University of California, San Diego

UNIVERSITY OF CALIFORNIA, SAN DIEGO

Regulation of Gene Expression by Peroxisome Proliferator Activated Receptors

A dissertation submitted in partial satisfaction of the requirements for the degree Doctor of Philosophy

Biology

Isaac R. Mehl

Committee in charge:

Professor Marc R. Montminy; Chair Professor Ronald M. Evans Professor James R. Feramisco Professor Amy E. Pasquinelli Professor Bing Ren

2007

The dissertation of Isaac R. Mehl is approved, and it is acceptable in quality and form for publication on microfilm:

______

______Chair

University of California, San Diego 2007

iii

Dedicated to my best friend Dinh Dinh Diep Without you this work would not be possible

TABLE OF CONTENTS

Signature Page ……………………………………………………………….. iii

Dedication ……………………………………………………………………. iv

Table of Contents …………………………………………………………….. v

List of Figures and Tables ……………………………………………………. vi

Acknowledgements …………………………………………………………... viii

Curriculum Vitae……………………………………………………………… xi

Abstract ………………………………………………………………………. xii

I. Introduction …………………………………………………………………... 1

II. Adipose tissue is responsible for the HDL raising effect of PPARδ ligands………………………………………….…………………………………. 19

III. The role of PPARα and PPARδ in caloric restriction……..………..………. 36

IV. Conclusions………………………………………………………………….. 60

LIST OF FIGURES AND TABLES

Chapter II

Figure II.1: GW501516 rapidly increases serum HDL cholesterol and apolipoprotein AI…………………………….……………… 29

Figure II.2: Apoa1 mRNA and protein are upregulated in white adipose tissue by PPARδ ligand treatment………………….…...... 30

Figure II.3: Apoa1 can be detected in adipose tissue by immunohistochemistry………………………………………. 31

Figure II.4: WAT transplantation can rescue the effect of PPARδ ligand in AZIP and PBN animals ……….……………..……….……... 33

Figure II.5. Apoa1 protein can be detected in apoa1 -/- animals with adipose transplant ……………………………………………. 34

Chapter III

Figure III.1: PPARα null animals have decreased weight loss during CR.. 48

Figure III.2: Thyroid hormone levels are dysregulated in PPARα null animals on CR………………………………………………. 49

Figure III.3: Seminal vesicles are greatly increased in WT animals on CR …………….....………………………………..………… 50

Figure III.4: Changes in liver gene expression of PPARα -/- animals on CR……………………………………………...... …………. 51

Figure III.5: Animals treated with PPARα ligand lose less weight after CR…………………………...………………………….….. 52

Table I: Metabolic consequences of CR with PPARα and PPARδ ligands………………………………………………………. 53

Figure III.6: The effect of WY treatment on weight loss is PPARα specific...... …………. 54

Figure III.7: PPARα ligand treatment causes decreased fasting glucose but increased serum insulin when coupled with CR………….... 55

Figure III.8: Activation of PPARα increases fatty acid synthesis ……..... 56

vii

ACKNOWLEDGEMENTS

During my five years in the Evans’ lab almost everyone contributed either directly or indirectly to my growth as a scientist and a person. First of all, I would like to thank my advisor, Ron, for creating a lab environment which was extremely challenging but at the same time a comforting place. Over the years I have made many mistakes but I am thankful that Ron allowed me to mature in a safe environment. The experience I gained in this lab will surely translate into a much more successful and fruitful life in the future.

I would specifically like to thank Dr. Chih-Hao Lee for all of his guidance and help. It was Lee who first suggested the role of PPAR delta ligands in cholesterol metabolism as a possible project and worked closely with me on this project ever since. Besides being “Mr. 100%” in the lab, Lee and I also shared many wonderful times outside of lab.

Another mentor of mine in the Evans’ lab was Dr. Junichiro Sonoda. Unlike many people, Jun has maintained that insatiable, innate curiosity which causes most of us to pursue a career in science in the first place. That curiosity and his ridiculous work ethic were a great inspiration to me in the last stretch of my thesis work. Jun also has the habit of always saying good things about others while still being an extremely critical scientist. We collaborated closely on all of the work presented in chapter three of this thesis. It was amazing how quickly we acquired new data working together something that also rekindled my love of science.

viii

Several other members of the Evans’ lab were responsible for making my graduate career unforgettable. Peter Olson was the leader for all of our outdoor adventures.

Because of him, I now have some of the greatest memories of hiking in the wilderness.

Also Russell Nofsinger was the first person I talked to in this lab and was also one of the biggest reasons I joined. Over the years I have enjoyed watching his metamorphosis from a typical Midwestern graduate student into a live-octopus-eating husband and postdoc. I have also had many good times with Mike Nelson especially during his brief stint of bachelorhood. Although we wasted many hours playing axis- and-allies together it was worth it knowing that the axis powers could not win. As a bonus he allowed me the experience of beating up on chubby 12-year-olds in LOTR

TCG.

Although often far away from me, my family has done much to support me through graduate school. Several times the trip to Columbus to visit my mother,

Mike, Racheal, Jacob, Dad, and Noah was just the thing needed to de-stress. I also enjoyed the trips to the Bay to see my sister Sarah and her trips to SD. Hopefully we will be able to catch a lobster soon. One of the best things that could have happened to me was my brother Lev moving to SD. It has allowed us to recapture the friendship of our early childhood.

Finally I cannot give enough thanks to my beautiful fiancé and soon-to-be wife

Dinh. My graduate career started with meeting her and it is ending with our marriage and the beginning of a new adventure. The last few months especially would have been impossible without her help and affection.

Dr. Chih-Hao Lee was a close collaborator on all of the work presented in chapter two of this thesis. Dr. Junichiro Sonoda was a close collaborator on all of the work presented in chapter three of this thesis.

CURRICULUM VITAE

Graduate Education University of California San Diego 2001- Present Doctor of Philosophy in Biology Salk Research Fellows research talk award winner Salk mobile science lab volunteer National Institute of Health, CMG Training Grant 2002-2005 Teaching Assistant, Immunology 2006 Teaching Assistant, Genetics 2003 Teaching Assistant, Animal Physiology 2002

Undergraduate Education The Ohio State University 1996-2000 Bachelor of Science in Molecular Genetics Graduated with distinction in Molecular Genetics Graduated with honors in the Liberal Arts National Science Foundation REU participant 1999 Battelle scholarship recipient Denman Research Competition honorable mention 2nd place Biological Sciences Undergraduate Research Colloquium High School outreach program organizer

Publications Mao, Y., I.R. Mehl, and M.T. Muller, Subnuclear distribution of topoisomerase I is linked to ongoing transcription and p53 status. Proc Natl Acad Sci U S A, 2002. 99(3): p. 1235-40.

Lee, C.H., et al., Peroxisome proliferator-activated receptor delta promotes very low-density lipoprotein-derived fatty acid catabolism in the macrophage. Proc Natl Acad Sci U S A, 2006. 103(7): p. 2434-9.

Lee, C.H., et al., PPARdelta regulates glucose metabolism and insulin sensitivity. Proc Natl Acad Sci U S A, 2006. 103(9): p. 3444-9.

International Work Experience 2000-2001 Spoken English Foreign Expert, BinHai College, Qingdao China

ABSTRACT OF THE DISSERTATION

Regulation of Gene Expression by Peroxisome Proliferator Activated Receptors

Isaac R. Mehl

Doctor of Philosophy in Biology

University of California, San Diego, 2007

Professor Marc R. Montminy, Chair

Nuclear hormone receptors are a model for ligand activated transcription. A subfamily of three nuclear receptors, the peroxisome proliferators-activated receptors, binds fatty acids to regulate a variety of biological processes. One of these, PPARδ, has a broad and even tissue distribution providing few clues as to its function.

Recently, a high affinity synthetic ligand to this receptor has greatly facilitated the study of gene regulation by this receptor.

xii

One of the hallmarks of PPARδ activation is an increase in serum high-density lipoprotein cholesterol levels. Because low HDLc is a hallmark of the metabolic syndrome and a significant risk factor for cardiovascular disease we attempted to discover the mechanism by which ligand activation of PPARδ raises HDLc in mice.

Surprisingly we found that PPARδ is able to regulate the main protein component of

HDL, apolipoprotein AI, in adipose tissue. Amazingly, adipose tissue transplantation of AZIP/F1 or PPARδ knockout animals was able to rescue the effect of ligand.

Recent studies have highlighted the role of PPARδ in oxidative metabolism.

To study the effect of PPARδ activation in a state of increased oxidative metabolism we utilized a regimen of caloric restriction. Paradoxically PPARα and PPARδ ligand administration during caloric restriction decreased weight loss. Analysis of gene expression revealed that activation of the fatty acid synthesis pathway in liver could be responsible for this phenotype.

xiii

I. Introduction

1 2

The nuclear receptor superfamily

The nuclear receptor (NR) family of transcription factors impacts almost every aspect of physiology in mammals including organogenesis, circadian clock regulation, metabolism, and cancer. There are 48 members in humans. In general, their structure can be broken into domains. At the most N-terminal is the AF1 domain the function of which is poorly defined. Next is the DNA binding domain or DBD. This consists of two zinc fingers which recognize the sequence AGGTCA. Nuclear receptors can bind DNA as either monomers, homodimers (type I receptors or classic steroid receptors that translocate to the nucleus) or as heterodimers with the promiscuous retinoid X receptor (RXR). The sequence of the DBD and heterodimer partner determines the specificity of binding to hormone response elements (HRE). The spacing and directionality (direct vs. inverted repeat) of half-sites determines binding specificity although many NRs share the same binding site at least in vitro. This domain is most highly conserved among different NRs and was used initially to clone the family members (Mangelsdorf et al. 1995).

The ligand binding domain (LBD) is responsible for activation of transcription by small molecules. Essentially this acts as a switch to turn transcriptional activity on or off dependent on the presence of ligand. The AF-2 domain, which has been shown to be critically involved in coregulator function, is partially contained within the LBD.

Amazingly, early work on NRs determined that individual domains are largely interchangeable and function autonomously. For example, swapping the LBD of the estrogen receptor with that of PPARα will allow clofibrate dependent induction of estrogen response elements (Issemann and Green 1990). Crystal structures of the

LBD have been solved for many NRs providing insight into the mechanism of ligand induced activity. 11 alpha-helices form the ligand “pocket”, a conserved structure among NRs with varying size and hydrophobicity specifying ligand binding. The entrance to this pocket is capped by the 12th alpha helix and is crucial for the function of the AF-2 domain.

Coregulators

Whether NRs activate or repress transcription is in large part determined by auxiliary proteins known as coactivators and corepressors. Coactivators ultimately promote transcriptional activation by enzymatic modification of chromatin structure.

Histone acetyltransferases (HATs) relax chromatin structure thereby allowing other proteins greater access to the DNA primary sequence (Hermanson et al. 2002).

Another group of proteins, the TRAP/DRIP/Mediator/ARC complex, serve as a bridging scaffold between NRs and the basal transcription machinery including RNA polymerase II (Malik and Roeder 2000). Finally, mammalian relatives of the yeast

SWI/SNF family utilize ATP to unwind the DNA helix (Lemon et al. 2001). Other proteins including nuclear receptor corepressor (N-CoR) (Horlein et al. 1995) and silencing mediator for retinoic acid and thyroid hormone receptors (SMRT) (Chen and

Evans 1995) bind to unliganded receptors and actively repress transcription.

