From the Department of Medicine, Division of Endocrinology and Metabolism, University of California, San Diego, La Jolla, California 92093 and the Department of Veterans Affairs, San Diego, California 92161
The nuclear receptor superfamily describes a related but diverse array of transcription factors, which include nuclear hormonereceptors (NHRs)1 and orphan nuclear receptors. NHRs are receptors for which hormonalligands have been identified, whereas orphan receptors are sonamed because their ligands are unknown, at least at the timethe receptor is identified. Unlike hormones for cell surface receptors,lipophilic hormones can traverse the plasma membrane to the cellinteriors where NHRs transduce signals from glucocorticoids, mineralocorticoids,the sex steroids (estrogen, progesterone, and androgen), thyroidhormones, and vitamin D3.
All of the nuclear receptors have common structural features (Fig. 1), which include a central DNA binding domain (DBD) responsiblefor targeting the receptor to highly specific DNA sequences comprisinga response element (1). The ligand binding domain (LBD) iscontained in the C-terminal half of the receptor and recognizesspecific hormonal and nonhormonal ligands directing specificityto the biologic response. These receptors contain variable N-terminaland C-terminal domains, as well as a variable length hinge regionbetween the DBD and LBD. Nuclear receptors can exist as homo-or heterodimers with each partner binding to specific RE sequencesthat exist as half-sites separated by variable length nucleotidespacers between direct or inverted half-site repeats. Severalyears ago Manglesdorf et al. (2) proposed four categories ofnuclear receptors in which Class 1 receptors include the knownsteroid hormone receptors, which function as homodimers bindingto half-site RE inverted repeats. Class 2 receptors exist as heterodimerswith RXR receptor partners and function in a ligand-dependentmanner. The second two classes include orphan receptors, whichfunction as homodimers binding to direct RE repeats (Class 3)or monomers binding to single site REs (Class 4).
Given the widespread relevance of the superfamily of nuclear receptors to almost all aspects of normal human physiology, therole of these receptors in the etiology of many human diseases,and their importance as therapeutic targets for pharmaceuticals,it is obvious that a detailed understanding of these systems hasmajor implications, not only for human biology but also for theunderstanding and development of new drug treatments. In composingthis minireview series, it is quite clear that it is not feasibleto comprehensively review any of the separate areas in great detail,and all of the minireviews in this series contain references tomore extensive review papers on particular topics. As in all minireviews,the purpose of this series is to highlight major themes and newdevelopments in these areas and to point out common molecularand biochemical principles that broaden our understanding of thesecomplex biologicalprocesses.
The first article in the series is entitled "Coregulator Codes of Transcriptional Regulation by Nuclear Receptors" authoredby Michael G. Rosenfeld and Christopher K. Glass. This reviewcovers the complex and ever-growing network of interactions betweencoregulatory proteins and nuclear receptors. These regulatoryproteins form multicomponent assemblies with the nuclear receptors,and these complexes can serve as coactivators or corepressors.The specific proteins in these complexes can bind to nuclear receptorsvia specific amino acid sequence motifs in a ligand-dependentor independent manner and can provide enzymatic or scaffoldingfunctions. These coregulators influence chromatin remodeling byhistone acetylation/deacetylation, methylation, and possibly otherevents. In general, ligand binding to nuclear receptors causesan exchange of coactivators for corepressors to facilitate transcription.This review also points out that the coregulatory proteins themselvesare subject to biochemical and functional regulation by varioussignaling pathways. Far more coregulatory molecules have beenidentified than can bind to a given nuclear receptor, and givenrecent findings of rapid turnover of these complexes on DNA, itis possible that they work in a combinatorial or sequential mannerto exert transcriptional control. Given the tissue specificityof some coregulators, their ability to be modified by variousother signaling molecules, and possible combinatorial functions,one can envision that a given nuclear receptor could exert diverseeffects depending on the environmental context of a given tissue,cell, or specificpromoter.
