Gaining weight: the Keystone Symposium on PPAR and LXR

Abstract

The nuclear receptor superfamily consists of 48 mammalian transcription factors that regulate nearly all aspects of development, inflammation, and metabolism. Two subclasses, the Peroxisome Proliferator-Activated Receptors (PPARs) and Liver X Receptors (LXRs), are lipid-sensing receptors that have critical roles in lipid and glucose metabolism. The parallel epidemics of obesity and diabetes shine a spotlight on the potential for therapeutic manipulation of PPARs and LXRs to combat these diseases. In recognition of this, a recent Keystone Symposium was devoted to these metabolic receptors. Here, we summarize some of the major highlights and future projections discussed at the meeting.

From April 12 to April 17, 2005, several hundred scientists from all over the world congregated in Whistler, British Columbia, Canada, for a Keystone Symposium devoted to Peroxisome Proliferator-Activated Receptors (PPARs) and Liver X Receptors (LXRs). PPARs and LXRs are members of the nuclear receptor (NR) superfamily of ligand-regulated transcription factors. NRs can be divided into three groups: (1) classical receptors for steroid and thyroid hormones (i.e., the glucocorticoid receptor, GR), (2) receptors for which nonclassical ligands have been discovered, and (3) “orphan” receptors, for which neither endogenous nor synthetic ligands have yet been identified. PPARs and LXRs are in the second group.

The PPAR subfamily consists of PPARα, PPARγ, and PPARδ (also known as β). These receptors initially became the focus of intense investigation following discoveries that PPARα and PPARγ are the molecular targets of major classes of drugs used to correct abnormalities of lipid and glucose homeostasis. Fibrates, effective in the treatment of hyperlipidemia, target PPARα, while thiazolidinediones (TZDs) are high-affinity ligands for PPARγ that have revolutionized the treament of type 2 diabetes by directly ameliorating tissue insulin resistance. Endogenous fatty acids are low-affinity ligands for PPARs, although the range of physiological ligands that regulate the activities of each PPAR in vivo remains to be clearly defined. Recent studies have suggested additional roles of PPARs in a diverse range of physiological and pathophysiological settings including inflammation, vascular wall biology, atherosclerosis, placental physiology, metabolic bone disease, and cancer.

A parallel of series of discoveries has emerged relating to the regulation of cholesterol and triglyceride homeostasis by the LXR subfamily of NRs, which consists of LXRα and LXRβ. The LXRs are activated by cholesterol metabolites and provide the basis for a feed-forward homeostatic control circuit in the maintenance of cellular cholesterol levels. Like PPARs, LXRs are increasingly recognized to play important roles in inflammation and atherosclerosis.

The human genome sequence encodes 48 NRs, each of which has been characterized to some extent. To those outside the NR field, an entire meeting focusing on just five of these might seem overly specialized. Ironically, before the first NR complementary cDNAs were cloned 20 years ago, the study of specific steroid and thyroid hormones were considered completely separate fields. The discoveries that receptors for glucocorticoids and estrogens were structurally related to each other, as well as to thyroid hormone, vitamin D, and retinoid receptors, changed the paradigm and forged the larger NR field. The recognition that such diverse ligands acted through a common mechanism, coupled with the tools of molecular biology, rapidly led to new, unifying principles that explain critical functions such as target gene recognition, ligand binding, and transcriptional regulation. The functional and structural similarities among NRs has been a theme of many international meetings.

Yet, the NR field has matured, with renewed recognition that each of the receptors has unique functions with physiological relevance, often in the context of metabolic regulation. Rising concern that obesity, insulin resistance, and hyperlipidemia are increasing in Westernized societies, leading to epidemics of diabetes and atherosclerosis, has fueled the interest in the role of individual NRs in the molecular mechanisms underlying metabolic disease. PPARs and LXRs have gained the most weight with maturity; hence, a diverse array of investigators from academia and industry came together in Whistler to focus mainly on this subset of metabolic NRs. This review attempts to provide a synthesis of the science presented at the meeting and define future directions for understanding and manipulating PPARs and LXRs so as to prevent and treat metabolic diseases.

