pubs.acs.org/jmc Article Development of a Highly Potent Transthyretin Amyloidogenesis Inhibitor: Design, Synthesis, and Evaluation Francisca Pinheiro, Irantzu Pallares̀ , Francesca Peccati, Adria ̀ Sánchez-Morales, Nathalia Varejão, Filipa Bezerra, David Ortega-Alarcon, Danilo Gonzalez, Marcelo Osorio, Susanna Navarro, Adrián Velázquez-Campoy, Maria Rosário Almeida, David Reverter, Félix Busqué, Ramon Alibés, Mariona Sodupe, and Salvador Ventura* Cite This: J. Med. Chem. 2022, 65, 14673−14691 Read Online ACCESS Metrics & More Article Recommendations * sı Supporting Information ABSTRACT: Transthyretin amyloidosis (ATTR) is a group of fatal diseases described by the misfolding and amyloid deposition of transthyretin (TTR). Discovering small molecules that bind and stabilize the TTR tetramer, preventing its dissociation and subsequent aggregation, is a therapeutic strategy for these pathologies. Departing from the crystal structure of TTR in complex with tolcapone, a potent binder in clinical trials for ATTR, we combined rational design and molecular dynamics (MD) simulations to generate a series of novel halogenated kinetic stabilizers. Among them, M-23 displays one of the highest affinities for TTR described so far. The TTR/M-23 crystal structure confirmed the formation of unprecedented protein−ligand contacts, as predicted by MD simulations, leading to an enhanced tetramer stability both in vitro and in whole serum. We demonstrate that MD-assisted design of TTR ligands constitutes a new avenue for discovering molecules that, like M-23, hold the potential to become highly potent drugs to treat ATTR. ■ INTRODUCTION Amyloid diseases constitute a diverse group of pathologies Pathogenic mutations accelerate TTR aggregation by thermodynamic or kinetic destabilization of the protein.11,12 characterized by protein misfolding, aggregation, and the To date, more than 130 mutations in the TTR gene have been buildup of insoluble fibrils in tissues and organs throughout the described, which result in autosomal dominant familial forms body.1 Transthyretin (TTR) is one of many proteins related of the disease.13 The majority of disease-associated variants are with these disorders.2 caused by missense mutations and display tissue specificity and The liver and the choroid plexus are the major sites of TTR pathology. Familial amyloid polyneuropathy (FAP) and synthesis. TTR transports the retinol-binding protein−retinol familial amyloid cardiomyopathy (FAC) are the most prevalent complex and functions as a backup carrier for thyroxine (T4) in presentations of ATTR, compromising the peripheral nervous the blood and as a main transporter of T4 in the cerebrospinal system and the heart, respectively.14−16 For some rare TTR fluid.3−5 The extracellular misfolding of TTR and subsequent mutations, central nervous system (CNS) involvement has accumulation of amyloid fibrils in a variety of tissues underlie been reported.17,18 the onset of a group of disorders known as transthyretin Aging is another risk factor for ATTR, and deposition of amyloidosis (ATTR).6 wild-type (WT) TTR, preferentially in the myocardium, causes Native TTR is a homotetramer comprising four β-sheet rich senile systemic amyloidosis (SSA), an underdiagnosed late- subunits of 127 amino acid residues each, termed A, B, C, and onset sporadic cardiomyopathy impacting up to 10−20% of D. The monomers associate via their edge β-strands, yielding the population over 65 years old.19,20 Notably, SSA is the two dimers (AB and CD) that further associate back-to-back to render the tetramer. The AB/CD dimer−dimer interface defines two identical funnel-shaped T4-binding sites at Received: July 25, 2022 opposite sides of the molecule.7,8 TTR-tetramer dissociation Published: October 28, 2022 at the T4-binding interface creates dimers that promptly dissociate into aggregation-prone monomers, representing the rate-limiting step during TTR misfolding and amyloid formation.9,10 © 2022 The Authors. Published by American Chemical Society https://doi.org/10.1021/acs.jmedchem.2c01195 14673 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article leading cause of mortality in human subjects aged over 110 tolcapone has been shown to inhibit TTR aggregation induced years.21 by proteolytic cleavage at physiological pH,42 a process that An alternative pathway for TTR amyloid formation in vivo might underlie TTR amyloidogenesis in vivo. Noteworthily, proposes that TTR aggregation is triggered by proteolytic tolcapone penetrates the blood−brain barrier43 and effectively cleavage. It is supported by the fact that the amyloid deposits inhibits the aggregation of the extremely destabilized and fast formed by most TTR variants in vivo contain the truncated 49- dissociating variants that cause the rare, but lethal, CNS 127 polypeptide.22,23 This mechanism might be especially amyloidosis.44 relevant in organs with substantial shear stress, such as the Thermodynamic analysis revealed that tolcapone binds and heart, where the physiological fluid flow, together with the stabilizes TTR more effectively than tafamidis, exhibiting hydrophobic forces acting on the protein, might increase its higher ex vivo anti-amyloidogenic activity. More enthalpically susceptibility to proteolytic cleavage.24,25 favorable binding to TTR, together with the lack of negative In the last years, new therapies have been developed for the cooperativity, which in the case of tafamidis significantly treatment of ATTR that aim to replace liver transplantation, reduces the affinity for the second T4 site, seems to underlie the standard therapy for hereditary ATTR.26,27 Current the higher potency of tolcapone. therapeutic approaches mainly rely on reducing amyloid Tolcapone was the best performer of our repurposed formation through TTR kinetic stabilization28 or inhibition compound library. However, this does not mean that the of TTR protein synthesis (e.g., inotersen29 and patisiran30). contacts it established with the TTR T4-binding cavities were The kinetic stabilizer strategy gained momentum upon the optimal. In a way, a new repurposed drug may be considered as identification of a TTR disease protective substitution, a hit to be further optimized to increase its target potency and T119M.31 The T119M mutation reduces the rate of TTR selectivity, especially if atomic structural information of the tetramer dissociation and, consequently, the amyloidogenic protein−drug complex is available. Here, we exploited this propensity of the native ensemble.32 concept with the purpose of generating a TTR kinetic stabilizer Thyroglobulin and albumin are the main transporters of T4 with improved binding affinity and anti-amyloidogenic activity in the blood, with only 1% of TTR being bound to the relative to tolcapone. hormone, and thus, T4 pockets are largely empty. Ligand Up to now, more than 300 TTR crystal structures are binding at the T4 sites kinetically stabilizes TTR, increasing the available in the Protein Data Bank,45,46 most of them energy barrier for tetramer dissociation. Thus, small molecules complexed to small-molecule ligands. A detailed structural displaying selectivity and affinity for docking at T4 cavities have study of a set of 23 high-resolution TTR structures comprising emerged as therapeutic options for treating ATTR (Figure 1). WT, non-amyloidogenic, and amyloidogenic variants con- cluded that they are almost exactly superimposable. Moreover, differences in the positioning of certain loops or in the side chain rotamers of some residues, including those of the T4- binding sites, were not significant.47 Accordingly, the binding of kinetic stabilizers does not result in significant TTR structural rearrangements, and indeed, the structures of TTR with or without ligands are essentially identical except for a reduced number of side chain rotamers in the vicinity of the binders (Figure 2A). This indicates that crystallographic structures correspond to static pictures where it is difficult to discern the dynamic impact of both mutations and ligands on the native TTR stability, challenging the use of structure− Figure 1. Chemical structures of T47 (in red), the natural ligand of activity relationships to evolve stronger TTR binders. TTR, and examples of T4-inspired kinetic stabilizers.33,37−39 Molecular dynamics (MD) simulations allow studying the structural dynamics of biological systems at atomic reso- So far, only the benzoxazole tafamidis33 (Vyndaqel and lution.48 MD simulations are especially useful in modeling and Vyndamax) has reached the market. Treatment with tafamidis assessing the binding capacity of small molecules to target was well tolerated and has shown to delay the progression of proteins since they not only provide atomic information on the neuropathy and cardiomyopathy in FAP34 and FAC,35 thus interaction but also allow estimation of the binding being approved for these indications. However, disease energetics.49,50 MD simulations have been used to investigate progression occurs in ∼30% of patients with familial the impact of mutations on TTR conformational flexibility51,52 ATTR,36 highlighting the need for developing alternative and to investigate the mechanism of TTR protection by TTR kinetic stabilizers. existent kinetic stabilizers.53−55 However, to the best of our More recently, our group repurposed tolcapone for the knowledge, MD-based methods for estimating binding treatment of ATTR. Tolcapone (Tasmar, 3,4-dihydroxy-4′- affinities have not been employed to assist in the design of methyl-5-nitrobenzophenone) is a potent inhibitor of catechol- novel molecules aimed to interact with T4-binding sites. O-methyltransferase (COMT) approved in the United States Here we combined rational design and MD simulations to and Europe as an adjunct to levodopa and carbidopa for the generate a series of tolcapone-inspired kinetic stabilizers. The treatment of Parkinson’s disease. Tolcapone binds with high candidates were chemically synthesized and experimentally affinity and specificity to TTR, stabilizing the tetramer and validated. As a result, we describe M-23, a noncooperative thus preventing amyloidogenesis and protecting from TTR kinetic stabilizer that binds TTR with an affinity >5-fold than cytotoxicity.39 In a Phase IIa clinical study for FAP,40,41 tolcapone. Thermodynamic analysis indicates that this high tolcapone completely stabilized plasmatic TTR in all patients affinity results from an optimized enthalpy of binding, whereas studied, and no adverse events were registered. In addition, the TTR/M-23 crystal structure demonstrates the effective 14674 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article tafamidis, in respect to diflunisal, is difficult to explain from the crystal structure in terms of specific protein−compound interactions and has been attributed to a stronger halogen bonding capability of the chloride moiety. In tolcapone (Figure 2D), the 4-methyl-phenyl ring occupies HBP3, and a specific hydrogen bond is established between the linker carbonyl group of the compound and the hydroxyl side chain of Thr119. The 3,4-dihydroxy-5-nitrophenyl ring of tolcapone is positioned in HBP1, forming a hydrogen bond with Lys15, which in turn stabilizes the ionic interactions between Lys15 and Glu54. These direct interactions in the outer face of the cavity, together with hydrogen bonding to Thr119, likely contribute to the favorable enthalpic binding of this molecule, explaining why it stabilizes TTR more effectively than tafamidis. In contrast to diflunisal and tafamidis, tolcapone does not possess halogen atoms, and although the methyl group is placed in a favorable environment in HBP3, it cannot establish the halogen bonds that characterize T4, diflunisal, and tafamidis. Halogenation has been shown to be effective at increasing the affinity of certain TTR ligands, like in the case of Figure 2. (A) Superposition of WT-TTR in Apo-form (light blue) iododiflunisal.57 Therefore, we decided to use MD simulations and complexed to diflunisal (olive), tolcapone (blue), and tafamidis to study if endorsing tolcapone with different halogen moieties, (orange). Overlaid tetramers are shown in cartoon, and diflunisal while trying to keep the optimal hydrogen bonding capability (green), tafamidis (yellow), and tolcapone (cyan) are represented as of the upper ring and the middle carbonyl group intact, might sticks. (B−D) Close-up view of the binding of diflunisal, tolcapone, result in optimized kinetic stabilizers. and tafamidis to WT-TTR. Compounds are colored as in panel A. Molecular Dynamics Simulations of Halogenated Residues interacting with TTR are shown as sticks (orange). Blue Tolcapone Analogues. All the new ligands developed in dashed lines signal the hydrogen bonds and the salt bridges. the present work are benzophenone derivatives with two Structures prepared from PDB structures (B) 3D2T,37 (C) 3TCT,33 and (D) 4D7B.39 phenyl moieties, one common to all ligands, i.e., the 3,4- dihydroxy-5-nitrophenyl ring, more exposed to the solvent, and a second one including different halogen substituents pointing existence of new contacts between the two moieties, as toward the inner binding pocket (Figure 3). Differences are predicted by MD simulations. These interactions result in a thus expected to arise from the distinct contacts between this significantly higher stabilization of the TTR tetramer in vitro inner aryl moiety and TTR at the binding site. and ex vivo relative to tolcapone, turning M-23 into a promising candidate for therapeutic intervention in ATTR. ■ RESULTS AND DISCUSSION Rationale for the Design of Tolcapone Analogues. Each TTR T4-binding site contains three pairs of symmetric depressions known as halogen-binding pockets (HBPs: HBP1 and