phenomenon has been replicated in both human and animal studies [3][4][5][6]. Literature searchers on "fructose AND hypertriglyceridemia" reveal that there are several plausible mechanisms that link fructose and HTG. In particular, this study will focus exclusively on the examination on the role of apolipoprotein CIII (APOC3) in promoting HTG in less than 24 hours.
The obesity epidemic that we are experiencing in the US is a national concern that affects all Americans. It does not discriminate among age, race, class, or gender.
In 1985 the rate of obesity across our country was under 14%; but by 2010 the rate exceeded 30% [7]. Furthermore, obesity is quickly becoming a leading cause of major health problems and death in the US [8]. The research community, along with policy makers and health care officials are grappling to find the causes of this relatively sudden increase in American's girth. Several researchers have targeted high-fructose corn syrup as a potential accomplice supporting the obesity epidemic [9][10][11][12]. They and their supporters have pointed out that fructose has become more prevalent in our diets over the past century. In 1900 the average fructose intake was 15g/day. Fructose was consumed mainly through eating fruits and vegetables, which have the added benefit of fiber. However, as of 2010, fructose consumption had risen to 73g/day and was being consumed in highly processed forms (13).
The undesirable HTG seen via fructose metabolism is best understood by comparing it to the non-lipogenic sugar, glucose. First, fructose metabolism is under minimal kinetic control when compared to glucose and can increase metabolites that favor triglyceride synthesis (de novo lipogenesis) [14]. This is in part due to the fact that all of fructose metabolism occurs in the liver and the liver is the site for triglyceride synthesis. Secondly, fructose has been shown to alter genetic expression of lipogenic genes that favor increased triglyceride synthesis and reduced plasma clearance when compared to glucose [9,[15][16][17]. Furthermore, fructose does not simulate a significant postprandial insulin release, which reduces the livers ability to restore homeostasis after a large bolus of fructose when compared to glucose [18].
Taken together, it becomes increasingly clearer how fructose can pose a threat to the metabolic homeostasis of the liver. that a transgenic mouse model that overexpressed or lacked the APOC3 gene resulted in hyperlipidemia or was protected from hyperlipidemia, respectively [30].
An attractive hypothesis to the lipogenic nature of fructose is in its ability to alter normal hepatic gene expression favoring an increase in lipogenic gene expression [19]. APOC3 has become an increasingly popular mechanistic approach that links fructose with HTG [17]. One study has replicated this phenomenon in chronic fructose feeding (>4 week) [20] but never has it been shown before in an acute fructose bolus (<24 hours). Keep in mind that it has been demonstrated already that fructose can elicit HTG in 24 hours or less [21]. The question addressed by the following experiment is "Is it APOC3 that is responsible for the HTG model in 24 hours or less?" Fructose can indirectly increase hepatic expression of APOC3. APOC-3 is produced in the liver and released into circulation attached to very low-density lipoproteins [22]. APOC3 is a potent inhibitor of lipoprotein lipase (LPL) [17]. In the presence of insulin LPL plays a pivotal role in the clearance of plasma triglycerides [23]. However, fructose alone does not stimulate a significant insulin release [18].
Thus, it seems that fructose is a dual threat to plasma triglyceride clearance; by its ability to activate APOC3 and its inability to produce significant insulin release.
The purpose of this study is to test various types of fructose-containing sugars (fructose, sucrose and high fructose corn syrup) with the intent of replicating our earlier observations that overnight access to fructose can promote HTG as well as to examine the role that apolipoprotein C3 (APOC3) in promoting HTG . Since pure fructose is the most lipogenic of all the sugars, we hypothesize that the treatment of pure fructose will elicit the greatest perturbation in APOC3 message and protein levels.

Animals
Adult male Sprague-Dawley (CD strain) rats (Charles River Laboratories, Wilmington, MA) with a mean weight of approximately 300 grams were used. Upon arrival, all animals were individually housed and maintained on a 12h light/dark cycle with a room temperature of 22° C ± 1° C. During the 1 wk acclimation period the rats were given free access to the control diet and water. Animals were weighed and 24h food intake and sugar solution intake (when appropriate) were measured daily at 0900h throughout the experiment.
All rats were killed by slow replacement of air in a specialized chamber with pure CO 2 followed by rapid decapitation and exsanguination. This method has been approved for use by the Panel on Euthanasia of the American Veterinary Medical Association as well as the UM IACUC. All procedures described herein are in compliance with the University of Maryland's IACUC guidelines.

Animal Diets
The control diet is a nutritionally complete low fat diet [Rodent diet 7012] prepared by Harlan Teklad (Bethlehem, PA) and provides 3.41 metabolizable kcal/g of diet. Per gram of chow, 2.14 kcal were derived from carbohydrate, 0.79 kcal was derived from protein, and 0.51 kcal was derived from fat. All animals were given free access to water throughout the experiment.