Lipid metabolism

Many of the presentations focused on roles of PPARs as regulators of lipid metabolism. Among the three members of the PPAR subfamily, PPARδ remains the most poorly understood. Emerging evidence reveals the distinct regulation of fatty-acid utilization by PPARδ in different tissues. Ron Evans (Salk Institute, La Jolla, CA) presented a detailed analysis of PPARδ function in mice that revealed markedly distinct regulatory networks in individual tissues. PPARδ promotes oxidative metabolism in adipose tissue and muscle, leading to reduced fat mass and enhanced aerobic capacity. Intriguingly, over-expression of a constitutively active form of PPARδ in muscle led to a significant increase in slow twitch fiber type and increased endurance in treadmill tests. However, in the liver, the major effect of PPARδ appeared to be promotion of glycolysis, leading to reduced hepatic glucose output and lower blood glucose levels.

PPARα is the molecular target of the fibrate class of drugs that are used in the treatment of hypertriglyceridemia. The fibrates in current clinical use are relatively low-affinity ligands and primarily act in liver because of their pharmacokinetics. A large number of new PPARα compounds have been developed that are of much higher affinity and more effectively regulate PPARα activity in peripheral tissues. Barbara Hansen (University of Maryland, Baltimore, MD) reported that treatment of obese hypertriglyceridemic monkeys with a selective PPARα agonist increased lipoprotein lipase activity in skeletal muscle, with a reciprocal decrease in adipose tissue. Thus, PPARα agonists reduce the uptake of triglycerides in adipose tissue while increasing uptake in muscle for fatty-acid oxidation.

The concept that high-density lipoproteins mediate the transport of cholesterol from peripheral tissues to the liver for biliary excretion was first proposed >30 years ago. While substantial indirect evidence supports this hypothesis, actually measuring reverse cholesterol transport in vivo has been difficult. Dan Rader (University of Pennsylvania, Phildelphia, PA) described a novel method of measuring reverse cholesterol transport by injecting [3H] cholesterol-labeled macrophages into the peritoneal cavities of recipient mice. Macrophage-derived cholesterol could be tracked in vivo and ultimately measured in feces, providing a direct measure of reverse cholesterol transport. Using this method, overall reverse cholesterol transport was shown not always to correlate simply with circulating HDL level, but was markedly increased by LXR agonists.

David Mangelsdorf (Southwestern Medical School, Dallas, TX) highlighted the yin/yang relationship between LXR and another NR, Farnesyl X/bile acid Receptor (FXR), in maintaining hepatic intracellular bile acid and cholesterol levels by regulating the expression of genes involved in cholesterol catabolism and biliary cholesterol secretion. Mice lacking FXR exhibited increased gallstone formation when fed a high-fat, cholic acid-enriched (“lithogenic”) diet. At the molecular level, this appeared to be explained by decreased expression of phospholipid and bile-salt transporters. Furthermore, treatment with FXR agonists protected wild-type mice from cholesterol precipitation and bile-stone formation. These studies raise the possibility of using FXR agonists as a novel therapeutic strategy for the prevention and management of gallstone disease.

Adipocyte biology

Elegant studies from several groups have previously shown that PPARγ is a master regulator of adipogenesis. Yu-Hua Tseng presented her work carried out in the laboratory of Ron Kahn (Joslin Diabetes Center, Boston, MA), identifying necdin, a protein previously found to cause neuronal growth arrest and differentiation, as key regulator of adipocyte differentiation. Necdin interacts with E2F4, leading to repression of PPARγ transcription via a CREB-dependent pathway. Steve Farmer (Boston University, Boston, MA) addressed the mechanism of cross-talk between PPARγ and the Wnt/β-catenin pathway, which was first shown by Ormond MacDougald (University of Michigan, Ann Arbor, MI) to modulate adipogenesis. Farmer demonstrated that ligand activation of PPARγ induces the proteasomal degradation of β-catenin in a glycogen synthase kinase 3β-dependent manner, and that this degradation could be rescued by a mutated, oncogenic form of β-catenin.

