Gaining ligand selectivity in thyroid hormone receptors via entropy.

Nuclear receptors are important targets for pharmaceuticals, but similarities between family members cause difficulties in obtaining highly selective compounds. Synthetic ligands that are selective for thyroid hormone (TH) receptor beta (TRbeta) vs. TRalpha reduce cholesterol and fat without effects on heart rate; thus, it is important to understand TRbeta-selective binding. Binding of 3 selective ligands (GC-1, KB141, and GC-24) is characterized at the atomic level; preferential binding depends on a nonconserved residue (Asn-331beta) in the TRbeta ligand-binding cavity (LBC), and GC-24 gains extra selectivity from insertion of a bulky side group into an extension of the LBC that only opens up with this ligand. Here we report that the natural TH 3,5,3'-triodothyroacetic acid (Triac) exhibits a previously unrecognized mechanism of TRbeta selectivity. TR x-ray structures reveal better fit of ligand with the TRalpha LBC. The TRbeta LBC, however, expands relative to TRalpha in the presence of Triac (549 A(3) vs. 461 A(3)), and molecular dynamics simulations reveal that water occupies the extra space. Increased solvation compensates for weaker interactions of ligand with TRbeta and permits greater flexibility of the Triac carboxylate group in TRbeta than in TRalpha. We propose that this effect results in lower entropic restraint and decreases free energy of interactions between Triac and TRbeta, explaining subtype-selective binding. Similar effects could potentially be exploited in nuclear receptor drug design.

Differential effects of TR ligands on hormone dissociation rates: evidence for multiple ligand entry/exit pathways.

Some nuclear receptor (NR) ligands promote dissociation of radiolabeled bound hormone from the buried ligand binding cavity (LBC) more rapidly than excess unlabeled hormone itself. This result was interpreted to mean that challenger ligands bind allosteric sites on the LBD to induce hormone dissociation, and recent findings indicate that ligands bind weakly to multiple sites on the LBD surface. Here, we show that a large fraction of thyroid hormone receptor (TR) ligands promote rapid dissociation (T(1/2)<2h) of radiolabeled T(3) vs. T(3) (T(1/2) approximately 5-7h). We cannot discern relationships between this effect and ligand size, activity or affinity for TRbeta. One ligand, GC-24, binds the TR LBC and (weakly) to the TRbeta-LBD surface that mediates dimer/heterodimer interaction, but we cannot link this interaction to rapid T(3) dissociation. Instead, several lines of evidence suggest that the challenger ligand must interact with the buried LBC to promote rapid T(3) release. Since previous molecular dynamics simulations suggest that TR ligands leave the LBC by several routes, we propose that a subset of challenger ligands binds and stabilizes a partially unfolded intermediate state of TR that arises during T(3) release and that this effect enhances hormone dissociation.

Human thyroid receptor forms tetramers in solution, which dissociate into dimers upon ligand binding.

Thyroid hormone nuclear receptors (TRs) bind to DNA and activate transcription as heterodimers with the retinoid X receptor (RXR) or as homodimers or monomers. RXR also binds to DNA and activates transcription as homodimers but can, in addition, self-associate into homotetramers in the absence of ligand and DNA templates. It is thought that homotetramer formation serves to sequester excess RXRs into an inactive pool within the cell. Here, we report systematic studies of the multimeric state of a recombinant human TRbeta1 truncation (hTRbeta1deltaAB) that encompasses the complete DNA binding domain and ligand binding domain in solution. Native gel electrophoresis, chemical crosslinking, gel filtration, and dynamic light scattering experiments reveal that hTRbeta1deltaAB forms a mixture of monomers, dimers, and tetramers. Like RXR, increasing protein concentration shifts the equilibrium between TR multimers toward tetramer formation, whereas binding of cognate thyroid hormone leads to dissociation of tetramers and increased formation of dimers. This work represents the first evidence that apo-hTRbeta1 forms homotetramers. The findings raise the possibility that tetramer formation provides an additional, and previously unsuspected, level of control of TR activity and that the capacity for homotetramer formation may be more widespread in the nuclear receptor family than previously thought.

Hammett analysis of selective thyroid hormone receptor modulators reveals structural and electronic requirements for hormone antagonists.

