Abstract
Half of the members of the nuclear receptors superfamily are so-called “orphan” receptors because the identity of their ligand, if any, is unknown. Because of their important biological roles, the study of orphan receptors has attracted much attention recently and has resulted in rapid advances that have helped in the discovery of novel signaling pathways. In this review we present the main features of orphan receptors, discuss the structure of their ligand-binding domains and their biological functions. The paradoxical existence of a pharmacology of orphan receptors, a rapidly growing and innovative field, is highlighted.
Introduction
The cloning of genes encoding the specific receptors for known hormones such as steroids, thyroid hormones, and vitamin-derived compounds such as retinoids and vitamin D3, as well as functional demonstration of their implication in fundamental biological processes of therapeutic interest, led to an intensive search for related proteins predicted to share similar features (Mangelsdorf et al., 1995; Chambon, 1996). The defining structural and functional features of nuclear receptors are a conserved zinc finger DNA-binding domain (DBD1) and a ligand-binding domain (LBD). The evolutionary combination of these functional domains led to the generation of a diverse family of ligand-activated transcription factors that regulate gene expression in response to ligand binding. The high degree of similarity among the first receptors identified, both at the structural and functional levels, set the stage for the search for other family members, initially by low stringency screening of cDNA libraries and polymerase chain reaction screens with degenerate primers (Giguere et al., 1988; Wang et al., 1989; Becker-Andre et al., 1993) and more recently by genome sequence analysis (Robinson-Rechavi and Laudet, 2003). These efforts led to the successful identification of the vast majority of known nuclear receptors (NRs) without prior knowledge of their ligand and defined the gene family (Blumberg and Evans, 1998). In humans, these proteins, referred to as orphan nuclear receptors, still represent half of the total number of NRs (24 of a total of 48 different genes in human).
The discovery of the orphan NRs has raised several questions concerning their physiological functions and the existence of specific ligand(s) and possibly new endocrine systems and has shifted “endocrinology into reverse” (Kliewer et al., 1999; Shiau et al., 2001). Thus, the search for biological function and ligands for orphan NRs has become the subject of intense investigation. In this introductory review we will briefly present these molecules and their diverse biological functions and discuss how the search for ligands has led to a refinement of our definition of a NR ligand.
What Are Orphan Receptors?
The definition of orphan receptors is a loose and paradoxical one because, by definition, orphan receptors are receptors for which no ligand is known. The term “receptor” itself implies that a physiological ligand should exist, even though there is still no consensus in the field as to whether this will be true for all orphan NRs. Because the absence of proof is not the proof of absence, it is extremely difficult to demonstrate that a given orphan NR truly has no endogenous ligand. Complicating the issue is the fact that once a natural ligand has been discovered for an orphan NR, the receptor is no longer considered to be an orphan, despite the fact that it may retain structural and functional features more similar to the other orphan NRs than to the classic steroid and thyroid hormone receptors. Two prime examples are the RXRs and PPARs, which were discovered as orphan NRs, but which are now clearly considered to be liganded receptors, although the precise identity of their physiological, endogenous ligands is somewhat controversial (Gottlicher et al., 1992; Heyman et al., 1992; Kitareewan et al., 1996; Lemotte et al., 1996; Mata de Urquiza et al., 2000; Willson et al., 2000; Lengqvist et al., 2004). Together with the RXRs and PPARs, the FXRs, LXRs, CAR, and PXR have been classified as a new type of NRs that are considered natural sensors (Janowski et al., 1996; Lehmann et al., 1998; Kawamoto et al., 1999; Makishima et al., 1999; Tzameli et al., 2000; Tzameli and Moore, 2001; Francis et al., 2003). The ligand-binding pocket of these receptors is larger than those of classic receptors (such as RARs, TRs, or steroid receptors), and they bind a large diversity of molecules with lower affinity (typically in the micromolar range) (Benoit et al., 2004). Even though some compounds were found inside the pocket of some orphan receptors such as HNF-4, RORs, or SF-1, they still are firmly part of the orphan receptor group because the regulatory role of the compound is unclear and/or the physiological relevance of the interaction with the receptor has not been clearly established (Dhe-Paganon et al., 2002; Kallen et al., 2004; Li et al., 2005; Stehlin et al., 2001; Wisely et al., 2002). Thus, the composition of the orphan receptor group is likely to continue to shrink in the future.