Interestingly, fully half of the NRs are orphan receptors and have no bona fide ligand. It is highly possible that many of these have no real ligand but rather are regulated by their own transcription, degradation, or by the availability of coactivators.

Transcriptional cascades are now being elucidated by gene expression profiling

(Barish et al. 2005; Fu et al. 2005; Bookout et al. 2006; Yang et al. 2006).

Cloning, nomenclature, and sequence homology.

In 1990, Isseman and Green cloned a gene from a mouse liver cDNA library using oligonucleotide probes based on the DBD of other nuclear receptors (Issemann and Green 1990). Amazingly they linked the tissue specific expression of this gene with the action of a class of compounds known to cause proliferation of liver peroxisomes in rodents. In quick succession other nuclear receptors with high sequence homology were cloned from frog (Dreyer et al. 1992), rat (Gottlicher et al.

1992), human (Schmidt et al. 1992), and mouse (Zhu et al. 1993; Kliewer et al. 1994;

Tontonoz et al. 1994). This early work firmly established the existence of three paralogous genes in vertebrates; a subfamily of NRs known as peroxisome proliferator-activated receptors (PPARs). Although the action of peroxisome proliferators has since been proven to be specific for the gene originally cloned by

Isselman and Green the name stuck for all three receptors. Full genome sequencing of several vertebrates confirms the existence of only three paralogous genes; PPARα,

PPARβ/δ, and PPARγ.

DNA Binding

Quickly after the initial cloning of the PPARs it was discovered that all three members heterodimerize with the retinoid X receptor to bind DR1 elements (Kliewer et al. 1992). Also known as peroxisome proliferator responsive elements (PPREs),

5 these sites consist of two direct repeats of the sequence AGGTCA spaced by one nucleotide. PPREs were initially identified upstream of genes involved in peroxisomal fatty acid beta-oxidation such as acyl-CoA oxidase (Dreyer, Krey et al.

1992; Tugwood et al. 1992) and the adipose specific gene aP2 (Tontonoz, Hu et al.

1994). Since this early work the list of functional PPREs has expanded greatly in the past 15 years. There is some evidence to suggest that the sequence immediately upstream of a PPRE confers additional binding specificity (Palmer et al. 1995; Juge-

Aubry et al. 1997) but this has never been confirmed in vivo. Furthermore, a genome wide analysis of PPREs did not find flanking sequence to improve predictive power

(Lemay and Hwang 2006). Conclusive evidence of PPAR specific site preference will have to await the results of genome-wide ChIP-chip experiments. Interestingly several other NR complexes bind DR1 elements to regulate target gene expression creating considerable cross-talk in vivo. Further obfuscation results from the fact that most orphan receptors bind AGGTCA half sites. Why evolution would favor the existence of many transcription factors that all recognize the same DNA element is currently a mystery.

Tissue localization

Examination of the expression pattern of PPARs in species as divergent as sea bass (Boukouvala et al. 2004) and homo sapiens has identified an essentially identical pattern of expression. PPARα is expressed most highly in oxidative tissues such as brown adipose tissue (Imamura et al.) and liver and to a lesser extent in heart and kidney. There are two isoforms of PPARγ termed PPARγ1 and PPARγ2. PPARγ1 is

6 expressed at low levels in most tissues. PPARγ2, which contains 30 additional residues at the amino terminus, is expressed exclusively in white and brown adipose tissue. A comparative analysis by RNAse protection assay (Escher et al. 2001) found

PPARδ to be the most highly expressed receptor in almost every tissue except WAT and BAT including liver. Recently a comprehensive analysis of both spatial and temporal expression of nuclear receptors was conducted by quatitative PCR (qPCR) in various mouse tissues (Yang, 2006 #62; Bookout, 2006 #63; www.nursa.org). This analysis revealed that PPARs are under circadian regulation in metabolic tissues.

PPARα

In the 1970’s a class of molecules known as fibrates were shown to be effective in treating hyperlipidemia. Fibrates belonged to a larger class of chemicals which caused peroxisome proliferation, hepatomegaly and eventually hepatocarcinoma in a rodent specific manner. It was only after two decades that

Issemann and Green (Issemann and Green 1990) cloned the receptor for these molecules and named it peroxisome proliferator-activated receptor (PPAR). Soon afterward DR1 response elements which bound PPARα were identified in the promoters of genes in the fatty acid beta-oxidation pathway.

The receptor was deleted from mice and proved unequivocally that peroxisome proliferators function through this receptor (Lee et al. 1995). It took another four years however to find the essential role for this receptor in the physiological response to fasting. PPARα null mice accumulate lipid in the liver after chronic high-fat feeding or a 24 hour fast. These mice also displayed severe hypoglycemia,

7 hypothermia and hypoketonemia after prolonged fasting (Kersten et al. 1999). It is now clear that a major role for PPARα is the conversion of fatty acids in the liver to substrates which can be used by peripheral tissues.

Interestingly a human splice variant exists which has dominant negative activity in vitro (Gervois et al. 1999). The role this variant plays in human biology is largely unknown but could explain some of the species specific differences between rodents and humans in the response to fibrates.

Clinical Usage

Fibrates exert their therapeutic effects through several pathways. Firstly,

PPARα has been shown to upregulate lipoprotein lipase (LPL) in skeletal muscle and liver (Staels et al. 1995; Schoonjans et al. 1996). LPL is a triglyceride lipase which hydrolyzes chylomicron and VLDL TG forming small dense LDL particles which can be more easily cleared by the LDL receptor. Fibrates also decrease the expression of

ApoCIII an inhibitor of LPL (Peters et al. 1997) further augmenting hydrolysis of triglyceride rich lipoproteins. Secondly, PPARα directly regulates the two major components of high-density lipoprotein (HDL) particles; Apoa1 and Apoa2 (Vu-Dac et al. 1995; Berthou et al. 1996). This makes fibrates one of the few drugs to increase

HDLc in humans routinely raising HDLc levels by 10-20%. Interestingly the PPRE which controls Apoa1 and Apoa2 expression in humans is not conserved and therefore

PPARα ligands do not increase apolipoprotein gene expression in rodents (Vu-Dac et al. 1998). Finally, activation of the fatty acid oxidation pathway decreases lipid synthesis and VLDL production further reducing serum triglycerides.

The efficacy and safety of fibrates have been born out by several million patient years. Additionally, the unique effects of fibrates have allowed the contribution of individual lipid parameters to cardiovascular disease (CVD) to be dissected. The Veterans Affairs High-Density Lipoprotein Intervention Trial (VA-

HIT) tested the effect of gemfibrozil on patients with low HDLc as the primary lipid abnormality (Robins et al. 2001). The results of this trial indicated that for every

5mg/dl increase in HDLc there was an 11% decrease in CVD events. These results are the best evidence that raising HDLc can have an independent beneficial effect on the development of CVD. Fibrates also seem to confer a greater benefit on patients with diabetes who are at an increased risk of developing CVD. In the Fenofibrate

Intervention and Event Lowering in Diabetes (FIELD) study almost 10,000 patients with type 2 diabetes mellitus were given either fenofibrate or placebo and major coronary events were tracked for 5 years (Keech et al. 2005). In this study there was a

24% reduction in myocardial infarction but no change in mortality. Whether fibrates are more effective alone or in combination with statins is currently being addressed in the Action to Control Cardiovascular Risk in Diabetes (Accord) study.

PPARα ligands also reduce lesion formation in mouse models of atherosclerosis. One possible mechanism is inhibition of foam cell formation; a process dependent on the liver X receptor (li 2004). Paradoxically, there was also less lesion formation in PPARα null animals (Tordjman et al. 2001). These results are just one example that often there is no simple relationship between activation and loss of receptor.

PPARδ

PPARδ was the only PPAR cloned and described before a high affinity ligand was already being used clinically to treat symptoms of the metabolic syndrome.

Therefore this might also be the only orphan receptor for which a clinically useful drug will be designed based on molecular cloning of the receptor. With a broad and even tissue expression pattern, very few clues pointed to the function of PPARδ in energy metabolism. Even after the publication of a high-affinity synthetic ligand in

2001, (Oliver et al. 2001) the function of this receptor is only beginning to be elucidated.

Adipogenesis

The role of PPARδ in adipogenesis and adipose function has yielded conflicting results. Initially, activation of PPARδ was thought to potentiate adipogenesis. Fatty acids and a non-metabolized fatty acid analog bromopalmitate induced adipocyte differentiation through PPARδ (Amri et al. 1995). Overexpression of PPARδ along with ligand increased differentiation of fibroblasts and a dominant- negative receptor inhibited differentiation (Bastie et al. 1999). Ectopically expressed

PPARδ and a ligand induced the differentiation of NIH-3T3 cells but not 3T3-L1 cells

(Hansen et al. 2001). However other work showed that overexpression of PPARδ in

3T3-L1 actually inhibited adipogenesis (Shi et al. 2002). Furthermore recent data failed to show an induction of PPARδ in 3T3-L1 cells during differentiation (Fu, Sun et al. 2005) and a fat specific knockout of PPARδ yielded no obvious phenotype

(Barak et al. 2002). Much of this confusion is probably due to different cell lines and

10 overexpression strategies as well as potential crosstalk with PPARγ. Whether or not a synthetic high affinity ligand for PPARδ, such as GW501516, activates adipogenesis in vitro has not been addressed. The only in vivo model of PPARδ’s function in adipose tissue is a constitutively activated receptor driven by the aP2 promoter (Wang et al. 2003). These animals have decreased weight gain on high fat diet and upregulation of the fatty acid beta oxidation pathway in brown adipose tissue.

Muscle function

Unlike other tissues, the function of PPARδ in muscle seems fairly clear.

Several groups have reported activation of genes involved in lipid catabolism in GW treated myotubes (Son et al. 2001; Dressel et al. 2003; Tanaka et al. 2003). These include pyruvate dehydrogenase kinase isoezyme 4 (PDK4), lipoprotein lipase (LPL), adipose differentiation related protein (ADRP), carnitine palmityltransferase 1 (CPT1) and members of the uncoupling protein family (UCP1, UCP2, UCP3). Luquet et al. overexpressed PPARδ specifically in skeletal muscle using a rather ingenious lox-cre system. They found a fiber type switch to more oxidative fibers and a decrease in body fat content (Luquet et al. 2003). A constitutively active PPARδ fused to VP16 activation domain expressed in muscle also caused an increase in oxidative muscle fiber content, decreased weight gain on HFD and increased running endurance (Wang et al. 2004). While it is clear that activation of PPARδ shifts muscle to a more oxidative fiber type the direct target genes responsible for this shift have yet to be identified.

Knockout phenotype

PPARδ has been knocked out by three independent groups all of which reported a high degree of embryonic lethality attributed to placental defects (Peters et al. 2000; Michalik et al. 2001; Barak, Liao et al. 2002). Surviving animals had greatly decreased adipose stores, altered myelination of the corpus callosum and increased skin inflammation. In general, tissue specific deletion of the PPARδ gene results in a rather disappointing lack of phenotype (Mehl and Zhang unpublished data).