Excess adipose tissue is a major risk factor for the development of insulin resistance. Although PPARγ is required for adipogenesis, PPARγ activators such as TZDs actually improve insulin sensitivity. Research over the past several years has made it increasingly clear that many of the actions of PPARγ on the mature adipocyte alter the secretion of adipocyte proteins that regulate insulin sensitivity. Takashi Kadowaki (University of Tokyo, Tokyo, Japan) focused on the insulin-sensitizing fat-cell hormone adiponectin, which circulates at reduced levels in obesity. He demonstrated that the insulin-sensitizing effects of TZDs were partly blunted in mice lacking adiponectin. Similar findings were reported by David Moller (Merck, Inc., Rahway, NJ), underlining the importance of adiponectin as a mediator of TZD-dependent sensitization.

In addition to its high level of expression in liver and macrophages, LXRα is also abundant in adipose tissue. Hilde Nebb (University of Oslo, Oslo, Norway) demonstrated that activated LXRs increase basal glucose uptake and incorporation of triglycerides into lipid droplets in adipose tissue. Insulin stimulated glucose uptake, and de novo lipogenesis is mediated through LXRs, indicating that LXRs are key regulators in the mediation of insulin actions in lipid synthesis in adipose tissue, potentially via the lipogenic transcription factor Sterol Response Element-Binding Protein (SREBP) 1.

Inflammation

Inflammation is a central component of several chronic human diseases, including atherosclerosis and type 2 diabetes. Many NRs have the ability to repress inflammatory responses, but molecular mechanisms remain poorly understood. In the case of atherosclerosis and type 2 diabetes, anti-inflammatory actions of PPARγ agonists may contribute to their protective/therapeutic effects in animal models and humans. Chris Glass (University of California, San Diego, CA) reported studies demonstrating that many inflammatory-response genes are occupied by N-CoR corepressor complexes under basal conditions, and that deletion of the N-CoR gene in macrophages resulted in a partially activated phenotype. Activation of these genes in response to lipopolysaccharide (LPS) resulted in proteolytic clearance of N-CoR complexes as a prerequisite to full transcriptional activation. Intriguingly, PPARγ agonists were demonstrated to prevent signal-dependent clearance of N-CoR from at least some of these inflammatory response genes, suggesting a new mechanism for transrepression.

In addition to the effects of PPARγ on the classical pathway of macrophage activation, Ajay Chawla (Stanford Medical School, Stanford, CA) presented evidence for roles of PPARγ and the PPARγ coactivator-1β (PGC-1β) in coordinating transcriptional programs involved in “alternative” macrophage activation by interleukin-4 (IL-4). IL-4 transcriptionally activates STAT6 to induce expression of PPARγ and PGC-1β. Interestingly, PGC-1β plays coactivator roles for both transcription factors, thereby contributing to induction of oxidative metabolism and maturation of the alternative phenotype, programs that may be required for host defense against parasitic organisms.

Peter Tontonoz (University of California, Los Angeles, CA) outlined a new role for LXRs in the innate immune response, demonstrating that LXRα is up-regulated in macrophages in response to Listeria infection. He presented evidence that LXR up-regulates Spa/AIM6, protecting against apoptosis during Listeria infection. Remarkably, LXR-deficient mice were significantly more susceptible to lethal Listeria infection than wild-type mice, establishing a critical role of LXRs in innate immunity. Evidence for coordination between the PPAR and LXR pathways was provided by Laszlo Nagy (University of Debrecen, Hungary), who reported that PPARγ activation induces a cytochrome p450 enzyme involved in the production of LXR ligands. Other NRs will likely play a role in macrophage activation. Grant Barish of the Evans lab used high-throughput quantitative PCR to document the expression of 28 of the known 48 NRs in macrophages. Many of the NRs, including some not previously documented to be expressed in macrophages, exhibited distinct temporal changes in expression following LPS stimulation. Application of this profiling method to other cell types may uncover a wealth of new NR-dependent programs of gene expression.