Following this definition, the orphan receptors form a highly diverse group. In fact, orphan receptors are not linked functionally or evolutionarily. In phylogenetic trees of NRs, they are scattered among the six defined subfamilies (Escriva et al., 2000). In addition, their structures are also highly diverse, not only at the structural level within the LBD as discussed below but also in the other domains (Fig. 1). Indeed, some orphan receptors have only one of the two characteristic domains of the NR superfamily. In vertebrates, DAX-1 and SHP, which contain only an LBD and lack a classic DBD with conserved cysteines as do the other receptors, are examples of such divergent orphan receptors (Zanaria et al., 1994; Burris et al., 1996; Seol et al., 1996). In other species (e.g., Drosophila or nematodes), there are several other examples of receptors containing only the LBD or only the DBD sequence. The size of the other domains is also variable; the A/B region of some orphan receptors is extremely short: 8 amino acids for some isoforms of RORβ and 14 amino acids in TLX, whereas in other cases this domain is quite long (250-280 amino acids for NGFI-B/NR4A group members). Like some liganded receptors, such as the RARs, the HNF-4 group members contain an F domain that modulates their transcriptional activities (Ruse et al., 2002).
The diversity of orphan receptors is also illustrated by various modes of binding to DNA. Although most of them seem to bind to DNA as homodimers on direct repeat elements (HNF-4, COUP-TFs, and TR2/4), some interact with RXRs (NGFI-B and NURR1) (Perlmann and Jansson, 1995), and probably the most singular example of a DNA-binding mechanism is the oligomerization of the orphan GCNF upon binding to a direct repeat AGGTCAAGGTCA (Gu et al., 2005c). This divergent DNA-binding mechanism of GCNF, hexamer formation, is probably a reflection of its being the only member in the distant sub-branch 6 of the superfamily. Importantly, the study of several orphan receptors (Reverbs, RORs, SF-1, NGFI-B, NURR1, NOR1, and ERRs) allowed definition of a new type of interaction with DNA, namely, monomer binding to half-site sequences (Wilson et al., 1993). Even though such an ability has been found in a few cases for liganded receptors (e.g., TRα), the functional relevance of monomeric binding is clear only for orphan receptors. In all cases, binding occurs on a conserved A/GGGTCA binding motif that is preceded by an A/T-rich region in 5′. The sequence of this A/T-rich region is variable from one receptor type to another. SF-1, LRH-1, and ERRs bind to TCAA/GGGTCA elements (called SFRE or ERRE) (Honda et al., 1993; Sladek et al., 1997), whereas NGFI-B/NR4A group members bind to AAA/GGGTCA elements (called NBRE) (Wilson et al., 1991). Lastly, Rev-erbs and RORs bind to a less constrained sequence, the consensus of which is A/TAA/TNTA/GGGTCA and is termed a RevRE or a RORE (Harding and Lazar, 1993; Giguère et al., 1994). In addition, Rev-erbs have been described to bind as homodimers to special DR2 elements, called RevDR2, in which the 5′ element, a RevRE, and the 3′ element, a classic A/GGGTCA, are separated by two bases, most often CT (Harding and Lazar, 1995). In all of these cases of monomeric binding to extended half-site sequences, the interaction between the receptor and DNA is in the A/GGGTCA motif, with a recognition helix at the C-terminal part of the first zinc finger interacting with the major groove of DNA and making specific contacts with the A/GGGTCA motif. A second helix in the second zinc finger stabilizes the interaction with DNA and allows dimerization with partners, when partners are present. In addition, the C-terminal part of the receptor is able to interact specifically with the extended 5′ element. Several detailed functional studies plus structural analyses, including one of the Rev-erb DBD associated with DNA, led to the identification of a region beyond the core DBD (C domain), called the A box, that forms a third α-helix of the DBD and is implicated in the recognition of the 5′ extension of the DNA element (Wilson, 1993; Rastinejad et al., 1995). In fact, it has now been shown that variations of this structural element can be found in liganded receptors, such as TRs or RXRs. This is a nice illustration of the impact that orphan receptors can have on the study and understanding of liganded receptors.