Unexpectedly, a cardiac specific knockout of PPARδ resulted in a very dramatic and lethal cardiomyopathy (Cheng et al. 2004). Perhaps a PPARδ dependent regulation of cardiac fatty acid oxidation is responsible for some of the embryonic lethality seen in the global knockout and compensating alleles were selected for during the generation of these lines. This would explain why cardiomyopathy is not seen in the global knockout.

Insulin resistance

Treatment with PPARδ ligands improves insulin sensitivity in both non-human primates (Oliver, Shenk et al. 2001) and mice (Tanaka, Yamamoto et al. 2003; Wang,

Zhang et al. 2004; Lee et al. 2006). However the tissues and gene expression responsible for this effect are largely unknown. Lee et al. attempted to discern the mechanism for insulin sensitivity in GW treated animals independent of weight loss

(lee PNAS 2006). Surprisingly they found upregulation of fatty acid synthesis in the livers of obese mice treated with PPARδ ligand. Based on these results they proposed a controversial mechanism of increased insulin sensitivity through shunting of glucose

12 into the pentose phosphate pathway. Whether induction of the fatty acid synthesis pathway in liver is a general mechanism of PPARδ ligands requires further studies.

Atherosclerosis

So far the role of PPARδ in macrophage biology and atherosclerosis is not clear. Early work showed that activation or overexpression of PPARδ facilitated lipid accumulation in macrophages (Vosper et al. 2001). Work in our lab found that addition of a PPARδ ligand or VLDL increased fatty acid catabolism in 264.7 RAW

(mouse monocyte) cells with stable expression of PPARδ (Lee et al. 2006).

Activation of PPARδ caused increased expression of aldehyde dehydrogenase 9

(Aldh9), mitochondrial carnitine palmityltransferase 1 (Cpt1a), carnitine/acylcarnitine translocase (Slc25a20), and a newly described lipase, adipose triglyceride lipase

(ATGL).

The effect of PPARδ ligands on atherosclerotic progression is also not clear.

Lee et al. found a decrease in lesion area of PPARδ -/- bone marrow transplanted into the ApoE -/- background (Lee et al. 2003). This effect was shown to be dependent on the ability of PPARδ to sequester a repressor of inflammation BCL-6. The only study to directly asses the effect of PPARδ ligand treatment on atherosclerosis was done in

LDLR-/- animals after 14 weeks of atherogenic diet (Li et al. 2004). This group saw no effect of PPARδ ligand on atherosclerotic progression. Indeed, activation of both

PPARα and PPARγ but not PPARδ led to reduction in plaque size. Whether PPARδ ligands are effective in reducing lesions in a less severe model of atherosclerosis or in humans requires further investigation.

HDL metabolism

An extremely consistent effect of PPARδ ligands is an increase in HDLc levels. The initial report of a high affinity ligand for PPARδ included treatment of obese rhesus macques. There was a dramatic beneficial change in serum lipoprotein levels including a decrease in the pro-atherogenic particles VLDL and LDL and an increase in the anti-atherogenic HDL. The effect of PPARδ ligand activation on

HDLc has also been observed in mice (Leibowitz et al. 2000; Lee, Olson et al. 2006) and another primate (Wallace et al. 2005). The safety and efficacy of the PPARδ ligand GW501516 on humans is currently being tested in clinical trials (Sprecher et al.

2007).

References cited

Alderman, J. D., et al. (1989). "Effect of a modified, well-tolerated niacin regimen on serum total cholesterol, high density lipoprotein cholesterol and the cholesterol to high density lipoprotein ratio." Am J Cardiol 64(12): 725-9.

Amri, E. Z., et al. (1995). "Cloning of a protein that mediates transcriptional effects of fatty acids in preadipocytes. Homology to peroxisome proliferator-activated receptors." J Biol Chem 270(5): 2367-71.

Barak, Y., et al. (2002). "Effects of peroxisome proliferator-activated receptor delta on placentation, adiposity, and colorectal cancer." Proc Natl Acad Sci U S A 99(1): 303-8.

Barish, G. D., et al. (2005). "A Nuclear Receptor Atlas: macrophage activation." Mol Endocrinol 19(10): 2466-77.

Bastie, C., et al. (1999). "Expression of peroxisome proliferator-activated receptor PPARdelta promotes induction of PPARgamma and adipocyte differentiation in 3T3C2 fibroblasts." J Biol Chem 274(31): 21920-5.

Berthou, L., et al. (1996). "Opposite regulation of human versus mouse apolipoprotein A-I by fibrates in human apolipoprotein A-I transgenic mice." J Clin Invest 97(11): 2408-16.

Bookout, A. L., et al. (2006). "Anatomical profiling of nuclear receptor expression reveals a hierarchical transcriptional network." Cell 126(4): 789-99.

Boukouvala, E., et al. (2004). Molecular characterization of three peroxisome proliferator-activated receptors from the sea bass (Dicentrarchus labrax). 39: 1085-92.

Chen, J. D. and R. M. Evans (1995). "A transcriptional co-repressor that interacts with nuclear hormone receptors." Nature 377(6548): 454-7.

Cheng, L., et al. (2004). "Cardiomyocyte-restricted peroxisome proliferator-activated receptor-delta deletion perturbs myocardial fatty acid oxidation and leads to cardiomyopathy." Nat Med 10(11): 1245-50.

Dressel, U., et al. (2003). "The peroxisome proliferator-activated receptor beta/delta agonist, GW501516, regulates the expression of genes involved in lipid catabolism and energy uncoupling in skeletal muscle cells." Mol Endocrinol 17(12): 2477-93.

Dreyer, C., et al. (1992). "Control of the peroxisomal beta-oxidation pathway by a novel family of nuclear hormone receptors." Cell 68(5): 879-87.

Escher, P., et al. (2001). "Rat PPARs: quantitative analysis in adult rat tissues and regulation in fasting and refeeding." Endocrinology 142(10): 4195-202.

Fu, M., et al. (2005). "A Nuclear Receptor Atlas: 3T3-L1 adipogenesis." Mol Endocrinol 19(10): 2437-50.

Gervois, P., et al. (1999). "Fibrates increase human REV-ERBalpha expression in liver via a novel peroxisome proliferator-activated receptor response element." Mol Endocrinol 13(3): 400-9.

Gottlicher, M., et al. (1992). "Fatty acids activate a chimera of the clofibric acid- activated receptor and the glucocorticoid receptor." Proc Natl Acad Sci U S A 89(10): 4653-7.

Hansen, J. B., et al. (2001). "Peroxisome proliferator-activated receptor delta (PPARdelta )-mediated regulation of preadipocyte proliferation and gene expression is dependent on cAMP signaling." J Biol Chem 276(5): 3175-82.

Hermanson, O., et al. (2002). "Nuclear receptor coregulators: multiple modes of modification." Trends Endocrinol Metab 13(2): 55-60.

Horlein, A. J., et al. (1995). "Ligand-independent repression by the thyroid hormone receptor mediated by a nuclear receptor co-repressor." Nature 377(6548): 397- 404.

Imamura, H., et al. (1996). "Relationship of cigarette smoking to blood pressure and serum lipids and lipoproteins in men." Clin Exp Pharmacol Physiol 23(5): 397- 402.

Issemann, I. and S. Green (1990). "Activation of a member of the steroid hormone receptor superfamily by peroxisome proliferators." Nature 347(6294): 645-50.

Juge-Aubry, C., et al. (1997). "DNA binding properties of peroxisome proliferator- activated receptor subtypes on various natural peroxisome proliferator response elements. Importance of the 5'-flanking region." J Biol Chem 272(40): 25252-9.

Keech, A., et al. (2005). "Effects of long-term fenofibrate therapy on cardiovascular events in 9795 people with type 2 diabetes mellitus (the FIELD study): randomised controlled trial." Lancet 366(9500): 1849-61.

Kersten, S., et al. (1999). "Peroxisome proliferator-activated receptor alpha mediates the adaptive response to fasting." J Clin Invest 103(11): 1489-98.

Kliewer, S. A., et al. (1994). "Differential expression and activation of a family of murine peroxisome proliferator-activated receptors." Proc Natl Acad Sci U S A 91(15): 7355-9.

Kliewer, S. A., et al. (1992). "Convergence of 9-cis retinoic acid and peroxisome proliferator signalling pathways through heterodimer formation of their receptors." Nature 358(6389): 771-4.

Lee, C. H., et al. (2003). "Transcriptional repression of atherogenic inflammation: modulation by PPARdelta." Science 302(5644): 453-7.

Lee, C. H., et al. (2006). "Peroxisome proliferator-activated receptor delta promotes very low-density lipoprotein-derived fatty acid catabolism in the macrophage." Proc Natl Acad Sci U S A 103(7): 2434-9.

Lee, C. H., et al. (2006). "PPARdelta regulates glucose metabolism and insulin sensitivity." Proc Natl Acad Sci U S A 103(9): 3444-9.

Lee, S. S., et al. (1995). "Targeted disruption of the alpha isoform of the peroxisome proliferator-activated receptor gene in mice results in abolishment of the pleiotropic effects of peroxisome proliferators." Mol Cell Biol 15(6): 3012-22.

Leibowitz, M. D., et al. (2000). "Activation of PPARdelta alters lipid metabolism in db/db mice." FEBS Lett 473(3): 333-6.

Lemay, D. G. and D. H. Hwang (2006). "Genome-wide identification of peroxisome proliferator response elements using integrated computational genomics." J Lipid Res 47(7): 1583-7.

Lemon, B., et al. (2001). "Selectivity of chromatin-remodelling cofactors for ligand- activated transcription." Nature 414(6866): 924-8.

Li, A. C., et al. (2004). "Differential inhibition of macrophage foam-cell formation and atherosclerosis in mice by PPARalpha, beta/delta, and gamma." J Clin Invest 114(11): 1564-76.

Luquet, S., et al. (2003). "Peroxisome proliferator-activated receptor delta controls muscle development and oxidative capability." Faseb J 17(15): 2299-301.

Malik, S. and R. G. Roeder (2000). "Transcriptional regulation through Mediator-like coactivators in yeast and metazoan cells." Trends Biochem Sci 25(6): 277-83.

Mangelsdorf, D. J., et al. (1995). "The nuclear receptor superfamily: the second decade." Cell 83(6): 835-9.

Michalik, L., et al. (2001). "Impaired skin wound healing in peroxisome proliferator- activated receptor (PPAR)alpha and PPARbeta mutant mice." J Cell Biol 154(4): 799-814.

Oliver, W. R., Jr., et al. (2001). "A selective peroxisome proliferator-activated receptor delta agonist promotes reverse cholesterol transport." Proc Natl Acad Sci U S A 98(9): 5306-11.

Palmer, C. N., et al. (1995). "Novel sequence determinants in peroxisome proliferator signaling." J Biol Chem 270(27): 16114-21.

Peters, J. M., et al. (1997). "Alterations in lipoprotein metabolism in peroxisome proliferator-activated receptor alpha-deficient mice." J Biol Chem 272(43): 27307-12.