The estrogen receptor (ER) is perhaps the most well defined nuclear receptor system from the point of view of biologic responsesand clinical implications. "Multifaceted Mechanisms of Estradioland Estrogen Receptor Signaling" by Julie M. Hall, John F. Crouse,and Kenneth S. Korach reviews the major features of this importantnuclear receptor system. There are two subtypes of the ER (ERaand -b), which are products of distinct genes but show differencesin tissue expression. Although quite similar in structure, thetwo ER subtypes display structural differences and can mediateoverlapping but different sets of biologic functions. This isbest exemplified in the ERb versus ERa knockout mice, which havequite different phenotypes. However, it is also clear that ERbcan substitute for ERa in some biologic pathways. Furthermore,ERb can interact with the same ERE as ERa, and the two ER subtypescan also form heterodimers, indicating that in cells that expressboth ER subtypes, the ratio of the two will effect estrogen action.The review provides a discussion of the classical mechanisms ofestrogen action mediated through the ER and EREs, as well as nonclassicalmechanisms in which ERs can be modulated by ligand independentmeans. Genomic actions of ERs can also be exerted in the absenceof direct DNA binding by mechanisms in which liganded ERs interactdirectly with other transcription factors such as Fos and Jun,influencing their function at AP-1 sites. To add to the complexityof ER action, it has now been proposed that estrogens can exertnongenomic effects by binding to plasma membrane receptors thatdirectly mediate biologic responses. Finally, this review providesan incisive discussion of the selective estrogen receptor modifierconcept, which holds that different ligands form specific three-dimensionalstructures with receptors that lead to tissue- and perhaps cell-specificbiologic effects. This concept has already had ramifications onthe clinical front, where it has been shown that different selectiveestrogen receptor modifier compounds (the Type 2 anti-estrogenRaloxifene and the Type 3 anti-estrogen, Tamoxifen) exert uniqueestrogenic effects in a tissue-specific manner. This is an importantarea of pharmaceutical discovery in which there is hope of developingagents that exert only the beneficial and not the potentiallyharmful effects ofestrogens.
Peroxisome proliferator-activated receptors (PPARs) exert diverse effects on fat and carbohydrate metabolism and are majortargets for therapeutic agents in metabolic diseases. This hasgenerated enormous interest in this class of NHRs leading to variousmolecular, physiologic, and clinical insights. Evan D. Rosen andBruce M. Spiegelman provide a review on this topic entitled "PPARg:a Nuclear Regulator of Metabolism, Differentiation, and Cell Growth."As suggested by the title, although some discussion of the otherPPAR members, PPARa and PPARd, is provided, the bulk of this paperfocuses on the PPARg receptor. Although potential endogenous ligandsfor this receptor have been proposed, definitive evidence foran endogenously made ligand is still lacking. Nevertheless, thiazoladinediones(TZDs), as well as other PPARg ligands, are used clinically asinsulin-sensitizing agents, and these pharmacologic ligands haveprovided a great deal of knowledge about the biologic functionof the PPARg receptor. This receptor clearly plays a criticalrole in adipogenesis, and the complex interactions between PPARgand other adipogenic transcription factors such as CCAAT/enhancer-bindingprotein a are explored. Because TZDs are clinically useful anti-diabeticinsulin-sensitizing agents, it is clear that PPARg is an importantfactor in the overall regulation of insulin action, and this area,including the tissue sites of action and the potential PPARg targetgenes that mediate insulin sensitization, are reviewed. Althoughthe effects of PPARg ligands in causing insulin sensitizationare the most well known, two other important areas of interestare reviewed, i.e. the roles of the PPARg receptor in atherosclerosisand oncogenesis. Evidence exists that PPARg receptors can modulatethe formation of foam cells in atherosclerotic plaques and thatTZD treatment may be antiatherogenic. Furthermore, because thisreceptor promotes differentiation, it is proposed that it mayinhibit oncogenic effects in various cell types. Consistent withthis, mutations and translocations of the PPARg receptor havebeen identified in human tumors, and this emerging area of PPARgbiology is examined and put into perspective in the review byRosen andSpiegelman.
Cholesterol and sterol homeostasis is another important regulatory system closely controlled by nuclear receptor function,and in this series, Timothy L. Lu, Joyce J. Repa, and David J. Mangelsdorf provide a review on this subject entitled "OrphanNuclear Receptors as eLiXiRs and FiXeRs of Sterol Metabolism"in which the two major nuclear receptors, LXR and FXR, involvedin this regulatory system are reviewed. The role of the LXR nuclearreceptor as a cholesterol sensor is discussed, including recentinformation covering target genes such as SREBP-I and the ATPbinding cassette transporters, which facilitate efflux of cholesterolfrom cells. In the enterocyte, increased function of these ATPbinding cassette transporters decreases cholesterol absorptionfrom the gastrointestinal tract, and in macrophages, impairedfunction of these proteins may promote atherogenesis. The FXRbile acid sensor also plays a key role in overall sterol metabolismby regulating transcription of an array of genes involved in bileacid metabolism. The function of these two nuclear receptors ishighly integrated, creating a complex but complementary physiologicnetwork for controlling various facets of cholesterol and sterolmetabolism across different tissues. Because of the importanceof cholesterol metabolism in the etiology of atherosclerosis,this regulatory system offers a number of potential pharmaceuticaltargets for the development of new drugs to control hypercholesterolemiaand favorably impact the process ofatherosclerosis.