Selective thyroid hormone modulators that function as isoform-selective agonists or antagonists of the thyroid hormone receptors (TRs) might be therapeutically useful in diseases associated with aberrant hormone signaling. The most potent thyroid hormone antagonist reported to date is NH-3. To explore the significance of the 5'-p-nitroaryl moiety of NH-3 and understand what chemical features are important to confer antagonism, we sought to expand the structure-activity relationship data for the class of 5'-phenylethynyl GC-1 derivatives. Herein, we describe an improved synthetic route utilizing palladium-catalyzed chemistry for efficient access to a series of 5'-phenylethynyl compounds with varying size and electronic properties. We prepared and tested sixteen analogues for TR binding and transactivation activity. Substitution at the 5'-position decreased binding affinity, but retained TRbeta-selectivity. In transactivation assays, the analogues displayed a spectrum of agonist, antagonist, and mixed agonist/antagonist activity that correlated with electronic character in a Hammett analysis between sigma substituent value and TR modulation. Analogues NH-5, NH-7, NH-9, NH-11, and NH-23 displayed full antagonist activity with reduced potency compared to NH-3, indicating the nitro group is not required for antagonism. However, para-substitution with strong electron withdrawing properties on the 5'-aryl extension is important for antagonist activity, and antagonist potency-but not ligand receptor binding-was found to correlate linearly with the sigma values for the electron withdrawing substituents.

Thyroxine-thyroid hormone receptor interactions.

Thyroid hormone (TH) actions are mediated by nuclear receptors (TRs alpha and beta) that bind triiodothyronine (T(3), 3,5,3'-triiodo-l-thyronine) with high affinity, and its precursor thyroxine (T(4), 3,5,3',5'-tetraiodo-l-thyronine) with lower affinity. T(4) contains a bulky 5' iodine group absent from T(3). Because T(3) is buried in the core of the ligand binding domain (LBD), we have predicted that TH analogues with 5' substituents should fit poorly into the ligand binding pocket and perhaps behave as antagonists. We therefore examined how T(4) affects TR activity and conformation. We obtained several lines of evidence (ligand dissociation kinetics, migration on hydrophobic interaction columns, and non-denaturing gels) that TR-T(4) complexes adopt a conformation that differs from TR-T(3) complexes in solution. Nonetheless, T(4) behaves as an agonist in vitro (in effects on coregulator and DNA binding) and in cells, when conversion to T(3) does not contribute to agonist activity. We determined x-ray crystal structures of the TRbeta LBD in complex with T(3) and T(4) at 2.5-A and 3.1-A resolution. Comparison of the structures reveals that TRbeta accommodates T(4) through subtle alterations in the loop connecting helices 11 and 12 and amino acid side chains in the pocket, which, together, enlarge a niche that permits helix 12 to pack over the 5' iodine and complete the coactivator binding surface. While T(3) is the major active TH, our results suggest that T(4) could activate nuclear TRs at appropriate concentrations. The ability of TR to adapt to the 5' extension should be considered in TR ligand design.

Towards selectively modulating mineralocorticoid receptor function: lessons from other systems.

Although there is clinical utility in blocking mineralocorticoid receptor (MR) action, the usefulness of available MR antagonists is limited because of cross-reactivity with the androgen and progesterone receptors (spironolactone) or possibly by low affinity for MR (eplerenone). MR binds aldosterone and physiologic glucocorticoids, such as cortisol, which both can act as MR agonists in epithelial tissues. However, in preliminary studies aldosterone and cortisol appear to induce different conformations in non-epithelial tissues; in the cardiomyocyte, cortisol usually acts as an MR antagonist, whereas in vascular smooth muscle cortisol mimics aldosterone actions if it can access MR, just as it does in the kidney. Thus, there are needs for improved MR antagonists with higher selectivity and potency and, if possible, for compounds that lock MR into specific desirable conformations. Efforts are underway to modulate selectively the action of many nuclear receptors, and insights from one nuclear receptor may be applicable to others given the similarities in structure and function. We have used traditional approaches aided by X-ray crystallography to obtain several classes of selective ligands that modulate thyroid receptor (TR) action. We describe the properties of these selective TR modulators here, and discuss the possibility that similar approaches to ligand design may yield MR interacting compounds with improved specificity and, possibly, tissue specificity.

Structural determinants of selective thyromimetics.