Peters, J. M., et al. (2000). "Growth, adipose, brain, and skin alterations resulting from targeted disruption of the mouse peroxisome proliferator-activated receptor beta(delta)." Mol Cell Biol 20(14): 5119-28.

Robins, S. J., et al. (2001). "Relation of gemfibrozil treatment and lipid levels with major coronary events: VA-HIT: a randomized controlled trial." Jama 285(12): 1585-91.

Schmidt, A., et al. (1992). "Identification of a new member of the steroid hormone receptor superfamily that is activated by a peroxisome proliferator and fatty acids." Mol Endocrinol 6(10): 1634-41.

Schoonjans, K., et al. (1996). "PPARalpha and PPARgamma activators direct a distinct tissue-specific transcriptional response via a PPRE in the lipoprotein lipase gene." Embo J 15(19): 5336-48.

Shi, Y., et al. (2002). "The peroxisome proliferator-activated receptor delta, an integrator of transcriptional repression and nuclear receptor signaling." Proc Natl Acad Sci U S A 99(5): 2613-8.

Son, C., et al. (2001). "Up-regulation of uncoupling protein 3 gene expression by fatty acids and agonists for PPARs in L6 myotubes." Endocrinology 142(10): 4189- 94.

Sprecher, D. L., et al. (2007). "Triglyceride:high-density lipoprotein cholesterol effects in healthy subjects administered a peroxisome proliferator activated receptor delta agonist." Arterioscler Thromb Vasc Biol 27(2): 359-65.

Staels, B., et al. (1995). "Fibrates downregulate apolipoprotein C-III expression independent of induction of peroxisomal acyl coenzyme A oxidase. A potential mechanism for the hypolipidemic action of fibrates." J Clin Invest 95(2): 705- 12.

Tanaka, T., et al. (2003). "Activation of peroxisome proliferator-activated receptor delta induces fatty acid beta-oxidation in skeletal muscle and attenuates metabolic syndrome." Proc Natl Acad Sci U S A 100(26): 15924-9.

Tontonoz, P., et al. (1994). "mPPAR gamma 2: tissue-specific regulator of an adipocyte enhancer." Genes Dev 8(10): 1224-34.

Tordjman, K., et al. (2001). "PPARalpha deficiency reduces insulin resistance and atherosclerosis in apoE-null mice." J Clin Invest 107(8): 1025-34.

Tugwood, J. D., et al. (1992). "The mouse peroxisome proliferator activated receptor recognizes a response element in the 5' flanking sequence of the rat acyl CoA oxidase gene." Embo J 11(2): 433-9.

Vosper, H., et al. (2001). "The peroxisome proliferator-activated receptor delta promotes lipid accumulation in human macrophages." J Biol Chem 276(47): 44258-65.

Vu-Dac, N., et al. (1998). "The nuclear receptors peroxisome proliferator-activated receptor alpha and Rev-erbalpha mediate the species-specific regulation of apolipoprotein A-I expression by fibrates." J Biol Chem 273(40): 25713-20.

Vu-Dac, N., et al. (1995). "Fibrates increase human apolipoprotein A-II expression through activation of the peroxisome proliferator-activated receptor." J Clin Invest 96(2): 741-50.

Wallace, J. M., et al. (2005). "Effects of peroxisome proliferator-activated receptor alpha/delta agonists on HDL-cholesterol in vervet monkeys." J Lipid Res 46(5): 1009-16.

Wang, Y. X., et al. (2003). "Peroxisome-proliferator-activated receptor delta activates fat metabolism to prevent obesity." Cell 113(2): 159-70.

Wang, Y. X., et al. (2004). "Regulation of muscle fiber type and running endurance by PPARdelta." PLoS Biol 2(10): e294.

Yang, X., et al. (2006). "Nuclear receptor expression links the circadian clock to metabolism." Cell 126(4): 801-10.

Zhu, Y., et al. (1993). "Cloning of a new member of the peroxisome proliferator- activated receptor gene family from mouse liver." J Biol Chem 268(36): 26817-20.

II. Adipose tissue is responsible for the HDL raising effect of PPARδ ligands

19 20

Abstract

Low levels of HDL cholesterol are a hallmark of the metabolic syndrome and a major, independent risk factor for the development of cardiovascular disease. PPARδ ligands are known to increase HDL cholesterol in both rodents and non-human primates. Although small molecule activators of PPARδ could be useful for the clinical treatment of low HDLc the mechanism for this effect is currently unknown.

Here we describe the adipose specific regulation of apoa1 by PPARδ. PPARδ ligand treatment increased both the mRNA and protein levels of apoa1 specifically in adipose tissue. Additionally, apoa1 protein could be found in the serum of Apoa1 null animals transplanted with WAT. These results identify PPARδ regulated gene expression in adipose tissue as a contributor to HDLc levels.

Introduction

Currently cardiovascular disease (CVD) is the leading cause of death worldwide. Almost one in every three deaths in the United States alone are due to

CVD (Mokdad et al. 2004). A subset of individuals with a cluster of related symptoms known as metabolic syndrome (MS) have an increased risk for developing

CVD (Ford et al. 2002; Lakka, Laaksonen et al. 2002). The National Cholesterol

Education Program’s Adult Treatment Panel III report (ATP III) defined the components of MS as hyperlipidemia, hypertension, obesity, insulin resistance, high

LDL cholesterol (LDLc) and low HDL cholesterol (HDLc) (Grundy et al. 2004).

HDLc levels in particular have a strong inverse correlation with CVD. The

Veterans Affairs HDL Intervention Trial (VA-HIT) was designed to specifically test

21 the effect of raising HDLc on clinical endpoints. The result of VA-HIT was that CVD events were reduced 11% for every 5mg/dl increase in HDLc (Robins et al. 2001).

Several animal models have shown that increasing the main protein component of

HDL, apolipoprotein AI (apoa1), reduces atherosclerotic lesion formation (Badimon et al. 1990; Rubin et al. 1991; Paszty et al. 1994; Miyazaki et al. 1995).

The beneficial effect of HDL on CVD is thought to occur through a process known as reverse cholesterol transport (RCT) (Rader 2003). Facilitated by the adenosine triphosphate–binding cassette protein A1 (ABCA1), cholesterol and phospholipids from peripheral tissues are loaded onto lipid-poor apoa1. Excess cholesterol is then exchanged in the liver via SR-BI and excreted into bile. The crucial role of apoa1 and ABCA1 in RCT is highlighted by the fact that deletion of these genes results in an almost complete lack of HDLc (Williamson et al. 1992;

McNeish et al. 2000).

Despite a wealth of in vitro data on the regulation of Apoa1 few compounds effectively increase HDL in humans. Recently a selective ligand for the nuclear receptor PPARδ was identified which increased HDLc levels by 80% (Oliver et al.

2001). In this paper we demonstrate that PPARδ regulates apoa1 expression in adipose tissue thereby increasing serum HDLc. These results implicate adipose tissue as a future target for the treatment of CVD.

Material and Methods

Animal Experiments. PPAR beta knockout mice and a control line were a kind gift of J. Peters (Peters et al. 2000). All other mutant animals and their

22 recommended controls used in this study were purchased from The Jackson

Laboratory. For HFD experiments animals were fed a high-fat diet containing 5.45 kcal/gm (F3282, Bio-Serv, Frenchtown, NJ) for 10 weeks.

Immunohistochemisty. Animals were perfused with PBS and then fixed in

10% neutral buffered formalin. Individual tissues were then processed by UCSD cancer center histology core, embedded in paraffin and stained with hematoxylin and eosin. For identification of apoa1 goat anit-mouse apoa1 primary antibody was used

1:200.

Apoa1 western blot and ELISA. Goat anti-mouse apoa1 was purchased from

Biodesign. For tissue blots 10ug of total protein was loaded onto a 12% SDS gel.

Primary antibody was used at 1:5000 and secondary (rabbit anti-goat-HRP, Santa

Cruz) at 1:10,000. For ELISA, serum was diluted 1:100 in binding buffer (100mM

Sodium Bicarbonate pH 9.6 + 1% BSA). 100ul of this dilution was then added to each well of a microtiter plate (maxisorb, Nunc). Serum protein was bound to the plate overnight at 4 degrees Celsius. Goat anti-mouse apoa1 was added at 1:1000 and incubated 2hrs room temperature with shaking. HRP conjugated secondary 1:10000.

HRP was developed with turbo TMB (Pierce) and stopped by addition of 1N HCL.

Signal was measure at 500nm and quatification was facilitated by comparison with a standard curve generated by serial dilution. Apoa1 -/- serum was used as negative background control.

Cholesterol and HDL assay. Serum total cholesterol was assayed in 96well microplates by enzymatic reaction (Infinity Cholesterol, Thermo Scientific) following the manufacture's instructions. HDL cholesterol was quantified after all non-HDL

23 lipoprotein particles were first precipitated with phosphotungstic acid (Precipitating

Reagent 1335-250; Thermo Electron).

WAT transplantation. Transplantation of epididymal adipose to donor mice was conducted essentially as previously described (Klebanov et al. 2005) with the following exceptions. An average of 1g of transplanted adipose was injected ventrally using a 16-gauge needle with ~100-150mg per injection site.

PPARδelta ligand GW501516. The PPARδelta ligand GW501516 was synthesized as described (ref.). GW501516 was first dissolved in DMSO at a concentration of 5mg/ml and then added daily to 0.5% carboxymethylcellulose

(Sigma) to a final concentration of 0.5mg/ml. Animals were gavaged daily with either

5mg/kg GW501516 or an equal amount of DMSO.

Gene Expression Analysis. Tissue RNAs were isolated by using TRIzol reagent (Invitrogen) followed by acid Phenol:Chloroform extraction (Ambion).

Contaminating DNA was first removed (TURBO DNA-free, Ambion) and then 1ug of

RNA was reverse-transcribed using oligo-dT (Superscript II, Invitrogen). Samples were run in technical triplicate using SYBR GreenER (Invitrogen) and normalized to

36B4 expression. WAT RNA was pooled from equal amounts of epididymal, ingunal, and perirenal adipose RNA.

Results

Serum Apoa1 protein is quickly induced by PPARδ ligand treatment

Although the ability of PPARδ ligands to raise HDLc levels in non-human primates and mice has been described previously (leibowitz 2000, oliver 2001, lee

2006) the mechanism responsible for this effect is not clear. To examine this phenomenon further we treated C57/Bl6 mice on a normal chow diet (NCD) and high- fat diet (HFD) daily with 5mg/kg GW501516 or DMSO and measured the change in serum cholesterol over time. We found that total cholesterol quickly reached maximal levels in 1-3 days and then leveled off in the GW treated group. The percent increase after 7 days was much more pronounced in HFD+GW animals (56%) than in the

NCD+GW group (15%) (fig.1a). Because of the short treatment duration these effects on cholesterol levels were independent of any effects on body weight, liver weight, or fat pad mass (data not shown). Analysis of total cholesterol at the end of the experiment revealed that only the HFD+GW group was significantly different from control (fig 1b). Precipitation of non-HDL particles and FPLC fractionation identified

(fig 1c and 1d). Interestingly there was a shift towards larger HDL particles in the

HFD+GW group reminiscent of human Apoa1 transgenic mice (Berthou et al. 1996).