The final installment of this series covers another class of transcriptional regulators termed "orphan receptors" belongingto this large superfamily. The orphan nuclear receptors are proteinsthat share a great deal of structural similarity to NHRs but donot have physiologic ligands that have been identified. At suchtime that a definitive ligand is identified, then that receptorwould lose its orphan status. It is now known that several ofthese orphan receptors respond to xenobiotics in the environmentthat includes foreign chemicals such as environmental pollutantsand prescription drugs. In response to xenobiotic compounds, thesereceptors mediate transcription of a variety of detoxifying enzymesthat are members of the supergene family of cytochrome P450 (CYP)molecules. As Wen Xie and Ronald M. Evans point out in their reviewon this topic entitled "Orphan Nuclear Receptors: the Exoticsof Xenobiotics," this class of nuclear receptors represents theregulatory interface between the human genome and the externalenvironment. This review discusses the major xenobiotic receptors,SXR, PXR, and CAR and points out that by inducing various CYPfamily members in response to specific xenobiotics, these receptorsdictate our ability to metabolize different pharmaceutical compounds.An understanding of the function of these receptors should providea mechanistic basis for drug interactions in which one drug altersthe metabolism of another. Interestingly, the human SXR receptorand its rodent PXR orthologue display differential sensitivityto various xenobiotic agents, providing the basis for speciesspecificity of xenobiotic responses. These workers go on to discussa humanized mouse model expressing SXR, which should prove quiteuseful in preclinical studies of metabolism and toxicology forcandidate pharmaceuticalagents.
As is clear from the scope of these reviews, nuclear receptors participate in the regulation of almost all biologic processes.Thus, understanding the function of these receptors should beuseful to a broad array of basic and clinicalscientists.
Because of their diverse biological effects, nuclear receptors have become major pharmaceutical targets in a host of diseasestates. Current pharmaceutical agents include natural hormonalligands or their analogs such as glucocorticoids, thyroid hormone,and estrogens, as well as ligands for the PPARa and -g receptors.Undoubtedly, many more therapeutically useful pharmaceutical agentsare on the horizon and will be entering the clinic in the nearfuture.
To whom correspondence should be addressed. Tel.: 858-534-6651; Fax: 858-534-6653; E-mail: [email protected]
Published, JBC Papers in Press, July 17, 2001, DOI 10.1074/jbc.R100047200
The abbreviations used are: NHR, nuclear hormone receptor; DBD, DNA binding domain; LBD, ligand binding domain; ER, estrogen receptor; ERE, estrogen response element; PPAR, peroxisome proliferator-activated receptor; TZD, thiazoladinedione.
|1.||Bourguet, W., Germain, P., and Gronemeyer, H. (2000) Trends Pharmacol. Sci. 21, 381-388|
|2.||Mangelsdorf, D. J., Thummel, C., Beato, M., Herrlich, P., Schültz, G., Umesono, K., Blumberg, B., Kastner, P., Mark, M., Chambon, P., and Evans, R. M. (1995) Cell 83, 835-839|
Source: J. Biol. Chem., Vol. 276, Issue 40, 36863-36864, October 5, 2001.
Fig. 1. Structure/function organization of nuclear receptors. The six domains (A-F) of nuclear receptors comprise regions of conserved function and sequence. All of the nuclear receptors contain a central DBD (region C), which is the most highly conserved domain and includes two zinc finger modules. A LBD (region E) is contained in the C-terminal half of the receptor. Situated between the DBD and LBD is a variable length hinge domain (region D), and variable N-terminal region (A/B) contains ampF-I activation function. Most receptors also contain a variable length C-terminal region F, the function of which is poorly understood. Many members of the nuclear receptor family form homo- or heterodimers, and amino acid sequences important for dimerization are contained within the DBD and LBD.
Source: J. Biol. Chem., Vol. 276, Issue 40, 36863-36864, October 5, 2001.