HBP1′, HBP2 and HBP2′, and HBP3 and HBP3′), wherein the four iodine atoms of the hormone reside. The innermost pocket is HBP3 and is established by Ser117, Leu110, Thr119, and Ala108 side chains. HBP1 is placed at the entrance of the T4-binding site and comprises Lys15, Leu17, Thr106, and Val121, whereas the central pocket HBP2 is Figure 3. (A−F) Chemical structures of tolcapone derivatives formed by Leu17, Ala108, Ala109, and Leu110 along with the synthesized and evaluated in this study . The second aryl ring is methylene carbons of Lys15. shown in red to highlight the differences between the compounds. Typically, TTR kinetic stabilizers have two aromatic rings, one ring substituted with halogens, placed at HBP2/HBP3, and the other displaying hydrophilic substituents, placed at In the first two ligands, the 4-CH3 group of tolcapone was HBP1. The nonsteroidal anti-inflammatory drug diflunisal substituted by either a −CF3 group or −F, rendering (Figure 2B), already in clinical trials,56 exemplifies these compounds M-14 and M-17, respectively (Figure 3A−C). properties, with the difluorophenyl group pointing to the inner The computed binding energies of tolcapone, M-14, and M-17 part of the channel and the two fluor atoms located in HBP2. to human WT-TTR and the main ligand−TTR hydrogen bond In HBP1, the carboxy group of diflunisal establishes a salt contacts (>10% frequency along the trajectory) are given in bridge with the amino group of Lys15. Table 1. Simulations are expected to provide the proper trends, In tafamidis (Figure 2C), the 3,5-dichloro moiety is although not absolute values. Note that for absolute affinities, surrounded by the residues in the HBPs 3/3′, whereas the the entropy term, particularly that owing to the decrease of carboxy end forms a water-mediated hydrogen bond with the translational and rotational freedom when the ligand binds to amino group of Lys15. The highest binding affinity of the protein, should be included. However, this entropic term 14675 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Table 1. Ligand-TTR Binding Energies (ΔGbind), Gas Phase appearing in less than 1% of the trajectory. This is due to the Binding Energies (ΔEgp), Ligand Solvation Energies presence of the −CF3 bulkier substituent, which introduces (ΔGL‑solv), in kcal mol−1, and Main Ligand−TTR H-Bond larger repulsive interactions that hinder the entrance into the Contacts cavity and leave the ligand more exposed to the solvent. compound ΔGbinda ΔEgpb ΔGL‑solv ligand···TTR contacts Indeed, the shortest distance between the two ligands in these two complexes, taken as the distance between the C atoms of tolcapone 59.7 81.5 −10.9 C�O···T119 M-14 55.4 76.6 −10.6 the two ligands’ C�O groups, is significantly larger in M-14 M-17 68.0 90.2 −11.1 C�O···T119, F···S117, F···T119 (19.9 Å) than in tolcapone (15.8 Å). Consequently, M-14 a exhibits a smaller calculated binding energy (55.4 kcal mol−1) ΔGbind = ΔEgp + 2 × ΔGL‑solv. bΔEgp = ETTR + 2EL − ETTR‑2L. than tolcapone (59.7 kcal mol−1). In contrast, for M-17, with a should not vary significantly between our candidate −F in the para position instead of a −CF3 group, the C�O··· molecules,58 and as such, it is not expected to affect ligand Thr119 interaction is maintained with an average frequency of comparisons. ∼50%, with the shortest C�O···HOT119 distance being 2.55 Å. As expected, all starting structures, derived from the crystal In addition, new contacts between the −F substituents and structure of the TTR−tolcapone complex (PDB: 4D7B), with both Thr119 and Ser117 appear, which further enhance the the two ligands related by a C2 symmetry axis, lose their initial calculated ligand−TTR binding energy (68.0 kcal mol−1) as symmetry along the simulation; i.e., the interactions between compared to tolcapone (59.7 kcal mol−1). each ligand with AB or CD, albeit similar, are not identical. Binding of Halogenated Tolcapone Analogues to Furthermore, in all cases, the specific interactions in the outer TTR. M-14 and M-17 were chemically synthesized as shown in binding pocket between the hydroxyl groups of the 3,4- Scheme 1. The synthesis procedure is explained in detail in the dihydroxy-5-nitrophenyl moiety and Lys15 or between Glu54 Experimental Section. and Lys15 identified in the crystal structure are only Isothermal titration calorimetry (ITC) was used to maintained in less than 5% of the trajectory due to temperature characterize their binding affinities for WT-TTR, and the effects. Thus, from now on, we will focus on the interaction thermodynamic parameters that describe the reaction were between the central C�O of the ligand and the hydroxyl determined. Ligand binding to TTR can be cooperative group of Thr119 and on the additional ligand−TTR specific (positive or negative cooperativity) or noncooperative.59−61 interactions resulting from the new substitutions on the phenyl moiety. The main structures, together with the H-bond Non- or positive cooperativity is desired; however, most evolution along the trajectories, are provided in Figure S1. reported ligands exhibit negative cooperativity,33,62,63 which Specific ligand−TTR contacts, with their frequency and implies a loss of affinity for the second binding site after shortest and average distances, are given in Table S1. binding to the first one. Tolcapone, M-14, and M-17 bound to Regarding tolcapone, the C�O···Thr119 interaction TTR without any cooperativity (Figure 4A and Table 2), appears with a frequency of 10%, with the minimum distance whereas tafamidis, included as a reference, exhibited the typical value being 2.63 Å, which resembles the one observed in the negative cooperative behavior (Kd1 = 9.9 nM and Kd2 = 260 crystal structure (2.55 Å). For M-14, this interaction is lost, nM). Scheme 1. Synthesis of Tolcapone Analogues 14676 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Figure 4. In vitro characterization of halogenated tolcapone analogues, M-14, and M-17. (A) Interaction of WT-TTR with tolcapone, M-14, and M-17 as assessed by ITC. Tafamidis was tested as a reference. The top panels represent the raw data (thermogram), while the lower panels correspond to the integrated heat changes upon binding plotted against the ligand/TTR concentration ratio (binding isotherm). The solid line describes the best fit according to a two-site binding model (with or without cooperativity) for each test compound. (B) WT-TTR (1.8 μM) urea- induced tetramer dissociation (6 M urea) in the absence or presence of tolcapone, M-14, or M-17 (at 3.6 μM), as measured by Trp fluorescence. The values correspond to mean ± SEM (n = 3). (C) Acid-mediated TTR (at a final assay concentration of 3.6 μM) aggregation as a function of inhibitor concentration determined by turbidity at 340 nm. The values refer to mean ± SEM (n = 3). Table 2. Thermodynamic Parameters Determined by ITC TTR Kinetic Stabilization by Halogenated Tolcapone for the Binding of M-14 and M-17 to WT-TTR Analogues. We addressed whether M-14 and M-17 kineti- cally stabilize TTR, inhibiting urea-induced tetramer dissoci- Kd ΔG ΔH −TΔS (nM) (kcal mol−1) (kcal mol−1) (kcal mol−1) ation. Using urea concentrations in the post-transition region tolcapone 34 −10.2 −12.8 2.6 for tertiary structural changes allows measuring tetramer tafamidis 9.9a −10.9a −6.0a −4.9a dissociation, as the monomers unfold in a few milliseconds 260b −9.0b −6.5b −2.5b and remain unfolded.11,64 Accordingly, TTR samples were M-14 310 −8.9 −9.2 0.3 incubated in the absence or presence of compounds, and TTR M-17 31 −10.3 −14.0 3.7 denaturation was triggered by adding 6 M urea. Tertiary a Correspond to the values for the first binding site of TTR. structural changes were monitored along time by tryptophan b Correspond to the values for the second binding site of TTR. (Trp) intrinsic fluorescence, which was used to determine the fraction of unfolded protein at any time (Figure 4B). Tolcapone was analyzed in parallel for comparative purposes. In agreement with the trend observed in the MD binding The three molecules significantly decreased the amount of energy calculations, M-14 exhibited a higher Kd (310 nM) and dissociated tetramer, as well as the rate of tetramer a lower enthalpic contribution to binding (ΔH = − 9.2 kcal dissociation, when present at equimolar levels relative to T4- mol−1) than tolcapone (Kd = 34 nM and ΔH = −12.8 kcal binding sites. Tolcapone protected up to 82.7 ± 3.1% of TTR mol−1), whereas the affinity and enthalpic binding term of M- molecules from urea induced unfolding; M-17 provided a 17 were higher (Kd = 31 nM and ΔH = −14.0 kcal mol−1). similar degree of protection, whereas M-14 was less effective. The strong binding of tolcapone and M-17, with dissociation TTR Anti-aggregation Activity of Halogenated Tolca- constants in the low nanomolar range, was entirely enthalpi- pone Analogues. The anti-amyloidogenic activity of M-14 cally driven (ΔH < 0; −TΔS > 0). As observed in MD and M-17 was evaluated using a well-established fibril- simulations, this indicates the formation of specific non- formation assay11,65 and compared with that of tolcapone. covalent interactions between the protein and the ligand. TTR (7.2 μM) was mixed with increasing concentrations of Enthalpy−entropy compensation effects resulted in very the compound (0−40 μM) for 30 min (pH 7.4, at 37 °C), and similar ΔG values for M-17 and tolcapone. then, the pH was lowered to 4.2, which is the most favorable 14677 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article pH for TTR fibrilization.66 After an additional incubation of 72 M-21 (−F) since the distance between the C atoms of the two h, the percentage of conversion of native TTR into amyloid ligands’ C�O groups is significantly larger in M-20 (19.4 Å) fibrils was calculated by measuring turbidity at 340 nm and is than in M-21 (15.3 Å) or tolcapone (15.8 Å). Because of this reported relative to TTR incubated in the same conditions in displacement toward the upper part of the T4 cavity, one of the the absence of an inhibitor (100%) in Figure 4C. −Cl atoms contacts Thr119 in M-20. As in M-17, the −F The three molecules displayed a strong anti-amyloidogenic atoms in M-21 establish additional interactions with Ser117 in activity, decreasing TTR aggregation in a concentration- the HBP3 pocket, whereas no such contacts were observed for dependent manner. In all cases, the protection was >60% at M-20. The interplay of interactions results in the calculated the equimolar total concentration (one molecule of the test binding energies for M-20 (57.0 kcal mol−1) and M-21 (74.0 compound bound per molecule of the TTR tetramer) and kcal mol−1) being lower and higher than the one of tolcapone ≥87.4 when the compound concentration was greater than or (59.7 kcal mol−1), respectively. equal to the one of T4-binding sites. These results indicate that The high binding energy of M-21 indicates that the contacts tolcapone is already a very potent inhibitor of TTR between the phenyl ring substituents and the inner Ser117 aggregation, reaching up to 92.8% inhibition at 20 μM; make important contributions to the molecule−TTR complex. accordingly, despite M-17 being a slightly better binder and With this idea in mind, we designed M-23 (Figure 3F), in kinetic stabilizer than the original molecule, these properties do which −F in R3 was substituted by an −OH to favor the not translate into an optimization of its anti-aggregation formation of a specific hydrogen bond with the side chain of properties, at least at acidic pH. Ser117. The MD simulations indicate that this interaction is Molecular Dynamics Simulations of Demethylated formed with a frequency that ranges from 15 to 90% (Table 3,5-Disubstituted and Halogenated Tolcapone Ana- S2), with the shortest distance being 2.52 Å. In addition, the logues. In M-17, replacing tolcapone −CH3 by −F resulted in C�O···Thr119 contact is maintained in M-23, with a increased affinity and a higher enthalpic contribution to the frequency of 37 and 49% for ligands 1 and 2, respectively, binding. This is in line with the suggestion that, in tafamidis, and a minimum distance of 2.59 Å. As a result, M-23 displays the −Cl atoms in the 3,5-dichloro moiety contribute the highest binding energy among the modeled compounds significantly to binding in the innermost TTR HBP3 pocket. (84.4 kcal mol−1) and is expected to be a significantly better Therefore, we designed M-20, a chimeric molecule in which binder than tolcapone (Table 3, Figure S2C, and Table S2). the 4-methyl-phenyl ring of tolcapone was replaced by the 3,5- Furthermore, and as found for M-21, a slightly larger dichloro-phenyl ring in tafamidis (Figure 3D). The idea was to internalization of the ligand is observed for M-23 compared combine the favorable noncovalent contacts of tolcapone in to tolcapone, with the shortest distance between the C atoms the outer and middle sections of the T4 cavity with those of the two ligands’ C�O groups being 15.3 Å for M-23 and established by tafamidis in the inner pocket. In addition, 15.8 Å for tolcapone. because −Cl and −F seemed to contribute differentially to Of note, in addition to the molecule−protein interactions, binding in HPB3,67 as deduced from the affinities for TTR of potential new residue−residue contacts induced by the binding diflunisal and tafamidis, we designed M-21, in which the lower of the compound might contribute to the tetramer stability. In tolcapone ring was substituted by a 3,5-difluoro-phenyl ring particular, the buried Ser117 residue in each TTR subunit can (Figure 3E). Thus, in a way, M-21 is a chimera of tolcapone establish an intersubunit hydrogen bond with Ser117 in the and diflunisal, although the −F substituents lay in different other monomer in the dimer (A−B or C−D contacts). MD relative positions. simulations indicate that these interactions are far more The computed binding energies of M-20 and M-21 to frequent when TTR is bound to the ligands than when the T4- human WT-TTR and the main ligand−TTR hydrogen bond binding cavity is empty (Table S3). Noticeably, among the contacts (>10% frequency along the trajectory) are provided in double substituted compounds, M-23 is the one rendering the Table 3. The main structures for all systems, together with the smaller average distances between the Ser117 residues in the weaker AB/CD dimer−dimer interface (Table S4). Table 3. Ligand−TTR Binding Energies (ΔGbind), Gas Binding of Demethylated 3,5-Disubstituted and Phase Binding Energies (ΔEgp), Ligand Solvation Energies Halogenated Tolcapone Analogues. Compounds M-20, (ΔGL‑solv), in kcal mol−1, and Main Ligand−TTR H-Bond M-21, and M-23 were obtained according to Scheme 1. The Contacts chemical synthesis is detailed in the Experimental Section. ITC compound ΔGbinda ΔEgpb ΔGL‑solv ligand···TTR contacts experiments indicated that the three molecules bound strongly to TTR without apparent cooperativity (Figure 5A and Table M-20 57.0 79.0 −11.0 Cl···T119 4). In excellent agreement with MD simulations, M-20 M-21 74.0 95.2 −10.6 C�O···T119, F···S117 exhibited a higher Kd (85 nM) and lower enthalpic M-23 84.4 116.2 −15.9 C�O···T119, OH···S117, F··· S117 contribution to binding (ΔH = − 11.7 kcal mol−1) than a tolcapone (Kd = 34 nM and ΔH = −12.8 kcal mol−1). The ΔGbind = ΔEgp + 2 × ΔGL‑solv. bΔEgp = ETTR + 2EL − ETTR‑2L. affinity of M-21 was higher (Kd = 26 nM) than that of tolcapone, whereas the enthalpic binding term was lower (ΔH H-bond evolution along the trajectories, are shown in Figure = −11.3 kcal mol−1). However, a lower entropic penalty for S2A,B. Specific ligand−TTR contacts, with their frequency and binding in M-21 results in a slightly higher ΔG (−10.4 kcal shortest and average distances, are given in Table S2. mol−1) relative to tolcapone (−10.2 kcal mol−1). For M-21, the C�O···Thr119 interaction appears with a The K d of M-23 is exceptionally low (6.2 nM), frequency of 20%, with the shortest C�O···HOT119 distance corresponding to a binding affinity >5-fold than that of being 2.63 Å. In M-20, this interaction is lost, occurring only in tolcapone, with a very high enthalpy for binding (ΔH = −16.6 less than 1% of the trajectory. As in M-14, this seems to kcal mol−1). The binding of M-23 is completely enthalpically respond to the bulkier substituents of M-20 (−Cl) relative to driven, with an entropy penalty (5.4 kcal mol−1) higher than 14678 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Figure 5. In vitro characterization of demethylated 3,5-disubstituted and halogenated tolcapone analogues, M-20, M-21, and M-23. (A) Interaction of WT-TTR with M-20, M-21, and M-23 as assessed by ITC. The top and the lower panels correspond to the thermogram and the binding isotherm, respectively. The solid line describes the best fit according to a two-site binding model without cooperativity for each compound. (B) WT-TTR (1.8 μM) urea-induced tetramer dissociation (6 M urea) in the absence or presence of M-20, M-21, or M-23 (at 3.6 μM), as measured by Trp fluorescence. The values correspond to mean ± SEM (n = 3). (C) Acid-mediated TTR (at a final assay concentration of 3.6 μM) aggregation as a function of inhibitor concentration determined by turbidity at 340 nm. The values represent mean ± SEM (n = 3). Table 4. Thermodynamic Parameters Determined by ITC perfectly agrees with MD simulations and demonstrates a for the Binding of M-20, M-21, and M-23 to WT-TTR significantly optimized TTR binding in M-23. TTR Kinetic Stabilization by Demethylated 3,5- Kd ΔG ΔH −TΔS (nM) (kcal mol−1) (kcal mol−1) (kcal mol−1) Disubstituted and Halogenated Tolcapone Analogues. tolcapone 34 −10.2 −12.8 2.6 Analysis of the kinetic stability induced by M-20, M-21, and M-20 85 −9.7 −11.7 2.0 M-23 by monitoring TTR tertiary structural changes in the M-21 26 −10.4 −11.3 0.9 presence of 6 M urea indicated that all of them significantly M-23 6.2 −11.2 −16.6 5.4 decreased the amount of dissociated tetramer, as well as the rate of tetramer dissociation, when present at an equimolar tolcapone (2.6 kcal mol−1), likely due to its higher polarity, but ratio relative to T4-binding sites (Figure 5B). M-20 and M-21 a still higher ΔG (−11.2 kcal mol−1). Compared with performed worse and equal to tolcapone, respectively. M-23 tafamidis, M-23 displays a higher affinity for the first and appears as the strongest kinetic stabilizer in the assay especially for the second binding site where its binding is >40- conditions, protecting as much as 87.0 ± 0.1% of TTR fold stronger. In addition, the enthalpy contribution for molecules from urea-induced unfolding. binding to any of the two sites is at least 2.5-fold higher in TTR Anti-aggregation Activity of Demethylated 3,5- M-23 than in tafamidis. Overall, the thermodynamic analysis Disubstituted and Halogenated Tolcapone Analogues. 14679 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article M-20, M-21, and M-23 effectively prevented TTR amyloid formation at acidic pH (Figure 5C). They decreased TTR aggregation in a concentration-dependent mode, with >60% reduction at 1:1 TTR/compound ratio and a decrease of >90% when the compound concentration was greater than or equal to that of T4-binding sites. As mentioned above, tolcapone is a very potent TTR aggregation inhibitor. The weaker binder in this series (M-20) equals its potency, while M-21 and M-23 perform better, completely abolishing amyloid fibril formation when present at 20 μM. M-23 Stabilizes TTR in Human Plasma. Altogether, the previous results suggested M-23 as the most promising compound in our set and encouraged us to investigate its activity further. First, we assessed the performance of M-23 in human plasma, a complex biological fluid where different issues, including unspecific binding to other plasma proteins, can compromise ligand efficacy. The capacity of M-23 to inhibit TTR tetramer dissociation in human plasma was evaluated by isoelectric focusing (IEF) electrophoresis under partially denaturing conditions (4 M urea). Tolcapone was used as a control. The assay allows one Figure 6. M-23 tetramer stabilization effect assessed by IEF and its to quantify the proportion of monomer and tetramer in the binding to TTR in human plasma. (A, B) Percentage of TTR tetramer sample and to calculate the extent of tetramer stabilization. stabilization upon incubation of M-23 with recombinant (A) WT- First, we wanted to assess if this assay can detect differences TTR or (B) human plasma as evaluated by IEF under partially in tetramer stabilization by M-23 and tolcapone when using denaturing conditions. M-23 was 5 or 10 times more concentrated purified recombinant proteins. TTR (6 μM) was incubated in than WT-TTR. Error bars represent SEM of mean values (n = 3 for recombinant protein; n = 6 for plasma samples); *p < 0.05; **p < the presence or absence of 30 and 60 μM compound overnight 0.01; ***p < 0.001. (C) Representative native gel electrophoresis at 4 °C. As shown in Figure 6A, M-23 exerted a higher showing the distribution of [125I]-T4 after incubation with human stabilizing effect than tolcapone at both tested concentrations. plasma in the absence or presence of compounds. In the absence of Then, the same assay was performed in human plasma (Figure compounds, three bands can be observed that correspond to the 6B). M-23 was significantly more effective than tolcapone, with major plasma T4 binding proteins: T4-binding globulin (TBG), a stabilizing effect >5-fold the one exerted by the original albumin (ALB), and TTR. Plasma incubated with DMSO was used as molecule. This indicates selective and tight binding of M-23 to negative control. (D) Fraction of T4 bound to TTR in the plasma of circulating TTR in plasma, as confirmed using the T4 binding individuals incubated with or without compounds as determined by densitometry. The values were normalized to the control, which competition assay. M-23 decreased the binding of T4 to TTR corresponds to the maximum. The values represent mean ± SEM (n = in human plasma by 84.6 ± 9.7%, outperforming tolcapone in 4). the same conditions (Figure 6C,D). The aqueous solubility of the two molecules was determined to rule out that differences Table 5. Solubility Experiments Performed for M-23 and in solubility could underlie the observed differences in binding Tolcaponea (Table 5). M-23 and tolcapone showed similar solubility values (0.045 mg/mL for M-23 and 0.056 mg/mL for weight weight water solubility (g) (mg) (mL) (mg/mL) tolcapone), suggesting that these parameters do not influence M-23 assay 1 0.0154 15.4 350 0.044 the results obtained. Importantly, in our assays, M-23 has a M-23 assay 2 0.0147 14.7 326 0.045 stabilizing effect in human plasma >10-fold the one exerted by M-23 assay 3 0.0150 15.0 333 0.045 tafamidis (Figure S3), the only marketed molecule for ATTR. M-23 average 0.045 The increased binding selectivity and stabilization potency tolcapone assay 1 0.0150 15.0 267 0.056 of M-23 in human plasma, relative to tolcapone, likely result tolcapone assay 2 0.0144 14.4 261 0.055 from its higher enthalpy for binding, as deduced from ITC tolcapone assay 3 0.0153 15.3 273 0.056 data. This is in accordance with recently reported data that tolcapone average 0.056 suggest that the correlation between ΔH and the selectivity a The experiments were performed in triplicate. and efficacy of TTR stabilizers in human plasma is greater than with ΔG or Kd,68 two parameters that in any case are also better in M-23. which is especially relevant in the context of this study, as M-23 Is Innocuous for Human Cells. Cytotoxicity tolcapone has been associated with cases of hepatotoxicity.74,75 analyses were performed for evaluating the potential M-23 In this study, HeLa and HepG2 cells were exposed to chemical toxicity to human cells. Two cell lines were chosen increasing concentrations of M-23 or tolcapone for 72 h at 37 for this purpose: HeLa, a human epithelioid cervix carcinoma °C using the PrestoBlue cell viability reagent. For both cell cell line, and HepG2, a human hepatocellular carcinoma cell lines, M-23 showed significantly lower toxicity than tolcapone line. These cell lines are well characterized and have been above 10 μM compound concentration (Figure 7). These widely used for in vitro assessment of compound toxicity.69−71 results are biologically relevant, particularly for HepG2, as they In particular, HepG2 is the most frequently used cell line in the recapitulate tolcapone’s in vivo hepatotoxicity and suggest that testing and investigation of drug-induced liver damage,72,73 M-23 is not only a better TTR stabilizer but might also 14680 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article T4-binding sites as responsible for the enhanced affinity and the high enthalpic contribution to binding in this molecule with respect to tolcapone. To inspect this possibility, we determined the crystal structure of TTR with M-23 at 1.2 Å resolution (Figure 8). The atomic coordinates have been deposited in the PDB (PDB code 7QC5). This high-resolution crystal allows one to unequivocally place M-23 in the AB/CD dimer−dimer interface in the forward binding mode. As a result of the twofold symmetry of the binding sites, M-23 Figure 7. (A) HeLa and (B) HepG2 cell viability of cells exposed to adopts two equivalent binding modes related by a 180° increasing concentrations of M-23 (striped bars) or tolcapone (empty rotation. As designed, M-23 sits deeper in the innermost bars) as measured by the PrestoBlue assay. The values correspond to mean ± SEM (n = 3). section of the hormone-binding cavity relative to tolcapone (Figure 8). As in tolcapone, the 3,4-dihydroxy-5-nitrophenyl ring of M- overcome one of the major concerns related with tolcapone 23 is oriented to the outer binding cavity, being surrounded by therapy. the hydrophobic residues from the HBPs 2/2′ and 1/1′ M-23 Stabilizes the TTR Dimer−Dimer Interface. The (Figure 8B). Remarkably, the ε-amino group of Lys15 is placed MD simulations, together with the thermodynamic analysis, between the two hydroxyl groups of the phenyl ring of M-23 point to a higher number and/or strength of interactions and the carboxylate group of Glu54, establishing important established between M-23 and the residues within the TTR electrostatic interactions. These interactions close the cavity Figure 8. Crystal structure of WT-TTR complexed with M-23. (A) Superposition of WT-TTR complexed with M-23 (orange) or tolcapone (green). Cα rmsd from 116 residues is 0.19 Å. A close-up view of one conformation of the superposed compounds is shown on the inset. (B) General view of WT-TTR bound to M-23 at 1.2 Å, represented as cartoon. The electron density maps of the two binding sites of M-23 are shown. Dashed line depicts the twofold symmetry axis of the dimer−dimer interface. The insets represent the close-up view of one of the WT-TTR T4- binding sites for M-23 and for tolcapone (PDB: 4D7B). Ligands and the most important interacting residues are illustrated as sticks. (C) M-23 binding at the WT-TTR dimer−dimer interface. M-23 and some of the TTR residues interacting with the ligand are represented by sticks. Dashed lines represent key interactions between M-23 and WT-TTR and between the hydroxyl groups of S117/S117′. 