Phylogenetic tree and schematic structure of orphan nuclear receptors present in human, mouse, and rat.
Biological Functions of Orphan Receptors
Given the wide diversity of orphan receptors, it is, of course, very difficult to summarize their biological functions (Giguere, 1999). Two points are nevertheless important to mention and to discuss. 1) All orphan receptors have a very important function that is specific to each one of them. Thus, they are not inert molecules, less important than classic receptors. Indeed, gene targeting in the mouse has revealed important, often essential, roles for orphan NRs in development and adult physiology. 2) Orphan receptors often play an important role in modulating the action of classic liganded receptors.
It is possible to generate a very short summary of the functions played by these molecules and because the function of most of them has been inactivated in the mouse or in other biological models, we have a fairly clear understanding of their role, even if, of course, many questions remain. Many orphan receptors are important players in development and cell differentiation. For example, HNF-4α is critical for early mouse development as well as for the development of the liver in vertebrates and arthropods (Watt et al., 2003). COUPTFs have a conserved fundamental role in nervous system development as illustrated in mouse, zebrafish, and even hydra (Cooney et al., 2001) as well as in organo-genesis of various organs (Park et al., 2003). The three NGFI-B members are also important players in brain development (Perlmann and Wallen-Mackenzie, 2004) and in T-cell biology (He, 2002). ERRα in zebrafish (Bardet et al., 2005) and ERRβ in mouse (Luo et al., 1997) have crucial roles in very early development, for example, during gastrulation. In addition ERRα has been proposed to regulate osteoblastic differentiation at later stages of mouse embryogenesis (Bonnelye et al., 1997). GCNF has been shown to play an important role in early mouse development (Chung et al., 2001). GCNF plays a pivotal role in the silencing of pluripotency gene expression at gastrulation and ES cell differentiation (Fuhrmann et al., 2001; Gu et al., 2005b). DAX-1 is specifically implicated in germ cell differentiation in mouse (Achermann, 2005; Zechel, 2005), whereas the TLX receptor and PNR also play an important role during development, for example, in retina formation (Kobayashi et al., 1999). SF-1 plays a central role in the development of steroidogenic tissues, the adrenals and gonads, whereas its close homolog LRH-1 plays a role in endoderm differentiation and maintenance of pluripotence in early embryos (Pare et al., 2004; Gu et al., 2005a).
In addition, orphan receptors have also an important role in adult physiology in regulating metabolism. This is the case for ERRα, which is important for adipogenesis and energy metabolism (Sladek and Giguere, 2000; Luo et al., 2003), but also for the previously mentioned LRH-1 and SF-1, which are critical players in the regulation of cholesterol metabolism in the liver as well as in steroidogenic tissues (Fayard et al., 2004). HNF-4s, COUP-TFs, Rev-erbs, and RORs also play a role in regulating metabolism (especially in cholesterol and fatty acid metabolism) (Jetten et al., 2001; Jetten, 2004; Laitinen et al., 2005), although their specific functions are not yet well understood.
Finally, an emerging, yet poorly characterized, role for orphan receptors is in the regulation of circadian rhythm, a function probably tightly linked to their role in metabolism (Inoue et al., 2005). Rev-erbs and RORs are prominent members of the circadian pacemaker in peripheral tissues as well as in the master clock organ, the suprachiasmatic nucleus (Alvarez and Sehgal, 2002; Preitner et al., 2002, 2003; Emery and Reppert, 2004; Jetten, 2004; Triqueneaux et al., 2004; Guillaumond et al., 2005). Other orphan receptors, such as ERRα, SHP, or EAR2, are also expressed in a circadian manner, and it is interesting to note that the knockout of EAR2 in the mouse exhibits a circadian phenotype (Horard et al., 2004; Warnecke et al., 2005).