The thyromimetic GC-1 shows a preference for binding the beta form of the thyroid hormone receptor (TR). GC-1 was designed as an analogue of the thyromimetic DIMIT, which has a lower affinity for TR and is not selective. GC-1 has a methylene group linking its two aromatic rings and an oxyacetic acid polar side chain, while DIMIT has an ether oxygen linking its aromatic rings and an l-alanine polar side chain. The structural features of GC-1 that confer its greater affinity and selectivity compared to DIMIT were analyzed with the preparation of analogues that bear only one of their two different structural features. The analogue of GC-1 with a biaryl ether has selectivity comparable to that of GC-1, while the analogue of DIMIT with a methylene group linking its aromatic rings is only slightly selective. These results demonstrate that the oxyacetic acid side chain of GC-1 is critical in conferring TR-beta selectivity.

Two resistance to thyroid hormone mutants with impaired hormone binding.

Resistance to hormones is commonly due to mutations in genes encoding receptors. Resistance to thyroid hormone is due mostly to mutations of the beta-form of the human (h) thyroid hormone receptor (hTRbeta). We determined x-ray crystal structures of two hTRbeta ligand-binding domains (LBDs), Ala 317 Thr and Arg 316 His. Amino acids 316 and 317 form part of the hormone-binding pocket. The methyl of Ala 317, contacting iodine, sculpts the T3 hormone-binding pocket. Arg 316 is not in direct contact with T3 and has an unknown role in function. Remarkably, the Arg forms part of an unusual buried polar cluster in hTRbeta. Although the identity of the amino acids changes, the polar cluster appears in all nuclear receptors. In spite of the differing roles of 316 and 317, both resistance to thyroid hormone mutants display decreased T3 affinity and weakened transcriptional activation. The two mutants differ in that the Arg 316 His receptor does not form TR-TR homodimers on DNA. 3,5,3'-Triiodothyroacetic acid is bound to both receptors. Thr 317 repositions 3,5,3'-triiodothyroacetic acid distending the face of the receptor that binds coregulators. Arg 316 forms two hydrogen bonds with helix 1. Both are lost with mutation to His displacing helix 1 of the LBD and disordering the loop after helix 1. The stability of the helix 1, deriving in part from the buried polar cluster, is important for hormone binding and formation of TR dimers. The observation that the Arg 316 His mutation affects these functions implies a role for helix 1 in linking hormone binding to the DNA-binding domain-LBD configuration.

Low resolution structures of the retinoid X receptor DNA-binding and ligand-binding domains revealed by synchrotron X-ray solution scattering.

Nuclear receptors are ligand-inducible transcription factors that share structurally related DNA-binding (DBD) and ligand-binding (LBD) domains. Biochemical and structural studies have revealed the modular nature of DBD and LBD. Nevertheless, the domains function in concert in vivo. While high-resolution crystal structures of nuclear receptor DBDs and LBDs are available, there are no x-ray structural studies of nuclear receptor proteins containing multiple domains. We report the solution structures of the human retinoid X receptor DBD-LBD (hRXRalphaDeltaAB) region. We obtained ab initio shapes of hRXRalphaDeltaAB dimer and tetramer to 3.3 and 1.7 nm resolutions, respectively, and established the position and orientation of the DBD and LBD by fitting atomic coordinates of hRXRalpha DBD and LBD. The dimer is U-shaped with DBDs spaced at approximately 2 nm in a head to head orientation forming an angle of about 10 degrees with respect to each other and with an extensive interface area provided by the LBD. The tetramer is a more elongated X-shaped molecule formed by two dimers in head to head arrangement in which the DBDs are extended from the structure and spaced at about 6 nm. The close proximity of DBDs in dimers may facilitate homodimer formation on DNA; however, for the homodimer to bind to a DNA element containing two directly repeated half-sites, one of the DBDs would need to rotate with respect to the other element. By contrast, the separation of DBDs in the tetramers may account for their decreased ability to recognize DNA.

Rational design and synthesis of a novel thyroid hormone antagonist that blocks coactivator recruitment.