14681 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article around M-23, protecting the compound and the interactions it M-23 is entirely enthalpy driven, displaying an enthalpic establishes with the protein, from the solvent. Although M-23 contribution to the binding significantly higher than those of is more buried into the cavity than tolcapone, the distances tafamidis and tolcapone, a parameter associated with the between Lys15 and the −OH are equal or shorter, resulting in molecule potency and selectivity. Consequently, M-23 strongly Lys15 moving inward the cavity, whereas Glu54 keeps binds to TTR in human plasma, exhibiting a higher TTR essentially the same position (Figure S4). stabilizing activity than the two reference molecules, thus As in the TTR−tolcapone structure, the M-23 central becoming a candidate for further preclinical and clinical carbonyl group establishes a specific H-bond with the hydroxyl investigation. Crucial to identify M-23 has been the application side chain of Thr119. As indicated by MD simulations, the 3- of MD simulations on top of rationally designed tolcapone fluoro-5-hydroxyphenyl ring projects deep within the inner variants since modeling the flexibility of the T4-binding cavity cavity where it participates in hydrophobic and van der Waals has allowed one to anticipate protein−compound interactions interactions with residues forming the two symmetrical T4- at atomic resolution and rank the molecules according to the HBPs, HBP2 and HBP3 (Ala108, Leu110, Ser117, and energetics of binding. This integral approach constitutes a Thr119). In addition, the M-23 5-OH group forms a short time- and cost-effective strategy to assist in the search of potent hydrogen bond with Ser117 (2.7/2.8 Å), and the 3-F ATTR disease modifiers, allowing the evaluation of small substituent is in contact with Ala108 (2.8/2.8 Å). Such a molecules rapidly and accurately. short F···Ala108 distance matches the shortest one observed in the MD simulation (2.76 Å). These two interactions are absent in the TTR−tolcapone and the TTR−tafamidis structures, confirming that, as designed, M-23 establishes a higher number ■ EXPERIMENTAL SECTION Computational Simulations. Binding energies have been of noncovalent H-bond contacts with the protein. They are estimated from gas-phase interaction energies of a collection of frames generated from a classical molecular dynamics (MD) expected to be stronger because they are buried in the low simulation and including solvent effects as the ligand solvation dielectric hydrophobic interior of the T4-binding cavity. These energy. This means that we assumed that the complex and the new contacts are likely the driving force for M-23 higher receptor have approximately the same solvation energy and that the affinity and exceptional enthalpy for binding. solvation contribution to the binding energy only arises from ligands’ Interestingly, the H-bond interactions of M-23 with Ser117 desolvation. MD simulations were performed with the Amber suite79 in HBP3 and HBP3′ mimic the ones observed between these using the ff14SB force field.80 Organic ligands were parametrized two residues in the kinetically stable T119M-TTR variant. using the gaff2 force field.81 These H-bond interactions in the two symmetric cavities help Production was run for 150 ns in the NPT ensemble at a constant bring the dimer subunits closer and strengthen the molecular temperature of 300 K, and binding energies were estimated from 140 contacts between them, increasing the energy barrier for structures evenly sampled from the last 140 ns. Ligand solvation energies were computed with the SMD continuum model82 at the dissociation.76,77 Indeed, the O−O distances between facing DFT (B3LYP/6-31+G(d,p)) level of theory with the Gaussian09 Ser117 residues in the two dimers are shorter in T119M-TTR program.83 (A−C = 4.7 Å, B−D = 4.8 Å) than in WT-TTR (A−C = 5.5 Å, Models were built by replacing tolcapone molecules with their B−D = 5.4 Å). In the same manner, these distances are shorter derivatives in the high-resolution X-ray diffraction tetrameric WT- in the TTR/M-23 structure (A−C = 5.0 Å, B−D = 4.8 Å) than TTR-tolcapone complex structure PDB: 4D7B.39 Complex structures in the TTR/tolcapone complex (A−C = 5.4 Å, B−D = 5.4 Å). were neutralized with the appropriate number of Na+ counterions, Actually, Ser117-compound contacts seem to hold the key for and water molecules were added up to a minimum distance of 8 Å the enthalpy-driven stabilization of TTR by AG10,38,68 a novel from the protein. It should be noted that while in the tetrameric WT- TTR ligand that has already shown effectivity in the clinic for TTR-tolcapone crystal structure the two ligands are symmetrically placed, simulations without any structural constraints led to an TTR cardiomyopathy.78 In one of the Ser117 side-chain asymmetric organization with one ligand at the center of the TTR conformations, they can establish a short hydrogen intradimer tetramer and the second in the binding channel, though further H-bond (2.5 Å), which might further contribute to stabilize the exposed to the solvent than in the crystallographic structure (Figure quaternary structure (Figure 8C). S5). This striking difference between our simulations and the X-ray ■ CONCLUSIONS We have developed M-23, a disubstituted and halogenated structure may be due either to (i) a dynamic disorder of the two ligands in the complex in solution, which collapses to a symmetric organization upon crystallization, or (ii) the inadequacy of the molecular mechanics model to properly describe the system’s tolcapone congener. This novel TTR kinetic stabilizer keeps structural features. Thus, we decided to enforce a small harmonic the interactions established by the parental molecule in the constraint on the protein backbone so that the structure does not drift outer and central sides of the binding cavity while being more significantly apart from the crystallographic one. After extensive internalized and establishing new and specific contacts with the testing, we found that when including a constraint of 2 kcal/mol−1 innermost residues. Under the same conditions, tolcapone Å−2 on the backbone, the symmetry of the two binding sites is exhibits a superior stabilizing and anti-aggregational activity preserved while allowing a reorganization of the ligand binding than tafamidis. The advantage of tolcapone is that, although it residues. Thus, this is the protocol adopted all along this work. binds worse than tafamidis to the first TTR T4-binding site, it Molecular Dynamics Protocol. The molecular dynamics protocol binds significantly better to the second one due to its lack of includes (i) a 200 ps equilibration run in the NVT ensemble, raising negative cooperativity. Here, we show that due to the unique the temperature from 0 to 100 K; (ii) a 2 ns equilibration run in the NPT ensemble, raising the temperature from 100 to 300 K; and (iii) a network of interactions it establishes with TTR, M-23 displays 150 ns production run in the NPT ensemble with the temperature a higher affinity than any of these two therapeutically relevant kept constant to 300 K. A Langevin thermostat and a Monte Carlo molecules for both binding sites, becoming one of the barostat were employed. Hydrogen bonds were calculated with strongest TTR ligands described so far. Accordingly, M-23 cpptraj default values (distance cutoff of 3 Å) in all cases except for stabilizes and protects TTR from aggregation in vitro at very Cl···X contacts for which the threshold was set to 3.5 Å. Visualization low compound concentrations. Furthermore, the binding of and postprocessing were done with VMD and MDtraj packages.84,85 14682 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article General Methods and Compound Characterization. Com- −COH), 7.83 (d, J6,2 = 1.8 Hz, 1H, H-6), 7.62 (d, J2,6 = 1.8 Hz, 1H, mercially available reagents were used as received. Solvents were dried H-2), 4.08 (s, 3H, CH3O-4), 4.00 (s, 3H, CH3O-3). The by distillation over the appropriate drying agents. All reactions were spectroscopic data were consistent with the literature.88 monitored by analytical thin-layer chromatography (TLC) using silica 1-Fluoro-3-iodo-5-methoxybenzene, 1e. To a solution of 3- gel 60 precoated aluminum plates (0.20 mm thickness). Flash column fluoro-5-methoxyaniline (1.02 g, 7.08 mmol) in H2O (3.3 mL) was chromatography was performed using silica gel Geduran SI 60 (40− added concentrated HCl (3.3 mL) at 0 °C. After stirring for 30 min, a 63 μm). 1H NMR and 13C NMR spectra were recorded at 250, 360, 1.8 M solution of NaNO2 in H2O (4.5 mL, 8.14 mmol) was added 400 MHz and 90, 100 MHz, respectively. 19F NMR spectra were dropwise. The resulting mixture was stirred for 15 min at 0 °C, and recorded at 250 MHz. Proton chemical shifts are reported in ppm (δ) then an ice-cold 3 M solution of KI (5 mL) was added slowly. The ice (CDCl3; δ 7.26, acetone-d6; δ 2.05, dimethyl sulfoxide (DMSO)-d6; δ bath was removed, and the reaction mixture was heated at the reflux 2.50, methanol-d4; δ 3.31). Carbon chemical shifts are reported in temperature for 1 h. The reaction mixture was allowed to cool to ppm (δ) (CDCl3; δ 77.16, acetone-d6; δ 29.84, DMSO-d6; δ 39.52 room temperature (rt) and extracted with EtOAc (3 × 33 mL). The methanol-d4; δ 49.00). NMR signals were assigned with the help of organic layer was washed with brine (2 × 50 mL), dried over HSQC, HMBC, and DEPT135. 1H NMR and 13C NMR spectra of anhydrous Na2SO4, and concentrated under a vacuum. The residue new compounds are shown in Figures S6-S27. Melting points were was purified by flash column chromatography (hexanes 100%) to determined on a hot stage and are uncorrected. HRMS was recorded afford 1e as a colorless oil (1.37 g, 5.41 mmol, 77% yield). 1H NMR using electrospray ionization. All final compounds are >95% pure by (400 MHz, CDCl3) δ 7.05−7.02 (m, 2H, H-2, H-4), 6.58 (dt, J6,F = HPLC analysis (Figures S28-S32). 10.6 Hz, J6,4 = 2.3 Hz, 1H, H-6), 3.78 (s, 3H, CH3O-5); 19F NMR General Synthesis Procedure to Prepare Compounds M-14 (250 MHz, CDCl3) δ −110.76 (s, F-3″); 13C NMR (100.5 MHz, to M-23. Tolcapone derivatives M-14 to M-23 were chemically CDCl3) δ 163.2 (d, J1,F = 250.0 Hz, C1), 161.4 (d, J5,F = 11.1 Hz, C5), synthesized as described in Scheme 1. The synthesis started with the 119.5 (d, J4,F = 3.2 Hz, C4), 117.4 (d, J2,F = 24.0 Hz, C2), 102.0 (d, J6,F coupling reaction of different preformed aryl Grignard reagents, from = 25.2 Hz, C6), 93.3 (d, J3,F = 11.0 Hz, C3), 55.9 (CH3O-5). IR the corresponding iodoarene derivatives 1a−e, to the known methyl (ATR) ν 2941, 1738, 1601, 1578, 1423, 1277, 1143 cm−1. diprotected catechol 2 to furnish alcohols 3a−e in 49−81% yield. The (3,4-Dimethoxy-5-nitrophenyl)[4-(trifluoromethyl)phenyl]- Grignard reagents were freshly prepared by an iodine−magnesium methanol, 3a. To a solution of 4-iodobenzotrifluoride, 1a (100 μL, exchange reaction86 with i-PrMgBr at −40 °C in THF of the 0.68 mmol), in dry THF (1 mL) at −40 °C, i-PrMgBr (1 M in THF, iodoarene derivatives 1a−e that were commercially available except 680 μL, 0.68 mmol) was added dropwise in 5 min under a nitrogen for 3-fluoro-5-iodomethoxybenzene 1e, obtained from 3-fluoro-5- atmosphere, and the reaction mixture was stirred at the same methoxyaniline through the appropriated arenediazonium chloride temperature for 1 h. Then, a solution of 3,4-dimethoxy-5-nitro- and an ion exchange reaction with KI in 77% yield.87 Catechol 2 was benzaldehyde, 2 (151 mg, 0.72 mmol), in dry THF (1.2 mL) was prepared in excellent yield by the methylation of commercially added. The final mixture was warmed to rt and stirred overnight available 5-nitrovainillin using a phase-transfer catalytic process.88,89 (ON). The reaction was quenched by slow addition of brine (4 mL), The following Dess−Martin periodinane oxidation on 3a−e provided and the aqueous layer was extracted with EtOAc (3 × 10 mL). All ketones 4a−e in 70−99% yield. organic layers were collected, dried over anhydrous Na2SO4, and Removal of both methyl protecting groups of the catechol moiety concentrated under a vacuum, obtaining a yellow oil that was purified was first attempted under conventional conditions using boron by flash column chromatography (hexanes/EtOAc, 2:1) to furnish 3a tribromide in CH2Cl2.90 However, under these standard conditions, a (167 mg, 0.47 mmol, 69% yield) as an orange oil. 1H NMR (400 complex mixture of products was obtained. The presence of the MHz, CDCl3) δ 7.61 (d, J3″,2″ = 8.1 Hz, 2H, H-3″), 7.48 (d, J2″,3″ = 8.1 relatively sensitive nitro-group prevents the use of other routine Hz, 2H, H-2″), 7.29 (dd, J6′,2 = 2.0 Hz, J6′,1 = 0.7 Hz, 1H, H-6′), 7.09 reagents. After some experimentation, it was found that by controlling (d, J2′,6 = 2.0 Hz, 1H, H-2′), 5.83 (s, 1H, H-1), 3.93 (s, 3H, CH3O- the reaction time and the equivalents of BBr3, a single ether cleavage 4′), 3.88 (s, 3H, CH3O-3′), 2.77 (br s, 1H, −OH); 19F NMR (250 was promoted delivering the expected monoprotected catechol. These MHz, CDCl3) δ −63.04 (s, −CF3); 13C NMR (100.6 MHz, CDCl3) δ optimized conditions were applied to the methyl deprotected 154.4 (C3′), 146.5 (C1″), 144.6 (C5′), 142.2 (C4′), 139.5 (C1′), 130.4 compounds 4a−d to furnish monoprotected catechols 5a−d in 50− (q, J4″,F = 32.2 Hz, C4″), 126.8 (C2″, C6″, C5″), 125.9 (q, J3″,F = 3.8 Hz, 91% yield. For compound 4e, the applied reaction conditions also led C3″), 124.1 (q, JCF3,F = 272.0 Hz, −CF3), 114.0 (C2′), 113.8 (C6′), to the removal of the methyl protecting group of R1, delivering 74.7 (C1), 62.1 (CH3O-4′), 56.5 (CH3O-3′). IR (ATR) ν 3422, 1534, compound 5e in 81% yield. 1325, 1165, 1124 cm−1. HRMS (ESI+) calcd for [C16H14F3NO5 + It has been described that the demethylation of ortho-hydroxy H]+ ([M + H]+) 358.2932, found 358.0900. nitroarylmethyl ethers can be accomplished in good yields by a milder (3,4-Dimethoxy-5-nitrophenyl)(4-fluorophenyl)methanol, 3b. procedure in the presence of aluminum chloride (AlCl3).91,92 Compound 3b was prepared as described for 3a by using a solution Accordingly, treatment of 5a−e with AlCl3 and pyridine in refluxing of 4-fluoroiodobenzene, 1b (200 μL, 0.90 mmol), in dry THF (2 chloroform smoothly provided the target tolcapone analogues M-14, mL), i-PrMgBr (1 M in THF, 900 μL, 0.90 mmol), and a solution of M-17, M-20, M-21, and M-23 in reasonably good yields (46−81%). 