Study Design
After an initial 1 wk acclimation period (during which they had ad lib access to food and water), the rats (N=40) were randomly assigned to one of five weightmatched groups (n=8). Rats assigned to the first group were given ad libitum access to control diet (Harlan rodent diet 7012) only, and served as the control group. Rats assigned to the second group had ad libitum access to the control diet and free access to a 16% weight/volume fructose (Tate & Lyle, Decatur, IL) solution. Rats assigned to the third group had ad libitum access to the control diet plus free access to a 16% glucose weight/volume (Sigma Aldrich, St Louis MO) solution. Rats assigned to the fourth group had ad libitum access to the control diet and free access to 16% high fructose corn syrup (HFCS) weight/volume (Tate & Lyle, Decatur, IL), Finally, rats assigned to the fifth group had ad libitum access to the control diet and free access to a 16% weight/volume sucrose (Domino Foods, Baltimore, MD) solution. All sugar solutions were prepared 24h in advance and stored at 4°C. The rats were maintained on their respective diets for 24h and then sacrificed. At the time of sacrifice, the lobus lateralis sinister of each liver was dissected and flash frozen for storage at -80°C.

Plasma Measures
A Rat/Mouse Insulin ELISA Kit was purchased from Milipore, (Cat.# EZRMI-13K) and used to quantify insulin in rat plasma. Plasma glucose concentrations were measured enzymatically (Smith-Kline Beecham Laboratories). A Dimension clinical chemistry system Flex reagent cartridge (Siemens Healthcare Diagnostics, Newark, DE) was used with a Dade Behring Dimension Xpand automated system to quantify plasma triglyceride levels. A rat APOC3 ELISA kit was purchased from NeobioLab (Cat.#RA0523) and used to quantify APOC3 protein levels in serum. All plasma measurements were run in duplicate.

RNA Extraction
Total RNA was extracted from each liver sample according to RNeasy Mini and RNeasy Lipid (QIAGEN) kit directions. The samples were purified with DNAfree (Ambion) and analyzed spectrophotometrically (Nanodrop) to determine concentration and check for quality. cDNA template was created for each sample from 500 ng of purified RNA using SuperScript III Reverse Transcriptase

Measurement of Gene Expression
Changes in gene expression were measured in each liver sample, n = 40, using RT-PCR array. IQ SYBER Green Supermix (BIORAD Cat # 170-8882) was used on a C1000 BIO-RAD thermal cycler (CFX96 Real-Time System) to quantify gene expression.
A "housekeeper" or reference gene is used in RT-PCR to enable quantification of a target gene. Numerous attempts were made to elucidate a "housekeeper" gene that was not influenced by our treatments, including beta-ACTIN, RPLP1, and GAPDH. Unfortunately, the Ct values of each identified "housekeeper" were influenced by the treatments. To correct/normalize the "housekeeper" gene in each group a weighting factor was assigned. The weighting factor that transformed each sample to match the beta-actin control groups "housekeeper" value was selected. In addition, the same weighting factor was applied to the target genes Ct values. A sample of how the weighting was conducted is provided in table 1.
Cycle threshold (Ct) values obtained through RT-PCR for all samples are used to calculate fold change of gene expression. Fold change was calculated as 2^(-ddCt).
Fold change of 0.5 or less or of 1.5 or greater were considered 'robust'.

Custom Primer Design and Optimization
Primers were designed using Beacon Designer 7 software default parameters (PREMIER Biosoft, Palo Alto, CA). These primers include beta-ACTIN, RPLP1, GAPDH, and APOC3.
Primers were evaluated using the CFX96 (BioRad) protocol, using a gradient that was centered around predicted optimum annealing temperature (TaOpt). The protocol for primer verification PCR is as follows: 95°C, 3 min, and 40 cycles at 95°C for 10 sec, at a gradient of temperatures for 30 sec with subsequent melt curve of 65°-95°C at increments of 0.5°C for 5 sec. Housekeepers were selected based on the least variation in cycles between groups. See 2 for primer sequences.

Body Weight
Twenty-four hour access to any of the four sugar solutions used failed to promote significant differences in body weight (p<0.05) (F=0.0571) (df=(4, 38)).
Refer to body weight data presented in Table . There was no significant correlation in the amount of sugar consumed and the observed body weight after treatment (R 2 = 0.12) (F= 3.9297).

Intake Chow
Food intake did not differ between groups during the six-day acclimation period leading up to the treatment day. The average amount of chow consumed by all animals during the acclimation period was 25 grams. However, during the treatment period when the rats had access to sugar, the average chow consumption plummeted to 18.5 grams. All treatment groups but high fructose corn syrup and fructose ate significantly less chow on the treatment day when compared to the control (water) group (P<0.05) (F = 4.1631)(df = 4, 37). There were no significant differences between any of the treatment groups. See Figure 1 for values.