Recent efforts have focused on the design and synthesis of thyroid hormone (T(3)) antagonists as potential therapeutic agents and chemical probes to understand hormone-signaling pathways. We previously reported the development of novel first-generation T(3) antagonists DIBRT, HY-4, and GC-14 using the "extension hypothesis" as a general guideline in hormone antagonist design.(1-3) These compounds contain extensions at the 5'-position (DIBRT, GC-14) of the outer thyronine ring or from the bridging carbon (HY-4). All of these compounds have only a modest affinity and potency for the thyroid hormone receptor (TR) that limits studies of their antagonistic actions. Here, we report the design and synthesis of a novel series of 5'-phenylethynyl derivatives sharing the GC-1 halogen-free thyronine scaffold.(4) One compound (NH-3) is a T(3) antagonist with negligible TR agonist activity and improved TR binding affinity and potency that allow for further characterization of its observed activity. One mechanism for antagonism appears to be the ability of NH-3 to block TR-coactivator interactions. NH-3 will be a useful pharmacological tool for further study of T(3) signaling and TR function.

Structure-based design and synthesis of a thyroid hormone receptor (TR) antagonist.

Antagonists have been developed for several nuclear receptors but not for others, including TRs. TR antagonists may have significant clinical utility for treating hormone excess states and other conditions. A structure derived "extension hypothesis" was applied to synthesize a TR antagonist. The principal design feature was to attach an extension group to a TR agonist whose structure would perturb formation of the TR coactivator-binding surface. The compound, 3,5-dibromo-4-(3',5'-diisopropyl-4'-hydroxyphenoxy)benzoic acid, has no (TRalpha) or very weak partial (TRbeta) TR agonist activity and blocks TR binding of T3, formation of the coactivator-binding surface, and both a positive T3 response on a thyroid hormone response element and a negative T3 response on the TSHbeta promoter in cultured cells. The results suggest that 3,5-dibromo-4-(3',5'-diisopropyl-4'-hydroxyphenoxy)benzoic acid is a TR antagonist for thyroid hormone response element-mediated responses, this approach can be used more generally to generate nuclear receptor antagonists, and this compound or analogues may have medical and research utility.

Synthesis and biological activity of novel thyroid hormone analogues: 5'-aryl substituted GC-1 derivatives.

Compounds that selectively modulate thyroid hormone action by functioning as isoform-selective agonists or antagonists of the thyroid hormone receptors (TRs) might be useful for medical therapy. We have synthesized a high affinity TRbeta-selective agonist ligand, GC-1, and optimized the synthetic route to provide large quantities of the compound for animal testing. In addition to an improvement in efficiency, the new synthetic route offers a chemical handle for selective modification of the thyronine skeleton to produce new derivatives. To explore the effect of GC-1 core structure modifications on binding to TR isoforms and activation of transcription, we developed here an efficient and flexible route to a new series of 5'-substituted GC-1 analogues. This route relies on ortho lithiation and in situ boration of the biarylmethane compound 1, a key intermediate of the revised GC-1 synthesis, followed by Suzuki cross-coupling. Using this approach we prepared and tested eleven 5'-substituted GC-1 analogues. Substitution at the 5'-position decreased binding affinity, but retained TRbeta-selectivity for most of the compounds. Transactivation assays reveal that most of these compounds function as thyroid hormone agonists, but one compound (GC-14) antagonizes the response to thyroid hormone.

Design of thyroid hormone receptor antagonists from first principles.

It is desirable to obtain TR antagonists for treatment of hyperthyroidism and other conditions. We have designed TR antagonists from first principles based on TR crystal structures. Since agonist ligands are buried in the fold of the TR ligand binding domain (LBD), we reasoned that ligands that resemble agonists with large extensions should bind the LBD, but would prevent its folding into an active conformation. In particular, we predicted that extensions at the 5' aryl position of ligand should reposition helix (H) 12, which forms part of the co-activator binding surface, and thereby inhibit TR activity. We have found that some synthetic ligands with 5' aryl ring extensions behave as antagonists (DIBRT, NH-3), or partial antagonists (GC-14, NH-4). Moreover, one compound (NH-3) represents the first potent TR antagonist with nanomolar affinity that also inhibits TR action in an animal model. However, the properties of the ligands also reveal unexpected aspects of TR behavior. While nuclear receptor antagonists generally promote binding of co-repressors, NH-3 blocks co-activator binding and also prevents co-repressor binding. More surprisingly, many compounds with extensions behave as full or partial agonists. We present hypotheses to explain both behaviors in terms of dynamic equilibrium of H12 position.