2 (210 mg, 0.99 mmol) in dry THF (2.2 mL). Purification by flash Synthesis Procedure. Tolcapone was purchased from Fisher. All column chromatography (hexanes/EtOAc, 4:1) afforded 3b (213 mg, tolcapone derivatives used in this study were prepared as described 0.69 mmol, 77% yield) as a colorless oil. 1H NMR (400 MHz, below. acetone-d6) δ 7.50 (br ddd, J2″,3 = J6″,5 = 9.0 Hz, J2″,F = J6″,1 = 5.5 Hz, 3,4-Dimethoxy-5-nitrobenzaldehyde, 2. To a solution of 5- J2″,1 = J6″,1 = 0.6 Hz, 2H, H-2″, H-6″), 7.41 (d, J2′,6 = 2.0 Hz, 1H, H- nitrovanillin (500 mg, 2.50 mmol), NaOH 3 M (1.7 mL, 5.00 2′), 7.38 (dd, J6′,2 = 2.0 Hz, H-6′), 7.09 (br t, J2″,F = J2′,6 = J3″,2 = J5″,6 = mmol), and TBAB (82 mg, 0.25 mmol) in a mixture of CH2Cl2/H2O 9.0 Hz, 1H, H-3″, H-5″), 5.91 (d, J1,OH = 3.3 Hz, 1H, H-1), 5.26 (d, (1:1, 10 mL), dimethyl sulfate (1.3 mL, 13.40 mmol) was added J1,OH = 3.3 Hz, 1H, −OH), 3.93 (s, 3H, CH3O-4′), 3.90 (s, 3H, slowly under a nitrogen atmosphere. The final mixture was stirred CH3O-3′); 19F NMR (250 MHz, CDCl3) δ −117.36 (m, F-4″); 13C vigorously for 24 h (TLC, hexane/EtOAc 3:2). The aqueous layer was NMR (100.6 MHz, acetone-d6) δ 162.9 (d, J4″,F = 243.6 Hz, C4″), extracted with CH2Cl2 (3 × 10 mL). Then, the organic layer was 154.8 (C3′), 145.8 (C5′), 142.9 (C1′), 141.7 (C3′), 141.5 (C1″), 129.3 concentrated under a vacuum and washed with water (20 mL), a 2 M (d, J2″,F = J6″,F = 7.8 Hz, C2″, C6″), 115.8 (d, J4″,F = 21.6 Hz, C3″, C5″), ammonia solution (20 mL), and a 2 M NaOH solution (20 mL) to 115.0 (C2′), 113.6 (C6′), 74.4 (C1), 62.0 (CH3O-4′), 56.9 (CH3O-3′). remove unreacted phenol and dimethyl sulfate. The organic layer was IR (ATR) ν 3416, 1531, 1359, 1280, 1223 cm−1. HRMS (ESI+) calcd dried over anhydrous Na2SO4 and concentrated under a vacuum. for [C15H14FNO5 + H]+ ([M + H]+) 308.0934, found 308.0931. Purification by flash column chromatography using a mixture of (3,5-Dichlorophenyl)(3,4-dimethoxy-5-nitrophenyl)methanol, hexane/EtOAc (3:2) furnished 2 as a pale brown solid (498 mg, 2.36 3c. Compound 3c was prepared as described for 3a by using a mmol, 93% yield). 1H NMR (250 MHz, CDCl3) δ 9.91 (s, 1H, solution of 3,5-dichloroiodobenzene, 1c (558 mg, 2.04 mmol), in dry 14683 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article THF (5 mL), i-PrMgBr (1 M in THF, 2.0 mL, 2.00 mmol), and a 2H, H-3″), 7.67 (d, J2′,6 = 2.0 Hz, 1H, H-2′), 7.65 (d, J6′,2 = 2.0 Hz, solution of 2 (360 mg, 1.70 mmol) in dry THF (5 mL). Purification 1H, H-6′), 4.08 (s, 3H, CH3O-4′), 4.00 (s, 3H, CH3O-3′); 19F NMR by flash column chromatography (hexanes/EtOAc, 1:1) furnished 3c (250 MHz, CDCl3) δ −63.58 (s, −CF3); 13C NMR (100.6 MHz, (427 mg, 1.19 mmol, 70% yield) as an orange solid. Mp 100−103 °C CDCl3) δ 192.7 (C1), 154.6 (C3′), 147.0 (C4′), 144.2 (C5′), 139.8 (from acetone). 1H NMR (400 MHz, acetone-d6) δ 7.50 (dd, J2″,4 = (C1″), 134.4 (q, J4″,F = 33.2 Hz, C4″), 131.7 (C1′), 130.1 (C2″), 125.9 J6″,4 = 1.9 Hz, J2″,1 = J6″,1 = 0.6 Hz, 2H, H-2″, H-6″), 7.48 (d, J2′,6 = 2.0 (q, J3″,F = 3.9 Hz, C3″), 123.6 (q, JCF3,F = 272.6 Hz, −CF3), 118.9 Hz, 1H, H-2′), 7.44 (dd, J6′,2 = 2.0 Hz, J6′,1 = 0.6 Hz, 1H, H-6′), 7.35 (C6′), 116.2 (C2′), 62.4 (CH3O-4′), 56.9 (CH3O-3′). IR (ATR) ν (t, J4″,2 = J4″,6 = 1.9 Hz, 1H, H-4″), 5.95 (d, J1,OH = 3.8 Hz, 1H, H-1), 3086, 2950, 2840, 1655, 1531, 1364, 1323, 1292, 1243, 1163 cm−1. 5.50 (d, JOH,1 = 3.9 Hz, 1H, −OH), 3.95 (s, 3H, CH3O-3′), 3.90 (s, HRMS (ESI+) calcd for [C16H12F3NO5 + H]+ ([M + H]+) 356.0746, 3H, CH3O-4′); 13C NMR (90.5 MHz, acetone-d6) δ 155.0 (C3′), found 356.0740. 149.5 (C1′), 145.9 (C5′), 142.0 (C4′), 141.7 (C1″), 135.5 (C3″, C5″), (3,4-Dimethoxy-5-nitrophenyl)(4-fluorophenyl)methanone, 4b. 127.9 (C4″), 125.9 (C2″, C6″), 115.0 (C2′), 113.7 (C6′), 73.9 (C1), 62.0 Compound 4b was prepared as described for 4a by using alcohol (CH3O-4′), 57.0 (CH3O-3′). IR (ATR) ν 3470, 2948, 1530, 1431, 3b (230 mg, 0.75 mmol) in CH2Cl2 (3 mL) and a solution of DMPI 1370 cm−1. HRMS (ESI−) calcd for [C15H13Cl2NO5-H2O]− ([M- (475 mg, 1.12 mmol) in CH2Cl2 (5 mL). Purification by flash column H2O]−) 340.0143, found 340.0145. chromatography (hexanes/EtOAc, 3:1) afforded 4b (226 mg, 0.74 (3,5-Difluorophenyl)(3,4-dimethoxy-5-nitrophenyl)methanol, mmol, 99% yield) as a brown solid. Mp 73−75 °C (from CHCl3). 1H 3d. Compound 3d was prepared as described for 3a by using a NMR (400 MHz, DMSO-d6) δ 7.89 (br dd, J2″,3 = J6″,5 = 8.9 Hz, J2″,F solution of 3,5-difluoroiodobenzene, 1d (555 mg, 2.31 mmol), in dry = J6″,F = 5.5 Hz, 2H, H-2″, H-6″), 7.70 (d, J6′,2 = 2.0 Hz, 1H, H-6′), THF (5 mL), i-PrMgBr (1 M in THF, 2.5 mL, 2.5 mmol), and a 7.66 (d, J2′,6 = 2.0 Hz, 1H, H-2′), 7.41 (br t, J3″,F = J5″,F = J3″,2 = J5″,6 = solution of 2 (582 mg, 2.78 mmol) in dry THF (5 mL). Purification 8.9 Hz, 2H, H-3″, H-5″), 3.97 (s, 3H, CH3O-4′), 3.96 (s, 3H, CH3O- by flash column chromatography (hexanes/EtOAc, 3:1) afforded 3d 3′); 19F NMR (250 MHz, DMSO-d6) δ −106.21 (F-4″); 13C NMR (612 mg, 1.88 mmol, 81% yield) as an orange oil. 1H NMR (400 (100.6 MHz, DMSO-d6) δ 191.8 (C1), 165.0 (d, J4″,F = 252.0 Hz, MHz, CDCl3) δ 7.25 (d, J6′,2 = 2.0 Hz, 1H, H-6′), 7.07 (d, J2′,6 = 2.0 C4″), 153.4 (C3′), 144.7 (C4′), 143.8 (C5′), 132.8 (d, J2″,F = J6″,F = 9.3 Hz, 1H, H-2′), 6.87 (m, J2″,F = J6″,F = 6.0 Hz, 2H, H-2″, H-6″), 6.70 Hz, C2″, C6″), 132.7 (C1″), 132.4 (C1′), 117.2 (C6′), 116.9 (C2′), (tt, J4″,2 = J4″,6″ = 2.4 Hz, J4″,F = 8.9 Hz, 1H, H-4″), 5.71 (br s, 1H, H- 115.9 (d, J3″,F = J5″,F = 22.1 Hz, C3″, C5″), 61.8 (CH3O-4′), 56.8 1), 3.92 (s, 3H, CH3O-4′), 3.88 (s, 3H, CH3O-3′), 3.10 (br s, 1H, (CH3O-3′). IR (ATR) ν 2954, 1650, 1598, 1531, 1229 cm−1. HRMS −OH); 19F NMR (250 MHz, CDCl3) δ −108.97 (m, F-3″, F-5″); 13C (ESI+) calcd for [C15H12FNO5 + Na]+ ([M + Na]+) 328.0597, found NMR (100.6 MHz, CDCl3) δ 163.2 (dd, J3″,F = J5″,F = 12.5 Hz, J3″,F = 328.0595. J5″,F = 249.7 Hz, C3″,C5″), 154.3 (C3′), 146.6 (t, J1″,F = 8.6 Hz, C1″), (3,5-Dichlorophenyl)(3,4-dimethoxy-5-nitrophenyl)methanone, 144.4 (C5′), 142.2 (C4′), 139.3 (C1′), 114.0 (C2′), 113.8 (C6′), 109.4 4c. Compound 4c was prepared as described for 4a by using alcohol (m, J2″,F = J6″,F = 18.7 Hz, C2″, C6″), 103.4 (t, J4″,F = 25.3 Hz, C4″), 3c (374 mg, 1.00 mmol) in CH2Cl2 (4.7 mL) and a solution of DMPI 74.2 (C1), 62.1 (CH3O-4′), 56.5 (CH3O-3′). IR (ATR) ν 3415, 1598, (505 mg, 1.60 mmol) in CH2Cl2 (4.5 mL). Purification by flash 1534, 1281 cm−1. HRMS (ESI+) calcd for [C15H13F2NO5 + Na]+ column chromatography (hexanes/EtOAc, 3:1) afforded 4c (268 mg, ([M + Na]+) 348.0659, found 348.0663. 0.75 mol, 73% yield) as a white solid. Mp 112−114 °C (from (3,4-Dimethoxy-5-nitrophenyl)(3-fluoro-5-methoxyphenyl)- acetone). 1H NMR (400 MHz, DMSO-d6) δ 7.96 (m, 1H, H-4″), methanol, 3e. Compound 3e was prepared as described for 3a by 7.75 (m, 2H, H-2″, H-6″), 7.75 (m, 1H, H-6′), 7.69 (d, J2′,6 = 2.0 Hz, using a solution of 1-fluoro-3-iodo-5-methoxybenzene, 1e (308 mg, H-2′), 3.98 (s, 3H, CH3O-4′), 3.96 (s, 3H, CH3O-3′); 13C NMR 1.22 mmol), in dry THF (3 mL), i-PrMgBr (1 M in THF, 1.9 mL, (100.6 MHz, DMSO-d6) δ 190.8 (C1), 153.5 (C3′), 145.3 (C4′), 143.9 1.90 mmol), and a solution of 2 (310 mg, 1.20 mmol) in dry THF (4 (C5′), 139.6 (C1″), 134.6 (C3″, C5″), 132.1 (C4″), 131.3 (C1′), 128.0 mL). Purification by flash column chromatography (hexanes/EtOAc, (C2″, C6″), 117.9 (C6′), 116.9 (C2′), 61.9 (CH3O-4′), 56.9 (CH3O- 5:1) afforded 3e (191 mg, 0.57 mmol, 49% yield) as a white solid. Mp 3′). IR (ATR) ν 3077, 1659, 1535, 1365, 1290 cm−1. HRMS (ESI+) 100−101 °C (from CHCl3). 1H NMR (400 MHz, CDCl3) δ 7.28 calcd for [C15H11Cl2NO5 + H]+ ([M + H]+) 356.0093, found (dd, J6′,2 = 2.0 Hz, J6′,1 = 0.6 Hz, 1H, H-6′), 7.11 (d, J2′,6 = 2.0 Hz, 1H, 356.0087. H-2′), 6.69 (m, H-6″), 6.65 (br dm, J2″,F = 9.1 Hz, J2″,4 = J2″,6 = 2.0 (3,5-Difluorophenyl)(3,4-dimethoxy-5-nitrophenyl)methanone, Hz, 1H, H-2″), 6.53 (dt, J4″,F = 10.5 Hz, J4″,2 = J4″,2 = 2.3 Hz, 1H, H- 4d. Compound 4d was prepared as described for 4a by using alcohol 4″), 5.70 (s, 1H, H-1), 3.94 (s, 3H, CH3O-4′), 3.88 (s, 3H, CH3O- 3d (233 mg, 716 μmol) in CH2Cl2 (4 mL) and a solution of DMPI 3′), 3.78 (s, 3H, CH3O-5″), 2.72 (br s, 1H, −OH); 19F NMR (250 (456 mg, 1.07 mmol) in CH2Cl2 (5 mL). Purification by flash column MHz, CDCl3) δ −110.76 (s, F-3″); 13C NMR (100.6 MHz, CDCl3) δ chromatography (hexanes/EtOAc, 3:1) afforded 4d (259 mg, 0.80 163.9 (d, J3″,F = 246.3 Hz, C3″), 161.3 (d, 3J5″,F = 11.2 Hz, C5″), 154.3 mmol, 75% yield) as a white solid. Mp 151−152 °C (from CH2Cl2). 1 (C3′), 145.8 (d, J1″,F = 8.8 Hz, C1″), 144.6 (s, C5′), 142.2 (s, C4′), H NMR (360 MHz, CDCl3) δ 7.67 (d, J6′,2 = 2.0 Hz, 1H, H-6′), 7.63 139.5 (C1′), 114.0 (C2′), 113.8 (C6′), 108.4 (d, J6″,F = 3.8 Hz, C6″), (d, J2′,6 = 2.0 Hz, 1H, H-2′), 7.28 (m, J2″,F = J6″,F = 5.2 Hz, 2H, H-2″, 105.7 (d, J2″,F = 22.7 Hz, C2″), 101.1 (d, J4″,F = 25.4 Hz, C4″), 74.8 (d, H-6″), 7.09 (tt, J4″,2 = J4″,6 = 2.2 Hz, J4″,F = 8.5 Hz, 1H, H-4″), 4.09 (s, J1,F = 2.3 Hz, C1), 62.1 (CH3O-4′), 56.6 (CH3O-3′), 55.7 (CH3O-5″). 3H, CH3O-4′), 4.00 (s, 3H, CH3O-3′); 19F NMR (250 MHz, CDCl3) IR (ATR) ν 3230, 2837, 1594, 1530, 1453, 1344, 1133 cm−1. HRMS δ −107.50 (t, J = 6.9 Hz, F-3″, F-5″); 13C NMR (90.5 MHz, CDCl3) (ESI−) calcd for [C16H16FNO6-H2O]− ([M-H2O]−) 320.0934, found δ 191.2 (br t, JCO,F = 2.3 Hz, C1), 163.0 (dd, J3″,F = J5″,F = 12.0 Hz, J3″,F 320.0936. = J5″,F = 252.5 Hz, C3″, C5″), 154.6 (C3′), 147.0 (C4′), 144.2 (C5′), (3,4-Dimethoxy-5-nitrophenyl)[4-(trifluoromethyl)phenyl]- 139.6 (t, J1″,F = 7.8 Hz, C1″), 131.4 (C1′), 118.7 (C6′), 116.2 (C2′), methanone, 4a. To a solution of alcohol 3a (167 mg, 0.52 mmol) in 112.9 (m, J2″,F = J6″,F = 18.6 Hz, C2″, C6″), 108.5 (t, J4″,F = 25.5 Hz, dry CH2Cl2 (2 mL) under a nitrogen atmosphere, a solution of DMPI C4″), 62.4 (CH3O-4′), 56.9 (CH3O-3′). IR (ATR) ν 3084, 1665, (338 mg, 1.07 mmol) in CH2Cl2 (3 mL) was added dropwise, and the 1589, 1437, 1320 cm−1. HRMS (ESI+) calcd for [C15H11F2NO5 + mixture was stirred ON at RT. The reaction was quenched with the Na]+ ([M + Na]+) 346.0503, found 346.0501. addition of 4 mL of a prepared solution of Na2S2O3 (1.13 g) in a (3,4-Dimethoxy-5-nitrophenyl)(3-fluoro-5-methoxyphenyl)- saturated aqueous solution of NaHCO3 (6 mL), and the mixture was methanone, 4e. Compound 4e was prepared as described for 4a by stirred for 15 min. The aqueous phase was extracted with CH2Cl2 (3 using alcohol 3e (287 mg, 0.85 mmol) in CH2Cl2 (4 mL) and a × 15 mL), and the combined organic extracts were dried over solution of DMPI (546 mg, 1.28 mmol) in CH2Cl2 (6 mL). anhydrous Na2SO4 and concentrated under a vacuum. Purification by Purification by flash column chromatography (hexanes/EtOAc, 5:2) flash column chromatography (hexanes/EtOAc, 3:1) of the resulting afforded 4e (199 mg, 0.59 mmol, 70% yield) as a white solid. Mp 95 residue provided ketone 4a (131 mg, 0.37 mmol, 70% yield) as a °C (from CHCl3). 1H NMR (400 MHz, CDCl3) δ 7.69 (d, J6′,2 = 2.0 white solid. Mp 118−119 °C (from CH2Cl2). 1H NMR (400 MHz, Hz, 1H, H-6′), 7.63 (d, J2′,6 = 2.0 Hz, 1H, H-2′), 7.08 (m, H-6″), 7.02 CDCl3) δ 7.87 (d, J2″,3 = 8.2 Hz, 2H, H-2″), 7.79 (d, J3″,2 = 8.2 Hz, (ddd, J2″,F = 8.4 Hz, J2″,4 = 2.3 Hz, J2″,6 = 1.4 Hz, 1H, H-2″), 6.86 (dt, 14684 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article J4″,F = 10.2 Hz, J4″,2 = J4″,2 = 2.3 Hz, 1H, H-4″), 4.07 (s, 3H, CH3O- d, 1H, H-2′), 7.28 (m, 2H, H-2″, H-6″), 7.10 (tt, J4″,2 = J4″,6 = 2.2 Hz, 4′), 3.99 (s, 3H, CH3O-3′), 3.85 (s, 3H, CH3O-5″); 19F NMR (250 J4″,F = 8.5 Hz, 1H, H-4″), 4.04 (s, 3H, CH3O-3′); 19F NMR (250 MHz, CDCl3) δ −110.1 (s, F-3″); 13C NMR (90.5 MHz, CDCl3) δ MHz, CDCl3) δ −110.05 (t, J = 7.6 Hz, F-3″, F-5″); 13C NMR (100.6 192.4 (d, J1,F = 2.8 Hz C1), 163.3 (d, J3″,F = 248.3 Hz, C3″), 161.1 (d, MHz, CDCl3) δ 191.1 (C1), 163.0 (dd, J3″,F = J5″,F = 12.0 Hz, J3″,F = J5″,F = 10.8 Hz, C5″), 154.4 (C3′), 146.7 (C4′), 144.2 (C5′), 139.0 (d, J5″,F = 252.5 Hz, C3″,C5″), 150.9 (C3′), 150.4 (C4′), 139.7 (t, J1″,F = 7.8 J1″,F = 8.4 Hz, C1″), 132.0 (C1′), 118.7 (C6′), 116.3 (C2′), 111.2 (d, Hz, C1″), 132.9 (C5′), 127.3 (C1′), 119.7 (C6′), 117.0 (C2′), 112.8 (m, J6″,F = 2.8 Hz, C6″), 109.1 (d, J2″,F = 23.1 Hz, C2″), 106.4 (d, J4″,F = J2″,F = J6″,F = 18.8 Hz, C2″, C6″), 108.4 (t, J4″,F = 25.4 Hz, C4″), 57.2 25.0 Hz, C4″), 62.4 (CH3O-4′), 56.8 (CH3O-3′), 56.1 (CH3O-5″). IR (CH3O-3′). IR (ATR) ν 3098, 1590, 1326, 1242 cm−1. HRMS (ATR) ν 2921, 1597, 1530, 1429, 1318, 1146 cm−1. HRMS (ESI+) (ESI−) calcd for [C14H9F2NO5-H]− ([M − H]−) 308.0371, found calcd for [C16H14FNO6 + Na]+ ([M + Na]+) 358.0703, found 308.0378. 358.0705. 3-Fluoro-5-hydroxyphenyl)(4-hydroxy-3-methoxy-5- (4-Hydroxy-3-methoxy-5-nitrophenyl)[4-(trifluoromethyl)- nitrophenyl)methanone, 5e. Compound 5e was prepared as phenyl]methanone, 5a. To a solution of 4a (359 mg, 1.01 mmol) in described for 5a by using ketone 4e (199 mg, 0.56 mmol) in CH2Cl2 (5.5 mL) was added dropwise BBr3 (1 M in CH2Cl2, 8.0 mL, CH2Cl2 (3 mL) and BBr3 (1 M in CH2Cl2, 4.5 mL, 4.5 mmol). 8.0 mmol) at −10 °C. The reaction was allowed to proceed at RT for Purification by flash column chromatography (CH2Cl2/MeOH, 20:1) 2 h. The reaction mixture was quenched carefully with water (5 mL), delivered 5e (139 mg, 0.45 mmol, 81% yield) as a yellow solid. Mp and the resulting aqueous layer was extracted with EtOAc (2 × 20 155−156 °C (from CHCl3). 