Sugar
No differences were observed in the amount of sugar consumed amongst each treatment group (P<0.05)(F=1.9691)(df=3, 30). See Table 4 for intake data.

Total Intake
No differences were observed in the amount of total caloric intake amongst each group, including control. See table 4 for total intake data (P<0.05) (F=0.2190) (df=4, 37).
It was discovered that our treatment groups could influence all of the housekeepers Ct values by a particular replicable pattern. The Ct values of both the Fructose and the control groups were always significantly higher than the other groups but did not differ from one another. This observation was the result of all available housekeepers tested. Refer to housekeeper selection data presented in Table   6.

APOC3 Gene Expression
After normalizing the housekeepers Ct values and adjusting the weighting factor for each group, there were some differences in gene expression among the treatment groups. The water group was our control and reference group for which APOC3 gene expression fold change was measured. The sucrose, & glucose groups had similar significantly positive fold changes at 2.68, and 2.59, respectively when compared to control (P<0.05)(F=7.1)(df= (4,37)). High-fructose corn syrup had a positive fold change of 2.40 but was not considered significant by the means ANOVA. Fructose elicited a negative fold change at 0.86 but was not considered significant but the means ANOVA. Refer to APOC3 gene expression presented in figure 3.

Discussion
There is an alarming obesity epidemic in this country. The obesity problem is particularly worrying not only in terms of human health and quality of life but also in terms of dollars currently spent and will be spent on healthcare in the treatment of obesity related comorbidities. Obesity has become the nation's most preventable health problem [24][25].
Despite it's prevalence, there is no known cure for this disease or consensus about how it develops. One hypothesis for the cause of obesity is that dietary intake patterns of Americans have significantly shifted over the past fifty years by including unprecedented amounts of refined sugar [26].
The purpose of this experiment was to examine this hypothesis by evaluating how sugars affect triglyceride metabolism in the liver. Animals were given 24-hour access to either a dilute glucose, sucrose, HFCS, or fructose sugar solution. Intake of both chow and sugar was measured. Hepatocyte tissue was collected and changes in APOC3 gene expression were measured using a RT-PCR assay for the different treatment groups.

Intake
Previous studies have shown that initial presentation of a sugar solution will result in rats consuming excess calories from sugar. Over time however this result washes out and total caloric intake returns to a level that resembles control [27][28].
These findings did not support previous studies that showed an increase in kcal upon initial presentation of sugar access. Our findings suggest that all animals were able to selectively regulate their overall total intake regardless off the type of sugar presented. All sugar-fed animals consumed the same amount of total calories and there were no differences in percent kcal from sugar for groups consuming fructosecontaining sugars (sucrose, HFCS, fructose).
The rats in this experiment did not increase intake of glucose-fructose disaccharides (sucrose and HFCS) in such a way as to match the amount of fructose consumed in the fructose group. In other words rats did not have a threshold for fructose consumption and compensate by consuming 50% more sucrose or 45% more HFCS to match the total consumption of fructose in the fructose fed group. We applied the same model as Collier and Bolles did in their 1968 sugar dilution studies.
Looking at intake of various sugar dilutions in rats, Collier and Bolles found that regardless of the type of sugar presented, rats will increase their intake of a dilute sugar so as to take in a certain threshold amount of total grams of sugar [29]. In the current study, this model was applied to dilutions of fructose intake. For example, sucrose fed animals had a solution that was 50% less fructose than the fructose animals. Replicating Collier and Bolles' findings, there was not set level for fructose intake as animals consuming sucrose and HFCS, which have less fructose, consumed the same amount of total kcal from sugar. These findings also show that fructose having a sweeter taste than glucose did not impact intake.

Plasma
Fructose, high-fructose corn syrup, and sucrose caused a significant and robust increase in triglycerides with only 24 hours of access. Others have demonstrated that fructose alone could elicit HTG in 24 hours but never has sucrose and high-fructose corn syrup resulted in HTG in only 24 hrs. The importance of this finding is that hypertriglyceridemia occurred with only 24-hour access to fructose and in addition with other fructose containing sugars.
No differences in plasma insulin, glucose, and APOC3 were observed, as all blood samples were collected under ad libitum feeding conditions.