We measured serum apolipoprotein AI (apoa1), the main protein component of HDL, and found a 4 fold increase in the HFD+GW group (fig. 1e).

Apoa1 expression is regulated in WAT by PPARδ

In order to find the gene expression changes responsible for this effect we examined previous microarray data of GW treated Db/Db animals. Surprisingly we found a marked upregulation of apoa1 mRNA in the WAT of GW treated animals.

This result was further confirmed by qPCR in an independent cohort of Db/Db

25 animals (fig 2a) and ApoE -/- animals (fig 2b). Importantly we did not find any change in the regulation of apoa1 in the liver or intestine of these animals by qPCR or array analysis. Apoa1 protein was easily detected in adipose tissue whole cell lysate by western blot and was increased ~2 fold by GW treatment (fig 2c). Unlike the mRNA expression which was hypervariable, apoa1 protein levels in WAT showed extremely low animal-to-animal variability.

Because the expression of apoa1 has never been described in WAT we examined apoa1 protein in this tissue by immunohistochemistry using liver as a positive control (fig 3c). Anit-apoa1 strongly stained the membrane of adipocytes (fig

3a). Importantly there was no staining of either liver or adipose from apoa1-/- animals

(fig 3b and 3d). This data shows that apoa1 protein is present in adipose tissue and may be localized to the membrane of adipocytes.

The effect of PPARδ ligand on serum cholesterol is blunted in AZIP animals and is rescued by fat transplantation

If the mechanism of action of PPARδ ligands on HDLc is mediated by adipose tissue we hypothesized that animals lacking adipose tissue would have a blunted response to GW. We used AZIP animals, a model of lypodystrophy shown previously to lack white adipose tissue (Moitra et al. 1998) for this purpose. As previously described (Colombo et al. 2003) WT mice had slightly lower cholesterol levels than

AZIP mice before GW treatment (WT = 99.6 ± 5 mg/dl, AZIP = 123 ± 6 mg/dl, p <

0.02). Treatment with GW caused a significantly higher increase in WT than in AZIP animals (fig 4a). Next we tested whether WAT transplantion would allow AZIP

26 animals to respond normally to PPARδ ligand. Indeed, transplantation of adipose tissue restored the response to PPARδ ligand to normal levels in AZIP mice (fig 4b).

Because PPARδ knockout animals do not respond to GW we transplanted WT adipose tissue into these animals. Again, transplantation of WT adipose tissue restored the response to GW in PPARδ knockout animals to WT levels (fig 4c.).

WAT transplantation can provide functional Apoa1 protein to Apoa1 -/- animals.

To conclusively determine whether adipose tissue is a source of functional apoa1 protein we transplanted WT adipose into apoa1 -/- animals (apoa1-TP). There was almost no detectable apoa1 protein in the serum of apoa1-TP animals two weeks after transplantation but apoa1 protein was readily detected around one month after transplant. The amount of protein in the serum declined rapidly three months after transplantation.

Discussion

In the last decade our understanding of adipose tissue function has shifted from that of a static energy depot to a dynamic endocrine organ. This concept is exemplified by the adipose specific hormone leptin (Zhang et al. 1994). Secretion of leptin allows adipose tissue to communicate levels of energy sufficiency to the central nervous system and modulate energy expenditure of peripheral tissues such as skeletal muscle and liver. Another such protein is adiponectin. Adipose specific expression of adiponectin ameliorates insulin resistance presumably through increased glucose utilization and fatty acid catabolism in other tissues (Kadowaki and Yamauchi 2005).

Adipocytes also have the ability to synthesize and secrete a number of cytokines; small peptides normally associated with cells of the immune system. Most notably adipose tissue and isolated adipocytes secrete TNFα which can then affect whole body energy homeostasis (Hotamisligil et al. 1993). Adipose tissue also secretes physiological amounts of IL6 and IL-1β (Fried et al. 1998).

We have found yet another level of plasticity in the transcriptional ability of adipose tissue. Activation of PPARδ in adipose tissue caused a dramatic increase in

Apoa1, ApoB, and ApoCIII mRNA levels. Previously the expression of these genes was thought to be completely restricted to either liver or intestine. Indeed, we found many other genes with a predominantly liver specific expression pattern to be induced by PPARδ ligand in adipose tissue (data not shown). The mechanism responsible for phenomenon is currently unknown but may involve induction of a potent, liver specific transcription factor such as HNF4α.

The contribution of adipose derived apoa1 in human serum remains to be examined. It is tempting to speculate that secretion from adipose is dysregulated in obese or insulin resistant states leading to lower levels of serum HDLc. In light of recent clinical data which saw a very modest increase in HDLc levels in healthy volunteers (Sprecher et al. 2007) our data suggests that PPARδ ligands might be more effective in the context of obesity or metabolic syndrome.

Acknowledgments

All work done in this chapter including analysis and figures were created in close collaboration with Dr. Chih-Hao Lee.

Figure II.1. GW501516 rapidly increases serum HDL cholesterol and apolipoprotein AI. (A-C) Percent increase in total cholesterol (A), total cholesterol (B), and HDL cholesterol (C) after 7 days of DMSO or 5mg/kg/day GW501516 (N = 4-5). (D) FPLC fractionation of serum cholesterol. (E) Serum apolipoprotein AI measured by ELISA. *, p < 0.05; **, p < 0.01. For percent increase student's paired t- test was used. In all other comparisons an unpaired t-test was used.

Figure II.2. Apoa1 mRNA and protein are upregulated in white adipose tissue by PPARδ ligand treatment. RNA from WAT of Db/Db (A) or ApoE-/- (B) animals treated for 7 days with control (DMSO) or the PPARδ ligand GW1516 analyzed by qPCR. (C) Protein extracts from animals on normal diet (NCD) or high fat diet (HFD) were run on a 12% SDS page gel and blotted with anti-apoa1. + indicates GW treatment. *, p < 0.05.

Figure II.3. Apoa1 can be detected in adipose tissue by immunohistochemistry. Tissues from WT (A and C) or apoa1 -/- (B and D) were fixed in 10% neutral buffered formalin, embedded in paraffin, sectioned and stained for apoa1 protein. Note the strong signal in WT (A) but not Apoa null (B) adipose tissue. WT (C) and apoa1 -/- (D) liver sections were used as positive and negative controls respectively.

Figure II.4. WAT transplantation can rescue the effect of PPARδ ligand in AZIP and PBN animals. (A) WT or AZIP animals were treated for 7 days with GW501516 at 5mg/kg/day. (B) AZIP animals were either sham operated or transplanted with WT adipose. After one month animals were treated for 7 days with GW501516 at 5mg/kg/day. (C) PPARδ null animals were either transplanted with WT adipose (PBN-WT) or adipose from PPARδ null animals (PBN-Null). After one month animals were treated for 7 days with GW501516 at 5mg/kg/day. WT animals treated with DMSO (WT-DMSO) or GW501516 (WT-GW) were included as negative and positive controls respectively. In all experiments serum cholesterol was assayed before and after drug treatment. Percent increase in cholesterol was calculated by (final-initial)/(initial) * 100.

Figure II.5. Apoa1 protein can be detected in apoa1 -/- animals with adipose transplant. Apoa1 -/- animals were transplanted with ~1g of WT adipose tissue. Serum was collected at the indicated timpoints and apoa1 protein was measured by western blotting. NS indicates a non-specific band.

References cited

Badimon, J. J., et al. (1990). "Regression of atherosclerotic lesions by high density lipoprotein plasma fraction in the cholesterol-fed rabbit." J Clin Invest 85(4): 1234-41.

Barak, Y., et al. (2002). "Effects of peroxisome proliferator-activated receptor delta on placentation, adiposity, and colorectal cancer." Proc Natl Acad Sci U S A 99(1): 303-8.

Berthou, L., et al. (1996). "Opposite regulation of human versus mouse apolipoprotein A-I by fibrates in human apolipoprotein A-I transgenic mice." J Clin Invest 97(11): 2408-16.

Colombo, C., et al. (2003). "Opposite effects of background genotype on muscle and liver insulin sensitivity of lipoatrophic mice. Role of triglyceride clearance." J Biol Chem 278(6): 3992-9.

Ford, E. S., et al. (2002). "Prevalence of the metabolic syndrome among US adults: findings from the third National Health and Nutrition Examination Survey." Jama 287(3): 356-9.

Fried, S. K., et al. (1998). "Omental and subcutaneous adipose tissues of obese subjects release interleukin-6: depot difference and regulation by glucocorticoid." J Clin Endocrinol Metab 83(3): 847-50.

Grundy, S. M., et al. (2004). "Definition of metabolic syndrome: Report of the National Heart, Lung, and Blood Institute/American Heart Association conference on scientific issues related to definition." Circulation 109(3): 433-8.

Hotamisligil, G. S., et al. (1993). "Adipose expression of tumor necrosis factor-alpha: direct role in obesity-linked insulin resistance." Science 259(5091): 87-91.

Kadowaki, T. and T. Yamauchi (2005). "Adiponectin and adiponectin receptors." Endocr Rev 26(3): 439-51.

Klebanov, S., et al. (2005). "Adipose tissue transplantation protects ob/ob mice from obesity, normalizes insulin sensitivity and restores fertility." J Endocrinol 186(1): 203-11.

Lakka, H. M., et al. (2002). "The metabolic syndrome and total and cardiovascular disease mortality in middle-aged men." Jama 288(21): 2709-16.

McNeish, J., et al. (2000). "High density lipoprotein deficiency and foam cell accumulation in mice with targeted disruption of ATP-binding cassette transporter-1." Proc Natl Acad Sci U S A 97(8): 4245-50.

Miyazaki, A., et al. (1995). "Intravenous injection of rabbit apolipoprotein A-I inhibits the progression of atherosclerosis in cholesterol-fed rabbits." Arterioscler Thromb Vasc Biol 15(11): 1882-8.

Moitra, J., et al. (1998). "Life without white fat: a transgenic mouse." Genes Dev 12(20): 3168-81.

Mokdad, A. H., et al. (2004). "Actual causes of death in the United States, 2000." Jama 291(10): 1238-45.

Oliver, W. R., Jr., et al. (2001). "A selective peroxisome proliferator-activated receptor delta agonist promotes reverse cholesterol transport." Proc Natl Acad Sci U S A 98(9): 5306-11.

Paszty, C., et al. (1994). "Apolipoprotein AI transgene corrects apolipoprotein E deficiency-induced atherosclerosis in mice." J Clin Invest 94(2): 899-903.

Rader, D. J. (2003). "Regulation of reverse cholesterol transport and clinical implications." Am J Cardiol 92(4A): 42J-49J.

Robins, S. J., et al. (2001). "Relation of gemfibrozil treatment and lipid levels with major coronary events: VA-HIT: a randomized controlled trial." Jama 285(12): 1585-91.

Rubin, E. M., et al. (1991). "Inhibition of early atherogenesis in transgenic mice by human apolipoprotein AI." Nature 353(6341): 265-7.

Williamson, R., et al. (1992). "Marked reduction of high density lipoprotein cholesterol in mice genetically modified to lack apolipoprotein A-I." Proc Natl Acad Sci U S A 89(15): 7134-8.