1H NMR (250 MHz, CDCl3) δ 8.11 (d, mL). The organic extracts were dried with Na2SO4, and the solvent J6′,2 = 1.9 Hz, 1H, H-6′), 7.69 (d, J2′,6 = 1.9 Hz, 1H, H-2′), 7.04 (m, was removed under a vacuum to give the crude product. Purification H-6″), 6.99 (ddd, J2″,F = 8.4 Hz, J2″,4 = 2.3 Hz, J2″,6 = 1.4 Hz, 1H, H- by flash column chromatography (CH2Cl2/MeOH, 30:1) afforded 5a 2″), 6.85 (dt, J4″,F = 9.4 Hz, J4″,2 = J4″,2 = 2.3 Hz, 1H, H-4″), 4.02 (s, (270 mg, 0.79 mmol, 78% yield) as a yellow solid. Mp 178−181 C 3H, CH3O-3′); 19F NMR (250 MHz, CDCl3) δ −110.01 (F-3″); 13C (from MeOH). 1H NMR (400 MHz, DMSO-d6) δ 7.95−7.93 (m, NMR (90.5 MHz, CDCl3) δ 192.3 (C1), 163.4 (d, J3″,F = 248.5 Hz, 4H, H-2″, H-3″, H-5″, H-6″), 7.78 (d, J6′,2 = 1.9 Hz, 1H, H-6′), 7.64 C3″), 157.5 (d, J5″,F = 11.8 Hz, C5″), 150.7 (C3′), 150.2 (C4′), 139.4 (d, J2′,6 = 1.9 Hz, 1H, H-2′), 3.96 (s, 3H, CH3O-3′); 19F NMR (250 (d, J1″,F = 8.1 Hz, C1″), 132.9 (C5′), 127.8 (C1′), 119.7 (C6′), 117.2 MHz, CDCl3) δ −63.51 (−CF3); 13C NMR (100.6 MHz, DMSO-d6) (C2′), 112.6 (d, J6″,F = 2.7 Hz, C6″), 109.3 (d, J2″,F = 23.1 Hz, C2″), δ 192.2 (C1), 149.7 (C3′), 146.9 (C4′), 140.5 (C1″), 136.5 (C5′), 132.0 107.8 (d, J4″,F = 25.0 Hz, C4″), 57.2 (CH3O-3′). IR (ATR) ν 3213, (q, J4″,F = 32.1 Hz, C4″), 130.1 (C2″, C6″), 125.9 (C1′), 125.6 (q, J3″,F = 2922, 1738, 1600, 1546, 1330, 1242, 1134 cm−1. HRMS (ESI−) calcd J5″,F = 4.2 Hz, C3″,C5″), 123.8 (q, JCF3,F = 272.4, −CF3), 120.0 (C2′), for [C14H10FNO6-H]− ([M − H]−) 306.0414, found 306.0424. 115.0 (C6′), 56.8 (CH3O-3′). IR (ATR) ν 3198, 2926, 1652, 1614, (3,4-Dihydroxy-5-nitrophenyl)-[4-(trifluoromethyl)phenyl]- 1551, 1287 cm−1. HRMS (ESI−) calcd for [C15H10F3NO5-H]− ([M methanone, M-14. To a mixture of monoprotected catechol 5a (20 − H]−) 340.0439, found 340.0439. mg, 62 μmol) and AlCl3 (15 mg, 0.11 mmol) in dry CHCl3 (1 mL) in (4-Fluorophenyl)(4-hydroxy-3-methoxy-5-nitrophenyl)- an argon atmosphere, pyridine (13 μL, 0.17 mmol) was added methanone, 5b. Compound 5b was prepared as described for 5a by dropwise at 0 °C. The reaction mixture was heated to the reflux using ketone 4b (259 mg, 0.85 mmol) in CH2Cl2 (4 mL) and BBr3 (1 temperature and stirred until the starting material was consumed. The M in CH2Cl2, 6.8 mL, 6.8 mmol). Purification by flash column orange suspension was concentrated under a vacuum, and HCl (5 M, chromatography (CH2Cl2/MeOH, 30:1) delivered 5b (225 mg, 0.77 2 mL) was added, keeping the temperature <25 °C. The aqueous mmol, 91% yield) as a green solid. Mp 120−122 °C (from CH2Cl2). layer was extracted with EtOAc (3 × 10 mL), and the combined 1 H NMR (400 MHz, methanol-d4) δ 7.92 (d, J6′,2 = 1.9 Hz, 1H, H- organic yellow extracts were dried over anhydrous Na2SO4 and 6′), 7.85 (br dd, J2″,3 = J6″,5 = 8.8 Hz, J2″,F = J6″,F = 5.4 Hz, 2H, H-2″, concentrated under a vacuum to give the crude product. Purification H-6″), 7.63 (d, J2′,6 = 1.8 Hz, 1H, H-2′), 7.28 (br t, J3″,F = J5″,F = J3″,2 = by recrystallization from acetone and water provided M-14 (15 mg, J5″,6 = 8.8 Hz, 2H, H-3″, H-5″), 3.98 (s, 3H, CH3O-3′); 19F NMR 45 μmol, 75% yield) as a pale brown solid. Mp 68−71 °C (from (250 MHz, methanol-d4) δ −108.63 (F-4″); 13C NMR (100.6 MHz, CH2Cl2). 1H NMR (360 MHz, acetone-d6) δ 8.02−8.00 (m, J2″,3 = methanol-d4) δ 194.0 (C1), 166.8 (d, J4″,F = 252.9 Hz, C4″), 151.8 J6″,5 = 8.1 Hz, 3H, H-2″, H-6″, H-6′), 7.94 (br d, J3″,2 = J6″,5 = 8.1 Hz, (C3′), 150.2 (C4′), 136.4 (C5′), 134.9 (d, J1″,F = 3.0 Hz, C1″), 133.5 (d, 2H, H-3″, H-5″), 7.68 (d, J2′,6 = 1.9 Hz, 1H, H-2′); 19F NMR (250 J2″,F = J6″,F = 9.2 Hz, C2″, C6″), 128.2 (C1′), 121.2 (C6′), 116.6 (d, J3″,F MHz, acetone-d6) δ −68.53 (−CF3); 13C NMR (90.5 MHz, acetone- = J5″,F = 22.3 Hz, C3″, C5″), 116.5 (C2′), 57.3 (CH3O-3′). IR (ATR) ν d6) δ 193.0 (C1), 148.9 (C3′), 148.2 (C4′), 141.6 (C1″), 135.3 (C5′), 3170, 2924, 1599, 1547, 1229 cm−1. HRMS (ESI−) calcd for 133.8 (q, J4″,F = 32.4 Hz, C4″), 130.9 (C2″, C6″), 128.5 (C1′), 126.4 (q, [C14H10FNO5-H]− ([M − H]−) 290.0475, found 290.0465. J3″,F = J5″,F = 3.8 Hz, C3″,C5″), 125.0 (q, JCF3,F = 286.7, −CF3), 121.7 (3,5-Dichlorophenyl)(4-hydroxy-3-methoxy-5-nitrophenyl)- (C2′), 119.3 (C6′). IR (ATR) ν 3247, 1655, 1545, 1408, 1247 cm−1. methanone, 5c. Compound 5c was prepared as described for 5a by HRMS (ESI−) calcd for [C14H8F3NO5-H]− ([M − H]−) 326.0276, using ketone 4c (130 mg, 0.37 mmol) in CH2Cl2 (2 mL) and BBr3 (1 found 326.0284. M solution in CH2Cl2, 2.9 mL, 2.9 mmol). Purification by (3,4-Dihydroxy-5-nitrophenyl)-(4-fluorophenyl)methanone, M- recrystallization from acetone furnished 5c (98 mg, 0.28 mmol, 17. Compound M-17 was prepared as described for M-14 by using 78% yield) as a yellow solid. Mp 204−205 °C (from acetone). 1H 5b (60 mg, 0.21 mmol), AlCl3 (57 mg, 0.41 mmol), and pyridine (50 NMR (400 MHz, DMSO-d6) δ 7.94 (t, J4″,2 = J4″,6 = 2.0 Hz, 1H, H- μL, 0.62 mmol) in dry CHCl3 (3 mL). Purification by digestion in 4″), 7.78 (d, J6′,2 = 2.0 Hz, 1H, H-6′), 7.73 (d, J2″,4 = J6″,4 = 2.0 Hz, hexane afforded M-17 (42 mg, 0.15 mmol, 70% yield) as a pale brown 2H, H-2″, H-6″), 7.59 (d, J2′,6 = 2.0 Hz, 1H, H-2′), 3.95 (s, 3H, solid. Mp 162−165 °C (from CH2Cl2). 1H NMR (360 MHz, acetone- CH3O-3′); 13C NMR (100.6 MHz, DMSO-d6) δ 190.5 (C1), 149.8 d6) δ 7.98 (d, J6′,2 = 2.0 Hz, 1H, H-6′), 7.91 (br dd, J2″,3 = J6″,5 = 8.7 (C3′), 147.3 (C4′), 140.3 (C5′), 136.6 (C1′), 134.4 (C3″, C5″), 131.6 Hz, J2″,F = J6″,F = 5.5 Hz, 2H, H-2″, H-6″), 7.65 (d, J2′,6 = 2.0 Hz, 1H, (C4″), 127.7 (C2″, C6″), 125.4 (C1″), 120.2 (C6′), 114.8 (C2′), 56.8 H-2′), 7.35 (br t, J3″,2 = J6″,5 = 8.7 Hz, J3″,F = J5″,F = 8.7 Hz, 2H, H-3″, (CH3O-3′). IR (ATR) ν 3190, 3079, 1543, 1405, 1285 cm−1. HRMS H-5″); 19F NMR (250 MHz, acetone-d6) δ −105.17 (F-4″); 13C (ESI−) calcd for [C14H9Cl2NO5-H]− ([M − H]−) 339.9780, found NMR (90.5 MHz, acetone-d6) δ 192.5 (C1), 166.2 (d, J3″,F = 252.0 339.9786. Hz, C4″), 148.6 (C3′), 147.7 (C4′), 135.1 (C5′), 134.5 (d, J1″,F = 2.8 (3,5-Difluorophenyl)(4-hydroxy-3-methoxy-5-nitrophenyl)- Hz, C1″), 133.4 (d, J2″,F = J6″,F = 9.1 Hz, C2″, C6″), 129.2 (C1′), 121.9 methanone, 5d. Compound 5d was prepared as described for 5a by (C2′), 118.9 (C6′), 116.4 (d, J3″,F = J5″,F = 22.0 Hz, C3″,C5″). IR (ATR) using ketone 4d (216 mg, 0.67 mmol) in CH2Cl2 (4 mL) and BBr3 (1 ν 3194, 2923, 1651, 1585, 1296 cm−1. HRMS (ESI−): calcd for M in CH2Cl2, 5.4 mL, 5.4 mmol). Purification by recrystallization [C13H8FNO5-H]− 276.0308 ([M − H]−); found 276.0315. from acetone afforded 5d (103 mg, 0.33 mmol, 50% yield) as a yellow (3,5-Dichlorophenyl)-(3,4-dihydroxy-5-nitrophenyl)methanone, solid. Mp 190−192 °C (from acetone). 1H NMR (400 MHz, CDCl3) M-20. Compound M-20 was prepared as described for M-14 by using δ 11.18 (s, 1H, −OH), 8.08 (br d, J6′,2 = 1.6 Hz, 1H, H-6′), 7.70 (br 5c (45 mg, 0.13 mmol), AlCl3 (35 mg, 0.26 mmol), and pyridine (50 14685 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article μL, 0.39 mmol) in dry CHCl3 (4 mL). Purification by recrystallization ometer (Jasco). Trp residues were excited at 295 nm, and emission from acetone and water provided M-20 (35 mg, 0.10 mmol, 81% spectra was collected from 310 to 400 nm. Trp exposure upon yield) as a green light solid. Mp 182−184 °C (from acetone). 1H denaturation red shifts the wavelength of maximum fluorescence from NMR (250 MHz, acetone-d6) δ 10.70 (br s, 1H, −OH), 8.02 (d, J6′,2′ 335 to 355 nm, approximately. The 355/335 fluorescence emission = 2.0 Hz, 1H, H-2′), 7.79 (t, J4″,2″ = J4″,6″ = 1.9 Hz, 1H, H-4″), 7.74 intensity was normalized from minimum (100% folded) to maximum (d, J2″,4″ = J2″,4″ = 1.9 Hz, 2H, H-2″, H-6″), 7.66 (d, J2′,6′ = 2.0 Hz, H- (0% folded) and plotted as a function of time. The TTR fluorescence 2′); 13C NMR (100.6 MHz, acetone-d6) δ 191.3 (C1), 148.9 (C3′), of the control sample after incubation at RT for 96 h in 6 M urea was 148.4 (C4′), 141.2 (C5′), 135.9 (C3″, C5″), 135.3 (C1′), 132.6 (C4″), considered the maximum. 128.7 (C2″, C6″), 128.2 (C1″), 121.7 (C2′), 119.3 (C6′). IR (ATR) ν TTR In Vitro Aggregation Assay. The anti-aggregational activity 3331, 1617, 1544, 1308 cm −1 . HRMS (ESI−) calcd for of TTR ligands was evaluated using the established acid-mediated [C13H7Cl2NO5-H]− ([M − H]−) 325.9623, found 325.9630. aggregation assay.44 In short, WT-TTR solutions (7 μM in 10 mM (3,5-Difluorophenyl)-(3,4-dihydroxy-5-nitrophenyl)methanone, sodium phosphate, 100 mM KCl, 1 mM EDTA, 1 mM DTT, pH 7.0) M-21. Compound M-21 was prepared as described for M-14 by using were mixed with increasing concentrations of test compounds 5d (100 mg, 0.32 mmol), AlCl3 (86 mg, 0.65 mmol), and pyridine (prepared in 100% DMSO). The percentage of DMSO was adjusted (105 μL, 1.29 mmol) in dry CHCl3 (10 mL). Purification by to 5% (v/v) in the final reaction assay mixture. After incubating for 30 recrystallization from acetone and water furnished M-21 (71 mg, 0.24 min at 37 °C, the pH of the samples was dropped to 4.2 by the mmol, 75% yield) as a yellow light solid. Mp 135−137 °C (from addition of acetate buffer (100 mM sodium acetate, 100 mM KCl, 1 acetone). 1H NMR (360 MHz, DMSO-d6) δ 10.96 (br s, 2H, 2 × mM EDTA, 1 mM DTT, pH 4.2), and the solutions were further −OH), 7.70 (d, J6′,2′ = 2.0 Hz, 1H, H-6′), 7.60 (br tt, J4″,2″ = J4″,6″ = incubated for 72 h at 37 °C. The extent of TTR aggregation was 2.2 Hz, J4″,F = 9.3 Hz, 1H, H-4″), 7.48 (d, J2′,6′ = 2.0 Hz, 1H, H-2′), assessed by measuring turbidity at 340 nm using a Varian Cary Eclipse 7.43 (m, J2″,F = J2″,F = 5.8 Hz, 2H, H-2″, H-6″); 19F NMR (250 MHz, Fluorescence Spectrophotometer (Agilent Technologies). As some of DMSO-d6) δ −110.07 (s, F-3″, F-5″); 13C NMR (90.5 MHz, DMSO- the compounds present dose-dependent absorbance at 340 nm, each d6) δ 190.7 (C1), 162.0 (dd, J3″,F = J5″,F = 12.6 Hz, J3″,F = J5″,F = 249.9 measurement was corrected with a buffer containing the same Hz, C3″,C5″), 147.9 (C3′), 146.6 (C4′), 140.4 (t, J1″,F = 8.0 Hz, C1″), concentration of the test compound but lacking TTR. For each 136.9 (C5′), 125.4 (C1′), 118.5 (C6′), 118.5 (C2′), 112.5 (m, J2″,F = inhibitor concentration, the percentage of TTR aggregation was given J6″,F = 18.5 Hz, C2″, C6″), 107.7 (t, J4″,F = 25.8 Hz, C4″). IR (ATR) ν by the ratio of the turbidity of the sample of interest to that of a 3363, 1617, 1587, 1436, 1323 cm−1. HRMS (ESI−) calcd for control sample incubated without compound multiplied by 100%. [C13H7F2NO5-H]− ([M − H]−) 294.0214, found 294.0221. TTR Stabilization Studies by Isoelectric Focusing. Isoelectric (3,4-Dihydroxy-5-nitrophenyl)-(3-fluoro-5-hydroxyphenyl)- focusing (IEF) under semidenaturing conditions was performed as methanone, M-23. Compound M-23 was prepared as described for described previously95 to evaluate the stabilizing effect of both M-14 by using 5e (158 mg, 0.52 mmol), AlCl3 (511 mg, 3.80 mmol), tolcapone and M-23 on recombinant TTR and in plasma TTR. and pyridine (360 μL, 4.45 mmol) in dry CHCl3 (10 mL). Recombinant WT-TTR was produced using an Escherichia coli Purification by flash column chromatography (CH2Cl2/MeOH, bacterial expression system, as detailed elsewhere.96 For the plasma 19:1) afforded catechol M-23 (69 mg, 0.24 mmol, 46% yield) as a assays, human plasma from control individuals (n = 6), carrying WT- yellow solid. Mp 215−217 °C (from acetone). 1H NMR (360 MHz, TTR (≈3.9 μM), was incubated ON at 4 °C with tafamidis, acetone-d6) δ 10.61 (br s, 1H, −OH), 9.28 (br s, 2H, 2 × −OH), 8.02 tolcapone, or M-23 at two different concentrations (19.5 and 39 μM). (d, J6′,2′ = 2.0 Hz, 1H, H-6′), 7.65 (d, J2′,6′ = 2.0 Hz, 1H, H-2′), 7.08 Similarly, recombinant WT-TTR (6 μM) was also treated ON at 4 °C (s, C-6″), 7.00 (br dt, J2″,F = 8.8 Hz, 1H, H-2″), 6.90 (br dt, J4″,2″ = 2.2 with the same compounds at concentrations of 30 and 60 μM. DMSO Hz, J4″,F = 10.3 Hz, 1H, H-4″); 19F NMR (250 MHz, acetone-d6) δ (5%) was used as vehicle. Control samples were incubated in similar −112.87 (t, J = 9.5 Hz, F-3″); 13C NMR (90.5 MHz, acetone-d6) δ conditions without the compounds. After incubation, the samples 192.4 (d, JCO,F = 2.7 Hz, C1), 164.1 (d, J3″,F = 254.0 Hz, C3″), 159.9 were loaded into a native PAGE, and the gel band containing TTR (d, J5″,F = 11.7 Hz, C5″), 148.8 (C3′), 148.1 (C4′), 140.6 (d, J1″,F = 8.4 was excised and applied to an IEF gel. The IEF gel contained 4 M Hz, C1″), 135.2 (C5′), 128.7 (C1′), 121.8 (C2′), 119.1 (C6′), 113.5 (d, urea (semidenaturing conditions) and 5% (v/v) ampholytes, pH 4− J6″,F = 2.2 Hz, C6″), 108.1 (d, J2″,F = 23.3 Hz, C2″), 107.4 (d, J4″,F = 6.5 (Sigma-Merck), and was run at 1200 V for 5 h. Proteins were fixed 24.2 Hz, C4″). IR (ATR) ν 3229, 1599, 1442, 1252 cm−1. HRMS and stained with Coomassie blue. The gels were scanned using a GS- (ESI−) calcd for [C13H8FNO6-H]− ([M − H]−) 292.0257, found 900 calibrated densitometer (Bio-Rad) and analyzed by densitometry 292.0265. using the QuantityOne software version 4.6.6 (Bio-Rad). The ratio of Isothermal Titration Calorimetry. The thermodynamic param- the TTR tetramer over total TTR (TTR tetramer + monomer) was eters that characterize the binding of TTR ligands to WT-TTR were calculated for each plasma sample, and the percentage of tetramer determined using a MicroCal Auto-iTC200 Calorimeter (MicroCal, stabilization was calculated as ((ratio treated sample − ratio control Malvern-Panalytical), as detailed before.44 A 100 μM solution of the sample)/ratio control sample) × 100. Treated and control plasma compound (in a PBS buffer pH 7.0 containing 100 mM KCl, 1 mM samples come from the same donor. EDTA and 2.5% DMSO) was titrated into an ITC cell containing a 5 T4 Binding Competition Assay. Binding competition assays with μM solution of WT-TTR in the same buffer at 25 °C. A stirring speed radioactive T4 were performed by incubation of 5 μL of human of 750 rpm and 2 μL injections were programmed, with a 150 s plasma samples of WT-TTR carriers (n = 4) with 1 μL of [125I]-T4 equilibration period between each injection to allow the calorimetric (specific radioactivity 1250 μCi/μg; concentration 320 μCi/mL; signal (thermal power) to return to baseline and a 10 μcal/s reference Perkin Elmer) in the presence of 39 μM compounds. The negative power. Two independent titrations were done for each TTR ligand. control was performed by adding the same percentage of DMSO in The experimental data were analyzed with a general model for a the samples. After 1 h incubation at rt, plasma proteins were protein with two ligand-binding sites93,94 implemented in Origin 7.0 fractionated by native PAGE,97 and the gels were dried and exposed (OriginLab) accounting for cooperative and noncooperative binding. to an autoradiography film. The films were scanned, and the intensity The best fit of the binding isotherm was attained with a model of the bands was determined by densitometry using Image Lab considering two identical binding sites (i.e., no cooperativity) for software version 5.2.1 (Bio-Rad). The ratio of T4 bound to TTR over tolcapone and its derivatives. total T4 (TBG + ALB + TTR) was calculated for each sample and Urea-Induced TTR Tetramer Dissociation Kinetics. TTR normalized to the maximum value, which corresponds to the negative solutions (1.8 μM in PBS) were incubated with the different TTR control sample. ligands (3.6 μM) for 30 min at RT, and 6 M urea was added. A Solubility Measurements. The solubility tests were carried out control sample containing the same amount of DMSO rather than the by weighing M-23 and tolcapone (solid samples). Then, distilled or compound was prepared. The process of unfolding was tracked by deionized water was added under gentle agitation until total intrinsic fluorescence spectroscopy using an FP-8200 Spectrofluor- dissolution. All the experiments were performed in triplicate. 