Gene Expression
The reoccurring pattern of group differences for each housekeeper inhibited our ability to quantify gene expression using traditional methods. Traditionally, a housekeeper is defined by not being affected by the treatment itself and is used as a reference in 2^(-ddCt) equation. Some collaboration with experts in using RT-PCR was required before concluding that a weighting factor was a viable option. The control group in our experiment became the reference Ct value that was needed to correct for all other grouped beta-actin & APOC3 values and eventually applied to the 2^(-ddCt) equation.
An obvious limitation to the APOC3 gene expression results includes the manipulation of data as seen in the weighting factor that was applied. However, many attempts were made to elucidate a housekeeper gene that would allow the use of the more traditional method in gene expression studies. Each housekeeper was tested twice with duplicate sample data. In addition, the selection of housekeepers to be tested came from literature searchers that suggested they are widely utilized in the scientific community.
However limited the results may be, they are much different than expected.
The available literature suggests that APOC3 is a possible link to eliciting hypertriglyceridemia via increased mRNA gene expression in liver that leads to increased protein expression in circulation. The literature also suggests chronic fructose consumption leads to an increase in APOC3 mRNA and Protein levels.
Therefore, the current studies hypothesis was that treatment groups with the highest fructose concentration would have highest levels of APOC3, both message and protein. Interestingly, the results from this study tell quite a different story. The groups containing glucose (glucose, sucrose, and high-fructose corn syrup) elicited the highest APOC3 mRNA concentration in the liver compared to the control. The pure fructose treatment group, which was thought to increase APOC3 message the most, deviated from the hypothesis by slightly decreasing the message compared to control.

Conclusion
This study contributed a few novel items to the literature on fructose-induced HTG. First, this study confirmed that fructose-induced HTG could be elicited in 24 hrs. or less in many of the readily available fructose containing sugars (fructose, sucrose, and high-fructose corn syrup). Not to mention that in the experimental design the animals had free access to sugars for the first time and could control their own intake, which makes it physiologically relevant. Secondly, this study tested the validity of APOC3 and its role in the fructose-induced HTG mechanism in 24 hrs or less. Contrary to other studies that suggest APOC3 is responsible for HTG, here we conclude that there is an alternative mechanism to elicit HTG in 24 hrs or less that has not yet been identified. To be certain a more comprehensive study needs to be designed that could measure the ratios of other lipoproteins involved in newly synthesized VLDL and test their relative proportionalities to APOC3. This could confirm or deny APOC3's involvement in the 24 h mechanism to elucidate HTG. Rat # is used for identification purposes. Ct value is obtained via RT-PCR and represents relative threshold amount of amplified beta-actin mRNA. The Group identifiers are W (water, control), S (sucrose, treatment), and HF (high fructose corn syrup, treatment). The water average Ct value represents the "housekeeper" value for the control group. Water avg. Ct / Ct value is represented as a percentage, which is used as the weighting factor for subsequent use. APOC3 Ct is multiplied by the weighting factor to produce the corrected APOC3 Ct value as seen in the last column.  Average body weight in grams. Rats were weighed prior to sacrifice. Averages that share a common superscript are not different from one another (F=0.0571; df= (4,38). Standard error of the mean (SEM) is presented below the averages for each group. No significant correlation existed between sugar intake (g) and body weight (g) (R 2 = 0.12). Table 3. Average body weight of rats following 24-hour sugar access. Mean intake of kilocalories derived from sugar and chow on treatment day. There are no significant differences in the amount of kcals derived from the sugar treated groups (F=1.9691)(df=3, 30) or total kcals from sugar + chow (F=0.2190) (df=4, 37). Standard error of the mean (SEM) is presented beside the averages for each group. Superscripts with different letters indicate a significant difference at the P<0.05 level. Group means are presented here for plasma insulin, triglycerides, glucose, and APOC3. Superscripts with different letters indicate a significant difference at the P<0.05 level. The treatments were only effective in eliciting a significant difference in the triglyceride assay as indicated by the subscripts (P<0.05)(F=13.69)(df=(4,38)). The statistics for insulin, glucose and APOC3 are (F=0.12)(df=(4,35)), (F=1.3)(df=(4,35)), and (F=1.35)(df=(4,38)), respectively. Standard error of the mean (SEM) is presented beside the averages for each group. Group means are presented here for three different housekeepers. The housekeepers that were tested include beta-actin (P<0.05)(F=18.5)(df=(4,37)), Ribosomal protein large protein 1 (RPLP1) (P<0.05)(F=20.9)(df=(4,37)), and glyceraldehyde 3phosphate dehydrogenase (GAPDH) (P<0.05)(F=18.9)(df=(4,37)). Superscripts with different letters indicate a significant difference at the P<0.05 level. Each housekeeper produced the same pattern for group differences. Standard error of the mean (SEM) is presented beside the averages for each group.   APOC3 mRNA gene expression is presented here with regards to fold changes using the 2^(-ddCt) equation. Only the glucose containing sugars sucrose, HFCS, and glucose were able to positively increase the mRNA expression of APOC3 in a significant manner (P<0.05)(F=7.1)(df=(4,36)).