Many experiments have demonstrated that orphan receptors regulate the activity of liganded receptors. This is the case for COUP-TFs, TR2, and TR4 (and also to a lesser extent for HNF-4s), which have been shown to repress the activation mediated by liganded receptors such as RAR, TR, or PPAR (Lee et al., 2002; Park et al., 2003). DAX-1 and in a broader sense SHP are regulators of the activity of other receptors (either orphan or liganded) by direct interaction with these receptors (Zhang and Dufau, 2004; Bavner et al., 2005). These highly unusual members of the NR superfamily can even be described as corepressors because they do not bind DNA and do not dimerize with other NRs through the canonical homo-/heterodimerization interface but rather through the cofactor interface. Another case is the connection that exists between ERRs and estrogen signaling. It has been shown that ERRs and ERs share both structural and functional attributes, such as synthetic ligands, interactions with coactivators, and binding to similar DNA sequences in vitro (Giguere, 2002). Many of these connections were found during the early days of orphan receptor research, when researchers were avidly searching for a functional role of these molecules. Thus, it has to be emphasized that many of these experiments were done in transient transfection assays and that in some cases their biological relevance in vivo still awaits confirmation.
Unorthodox LBD for Unorthodox NRs
Two main strategies were developed to search for orphan receptor ligands. These were based either on the search for the ligand per se by focused or random screening of naturally occurring or synthetic compounds (Chawla et al., 2001) or alternatively through the resolution of the structure of the NR LBD by X-ray crystallography. The successful identification of fatty acids, oxysterols, and bile acids as naturally occurring agonists of the PPARs (Gottlicher et al., 1992), the LXRs (Janowski et al., 1996), and the FXR (Makishima et al., 1999; Parks et al., 1999), respectively, led to the suggestion that all NRs may be ligand-regulated. However, it seems that some orphan NRs were resistant to the traditional screening approaches, especially those displaying some level of constitutive activation (i.e., NGFI-B and NURR1) or repression (i.e., Rev-erbs). More recently, evolutionary studies have suggested and structural studies have shown that there are orphan NRs, in which the LBD can carry out its regulatory functions without the need for a ligand.
In contrast to classic liganded receptors, many orphan receptors show a “constitutive” AF-2-dependent transcriptional activity in different biological systems. How this group of receptors (RORs, ERRs, HNF4s, NURRs, and LRH-1) achieves this constitutive activity is of considerable interest to the understanding the function of these orphan NRs. This question was recently elucidated by analysis of the crystal structure of their LBDs. Interestingly, the answers seem to be just as diverse as the LBD structure is conserved, further indicating that the NR family evolved in multiple directions and took advantage of a single structure, namely the LBD, to achieve different physiological functions. In contrast to the screening approaches, which focused on the ligand only, the characterization of the crystal structure of NR LBDs yielded insights into the capacity of a given NR to be regulated by a ligand and in some cases even led to the direct identification of cocrystallized chemical compounds. In addition, and most importantly, the resolution of these structures has led to a refinement of the definition of what is a ligand of a NR. There are four different possibilities for orphan receptors and their potential ligand: 1) receptors with no ligand-binding pocket at all; 2) receptors with empty ligand-binding pockets; 3) receptors with structural ligands; and 4) receptors regulated by ligands, but the physiological relevance of those remains an open question. There are still many receptors, for which we simply have no clear information. This fifth category is the largest one and contains the COUP-TFs, GCNF, TLX, PNR, TR2/4, and DAX/SHP. We will now briefly examine the four possibilities.