14686 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Cytotoxicity Assay. Cytotoxicity analyses were performed for Molecular formula strings for the new compounds evaluating the potential M-23 chemical toxicity to human cells. described in the study (CSV) Tolcapone was used for comparison. Two cell lines, HeLa and HepG2 cells, were cultured in MEM ALPHA (Gibco) and 10% FBS at 37 °C Accession Codes in 5% CO2 humidified atmosphere. HeLa cells were seeded at 3500 The atomic coordinates have been deposited into the Protein cells/well and HepG2 cells at 4500 cells/well in 96-well plates and Data Bank with accession code 7QC5. The authors will release incubated with increasing concentrations of compound at a range of the atomic coordinates and experimental data upon article concentrations from 2 to 100 μM during 72 h at 37 °C. Controls were publication. performed with the equivalent amount of DMSO relative to each concentration of compound diluted in MilliQ water. Then, cell viability was determined by adding 10 μL of the PrestoBlue reagent (Thermo Fisher Scientific), and after an incubation period of 15 min ■ AUTHOR INFORMATION Corresponding Author at 37 °C, the fluorescence intensity was collected using a 590/20 filter Salvador Ventura − Institut de Biotecnologia i Biomedicina with an excitation wavelength of 535 in a Victor3 Multilabel Reader and Departament de Bioquímica i Biologia Molecular, (PerkinElmer). Experiments were carried out in triplicate, and the Universitat Autònoma de Barcelona, Barcelona 08193, percentage of cell viability for each well was calculated as (intensity Spain; ICREA, E-08010 Barcelona, Spain; orcid.org/ sample − mean intensity blank)/(mean intensity control − mean 0000-0002-9652-6351; Email: salvador.ventura@uab.es intensity blank) × 100, where ″mean intensity blank″ corresponds to the mean intensity of wells with PrestoBlue alone and ″mean intensity Authors control″ is the mean intensity of wells that contain the corresponding Francisca Pinheiro − Institut de Biotecnologia i Biomedicina percentage of DMSO. and Departament de Bioquímica i Biologia Molecular, Crystallization and Structure Determination of the WT-TTR/ M-23 Complex. Co-crystals of WT-TTR/M-23 were obtained as Universitat Autònoma de Barcelona, Barcelona 08193, Spain described previously.44 In short, WT-TTR (85 μM) was mixed with Irantzu Pallarès − Institut de Biotecnologia i Biomedicina and 10-fold molar excess of M-23 and co-crystallized at 18 °C by the Departament de Bioquímica i Biologia Molecular, Universitat hanging-drop vapor diffusion method (1:1, complex and reservoir Autònoma de Barcelona, Barcelona 08193, Spain; solution). Crystals were grown from a solution containing 30% PEG orcid.org/0000-0002-8205-2060 400, 200 mM CaCl2, and 100 mM HEPES (pH 8.0) and directly Francesca Peccati − Departament de Química, Universitat flash-frozen in liquid nitrogen before analysis. Diffraction data were Autònoma de Barcelona, Barcelona 08193, Spain; Present recorded from PEG400 cryo-cooled crystals (100 K) at the ALBA Address: Center for Cooperative Research in Biosciences Synchrotron in Barcelona (BL13-XALOC beamline).98 Data were integrated and merged using XDS99 and scaled, reduced, and further (CIC bioGUNE), Basque Research and Technology analyzed using Ccp4.100 The structure of TTR/M-23 complex was Alliance (BRTA), 48160 Derio, Spain; orcid.org/0000- determined by molecular replacement with Phenix (version 1.19.2- 0002-7813-8216 4158)101 using the crystal structure of TTR (PDB code 1F41) as a Adrià Sánchez-Morales − Departament de Química, search model. Refinement was performed with Phenix, and model Universitat Autònoma de Barcelona, Barcelona 08193, Spain building was performed with Coot.102 Refinement and data statistics Nathalia Varejão − Institut de Biotecnologia i Biomedicina are provided in Table S5. Figures were prepared with Pymol (The and Departament de Bioquímica i Biologia Molecular, PyMOL Molecular Graphics System, Version 2.0, Schrödinger, LLC). Universitat Autònoma de Barcelona, Barcelona 08193, Spain Statistical Analysis. All the graphs were generated with Filipa Bezerra − Molecular Neurobiology Group, GraphPad Prism software version 6.0 (GraphPad Software, La Jolla, California, USA). Data are shown as means ± standard error of mean i3S−Instituto de Investigação e Inovação em Saúde, (SEM). The results obtained from TTR stabilization studies in human IBMC−Instituto de Biologia Molecular e Celular, plasma were analyzed by one-way ANOVA Tukey’s test using Universidade do Porto, 4200-135 Porto, Portugal; GraphPad Prism 6.0. p < 0.05 was considered statistically significant Departamento de Biologia Molecular, ICBAS−Instituto de (* statistically significant at p < 0.05; ** statistically significant at p < Ciências Biomédicas Abel Salazar, Universidade do Porto, 0.01; *** statistically significant at p < 0.001). 4050-313 Porto, Portugal ■ * ASSOCIATED CONTENT sı Supporting Information David Ortega-Alarcon − Department of Biochemistry and Molecular & Cellular Biology, and Institute for Biocomputation eand Physics of Complex Systems (BIFI), The Supporting Information is available free of charge at Joint Unit GBsC-CSIC-BIFI, Universidad de Zaragoza, https://pubs.acs.org/doi/10.1021/acs.jmedchem.2c01195. 50018 Zaragoza, Spain; Aragon Institute for Health Main structures together with H-bond evolution along Research, 50009 Zaragoza, Spain; Biomedical Research the MD trajectories for all the systems; TTR tetramer Network Center in Hepatic and Digestive Diseases stabilization effect of tafamidis, tolcapone, and M-23 in (CIBERehd), 28029 Madrid, Spain human plasma as assessed by IEF; overlay of the WT- Danilo Gonzalez − Departament de Química, Universitat TTR binding pocket when bound to M-23 and Autònoma de Barcelona, Barcelona 08193, Spain tolcapone; tolcapone−TTR binding poses from MD Marcelo Osorio − Departament de Química, Universitat simulations in the absence of backbone constrains; 1H Autònoma de Barcelona, Barcelona 08193, Spain and 13C NMR spectra of all synthesized compounds; Susanna Navarro − Institut de Biotecnologia i Biomedicina and the HPLC traces for the final compounds (PDF) and Departament de Bioquímica i Biologia Molecular, Universitat Autònoma de Barcelona, Barcelona 08193, PDB coordinates for tolcapone (PDB) Spain; orcid.org/0000-0001-8160-9536 PDB coordinates for M-14 (PDB) Adrián Velázquez-Campoy − Department of Biochemistry PDB coordinates for M-17 (PDB) and Molecular & Cellular Biology, and Institute for PDB coordinates for M-20 (PDB) Biocomputation eand Physics of Complex Systems (BIFI), PDB coordinates for M-21 (PDB) Joint Unit GBsC-CSIC-BIFI, Universidad de Zaragoza, PDB coordinates for M-23 (PDB) 50018 Zaragoza, Spain; Aragon Institute for Health 14687 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Research, 50009 Zaragoza, Spain; Biomedical Research tional Society of Amyloidosis 2016 Nomencleature Guidelines. Network Center in Hepatic and Digestive Diseases Amyloid 2016, 23, 209−213. (CIBERehd), 28029 Madrid, Spain; orcid.org/0000- (3) Kanai, M.; Raz, A.; Goodman, D. S. Retinol binding-protein: the 0001-5702-4538 transport protein for vitamin A in human plasma. J. Clin. Invest. 1968, 47, 2025−2044. Maria Rosário Almeida − Molecular Neurobiology Group, (4) Schreiber, G.; Richardson, S. J. The Evolution of Gene i3S−Instituto de Investigação e Inovação em Saúde, Expression, Structure and Function of Transthyretin. Comp. Biochem. IBMC−Instituto de Biologia Molecular e Celular, Physiol. Part B Biochem. Mol. Biol. 1997, 116, 137−160. Universidade do Porto, 4200-135 Porto, Portugal; (5) Hagen, G. A.; Elliott, W. J. Transport of thyroid hormones in Departamento de Biologia Molecular, ICBAS−Instituto de serum and cerebrospinal fluid. J. Clin. Endocrinol. Metab. 1973, 37, Ciências Biomédicas Abel Salazar, Universidade do Porto, 415−422. 4050-313 Porto, Portugal (6) Benson, M. D. Transthyretin amyloidosis: a little history of David Reverter − Institut de Biotecnologia i Biomedicina and hereditary amyloidosis. Amyloid 2017, 24, 76−77. Departament de Bioquímica i Biologia Molecular, Universitat (7) Blake, C. C. F.; Geisow, M. J.; Swan, I. D. A.; Rerat, C.; Rerat, B. Autònoma de Barcelona, Barcelona 08193, Spain Structure of human plasma prealbumin at 2-5 A resolution. A Félix Busqué − Departament de Química, Universitat preliminary report on the polypeptide chain conformation, quaternary structure and thyroxine binding. J. Mol. Biol. 1974, 88, 1−12. Autònoma de Barcelona, Barcelona 08193, Spain (8) Blake, C. C. F.; Geisow, M. J.; Oatley, S. J.; Rérat, B.; Rérat, C. Ramon Alibés − Departament de Química, Universitat Structure of prealbumin: secondary, tertiary and quaternary Autònoma de Barcelona, Barcelona 08193, Spain; interactions determined by Fourier refinement at 1.8 A. J. Mol. Biol. orcid.org/0000-0002-7997-2691 1978, 121, 339−356. Mariona Sodupe − Departament de Química, Universitat (9) Foss, T. R.; Wiseman, R. L.; Kelly, J. W. The pathway by which Autònoma de Barcelona, Barcelona 08193, Spain; the tetrameric protein transthyretin dissociates. Biochemistry 2005, 44, orcid.org/0000-0003-0276-0524 15525−15533. (10) Lai, Z.; Colón, W.; Kelly, J. W. The acid-mediated denaturation Complete contact information is available at: pathway of transthyretin yields a conformational intermediate that can https://pubs.acs.org/10.1021/acs.jmedchem.2c01195 self-assemble into amyloid. Biochemistry 1996, 35, 6470−6482. (11) Hammarström, P.; Jiang, X.; Hurshman, A. R.; Powers, E. T.; Author Contributions Kelly, J. W. Sequence-dependent denaturation energetics: A major The manuscript was written through contributions of all determinant in amyloid disease diversity. Proc. Natl. Acad. Sci. U. S. A. authors. All authors have given approval to the final version of 2002, 99, 16427−16432. the manuscript. (12) Hurshman Babbes, A. R.; Powers, E. T.; Kelly, J. W. Notes Quantification of the thermodynamically linked quaternary and tertiary structural stabilities of transthyretin and its disease-associated The authors declare no competing financial interest. variants: the relationship between stability and amyloidosis. ■ ACKNOWLEDGMENTS This work was funded by the Spanish Ministry of Economy Biochemistry 2008, 47, 6969−6984. (13) Finsterer, J.; Iglseder, S.; Wanschitz, J.; Topakian, R.; Löscher, W. N.; Grisold, W. Hereditary transthyretin-related amyloidosis. Acta and Competitiveness (BIO2016-78310-R to S.V. and Neurol. Scand. 2019, 139, 92−105. BFU2016-78232-P to A.V.C.), the Spanish Ministry of Science (14) Jacobson, D. R.; Pastore, R. D.; Yaghoubian, R.; Kane, I.; Gallo, G.; Buck, F. S.; Buxbaum, J. N. Variant-sequence transthyretin and Innovation (PID2019-105017RB-I00/PDC2021-120914- (isoleucine 122) in late-onset cardiac amyloidosis in black Americans. I00) to S.V., and by ICREA, ICREA-Academia 2015 and 2020 N. Engl. J. Med. 1997, 336, 466−473. to S.V. R.A. acknowledges the financial support from (15) Andrade, C. A peculiar form of peripheral neuropathy; familiar MINECO/FEDER (projects CTQ2016-75363-R/PID2019- atypical generalized amyloidosis with special involvement of the 106403RB-IOO). F.P. and A.S. acknowledge the Universitat peripheral nerves. Brain 1952, 75, 408−427. Autònoma de Barcelona for their doctoral grant. This work was (16) Saraiva, M. J.; Birken, S.; Costa, P. P.; Goodman, D. S. Amyloid funded by FEDER through COMPETE 2020 and N2020 fibril protein in familial amyloidotic polyneuropathy, Portuguese type. through PT2020 and HEALTH-UNORTE: NORTE-01-0145- Definition of molecular abnormality in transthyretin (prealbumin). J. FEDER-000039 to M.R.A. F.B. was supported by FCT� Clin. Invest. 