Cardiovascular disease

Several rodent and clinical studies have suggested that PPARγ ligands such as TZDs protect against the development of atherosclerosis by globally improving insulin action and by acting directly on cells in the vascular wall. Kadowaki provided evidence that part of the beneficial effect of TZDs is due to the induction of adiponectin expression in adipose tissue that, in turn, acts on cells in the artery wall, suggesting that vascular effects result from a combination of local and distant actions of PPARγ. The role of PPARδ in atherosclerosis remain unclear. On the one hand, PPARδ agonists clearly exert anti-inflammatory effects in the artery wall, and a study presented by Evans demonstrated atheroprotective effects in an apo E knockout mouse model. However, Glass and colleagues did not find that PPARδ agonists offered atheroprotection in the mouse LDL receptor knockout model. Further studies will be needed to explain the differences observed in these models and to determine the potential impact of PPARδ agonists on human atherosclerosis.

PPARα ligands have been established to reduce the incidence of myocardial infarction in humans and are likely to act both locally in the artery wall and systemically to control lipid homeostasis. The development of complex atherosclerotic lesions is accompanied by the migration of smooth muscle cells into the arterial intima and their proliferation and conversion into foam cells. In addition, following the treatment of atherosclerotic contstrictions in coronary vessels using balloon catheters, the proliferation of smooth muscle cells can lead to restenosis. Bart Staels (University of Lille, France) demonstrated that PPARα controls vascular smooth muscle cell proliferation by inducing p16^INK. This results in recruitment of p16INK to CDK4 complexes and inhibition of hyperphosphorylation of Rb protein. Smooth muscle cells obtained from PPARα^-/- mice exhibited lower levels of p16^INK and were hyperproliferative, suggesting an important role of PPARα in the physiological role of smooth muscle cell proliferation. These findings also raise the possibility that PPARα agonists might be useful in preventing the troublesome clinical problem of renarrowing (or “restenosis”) of the arterial lumen after coronary angioplasty.

Synthetic LXR agonists have been shown to inhibit the development of atherosclerosis in mouse models, and this effect has primarily been attributed to their effects on LXR in macrophages. However, Mitch Lazar (University of Pennsylvania, Philadelphia, PA) reported that increased hepatic LXRα expression had favorable effects on serum lipids and reduced atherosclerosis in mice lacking LDL receptors when placed on a high-fat, high-cholesterol, Western-style diet. This may reflect differential receptor activation by less potent LXR ligand(s) that are either present in or generated by the Western diet; in this model, potent synthetic LXR agonists have a dominant effect by stimulating production of very low-density lipoprotein (VLDL) and dramatically raising serum triglyceride levels. These findings may provide an important clue to the development of selective LXR modulators that retain anti-atherogenic activities, but do not cause hypertriglyceridemia. To this end, Rich Heyman (Exelixis, Inc., San Diego, CA) reported the development of less potent LXR compounds that exhibit these characteristics in model systems.

PPARs also play an important role in the heart. Dan Kelly (Washington University, St. Louis, MO) showed that the coactivator PPARγ-coactivator (PGC)-1 docks on the orphan NR ERRα, inducing PPARα, and thereby enhancing PPARα-dependent pathways of gene expression. He also reported that in the diabetic heart, PPARα is chronically activated, resulting in cardiac dysfunction through increased uptake and oxidation of fatty acids. Although PPARγ is expressed at very low levels in the heart, Willa Hsueh (University of California, Los Angeles, CA) reported that mice lacking PPARγ in cardiac myocytes exhibited increased susceptibility to angiotensin II-induced cardiac fibrosis. These findings suggest a previously overlooked role of PPARγ in the heart.

Natural ligands

Despite the great advances in developing therapeutic ligands for PPARs, the identity of their natural ligands has remained elusive and has been a subject of controversy. Jorge Plutzky (Brigham and Women's Hospital, Boston, MA) has pursued potential endogenous ligands for PPARs by focusing on candidate pathways as opposed to studying specific candidate molecules. Using this approach, he suggested that PPARα ligands are generated by the action of lipoprotein lipase, a key enzyme in triglyceride metabolism, on VLDL, and by endothelial lipase-mediated hydrolysis of HDL.

