PPARs…in the 21^st Century

by Kirk Maxey, M.D. and Beth Meade, Ph.D.

It was noted nearly 40 years ago that certain 2,2'-dimethyl carboxylic acids and their esters, exemplified by clofibrate (see Fig. 3), induced a proliferation of the organelle within liver cells where fatty acid oxidation takes place. These peroxisomes increased both in size and in number in response to clofibrate treatment, with a corresponding drop in circulating plasma lipids. There was also an induction of the expression of an entire suite of genes associated with fatty acid synthesis, transport, and catabolism. Cloning studies eventually revealed three ligand-dependent transcription factors that were responsible for this induction. These are the Peroxisome Proliferator Activated Receptors (PPARs) α, β/δ, and γ. (PPAR β/δ will be referred to as PPARβ hereafter.) Homology analysis showed that these orphan receptors were members of the superfamily of nuclear hormone receptors which includes the retinoic acid and thyroid hormone receptors. (For a brief description of the domain structure and action of nuclear receptors see Fig. 2.)

The clinical implications of the PPARs were readily apparent. PPARα agonists were potent hypolipidemic agents, and this led to the development of the widely prescribed fibrate class of drugs. By lowering plasma lipids, these drugs helped to reduce the incidence of atherosclerosis. In addition, PPARγ agonists increased the patients' responsiveness to insulin, and were therefore potential new drugs for adult onset (Type II) diabetes, where insulin levels are often high but the normal response to insulin is blunted. The thiazolidinediones, typified by rosiglitazone, (see Fig. 3) were quickly developed as antidiabetic drugs.

The presence of these potent and relatively selective agonists for the different PPARs did little to help identify their natural ligands and the role of these receptors in normal growth and development. More than a dozen fatty acids are ligands for one or all three of the main PPAR receptor subtypes. The search for a more potent, selective autacoid agonist has been frustrating. Although candidate biomolecules have been proposed as "endogenous" ligands for each receptor, there are substantial questions surrounding each of them. These can be best examined individually for each receptor subtype.

PPARα

PPARα The PPARα receptor is expressed most strongly in brown adipose tissue and liver. The next highest are tissues with constant high metabolic demand, as in heart, skeletal muscle, and kidney. The PPARα receptor gene can be disrupted with no apparent effect on the homozygous null mutants. However, when starved these animals are unable to efficiently switch to fat catabolism to meet their energy needs. The role best attributed to PPARα is thus as a functional metabolic switch which controls the utilization of adipose reserves. The many saturated, polyunsaturated, and branched chain fatty acids which act as ligands for PPARα are thus intuitively reasonable endogenous ligands.

PPARβ

PPARβ is the least studied of the PPAR receptor subtypes. It is notable for its ubiquitous expression throughout the body, without any predisposition for adipose tissue. It is also distinguished by reports that prostacyclin may be a ligand, in addition to the usual promiscuous affinity for fatty acids. Null mutants for PPARβ have not been reported.

PPARγ

The γ-isoform of the PPAR receptor is the one most closely associated with fat. There are two PPARγs, PPARγ1 and PPARγ₂. PPARγ₂ is a splice variant with 42 additional amino acids appended to the PPARγ₁ sequence. Both PPARγ₁ and PPARγ₂ bind ligands with the same apparent affinity, so their differences may reside more in tissue-specific induction and expression. PPARγ₂ is expressed very strongly in white adipose, and most of the genes whose transcription are induced by PPARγ are involved in lipogenesis. PPARγ agonists induce fibroblasts to differentiate into adipocytes, providing a mechanism whereby PPARγ controls the relative fat content of the body. Knockout mice have confirmed this critical role, for null mutant mice suffer embryonic death. The embryos have no detectable fat and accumulate lipids in the liver. The role of PPARγ thus seems to be the establishment and provisioning of the adipose storage pool. In this way it complements PPARα, which controls metabolic withdrawal from this pool during starvation. Although the compound 15-deoxy-Δ^12,14-PGJ₂ is clearly a potent PPARγ ligand, it is not clear if it is endogenous (see Fig. 1, below and Ref. 7).

The compound 15-deoxy-Δ^12,14-PGJ₂ has been reported to be a naturally occuring, endogenous substance. This claim is controversial, because the compound has never been isolated from a natural source. In addition, the synthetic compound exists as a mixture of at least five different isomers (see Fig. 1). It is not clear which of these isomers the "endogenous" ligand might be.

Separation of 15-deoxy-Δ^12,14-PGJ₂ (15-ΔJ₂) isomers obtained from the base catalyzed decomposition of PGD₂. The crude 15-ΔJ₂ was subjected to normal phase HPLC (column: silica, 5, 4.6 × 250 mm; detection: 306 nm) to give five pure compounds. The peak area shows the relative abundance of each isomer in mAU at 306 nm. The shaded bar represents the relative antiproliferative potency of each pure isomer tested separately on cultured MDA-MB-231 breast carcinoma cells, as percent of control in growth arrest. The vertical bar superimposed on each peak represents the relative PPARγ ligand binding affinity of that isomer in a luciferase reporter assay, expressed as RLU.

The presence of endogenous ligands for the PPAR receptors other than fatty acids is controversial (see Fig. 3 below). An antiinflammatory role has been postulated for PPARα, based on the fact that PPARα knockout mice have a prolonged inflammatory response. 5-LO knockout mice that cannot make LTB₄, the putative PPARα ligand, contradict this notion. They have attenuated inflammatory responses, which is opposite what this model would predict. The affinity of LTB₄ for the PPARα receptor is so low (K_d ~ 100 nM) that proponents have been forced to postulate "locally enhanced" nuclear concentrations. This is an interesting, recurring, and so far untestable hypothesis. Prostacyclin, the proposed endogenous ligand for PPARβ, and 15-deoxy-Δ^12,14-PGJ₂, the proposed PPARγ ligand, share with LTB₄ a vanishingly short half-life in cell culture on the order of 30 seconds. The reasons for their transient natures are all different; prostacyclin simply reacts with water and is hydrolyzed to inactive 6-keto PGF_1α. LTB₄ is rapidly inactivated by ω-oxidation in a cell-specific, enzymatic process. 15-deoxy-Δ^12,14-PGJ₂ is efficiently scavenged by glutathione-dependent transferases, but also reacts with thiol-containing proteins non-enzymatically. Thus, proponents of these "endogenous" PPAR ligands must invoke unconventional arguments to explain a prolonged association with the DNA-bound transcriptional complex in the nucleus. Testing these arguments presents a challenge to investigators in this promising field.

References

1. Forman, B.M., Tontonoz, P., Chen, J., et al. Cell 83, 803-812 (1995).
2. Tontonoz, P., Hu, E., and Spiegelman, B.M. Cell 79, 1147-1156 (1994).
3. Kliewer, S.A., Lenhard, J.M., Willson, T.M., et al. Cell 83, 813-819 (1995).
4. Dreyer, C., Krey, G., Keller, H., et al. Cell 68, 879-887 (1992).
5. Greene, M.E., Blumberg, B., McBride, O.W., et al. Gene Expression 4, 281-299 (1995).
6. Aubert, J., Ailhaud, G., and Negrel, R. FEBS Lett. 397, 117-121 (1996).
7. Maxey, K.M., Hessler, E., MacDonald, J., et al. Prostaglandins 62, 1-90 (2000).