Zhang, Y., et al. (1994). "Positional cloning of the mouse obese gene and its human homologue." Nature 372(6505): 425-32.

III. The role of PPARα and PPARδ in caloric restriction

36 37

Abstract

Despite extensive research into pharmacological manipulation of weight gain caloric restriction remains the most effective method for treating obesity. Here we show that the nuclear receptor PPARα plays a central role in caloric restriction mediated weight loss. Both loss of the receptor in null animals and activation by high affinity ligand resulted in decreased weight loss during caloric restriction. In PPARα null animals, decreased weight loss was associated with abnormalities of the hypothalamic-pituitary-thyroid axis. Activation of PPARα, and to a lesser extent

PPARδ, during caloric restriction resulted in an increase of the fatty acid synthesis pathway and liver triglyceride content. Our data demonstrate an unexpected role for

PPARα in caloric restriction and the regulation of liver gene expression.

Introduction

The metabolic syndrome (MS) is a cluster of disorders which subsequently results in a greatly increased risk for cardiovascular disease (Mokdad et al. 2004)

(Flegal et al. 2005; Gregg et al. 2005). A central and possibly causative component to

MS is obesity (Kahn and Flier 2000). This is illustrated by the fact that only around

5% of normal weight individuals have MS whereas 50% of obese are diagnosed with

MS. Consequently, weight loss through caloric restriction (Kastelein et al.) is extremely effective in treating almost every facet of MS (Wadden 1993). In primates and humans CR is associated with decreases in serum insulin, TNF (Kern et al.

1995),adiponectin (Xydakis et al. 2004) and of course obesity (Fontana et al. 2004)

(Heilbronn et al. 2006). CR has also been demonstrated to ameliorate insulin

38 resistance (Ramsey et al. 2000), cancer and hypertension (Fontana, Meyer et al. 2004).

Furthermore, CR has been shown to increase lifespan in yeast (Lin et al. 2002), worms, flies, rodents (McCay et al. 1989) and nonhuman primates (Hansen et al.

1999). Transcriptional profiling has identified numerous gene expression changes that occur with aging and CR has been demonstrated to reverse many of them (Dhahbi et al. 2004). However, the transcription factors and regulatory networks involved have been poorly defined (Koubova and Guarente 2003).

Although homeostatic mechanisms exist to maintain weight in a fairly narrow range, the global obesity epidemic seems to be proof positive that weight gain is the stronger urge. It is tempting to speculate that weight gain is actually the "default" state of human physiology. This makes sense evolutionarily since caloric availability would always be a limiting resource and metabolic efficiency would therefore be a selective advantage. In this situation pressure to maintain "thrifty genes" in a population would be high. The genetic contribution to MS can be seen most readily in ethnic populations where food has shifted rapidly from being a scarce resource to an overabundant resource (Diamond 2003). Weight loss, on the other hand, is extremely difficult due in part to a compensatory decrease in basal metabolic rate (MacLean).

The identification of genes involved in these processes would ultimately help to better understand and treat MS.

The peroxisome proliferator-activator receptor (PPAR) sub-family of nuclear receptors are critical regulators of metabolic homeostasis. Whereas PPARγ plays a central role in energy storage and adipogenesis, PPARα and PPARδ regulate the dissipation of energy through the beta-oxidation pathway. PPARα in particular has

39 been shown to be required for the proper physiological response to fasting (Kersten et al. 1999). Because of their central role in energy homeostasis we investigated the role of PPARα and PPARδ in the response to CR.

Methods

Metabolic Cages. Metabolic cages were from Columbus Instruments. Mice were measured in 3 batches of 8 animals for 2 consecutive days each. Oxygen consumption, CO2 production, total movement and ambulatory movement was measured for 22 hours starting at 6pm. Average respiratory exchange ratio (RER) was calculated as the total volume of CO2 produced divided by the total volume of oxygen consumed.

Thyroid Hormone and Insulin Levels. Insulin was measured by using an

ELISA kit (Linco, St. Charles, MO). Levels of T3, T4, and TSH were quantified by custom ELISA assay.

Serum Cholesterol, Triglycerides and Glucose. Animals were euthanized in the fasted state and whole blood was obtained by heart punch followed by coagulation to purify serum. Total cholesterol (TC) and triglycerides (TG) were determined by using enzymatic reactions (Thermo DMA, Arlington, TX). Glucose was measured in either whole blood through the tail tip or in purified serum using the OneTouch

(Lifescan, Milpitas, CA) glucose-monitoring system.

Gene Expression. Tissue RNAs were isolated by using TRIzol reagent

(Invitrogen) followed by acid Phenol:Chloroform extraction (Ambion).

Contaminating DNA was first removed (TURBO DNA-free, Ambion) and then 1ug of

RNA was reverse-transcribed using oligo-dT (Superscript II, Invitrogen). Samples were run in technical triplicate using SYBR GreenER (Invitrogen) and normalized to

36B4 expression.

Animal usage and Caloric restriction. PPARα null animals on a C57BL6 background, C57BL6 control animals, and Db/Db mice where all purchased from The

Jackson Laboratories. To begin caloric restriction animals were switched to precision pellet diet (Bioserv) and food consumption was monitored for 15 days. Average weekly food consumption was calculated and mice were given 60% of this value.

Animals were fed daily at the beginning of the dark cycle (1800 hours).

Results

PPARα knockout animals have decreased weight loss during CR

Because PPARα is essential for the acute adaptation to fasting (Kersten,

Seydoux et al. 1999) we wanted to study the role of this nuclear receptor in the context of caloric restriction (CR). To see if loss of PPARα would affect weight loss during

CR we compared PPARα knockout mice (aNull) to wild-type (Blanc et al.) animals on a C57Bl6 background. Initially we measured the basal food consumption at three points over 15 days. At first WT animals ate significantly less than aNull (15.9g vs.

20.3g, p<0.05). At the second time point consumption was equal (WT = 18.8g, aNull

= 18.3g, p<0.8). In the final time point WT animals consumed significantly more than aNull (20.7g vs. 17.8g, p<0.05). However, average consumption per day over the entire 15 days was not significantly different between genotypes (WT = 3.69g/day, aNull = 3.71g/day, p = 0.9). This data highlights the variability that can be introduced

41 by switching diets. We then fed animals 60% of the 15 day average consumption for three weeks and tracked the change in percent weight loss over time. Within the first

7 days we found a difference in both absolute weight loss (WT = -6.53g ± 0.45, aNull

= -4.07g ± 0.25, p <0.001) and percentage weight loss (WT = 20.0% ± 1.5, aNull =

13.0% ± 1.1, p < 0.01) between WT and aNull animals (figure 1). This difference in weight was maintained but did not increase for the rest of the experiment. Overall starting weight was not predictive of percent weight loss (R2 = 0.07, f = 0.83) however starting weight was correlated to the amount of weight loss in aNull animals

(R2 = 0.31, f = 1.80).

Levels of thyroid hormone are indicative of basal metabolism and are known to be reduced during CR (Weinsier et al. 2001). To see if changes in thyroid hormone levels were responsible for the decreased weight loss in aNull animals we measured

T3, T4, and TSH levels in this cohort. Although there was no difference between genotypes before CR the level of T4 in aNull animals was 25% greater after CR.

Furthermore, WT animals had the expected reduction in T4 levels after CR but aNull animals failed to reduce circulating levels of T4 (Fig 2a). Also, circulating T3, the most active form of thyroid hormone, was 53% greater in aNull animals (Fig 2b).

Other differences included a 2.7 fold decrease in TSH in the aNull animals (Fig 2c).

This data provides evidence that aNull animals on CR are relatively hyperthyroid when compared to control animals. Since increased levels of thyroid hormone should equate to increased metabolic rate, the hypothalamic-pituitary-thyroid axis is misregulated in aNull animals.

We found no difference in normalized liver weight (WT = 0.043 ± 0.002, aNull

= 0.043 ± 0.001, p = 0.88). However we did find a striking difference in the size of the seminal vesicles with WT being almost 6 fold larger. Also, WT animals were completely devoid of epididymal adipose (Fig 3a and 3b).

Next we undertook gene expression profiling of liver samples. Since thyroid hormone levels differed greatly by genotype we looked at a number of thyroid hormone conjugating enzymes and members of the cytochrome p450 family in the liver known to be important for the conversion of T4 to the more potent T3 (Maglich et al. 2004). As expected PPARα and its canonical target genes CYP4a10 and AOX were reduced in the livers of aNull animals. Although there were also significant changes in several conjugating enzymes no coordinated pathway was regulated (figure

4).

PPARα ligand treatment and CR also results in less weight loss

Since the absence of PPARα caused animals to lose less weight on CR we wanted to test the effect of PPARα activation in the context of CR. We used Db/Db animals as a model of obesity and treatment with either the high affinity PPARα ligand WY14643 (Wallace et al.) or vehicle (CON) along with CR for 3 weeks.

Unexpectedly WY treated animals actually lost less weight than controls after three weeks of CR (fig. 5). This difference was evident and significant within 7 days and continued to increase throughout the experiment.

One of the most potent effects of PPARα ligands is a decrease in circulating serum triglycerides. As expected we observed a 3.4 fold decrease in serum

43 triglyceride levels in the WY group indicating that the PPARα pathway was activated by drug treatement. We also observed a 23% decrease in serum glucose (CON = 1595

± 118, WY = 1226 ± 113, p < 0.05), and a 33% increase in serum cholesterol (CON =

113 ± 7.3mg/dl, WY = 150 ± 11.7mg/dl, p < 0.03). The livers of WY treated animals were almost 2 fold larger (CON = 2.11 ± 0.11g, WY = 4.00 ± 0.07g, p < 1.5E-09).

The large increase in liver size was expected since treatment with PPARα ligands is known to cause hepatomegaly in rodents, however visually WY treated livers were markedly paler in color a hallmark of triglyceride accumulation. Surprisingly, quantification revealed a 2.7 fold increase in the triglyceride content of WY treated livers (Table 1).

To test the specificity of this effect we repeated the experiment and included a

PPARδ ligand GW501516. Db/Db animals were given the indicated ligand and restricted to 60% caloric intake for 4 weeks. Treatment with WY gave essentially the same result with significantly less weight loss than DMSO treated animals. Although

GW treated animals tended to lose less weight over the course of the experiment they were never significantly different from control animals at any of the time points.

Because WY+CR caused a decrease in fasting glucose in our previous experiment we decided to look at the effect of CR and PPAR ligands on glucose homeostasis in more detail. Fasting glucose levels were measured at the start of CR and every week afterwards. There was an increase in mean fasting glucose during the course of the experiment in DMSO (+120.7 ± 45.6 mg/dl, p < 0.04) and GW (+112.7

± 30.2 mg/dl, p < 0.01) treatment but WY caused a decrease in fasting glucose (-132.4

± 67.7, p < 0.08) (figure 7A). There was also a significant effect on fasting insulin

44 levels with a 3.4 fold increase in WY treatment (Fig. 7B). This data suggests that although WY+CR animals have increased serum insulin they are more sensitive and therefore have lower fasting glucose levels.