Beatrice Desvergne (University of Lausanne, Switzerland) presented a pathway in skin whereby pro-inflammatory cytokines such as TNFα, which are produced during wound healing, generate endogenous PPARδ ligands that allow induction of Akt1 expression and prevent apoptosis. Furthermore, similar activation of PPARδ in kidney was protective in a renal ischemia-reperfusion model. Barry Forman (City of Hope Medical Center, Duarte, CA) reported on a combined affinity and mass spectrometry approach to identify natural ligands. Using this strategy, the native form of the orphan NR hepatocyte nuclear factor 4 (HNF4) in mouse liver was found to be bound to a fatty acid. These findings suggest a physiological relevance of published crystal structures, in which a fatty acid was noted in the ligand-binding pocket of HNF4.

Lessons from human genetics

Helen Hobbs (Southwestern Medical Center, Dallas, TX) discussed her group's discovery that β-sitosterolemia, a disease characterized by elevated plasma levels of plant sterols, results from mutations in the LXR-responsive ABC transporters ABCG5 and ABCG8. These proteins are expressed in the intestine, where they limit absorbtion of dietary plant-derived sterols including sitosterol, campesterol, and stigmasterol. ABCG5 and ABCG8 function as obligate heterodimers, so that loss-of-function mutations in either gene results in markedly elevated circulating and tissue plant sterol levels. Intriguingly, the accumulation of certain plant sterols in mice lacking both transporters profoundly perturbs cholesterol homeostasis by both activating LXRα and inhibiting the processing of SREBP2. These findings suggest a physiological role for the active export of plant sterols from the body. Krishna Chatterjee (University of Cambridge, UK) discussed a syndrome associated with diverse mutations in human PPARγ, which was manifested by partial lipodystrophy together with severe insulin resistance, early-onset hypertension, and dyslipidemia. Studies of these mutant receptors revealed novel mechanisms for their dominant negative action, including the recruitment of corepressors or sequestration of cofactors to limit wild-type PPARγ function.

Metabolic regulation by coregulators of PPAR and LXR

NRs activate transcription by recruiting one or more coactivators, whose docking is generally regulated by a ligand-induced conformational change. Many NRs, including PPAR and LXR, also repress transcription by recruiting corepressors in the absence of ligand. Bruce Spiegelman (Dana-Farber Cancer Institute, Boston, MA) presented an overview of the PGC1 family of coactivators. In general, PGC1 promotes programs of gene expression that mediate oxidative metablolism of fatty acids. For example, PGC1α is crucial for mitochondrial uncoupling and cold adaptation in brown fat, and is necessary for exercise responses in cardiomyocytes and skeletal muscle. PGC1β shares specific features with the α isoform, including promotion of mitochondrial proliferation. In addition, PGC1β is involved in a novel pathway, whereby saturated and trans-fatty acids elevate circulating triglyceride and cholesterol levels. Specifically, fatty acids in the diet of mice elevate liver PGC-1β levels, along with SREBP1c. PGC-1β binds to and coactivates the entire family of SREBP transcription factors as well as LXR, thus up-regulating the expression of genes involved in both cholesterol and triglyceride biosynthesis. Ectopic PGC-1β expression in the liver also leads to synthesis and secretion of VLDL lipoproteins into the blood. Thus, PGC-1β is an important component in a pathway that links saturated and trans-fatty acids in the diet with plasma trigylcerides and cholesterol.

Pere Puigserver (Johns Hopkins Medical School, Baltimore, MD) reported that the NAD-activated histone deacetylase SIRT1 deacetylates PGC1α during times of fasting, thereby enhancing its transcriptional activity and resulting in increased gluconeogenesis. He also described GCN5-dependent acetylation of PGC1α, and proposed that PGC1α activity is regulated by a balance between the acetylation and deacetylation pathways. Jan Reddy (Northwestern Medical School, Chicago, IL) presented data on the role of the transcriptional coactivator PBP/TRAP220/MED1 in PPARα-regulated gene expression in liver and acetaminophen-induced hepatotoxicity. Targeted deletion of PBP in liver parenchymal cells resulted in the near abrogation of PPARα ligand-induced peroxisome and liver cell proliferation, as well as the induction of PPARα-regulated genes. Chromatin immunoprecipitation (ChIP) studies revealed reduced association of cofactors, specifically, CBP and TRAP130, to PPARα target genes in PBP deficient livers, suggesting that PBP is required for the stabilization of multiprotein cofactor complexes.