The case of NURR1 is probably the most convincing for a receptor containing no ligand-binding pocket (LBP) (Wang et al., 2003). Because the Drosophila homolog of NURR1 (Baker et al., 2003), called DHR38, has the same features as NURR1, it has been suggested that the two other members of the group are characterized by a comparable structure. This hypothesis was since verified for NGFI-B (Flaig et al., 2005). The NGFI-B/NR4A group members form a branch of nuclear receptor homologs expressed in various cell types. When transfected into mammalian cells, all NR4A family members act as constitutively active transcription factors, and all early attempts to define ligands for them have failed. Interestingly, the crystallographic analysis of the NURR1 and DHR38 LBDs reveals that these proteins lack a ligand-binding pocket. The overall structures are very similar to the canonical LBD fold (Wurtz et al., 1996), but bulky amino acid side chains occupy the space that would normally form the LBP. In the crystal structure of the NURR1 LBD, helix 12 is in the active conformation explaining the known constitutive activity of NR4A receptors. It was also revealed that residues conserved between canonical receptors forming the so-called “charge clamp” region, which is required for coactivator binding, are substituted in the NURR1 LBD. These observations and the fact that the NURR1 LBD does not interact with classic p160 coactivators raise questions about the mechanism by which NURR1 activates transcription. More recently, several studies revealed the existence of an alternative coactivator binding cleft (Codina et al., 2004, 2005; Flaig et al., 2005; Volakakis et al., 2006), but cofactors for NR4A family members that use that cleft have not yet been identified. It is now evident that despite obvious similarity to canonical LBDs, NR4A family members are not receptors and are regulated at the level of their expression or via post-translational modifications triggered by intracellular signaling pathways. The Rev-erbs, which are potent transcriptional repressors, are also good candidates to be orphan receptors with no LBP, because modeling studies have suggested that, as for NURR1, the LBP is filled with amino acid side chains (Renaud et al., 2000). A note of caution is nevertheless required here, because this result is only based on modeling studies, not on the experimental determination of the structure. Interestingly, the Drosophila homolog of Rev-erb, E75, is regulated by a unique mechanism. E75, in fact, contains a heme prosthetic group in the LBP, which by controlling the oxidation state of the heme iron, gases, such as nitric oxide or carbon monoxide, controls the activity of the receptor (Reinking et al., 2005). This exemplifies the unexpected diversity of mechanisms that regulate orphan NR activity.
The second interesting case is represented by orphan receptors that have a constitutive activity but in which an empty LBP has been observed. This is the case for ERRγ, for which the structure of the LBD in complex with the SRC-1 peptide was determined (Greschik et al., 2002). This structure reveals a helix 12 in active conformation and a small, but empty, LBP. Similar results were obtained more recently with the structure of the ERRα LBD in complex with a PGC1 peptide (Kallen et al., 2004). There are known examples of compounds able to block the constitutive activity of ERRγ. Notably, these are widely used synthetic antiestrogens, such as 4-hydroxytamoxifen (Coward et al., 2001). Interestingly, and in contrast to other cases such as RORβ (see below), the activity of ERRγ does decline after LBP is blocked with bulky amino acid side chains. Antiestrogens, however, no longer inactivate such mutants. In summary, ERRγ and possibly ERRα and ERRβ are nuclear receptors that are activated by default, but which are also capable of responding to deactivating ligands (Coward et al., 2001; Tremblay et al., 2001a,b), although the physiological relevance of these ligands has not yet been demonstrated. Mouse LRH-1 is another example of an orphan receptor with a large and empty hydrophobic pocket (Sablin et al., 2003), but because this feature seems to be different for the human LRH-1, it will be further discussed below.
Several receptors contain ligands that are unable to leave the receptor and are in fact part of the structure itself. These “structural ligands” that behave as prosthetic groups are usually fatty acids or fatty acid derivatives. This fact is illustrated by the HNF-4s, which form another distinct group of constitutively active nuclear receptors (Dhe-Paganon et al., 2002; Wisely et al., 2002). Analysis of the structure of the HNF-4γ LBD revealed the presence of fatty acids, which could not be displaced from the LBP without protein denaturation (Wisely et al., 2002a). The helix 12 was in an active conformation, and mutations of the LBP designed to prohibit the binding of fatty acids reduced the constitutive activity of the receptor. Thermal denaturation studies of mutated HNF-4α derivatives, however, indicated reduced stability of variants unable to bind fatty acids. Taken together, these studies suggest that HNF4s evolved to position their AF-2 into active conformation without the involvement of the ligand but instead need the help of lipophilic fatty acids to globally fold the LBD. Therefore, HNF-4s may not be nuclear receptors in the classic sense. However, the activity of HNF-4s is regulated at the transcriptional level and by coexpression of regulatory factors, such as SHP. Two other examples in insects, Drosophila USP (Billas et al., 2001; Clayton et al., 2001) and E75 (as mentioned above), are also cases of receptors bound to prosthetic groups, namely, a phospholipid and a heme, respectively.