Metabolic cage analysis

After two weeks of CR mice were placed into metabolic cages to measure oxygen consumption, CO2 production and activity. This time point was chosen as there was no appreciable difference in overall weight between groups (con=37.8g, gw=37.9g, wy=38.2g). Using linear mixed-effects models (r-project.org) to control for experimental batch, cage and day effects, there was no effect of either PPARδ or

PPARα ligand on total oxygen consumption, total carbon dioxide production or activity. WY treatment was suggestive of a slight increase in RER (+0.03, p < 0.08) when oxygen consumption was factored into the model as a covariate.

CR with PPARα ligand upregulates the fatty-acid synthesis pathway

Gene expression analysis confirmed that the PPARα pathway was induced by

WY treatment. Liver fatty-acid binding protein, acetyl-CoA oxidase and lipoprotein lipase were all robustly induced in the livers of WY treated animals. Surprisingly though almost the entire suite of fatty acid synthesis (FAS) genes were also upregulated consistent with the phenotypic increase in liver triglycerides.

Interestingly GW treatment did not activate PPARα target genes but did cause an induction of the FAS pathway. GW treatment caused an intermediate upregulation of

45 genes such as fatty-acid synthase, malic enzyme, and acetyl-CoA carboxylase which correlated with the amount of liver triglyceride found in this group.

Conclusions

A surprising result of this study was the induction of fatty acid synthesis by

PPARα ligand treatment. The PPARα dependent control of individual genes in the fatty acid synthesis pathway has been described previously. In fact malic enzyme

(Mod1), 3-hydroxy-3-methylglutaryl-CoA synthase (Hmgcs), long-chain acyl- coenzyme A synthetase (Acsl1), and stearoyl-CoA desaturase (SCD1) were among the first genes to have peroxisome proliferator response elements (PPREs) identified in their promoters (Castelein et al. 1994; Rodriguez et al. 1994; Schoonjans et al. 1995;

Miller and Ntambi 1996). However this is the first report of PPARα ligand treatment causing an increase in the entire fatty acid synthesis pathway and resulting in liver triglyceride accumulation. From this data it is impossible to determine whether this was a liver specific effect or due to PPARα activation in other metabolic tissues such as cardiac muscle or brown adipose tissue. The use of a liver specific PPARα knockout would address this question. Another caveat is that these studies were done in animals lacking the leptin receptor and therefore the phenotypic changes might be dependent on aberrant leptin signalling. It would be interesting to see if the same effect is seen in caloric restriction after diet induced obesity.

Sterol response element binding protein (SREBP) is one of the key regulators of fatty acid and cholesterol synthesis (Brown and Goldstein 1997). The SREBP pathway is known to be regulated at both the transcriptional and post-transcriptional

46 level. Another nuclear receptor, liver X receptor (LXR) is known to activate the transcription of SREBP and lead to fatty acid synthesis in the liver (Repa et al. 2000).

We do not believe that LXR is responsible for this phenotype since there is no increase in SREBP mRNA or two other canonical LXR targets ABCG5 and ABCG8 (data not shown) (Repa et al. 2002). Recently another mechanism of SREBP regulation has been described which involves the co-activator PGC1β (Lin et al. 2005). It is possible that in the context of caloric restriction a PPARα → PGC1β → SREBP pathway exists to regulate fatty acid synthesis.

Metabolic rate as measured by oxygen consumption and CO2 production and its link to CR is controversial. Initially, CR was thought to decrease energy expenditure (EE) but subsequent studies have had conflicting results (Ramsey,

Colman et al. 2000). In lower organisms CR actually increased respiration in yeast

(Lin, Kaeberlein et al. 2002) and caused no change in flies (Hulbert et al. 2004) or C. elegans (Houthoofd et al. 2002). In our study there was no clear link between metabolic rate and changes in weight loss. This highlights the fact that very small changes in oxygen consumption or respiratory exchange ratio can result in significant changes in energy balance. Designing experiments with enough power to detect these changes is a major challenge for the field of energy metabolism in general. Recently the use of linear regression to model metabolism has gained popularity. In the current study we also employed linear mixed models but were unable to find a model which significantly predicted final weight loss. Alternatively, there could exist an as yet unidentified parameter which predicts weight loss much better than indirect calorimetry.

Acknowledgments

All of the work done in this chapter including analysis and figures were created in close collaboration with Dr. Junichiro Sonoda.

Figure III.1. PPARα null animals have decreased weight loss during CR. Wild- type (Blanc, Schoeller et al.) or PPARα knockout (PPARα -/-) animals were subjected to 60% of their normal caloric intake. Percent weight loss is given as weight lost divided by initial weight. Data is presented as the average of 6 or 7 animals ± SEM. Significance at each time point was evaluated by Student’s T-test where **, p < 0.01.

Figure III.2: Thyroid hormone levels are dysregulated in PPARα null animals on CR. Serum was collected before and after caloric restriction. Levels of T4 (A), T3 (B), and TSH (C) were measured by ELISA as described in materials and methods. The change in T4 levels before and after CR was evaluated by paired Student’s T-test all other comparisons are unpaired. *, p < 0.05; **, p < 0.01.

Figure III.3: Seminal vesicles are greatly increased in WT animals on CR. PPARα null (A) and WT (B) animals were dissected after 3 weeks of caloric restriction. Red arrows indicate seminal vesicles and white arrows epididymal adipose. Note the complete lack of adipose in WT animals.

Figure III.4. Changes in liver gene expression of PPARα -/- animals on CR. Liver RNA was prepared as described in materials and methods. Bars represent the average of 6 or 7 animals ± SEM. *, p < 0.05. PXR, pregnane X receptor; CAR, Constitutive androstane receptor; Ugt, UDP glucuronosyltransferase; Sult, sulfotransferase family; GST, glutathione S-transferase; Cyp, cytochrome P450 family; AOX-1, acyl-coenzyme A oxidase.

Figure III.5. Animals treated with PPARα ligand lose less weight after CR. Db/Db animals were treated with vehicle (DMSO) or WY14643 (Wallace, Schwarz et al.) while being fed 60% of their normal caloric intake. Animals were weighed every week for 3 weeks. Percent weight loss is given as weight lost divided by initial weight. Data is presented as the average of 7 or 8 animals ± SEM. Significance at each time point was evaluated by Student’s T-test where *, p < 0.05; **, p < 0.01.

Table 1. Metabolic consequences of CR with PPARα and PPARδ Ligands.

DMSO WY p.value TG (mg/dl) 272.6 ± 35.9 79.0 ± 15.8 p < 0.0002 TC (mg/dl) 113.0 ± 7.3 149.9 ± 11.7 p < 0.03 Glucose (AU) 1595 ± 118 1226 ± 113 p < 0.05

Epididymal WAT (g) 1.38 ± 0.18 1.47 ± 0.13 p < 0.07 Liver weight (g) 2.11 ± 0.11 4.00 ± 0.07 p < 1.5E-09 Liver TG (AU) 0.69 ± 0.12 1.88 ± 0.10 p < 6E-06

Figure III.6. The effect of WY treatment on weight loss is PPARα specific. Db/Db animals were treated with vehicle (DMSO), GW501516 (Tugwood et al.) or WY14643 (Wallace, Schwarz et al.) while being fed 60% of their normal caloric intake. Animals were weighed every week for 4 weeks. Percent weight loss is given as weight lost divided by initial weight. Data is presented as the average of 7 animals ± SEM. Significance at each time point was evaluated by Student’s T-test where #, p < 0.1; *, p < 0.05.

A B 6 * DMSO GW WY

* #

# * ** 2 Serum insulin (ng/dl) insulin Serum **

Figure III.7. PPARα ligand treatment causes decreased fasting glucose but increased serum insulin when coupled with CR. Fasting glucose (A) and serum insulin (B) were measured in vehicle (DMSO), PPARδ ligand (Tugwood, Issemann et al.), and PPARα ligand (Wallace, Schwarz et al.) treated animals subjected to CR. #, p < 0.1; *, p < 0.05; **, p < 0.01. Note; markers of significance are located below points.

Figure III.8. Activation of PPARα increases fatty acid synthesis. Liver RNA was prepared as described in materials and methods. Expression of known PPARα target genes (A) or fatty-acid synthesis pathway (B) were quantified. All values are normalized by 36B4. Bars represent the average of 7 animals ± SEM. *, p < 0.05; **, p < 0.01; ***, p < 0.001. Fabp1, liver fatty acid binding protein; Acox, acyl-CoA oxidase; Lpl, lipoprotein lipase; Acadm, medium-chain acyl-CoA dehydrogenase; Pck1, phosphoenolpyruvate carboxykinase 1; Apoa1, apolipoprotein AI; Fdft1, farnesyl diphosphate farnesyl transferase 1; Fasn, fatty acid synthase; Mod1, malic enzyme; Ldlr, low density lipoprotein receptor; Hmgcr, 3-hydroxy-3-methylglutaryl- Coenzyme A reductase; Hmgcs1, 3-hydroxy-3-methylglutaryl-Coenzyme A synthetase; Acaca, acetyl-CoA carboxylase alpha; Gpam, glycerol-3-phosphate acyltransferase, mitochondrial; Scd1, stearoyl-CoA desaturase 1; Insig2, insulin induced gene 2.

References cited

Barak, Y., et al. (2002). "Effects of peroxisome proliferator-activated receptor delta on placentation, adiposity, and colorectal cancer." Proc Natl Acad Sci U S A 99(1): 303-8.

Blanc, S., et al. (2003). "Energy expenditure of rhesus monkeys subjected to 11 years of dietary restriction." J Clin Endocrinol Metab 88(1): 16-23.

Brown, M. S. and J. L. Goldstein (1997). "The SREBP pathway: regulation of cholesterol metabolism by proteolysis of a membrane-bound transcription factor." Cell 89(3): 331-40.

Castelein, H., et al. (1994). "The peroxisome proliferator activated receptor regulates malic enzyme gene expression." J Biol Chem 269(43): 26754-8.

Dhahbi, J. M., et al. (2004). "Temporal linkage between the phenotypic and genomic responses to caloric restriction." Proc Natl Acad Sci U S A 101(15): 5524-9.

Diamond, J. (2003). "The double puzzle of diabetes." Nature 423(6940): 599-602.

Flegal, K. M., et al. (2005). "Excess deaths associated with underweight, overweight, and obesity." Jama 293(15): 1861-7.

Fontana, L., et al. (2004). "Long-term calorie restriction is highly effective in reducing the risk for atherosclerosis in humans." Proc Natl Acad Sci U S A 101(17): 6659-63.

Gregg, E. W., et al. (2005). "Secular trends in cardiovascular disease risk factors according to body mass index in US adults." Jama 293(15): 1868-74.

Hansen, B. C., et al. (1999). "Calorie restriction in nonhuman primates: mechanisms of reduced morbidity and mortality." Toxicol Sci 52(2 Suppl): 56-60.

Heilbronn, L. K., et al. (2006). "Effect of 6-month calorie restriction on biomarkers of longevity, metabolic adaptation, and oxidative stress in overweight individuals: a randomized controlled trial." Jama 295(13): 1539-48.

Houthoofd, K., et al. (2002). "No reduction of metabolic rate in food restricted Caenorhabditis elegans." Exp Gerontol 37(12): 1359-69.