Moderation in all ligands: when less is more

Despite their great success in the treatment of type 2 diabetes, TZDs have several side effects that have resulted in their limited clinical use. Some of these effects include weight gain due to a combination of increased adiposity and edema. Tim Willson (Glaxo SmithKline, Inc., Research Triangle Park, NC) illustrated a combination of structural and biochemical approaches to devise and evaluate newer PPARγ ligands, whose mode of binding is different from TZDs. These compounds all bind the receptor in the ligand-binding pocket, but have varying potencies in terms of target gene induction and selective effects on cofactor recruitment. These compounds have been characterized as full, partial, or inverse agonists, depending on their ability to recruit coactivators or corepressors and activate or repress target genes. Remarkably, despite the repressive effect on certain PPARγ-dependent genes, an inverse agonist was found to improve insulin sensitivity in mice.

Consistent with this, Moller defined different classes of novel PPARγ selective modulators, which were selected on the basis of attenuated gene regulation of “classic” PPARγ target genes. Similar to the compounds described by Willson, these selective and somewhat less potent PPARγ modulators were effective in treating rodent models of diabetes with less toxicity than the more potent, classical PPARγ agonists such as TZDs. The ability of selective PPARγ modulators to differentially recruit coregulators dovetailed nicely with Lazar's finding of two sets of PPARγ-induced genes in adipocytes: those induced during normal adipogenesis in the absence of exogenous ligand, and those only moderately expressed in adipocytes unless induced by potent agonists. The latter gene set, including several regulators of adipocyte lipid metabolism, is occupied by corepressors N-CoR and SMRT in the absence of ligand, which induces the classical exchange for coactivators.

Despite their therapeutic potential for hypercholesterolemia and atherosclerotic vascular disease, potent synthetic LXR ligands have problematic unwanted side effects, such as hypertriglyceridemia in rodent models, largely due to activation of SREBP-1c. Consistent with this, treatment with a potent agonist abrogated the improved serum triglycerides in Lazar's hepatic LXR over-expression model. Moreover, the dissociation between in vitro potency and in vivo efficacy among PPARγ ligands was paralleled, remarkably, in the case of LXR. Heyman described LXR ligands that function as full, partial, and inverse agonists, again reflecting their potential to recruit either coactivators and/or corepressors. A new partial agonist provided atheroprotection similar to that of full agonists, but resulted in only partial induction of SREBP1 and, unlike the more potent agonist, did not cause hypertriglyceridemia.

New tools/new frontiers

The PPAR/LXR Keystone meeting was held simultaneously with the Bioactive Lipids and Lipidomics meeting organized by Garret FitzGerald and Ian Blair (both at the University of Pennsylvania, Philadelphia, PA) along with Ron Evans. Lipidomics will be a critical tool for the discovery and validation of novel NR ligands and for the systemic linkage between metabolic pathway gene-expression data and the metabolites themselves. In joint sessions, Alex Brown (Vanderbilt Medical School, Nashville, TN) and Gerhard Liebisch (University of Regensburg, Germany) described evolving lipidomic technologies, and demonstrated their use in studying lipid metabolism in macrophages. Frank Gonzalez (National Cancer Institute, Bethesda, MD) presented data using PPARα-humanized mice, establishing that the species difference in response to PPARα ligands is due to molecular differences between the mouse and human receptors, and described a lipidomic approach using these mouse models to identify biomarkers for the clinical and toxic effects of PPARα ligands

Bioinformatics and new technologies such as chromosome-wide ChIP have the promise to lead to new insights into gene-regulatory networks and mechanisms of gene regulation. Qianfei Jeffrey Wang (Lawrence Berkeley Laboratory, Berkeley, CA) described a comparative computational analysis of LXRα target genes in multiple primate species, identifying three elements that were conserved in humans, but otherwise divergent from rodents. These elements were hypersensitive, transcriptional active, and provided enhancer activity in transient transfection reporter assays. This represents an intriguing approach for the understanding of species differences.