Finally, researchers are currently in the process of identifying ligands for some orphan NRs such as RORs, LRH-1 and SF-1, the members of subfamily 5. RORα is constitutively active, and its LBD can efficiently recruit p300 and glucocorticoid receptor interacting protein co-activators, which shifts the AF-2 into the active conformation. A surprising feature of the RORα LBP is that this domain copurifies with a cholesterol molecule inside (Kallen et al., 2002, 2004). Interestingly, cholesterol bound in the LBP can be exchanged with cholesterol sulfate, which, using structural predictions, is expected to be a more potent ligand. In addition, changes in the intracellular level of cholesterol modulate RORα transcriptional activity. These results suggest that RORα could potentially serve as a cellular cholesterol “sensor” (Willson, 2002). Cholesterol fills the RORα LBP either to stabilize helix 12 in an active conformation or to globally assist the folding of the LBD. The latter possibility is compatible with the cholesterol receptor hypothesis for RORα. In summary, RORα differs from canonical nuclear receptors in that it is bound to its ligand constitutively, but reversibly. The LBP of the RORβ nuclear receptor, like its homolog RORα, was originally crystallized together with a fortuitously captured molecule of stearic acid (Stehlin et al., 2001). Additional experiments established that stearate did not fulfill the criteria for a true RORβ ligand. Because mutagenesis studies designed to block the RORβ ligand-binding pocket yielded inactive receptors, the search for a RORβ ligand continued. This resulted in the discovery that the well-known RAR natural agonist all-trans-retinoid acid binds the RORβ LBD with low, but biologically relevant, affinity (Stehlin-Gaon et al., 2003). In addition, all-trans-retinoic acid acts as a partial cell-type specific antagonist for RORβ. Many questions remain concerning the in vivo relevance of these interesting observations, and more work is needed before RORα can be considered a real cholesterol sensor and RORβ a third type of retinoic acid receptor.
An even more striking scenario is represented by LRH-1 and SF-1. The mouse LRH-1 LBD assembles into the active conformation with a large, but empty, LBP (Sablin et al., 2003). The ability of helix 12 of LRH-1 to associate with the LBD core was attributed to an unusual helix 2 structure, which forms a unique fourth outer layer of the LBD and actively contributes to the maintenance of the basal activity of the receptor as demonstrated by site-directed mutagenesis (Sablin et al., 2003). Additional experiments aimed at artificially “filling” the LRH-1 LBP with bulky amino acid side chains resulted in an increase of basal activity of the receptor, suggesting that mouse LRH-1 is still ligand-responsive. Strikingly, the determination of the structure of the human LRH-1 as well as of mouse and human SF-1 shows that these receptors, in contrast to mouse LRH-1, bind phosphatidyl inositol second messengers and that ligand binding is required for maximal activity (Krylova et al., 2005; Li et al., 2005; Wang et al., 2005). In line with these findings, mutations of specific amino acids that are part of the LBP of mouse SF-1 induce a loss of activity. The question, of course, remains whether these “fortuitous” ligands that were discovered because they were captured in the LBP during overexpression in bacteria are natural ligands or are at least indicative of the existence of natural ligands. An important and still unanswered question is whether these ligands can really enter and leave the LBD freely (i.e., act as bona fide signaling molecules) and by doing so regulate its transcriptional activity.
All of these observations illustrate the tremendous diversity that exists for orphan receptors with respect to their relationships with small molecules and allows us to redefine the term “ligand” for nuclear receptors. Historically, since the discovery of the superfamily started with the characterization of steroid receptors (i.e., receptors with nanomolar affinity for very selective ligands that are typically hormones synthesized in specific tissues in the organism), it was thought that most, if not all, NRs should have ligands with similar characteristics. The characterization of “sensors” such as PPARs, LXRs, FXRs, PXR/CAR, and even RXRs has prompted a reevaluation of this definition because the LBDs of these receptors bind a large number of molecules, often derived from food or intermediate metabolism products and present at very high physiological concentrations relative to steroid hormones, with a much lower affinity (typically in the micromolar range). Thus, it became clear that NRs do bind not only hormones or morphogens, such as retinoic acid, but also a much broader set of small molecules. If we consider all orphan receptors, we can see that there is a continuum between small molecules forming prosthetic groups tightly linked to the receptor and exchangeable molecules with signaling activities. The main challenge for future work will be to decipher which of the newly discovered “ligands” of orphan receptors are physiologically relevant molecules with a signaling role (i.e., carrying biological information). For this, the emphasis will have to shift from structural studies that were extremely powerful in revealing the nature of these molecules to in vivo analyses of their biological role.