Hulbert, A. J., et al. (2004). "Metabolic rate is not reduced by dietary-restriction or by lowered insulin/IGF-1 signalling and is not correlated with individual lifespan in Drosophila melanogaster." Exp Gerontol 39(8): 1137-43.

Kahn, B. B. and J. S. Flier (2000). "Obesity and insulin resistance." J Clin Invest 106(4): 473-81.

Kastelein, J. J., et al. (2000). "Comparison of effects of simvastatin versus atorvastatin on high-density lipoprotein cholesterol and apolipoprotein A-I levels." Am J Cardiol 86(2): 221-3.

Kern, P. A., et al. (1995). "The expression of tumor necrosis factor in human adipose tissue. Regulation by obesity, weight loss, and relationship to lipoprotein lipase." J Clin Invest 95(5): 2111-9.

Kersten, S., et al. (1999). "Peroxisome proliferator-activated receptor alpha mediates the adaptive response to fasting." J Clin Invest 103(11): 1489-98.

Koubova, J. and L. Guarente (2003). "How does calorie restriction work?" Genes Dev 17(3): 313-21.

Lin, J., et al. (2005). "Hyperlipidemic effects of dietary saturated fats mediated through PGC-1beta coactivation of SREBP." Cell 120(2): 261-73.

Lin, S. J., et al. (2002). "Calorie restriction extends Saccharomyces cerevisiae lifespan by increasing respiration." Nature 418(6895): 344-8.

MacLean, P. S. (2005). "A peripheral perspective of weight regain." Am J Physiol Regul Integr Comp Physiol 288(6): R1447-9.

Maglich, J. M., et al. (2004). "The nuclear receptor CAR is a regulator of thyroid hormone metabolism during caloric restriction." J Biol Chem 279(19): 19832- 8.

McCay, C. M., et al. (1989). "The effect of retarded growth upon the length of life span and upon the ultimate body size. 1935." Nutrition 5(3): 155-71; discussion 172.

Miller, C. W. and J. M. Ntambi (1996). "Peroxisome proliferators induce mouse liver stearoyl-CoA desaturase 1 gene expression." Proc Natl Acad Sci U S A 93(18): 9443-8.

Mokdad, A. H., et al. (2004). "Actual causes of death in the United States, 2000." Jama 291(10): 1238-45.

Ramsey, J. J., et al. (2000). "Dietary restriction and aging in rhesus monkeys: the University of Wisconsin study." Exp Gerontol 35(9-10): 1131-49.

Repa, J. J., et al. (2002). "Regulation of ATP-binding cassette sterol transporters ABCG5 and ABCG8 by the liver X receptors alpha and beta." J Biol Chem 277(21): 18793-800.

Repa, J. J., et al. (2000). "Regulation of mouse sterol regulatory element-binding protein-1c gene (SREBP-1c) by oxysterol receptors, LXRalpha and LXRbeta." Genes Dev 14(22): 2819-30.

Rodriguez, J. C., et al. (1994). "Peroxisome proliferator-activated receptor mediates induction of the mitochondrial 3-hydroxy-3-methylglutaryl-CoA synthase gene by fatty acids." J Biol Chem 269(29): 18767-72.

Schoonjans, K., et al. (1995). "Induction of the acyl-coenzyme A synthetase gene by fibrates and fatty acids is mediated by a peroxisome proliferator response element in the C promoter." J Biol Chem 270(33): 19269-76.

Tugwood, J. D., et al. (1992). "The mouse peroxisome proliferator activated receptor recognizes a response element in the 5' flanking sequence of the rat acyl CoA oxidase gene." Embo J 11(2): 433-9.

Wadden, T. A. (1993). "Treatment of obesity by moderate and severe caloric restriction. Results of clinical research trials." Ann Intern Med 119(7 Pt 2): 688-93.

Wallace, J. M., et al. (2005). "Effects of peroxisome proliferator-activated receptor alpha/delta agonists on HDL-cholesterol in vervet monkeys." J Lipid Res 46(5): 1009-16.

Weinsier, R., et al. (2001). "Metabolic response to weight loss." Am J Clin Nutr 73(3): 655-8.

Xydakis, A. M., et al. (2004). "Adiponectin, inflammation, and the expression of the metabolic syndrome in obese individuals: the impact of rapid weight loss through caloric restriction." J Clin Endocrinol Metab 89(6): 2697-703.

IV. Conclusions

60 61

Over the past twenty-five years the PPARs have firmly established themselves as master regulators of metabolic homeostasis. While PPARα is the founding member of the family new roles for this receptor are still being discovered. The function of

PPARδ, on the other hand, is just beginning to be elucidated. Whether this receptor will become just as important as PPARα and PPARγ clinically for treating components of the metabolic syndrome has yet to be seen.

PPARα and caloric restriction

Americans spend over 30 billion dollars annually on weight loss related services and more than two thirds of Americans reported that they are actively trying to maintain or lose weight (Serdula et al. 1999). Despite this the rate of obesity in the

United States and around the world continues to rise. Oddly though, the genes responsible for augmenting weight loss are largely unknown.

We have identified PPARα as a critical component of CR mediated weight loss. From this study there are two main interesting results which require further examination. Firstly, PPARα null animals lose less weight during CR despite having higher thyroid hormone levels. To our knowledge this is the only model where CR did not result in decreased thyroid hormone levels suggesting that the hypothalamic- pituitary-thyroid axis is disrupted in these animals. A detailed analysis of metabolic parameters including body temperature would determine if energy expenditure is actually decreased in the PPARα null animals. Secondly, PPARα ligand treatment along with CR results in decreased weight loss and fatty liver. This result is extremely surprising given that PPARα is known to activate fatty acid beta oxidation. Indeed,

62 the traditional PPARα targets acyl-CoA oxidase (AOX) and lipoprotein lipase (LPL) were still induced in the livers of WY treated animals. It would seem that coordinated activation of fatty acid synthesis and fatty acid activation would create a futile cycle of energy usage but the phenotypic inhibition of weight loss we observed argues against this hypothesis. An alternative explanation is that PPARα ligand treatment coupled with CR causes a fundamental shift in substrate usage whereby glucose is converted to fatty acids which are then efficiently converted to energy by the beta oxidation pathway.

Regulation of apoa1 in adipose tissue by PPARδ

Adipose is actually an extremely heterologous tissue made up of several cell types. These include mature adipocytes (MA), preadipocytes, endothelial cells, and cells of the nervous system. Even within the MA population there is considerable heterogeneity in cell size. Recent research has discovered that differently sized adipocytes can have vastly different funtions with adipocyte hypertrophy being associated with insulin resistance.

Although we were easily able to detect apoa1 mRNA and protein levels in intact adipose tissue this signal completely disappeared after separation into mature adipocytes and a stromal vascular fraction (SVF). Because of this we were never able to definitively identify the cell type which expressed apoa1. However dissociation of adipose tissue by collagenase treatment is known to rapidly change the expression profile of adipocytes (Ruan et al. 2003). Even a brief collagenase treatment quickly induces the expression of inflammatory cytokines and downregulates adipose specific

63 genes (Mehl unpublished data). Because we observe strong staining of apoa1 in mature adipoctyes by immunohistochemistry we believe this to be the main site of apoa1 expression.

The role of adipose derived apoa1 protein in normal human physiology is of some interest. Several physiological conditions are correlated with either increased or decreased levels of HDLc. Obesity and insulin resistance in particular are correlated with decreased levels of HDLc (Denke et al. 1993; Wakabayashi 2004). It is tempting to speculate that in the obese or insulin resistant state adipose production of apoa1 is inhibited. Paradoxically a diet high in saturated fat raises HDLc (Brinton et al. 1990;

Meksawan et al. 2004). It is possible that this diet activates PPARδ in adipose tissue to raise HDLc. Exercise (King et al. 1995; Couillard et al. 2001) and alcohol consumption (Gaziano et al. 1993; Ellison et al. 2004) have a positive affect on HDLc.

Additionally, niacin is known to substantially increase HDLc (Alderman et al. 1989;

Kashyap 1998) the mechanism of which is not known. Future studies will determine whether or not PPARδ modulates the effect of these environmental factors by adipose specific expression of apoa1.

Recently an inhibitor of the cholesterol ester transfer protein (CETP) was shown to be effective in raising HDLc levels in humans (Brousseau et al. 2004).

Unexpectedly this drug, torcetrapib, was pulled from a massive stage III clinical trial when it was discovered that there was a small but significant increase in mortality in the treated group. Whether increased mortality is a general attribute of CETP inhibitors will have to await additional clinical trials. However this data highlights the fact that increasing HDLc levels does not automatically confer a therapeutic benefit.

Probably stimulation of reverse cholesterol transport is more important for reducing the progression of atherosclerotic lesions. If and how PPARδ ligands stimulate this pathway is completely unknown.

References cited

Brinton, E. A., et al. (1990). "A low-fat diet decreases high density lipoprotein (HDL) cholesterol levels by decreasing HDL apolipoprotein transport rates." J Clin Invest 85(1): 144-51.

Brousseau, M. E., et al. (2004). "Effects of an inhibitor of cholesteryl ester transfer protein on HDL cholesterol." N Engl J Med 350(15): 1505-15.

Couillard, C., et al. (2001). "Effects of endurance exercise training on plasma HDL cholesterol levels depend on levels of triglycerides: evidence from men of the Health, Risk Factors, Exercise Training and Genetics (HERITAGE) Family Study." Arterioscler Thromb Vasc Biol 21(7): 1226-32.

Denke, M. A., et al. (1993). "Excess body weight. An underrecognized contributor to high blood cholesterol levels in white American men." Arch Intern Med 153(9): 1093-103.

Ellison, R. C., et al. (2004). "Lifestyle determinants of high-density lipoprotein cholesterol: the National Heart, Lung, and Blood Institute Family Heart Study." Am Heart J 147(3): 529-35.

Gaziano, J. M., et al. (1993). "Moderate alcohol intake, increased levels of high- density lipoprotein and its subfractions, and decreased risk of myocardial infarction." N Engl J Med 329(25): 1829-34.

Kashyap, M. L. (1998). "Mechanistic studies of high-density lipoproteins." Am J Cardiol 82(12A): 42U-48U; discussion 85U-86U.

King, A. C., et al. (1995). "Long-term effects of varying intensities and formats of physical activity on participation rates, fitness, and lipoproteins in men and women aged 50 to 65 years." Circulation 91(10): 2596-604.

Meksawan, K., et al. (2004). "Effect of low and high fat diets on nutrient intakes and selected cardiovascular risk factors in sedentary men and women." J Am Coll Nutr 23(2): 131-40.

Mokdad, A. H., et al. (2004). "Actual causes of death in the United States, 2000." Jama 291(10): 1238-45.

Ruan, H., et al. (2003). "Standard isolation of primary adipose cells from mouse epididymal fat pads induces inflammatory mediators and down-regulates adipocyte genes." J Biol Chem 278(48): 47585-93.

Serdula, M. K., et al. (1999). "Prevalence of attempting weight loss and strategies for controlling weight." Jama 282(14): 1353-8.

Wakabayashi, I. (2004). "Relationships of body mass index with blood pressure and serum cholesterol concentrations at different ages." Aging Clin Exp Res 16(6): 461-6.