Bing Ren (University of California, San Diego, CA) discussed chromatin immunoprecipitation-chromosome microarray (ChIP-on-Chip) technology and its utilization in mapping preinitiation complexes marked by RNA polymerase II or TAF250. In the large majority of cases, promoter occupancy of TAF250 correlated with gene expression, although some discordance (gene expression without TAF250 or TAF250 without gene expression) was noted, in part art factual, due to sensitivity differences between the two readouts and in part due to biological reality that needs to be better understood. Using estrogen receptor (ER) as a model NR, Myles Brown (Dana-Farber Cancer Institute, Boston, MA) illustrated the power of ChIP-on-Chip. Most ER-binding sites contained classical estrogen-responsive elements (EREs), but only a small fraction of chromatinized EREs were occupied in breast cancer cells. Bioinformatic analysis revealed that many of the functional EREs were in the vicinity of FOXA1-binding sites, and FOXA1 proved to be important for ligand-dependent ER recruitment to these genes. Interestingly, many ER-binding sites were localized far upstream of the transcriptional start sites of target genes. These bioinformatic and profiling methods have great potential to identify new target genes, regulatory sequences, and genetic networks for PPAR, LXR, and other NRs.

The bigger picture: circadian rhythms and aging

Metabolism is complicated. PPARs and LXRs represent a subset of transcription factors regulating gene networks involved in the metabolism of glucose, lipids, and amino acids. Many substrates and metabolites are at the intersection of several pathways, and much of the enzymatic regulation occurs post-transcriptionally. Superimposed on these intricate systems are additional levels of regulation related to time. At one extreme is aging. Keynote speaker Cynthia Kenyon (University of California, San Francisco, CA) presented an elegant overview on the genetic determination of longevity. Using the model system of Caenorhabditis elegans, she demonstrated gene-dependent control in aging. Mutations in the daf-2 gene, the homolog of mammalian insulin and IGF-1 receptors, doubled the life span of worms; similarly, correlations between defective insulin signaling and longevity pertain in flies and mice. Interestingly, Kenyon showed that tissue-specific knockdown of daf-2 increased the life span of the whole organism, suggesting a humoral (or hormonal) control mechanism. The specific hormones are unknown, but could be ligands for NRs such as daf-12, whose mutation has a synergistic effect on increased daf-2 adult longevity.

Another time factor occurs on a daily basis. The mammalian circadian timing system has a hierarchical structure, in that a master pacemaker in the suprachiasmatic nucleus (SCN) of the brain's hypothalamus serves as a higher order synchronizer of gene oscillatory networks that exist in most peripheral body cells. These cycles in clock gene expression are then translated into daily rhythms of cellular physiology. Ueli Schibler (University of Geneva, Switzerland) described one such clock output pathway in liver cells, in which the rhythmic expression of key PAR bZip transcriptional regulatory proteins is driven directly by core components of the molecular oscillator (BMAL1, CLOCK, CRY1/2, PER1). In turn, the circadian expression of PAR bZip proteins elicits daily accumulation cycles of enzymes involved in detoxification, such as cytochrome P450 enzymes. The NRs are likely to play a major role as well, since many P450 enzymes are also regulated by NRs and, as revealed in Evan's presentation at this meeting, many NRs display marked circadian gene regulation in metabolic tissues.

Conclusions

PPARs and LXRs are clearly major regulators of metabolism and are promising candidates for the pharmacological defense against advancing epidemics of obesity, diabetes, and cardiovascular disease. This justified the Keystone Symposium focusing on these NRs, highlighting the fact that the work in this area is approaching maturity and gaining weight as a field. Yet we've only known about PPARs and LXRs for <20 years. While there is good evidence that PPARs and LXRs function as lipid sensors, the “true” endogenous ligands remain unknown. PPARs and LXRs are essential for normal metabolism in mice (and almost certainly in man), yet, we are just defining the scope and interactions of the pathways they regulate. These are “druggable” transcriptional regulators, with compounds in hand, and in some cases in the clinic, but the therapeutic index of PPAR and LXR modulators needs to be improved. The most recent insights from ligand development is that reduced potency may be a good thing in the balance between therapeutic benefit and unwanted side effects. But why? It could be differential coregulator recruitment or dissociation, selective tissue or target gene regulation, or some other mechanism. The systems biology approach to understanding the simultaneous regulation of gene targets in the chromatin milieu promises to answer these questions and raises new ones to be addressed at future international meetings—in a field to be named later.