However, in trying to physiologically link a new ligand to an orphan receptor, an instructive paradigm to look at is the estrogen receptors and the enzyme, aromatase, that produces their ligand. The reproductive roles of the estrogen receptors α and β were succinctly determined by gene targeting (Lubahn et al., 1993; Krege et al., 1998; Dupont et al., 2000). The reproductive phenotype of the aromatase knockout only reinforced its role in producing the signal that regulates the estrogen receptors (Fisher et al., 1998; Honda et al., 1998; Nemoto et al., 2000). Thus, inactivation of an enzyme that produces a putative ligand should phenocopy part or all of a receptor phenotype.
The Paradox: Toward Pharmacology of Orphan Receptors?
To conclude, one cannot help but comment that this wide diversity of mechanisms is very good news for pharmacologists. Examples such as ERRγ clearly show that even if an orphan NR has apparently no ligand and an empty pocket, it can still be a valid pharmaceutical target, potentially bound and regulated by drug molecules. The same is likely to be true for other receptors, such as LRH-1, SF-1, and the RORs, making them promising pharmaceutical targets as well.
Finally, it is also important to note that numerous orphan receptors form heterodimers with RXR (Mangelsdorf and Evans, 1995). This functional property seems to be critical for the true orphans of the NR4A subfamily (NGFI-B and NURR1 but not NOR1) (Perlmann and Jansson, 1995; Zetterstrom et al., 1996), because these heterodimers were shown to be responsive to RXR specific ligands, adding yet another mechanism by which the transcriptional activity of these physiologically essential receptors can be regulated.
Tables 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25 summarize the functions, biologic activities, structural properties, and ligands of these receptors.
DAX-1
SHP
Rev-erbα
Rev-erbβ
RORα
RORβ
RORγ
HNFα
HNF-4γ
TR2
TR4
TLX
PNR
COUP-TFI
COUP-TFII
EAR2
ERRα
ERRβ
ERRγ
NGFI-B
NURR1
NOR1
SF-1
LRH-1
GCNF
Footnotes
-
↵1 Abbreviations: DBD, DNA-binding domain; LBD, ligand-binding domain; NR, nuclear receptor; RXR, retinoid X receptor; PPAR, peroxisome proliferator-activated receptor; FXR, farnesoid X receptor; LXR, liver X receptor; CAR, constitutive androstane receptor; PXR, pregnane X receptor; RAR, retinoic acid receptor; TR, thyroid hormone receptor; HNF, hepatocyte nuclear factor; ROR, retinoid-related orphan receptor; SF-1, steroidogenic factor 1; DAX-1, dosage-sensitive sex reversal-adrenal hypoplasia congenital critical region on the X chromosome protein 1; SHP, small heterodimer partner; TLX, tailless; NGFI-B, nerve growth factor-induced clone B; COUP-TF, chicken ovalbumin upstream promoter transcription factor; TR2, testicular receptor 2; TR4, testicular receptor 4; NURR1, Nur-related factor 1; GCNF, germ cell nuclear factor; NOR1, neuron-derived orphan receptor 1; ERR, estrogen-related receptor; LRH, liver receptor homolog; DR, direct repeat; PNR, photoreceptor-specific nuclear receptor; EAR, ErbA-related protein; ER, estrogen receptor; AF, activation factor; LBP, ligand-binding pocket.
-
J.-P.R. is President and CEO/CSO of AliX, S.A., a biopharmaceutical company working in the field of orphan nuclear receptors.
-
Article, publication date, and citation information can be found at http://pharmrev.aspetjournals.org.
-
doi:10.1124/pr.58.4.10.
- The American Society for Pharmacology and Experimental Therapeutics