1984, 74, 104−119. Fundação para a Ciência e Tecnologia/MEC�Ministério da (17) Goren, H.; Steinberg, M. C.; Farboody, G. H. Familial Educaçõo e Ciência with a PhD fellowship (SFRH/BD/ oculoleptomeningeal amyloidosis. Brain 1980, 103, 473−495. (18) Sekijima, Y.; Hammarstr̈om, P.; Matsumura, M.; Shimizu, Y.; 123674/2016). Iwata, M.; Tokuda, T.; Ikeda, S.; Kelly, J. W. Energetic characteristics ■ ABBREVIATIONS ATTR, transthyretin amyloidosis; TTR, transthyretin; T4, of the new transthyretin variant A25T may explain its atypical central nervous system pathology. Lab. Invest. 2003, 83, 409−417. (19) Christmanson, L.; Betsholtz, C.; Gustavsson, A.; Johansson, B.; thyroxine; FAP, familial amyloid polyneuropathy; FAC, Sletten, K.; Westermark, P. The transthyretin cDNA sequence is familial amyloid cardiomyopathy; SSA, senile systemic normal in transthyretin-derived senile systemic amyloidosis. FEBS amyloidosis; HBP, halogen binding pocket; ITC, isothermal Lett. 1991, 281, 177−180. (20) Buxbaum, J. N.; Reixach, N. Transthyretin: the servant of many titration calorimetry; IEF, isoelectric focusing masters. Cell. Mol. Life Sci. 2009, 66, 3095−3101. ■ REFERENCES (1) Eisenberg, D.; Jucker, M. The amyloid state of proteins in (21) Coles, L. S.; Young, R. D. Supercentenarians and transthyretin amyloidosis: The next frontier of human life extension. Prev. Med. 2012, 54, S9−S11. human diseases. Cell 2012, 148, 1188−1203. (22) Mangione, P. P.; Porcari, R.; Gillmore, J. D.; Pucci, P.; Monti, (2) Sipe, J. D.; Benson, M. D.; Buxbaum, J. N.; Ikeda, S.-i.; Merlini, M.; Porcari, M.; Giorgetti, S.; Marchese, L.; Raimondi, S.; Serpell, L. G.; Saraiva, M. J. M.; Westermark, P. Amyloid fibril proteins and C.; Chen, W.; Relini, A.; Marcoux, J.; Clatworthy, I. R.; Taylor, G. W.; amyloidosis: chemical identification and clinical classification Interna- Tennent, G. A.; Robinson, C. V.; Hawkins, P. N.; Stoppini, M.; 14688 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Wood, S. P.; Pepys, M. B.; Bellotti, V. Proteolytic cleavage of Coelho, T.; Kelly, J. W. Predictive model of response to tafamidis in Ser52Pro variant transthyretin triggers its amyloid fibrillogenesis. Proc. hereditary ATTR polyneuropathy. JCI Insight 2019, 4, Natl. Acad. Sci. U. S. A. 2014, 111, 1539−1544. DOI: 10.1172/jci.insight.126526. (23) Ihse, E.; Rapezzi, C.; Merlini, G.; Benson, M. D.; Ando, Y.; (37) Adamski-Werner, S. L.; Palaninathan, S. K.; Sacchettini, J. C.; Suhr, O. B.; Ikeda, S.; Lavatelli, F.; Obici, L.; Quarta, C. C.; Leone, Kelly, J. W. Diflunisal analogues stabilize the native state of O.; Jono, H.; Ueda, M.; Lorenzini, M.; Liepnieks, J.; Ohshima, T.; transthyretin. Potent inhibition of amyloidogenesis. J. Med. Chem. Tasaki, M.; Yamashita, T.; Westermark, P. Amyloid fibrils containing 2004, 47, 355−374. fragmented ATTR may be the standard fibril composition in ATTR (38) Penchala, S. C.; Connelly, S.; Wang, Y.; Park, M. S.; Zhao, L.; amyloidosis. Amyloid 2013, 20, 142−150. Baranczak, A.; Rappley, I.; Vogel, H.; Liedtke, M.; Witteles, R. M.; (24) Marcoux, J.; Mangione, P. P.; Porcari, R.; Degiacomi, M. T.; Powers, E. T.; Reixach, N.; Chan, W. K.; Wilson, I. A.; Kelly, J. W.; Verona, G.; Taylor, G. W.; Giorgetti, S.; Raimondi, S.; Sanglier- Graef, I. A.; Alhamadsheh, M. M. AG10 inhibits amyloidogenesis and Cianférani, S.; Benesch, J. L.; Cecconi, C.; Naqvi, M. M.; Gillmore, J. cellular toxicity of the familial amyloid cardiomyopathy-associated D.; Hawkins, P. N.; Stoppini, M.; Robinson, C. V.; Pepys, M. B.; V122I transthyretin. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 9992− Bellotti, V. A novel mechano-enzymatic cleavage mechanism underlies 9997. transthyretin amyloidogenesis. EMBO Mol. Med. 2015, 7, 1337−1349. (39) Sant’Anna, R.; Gallego, P.; Robinson, L. Z.; Pereira-Henriques, (25) Slamova, I.; Adib, R.; Ellmerich, S.; Golos, M. R.; Gilbertson, J. A.; Ferreira, N.; Pinheiro, F.; Esperante, S.; Pallares, I.; Huertas, O.; A.; Botcher, N.; Canetti, D.; Taylor, G. W.; Rendell, N.; Tennent, G. Almeida, M. R.; Reixach, N.; Insa, R.; Velazquez-Campoy, A.; A.; Verona, G.; Porcari, R.; Mangione, P. P.; Gillmore, J. D.; Pepys, Reverter, D.; Reig, N.; Ventura, S. Repositioning tolcapone as a M. B.; Bellotti, V.; Hawkins, P. N.; Al-Shawi, R.; Simons, J. P. Plasmin potent inhibitor of transthyretin amyloidogenesis and associated activity promotes amyloid deposition in a transgenic model of human cellular toxicity. Nat. Commun. 2016, 7, 10787. transthyretin amyloidosis. Nat. Commun. 2021, 12, 7112. (40) Games, J.; Salvadó, M.; Reig, N.; Suñé, P.; Casanovas, C.; (26) Okumura, K.; Yamashita, T.; Masuda, T.; Misumi, Y.; Ueda, A.; Rojas-Garcia, R.; Insa, R. Transthyretin stabilization activity of the Ueda, M.; Obayashi, K.; Jono, H.; Yamashita, S.; Inomata, Y.; Ando, catechol-O-methyltransferase inhibitor tolcapone (SOM0226) in Y. Long-term outcome of patients with hereditary transthyretin V30M hereditary ATTR amyloidosis patients and asymptomatic carriers: amyloidosis with polyneuropathy after liver transplantation. Amyloid proof-of-concept study. Amyloid 2019, 26, 74−84. 2016, 23, 39−45. (41) Reig, N.; Ventura, S.; Salvadó, M.; Gámez, J.; Insa, R. (27) Banerjee, D.; Roeker, L. E.; Grogan, M.; Swiecicki, P.; SOM0226, a repositioned compound for the treatment of TTR Poterucha, J.; Heimbach, J.; Zeldenrust, S.; Gertz, M.; Edwards, B.; amyloidosis. Orphanet J. Rare Dis. 2015, 10, P9. Daly, R.; Klarich, K. W.; Dispenzieri, A. Outcomes of Patients With (42) Verona, G.; Mangione, P. P.; Raimondi, S.; Giorgetti, S.; Familial Transthyretin Amyloidosis After Liver Transplantation. Prog. Faravelli, G.; Porcari, R.; Corazza, A.; Gillmore, J. D.; Hawkins, P. N.; Transplant. 2017, 27, 246−250. Pepys, M. B.; Taylor, G. W.; Bellotti, V. Inhibition of the mechano- (28) Johnson, S. M.; Connelly, S.; Fearns, C.; Powers, E. T.; Kelly, J. enzymatic amyloidogenesis of transthyretin: role of ligand affinity, W. The Transthyretin Amyloidoses: From Delineating the Molecular binding cooperativity and occupancy of the inner channel. Sci. Rep. Mechanism of Aggregation Linked to Pathology to a Regulatory- 2017, 7, 182. Agency-Approved Drug. J. Mol. Biol. 2012, 421, 185−203. (43) Russ, H.; Müller, T.; Woitalla, D.; Rahbar, A.; Hahn, J.; Kuhn, (29) Ackermann, E. J.; Guo, S.; Benson, M. D.; Booten, S.; Freier, S.; W. Detection of tolcapone in the cerebrospinal fluid of parkinsonian Hughes, S. G.; Kim, T. W.; Jesse Kwoh, T.; Matson, J.; Norris, D.; Yu, subjects. Naunyn-Schmiedeberg’s Arch. Pharmacol. 1999, 360, 719− R.; Watt, A.; Monia, B. P. Suppressing transthyretin production in 720. mice, monkeys and humans using 2nd-Generation antisense (44) Pinheiro, F.; Varejão, N.; Esperante, S.; Santos, J.; Velázquez- oligonucleotides. Amyloid 2016, 23, 148−157. Campoy, A.; Reverter, D.; Pallarès, I.; Ventura, S. Tolcapone, a potent (30) Kristen, A. V.; Ajroud-Driss, S.; Conceiçaõ , I.; Gorevic, P.; aggregation inhibitor for the treatment of familial leptomeningeal Kyriakides, T.; Obici, L. Patisiran, an RNAi therapeutic for the amyloidosis. FEBS J. 2021, 288, 310−324. treatment of hereditary transthyretin-mediated amyloidosis. Neuro- (45) Goodsell, D. S.; Zardecki, C.; Di Costanzo, L.; Duarte, J. M.; degener. Dis. Manag. 2019, 9, 5−23. Hudson, B. P.; Persikova, I.; Segura, J.; Shao, C.; Voigt, M.; (31) Hammarstrom, P.; Schneider, F.; Kelly, J. W. Trans-suppression Westbrook, J. D.; Young, J. Y.; Burley, S. K. RCSB Protein Data Bank: of misfolding in an amyloid disease. Science 2001, 293, 2459−2462. Enabling biomedical research and drug discovery. Protein Sci. 2020, (32) Coelho, T.; Chorão, R.; Sousa, A.; Alves, I.; Torres, M. F.; 29, 52−65. Saraiva, M. J. M. Compound heterozygotes of transthyretin Met30 (46) Palaninathan, S. K. Nearly 200 X-ray crystal structures of and transthyretin Met119 are protected from the devastating effects of transthyretin: what do they tell us about this protein and the design of familial amyloid polyneuropathy. Neuromuscul. Disord. 1996, 6, S20. drugs for TTR amyloidoses? Curr. Med. Chem. 2012, 19, 2324−2342. (33) Bulawa, C. E.; Connelly, S.; Devit, M.; Wang, L.; Weigel, C.; (47) Hörnberg, A.; Eneqvist, T.; Olofsson, A.; Lundgren, E.; Sauer- Fleming, J. A.; Packman, J.; Powers, E. T.; Wiseman, R. L.; Foss, T. Eriksson, A. E. A comparative analysis of 23 structures of the R.; Wilson, I. A.; Kelly, J. W.; Labaudiniere, R. Tafamidis, a potent amyloidogenic protein transthyretin. J. Mol. Biol. 2000, 302, 649−669. and selective transthyretin kinetic stabilizer that inhibits the amyloid (48) Shaw, D. E.; Maragakis, P.; Lindorff-Larsen, K.; Piana, S.; Dror, cascade. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 9629−9634. R. O.; Eastwood, M. P.; Bank, J. A.; Jumper, J. M.; Salmon, J. K.; (34) Coelho, T.; Maia, L. F.; da Silva, A. M.; Cruz, M. W.; Planté- Shan, Y.; Wriggers, W. Atomic-level characterization of the structural Bordeneuve, V.; Suhr, O. B.; Conceiçao, I.; Schmidt, H. H. J.; Trigo, dynamics of proteins. Science 2010, 330, 341−346. P.; Kelly, J. W.; Labaudinière, R.; Chan, J.; Packman, J.; Grogan, D. R. (49) Liu, X.; Shi, D.; Zhou, S.; Liu, H.; Liu, H.; Yao, X. Molecular Long-term effects of tafamidis for the treatment of transthyretin dynamics simulations and novel drug discovery. Expert Opin. Drug familial amyloid polyneuropathy. J. Neurol. 2013, 260, 2802−2814. Discov. 2018, 13, 23−37. (35) Maurer, M. S.; Schwartz, J. H.; Gundapaneni, B.; Elliott, P. M.; (50) Ferreira, L. G.; Dos Santos, R. N.; Oliva, G.; Andricopulo, A. D. Merlini, G.; Waddington-Cruz, M.; Kristen, A. V.; Grogan, M.; Molecular docking and structure-based drug design strategies. Witteles, R.; Damy, T.; Drachman, B. M.; Shah, S. J.; Hanna, M.; Molecules 2015, 20, 13384−13421. Judge, D. P.; Barsdorf, A. I.; Huber, P.; Patterson, T. A.; Riley, S.; (51) Yang, M.; Lei, M.; Huo, S. Why is Leu55–>Pro55 transthyretin Schumacher, J.; Stewart, M.; Sultan, M. B.; Rapezzi, C. Tafamidis variant the most amyloidogenic: insights from molecular dynamics Treatment for Patients with Transthyretin Amyloid Cardiomyopathy. simulations of transthyretin monomers. Protein Sci. 2003, 12, 1222− N. Engl. J. Med. 2018, 379, 1007−1016. 1231. (36) Monteiro, C.; Mesgazardeh, J. S.; Anselmo, J.; Fernandes, J.; (52) Das, J. K.; Mall, S. S.; Bej, A.; Mukherjee, S. Conformational Novais, M.; Rodrigues, C.; Brighty, G. J.; Powers, D. L.; Powers, E. T.; flexibility tunes the propensity of transthyretin to form fibrils through 14689 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article non-native intermediate states. Angew. Chem. Int. Ed. Engl. 2014, 53, M.; Chan, W.; Graef, I.; Zamboni, R.; Kumar, N.; Fox, J.; Sinha, U.; 12781−12784. Alhamadsheh, M. Enthalpy-Driven Stabilization of Transthyretin by (53) Morris, K. F.; Geoghegan, R. M.; Palmer, E. E.; George, M., Jr.; AG10 Mimics a Naturally Occurring Genetic Variant That Protects Fang, Y. Molecular dynamics simulation study of AG10 and tafamidis from Transthyretin Amyloidosis. J. Med. Chem. 2018, 61, 7862−7876. binding to the Val122Ile transthyretin variant. Biochem. Biophys. Rep. (69) Ekwall, B. Toxicity to HeLa cell of 205 drugs as determined by 2020, 21, 100721. the metabolic inhibition test supplemented by microscopy. Toxicology (54) Yee, A. W.; Aldeghi, M.; Blakeley, M. P.; Ostermann, A.; Mas, 1980, 17, 273−295. P. J.; Moulin, M.; de Sanctis, D.; Bowler, M. W.; Mueller-Dieckmann, (70) Thabrew, M. I.; Hughes, R. D.; McFarlane, I. G. Screening of C.; Mitchell, E. P.; Haertlein, M.; de Groot, B. L.; Boeri Erba, E.; hepatoprotective plant components using a HepG2 cell cytotoxicity Forsyth, V. T. A molecular mechanism for transthyretin amyloido- assay. J. Pharm. Pharmacol. 1997, 49, 1132−1135. genesis. Nat. Commun. 2019, 10, 925. (71) Schoonen, W. G. E. J.; Westerink, W. M. A.; de Roos, J. A. D. (55) Zhou, S.; Ge, S.; Zhang, W.; Zhang, Q.; Yuan, S.; Lo, G. V.; M.; Debiton, E. Cytotoxic effects of 100 reference compounds on Dou, Y. Conventional Molecular Dynamics and Metadynamics Hep G2 and HeLa cells and of 60 compounds on ECC-1 and CHO Simulation Studies of the Binding and Unbinding Mechanism of cells. I mechanistic assays on ROS, glutathione depletion and calcein TTR Stabilizers AG10 and Tafamidis. ACS Chem. Neurosci. 2020, 11, uptake. Toxicol. In Vitro 2005, 19, 505−516. 3025−3035. (72) O’Brien, P. J.; Irwin, W.; Diaz, D.; Howard-Cofield, E.; Krejsa, (56) Berk, J. L.; Suhr, O. B.; Obici, L.; Sekijima, Y.; Zeldenrust, S. R.; C. M.; Slaughter, M. R.; Gao, B.; Kaludercic, N.; Angeline, A.; Yamashita, T.; Heneghan, M. A.; Gorevic, P. D.; Litchy, W. J.; Bernardi, P.; Brain, P.; Hougham, C. High concordance of drug- Wiesman, J. F.; Nordh, E.; Corato, M.; Lozza, A.; Cortese, A.; induced human hepatotoxicity with in vitro cytotoxicity measured in a Robinson-Papp, J.; Colton, T.; Rybin, D. V.; Bisbee, A. B.; Ando, Y.; novel cell-based model using high content screening. Arch. Toxicol. Ikeda, S.; Seldin, D. C.; Merlini, G.; Skinner, M.; Kelly, J. W.; Dyck, P. 2006, 80, 580−604. J.; Diflunisal Trial Consortium. Repurposing diflunisal for familial (73) Gonzalez, L. T.; Minsky, N. W.; Espinosa, L. E.; Aranda, R. S.; amyloid polyneuropathy: a randomized clinical trial. JAMA 2013, 310, Meseguer, J. P.; Perez, P. C. In vitro assessment of hepatoprotective 2658−2667. agents against damage induced by acetaminophen and CCl4. BMC (57) Cotrina, E. Y.; Pinto, M.; Bosch, L.; Vilà, M.; Blasi, D.; Complement. Altern. Med. 2017, 17, 39. Quintana, J.; Centeno, N. B.; Arsequell, G.; Planas, A.; Valencia, G. (74) Olanow, C. W. Tolcapoene and hepatotoxic effects. Tasmar Modulation of the fibrillogenesis inhibition properties of two Advisory Panel. Arch. Neurol. 2000, 57, 263−267. transthyretin ligands by halogenation. J. Med. Chem. 2013, 56, (75) Smith, K. S.; Smith, P. L.; Heady, T. N.; Trugman, J. M.; 9110−9121. Harman, W. D.; Macdonald, T. L. In vitro metabolism of tolcapone to (58) Genheden, S.; Ryde, U. The MM/PBSA and MM/GBSA reactive intermediates: relevance to tolcapone liver toxicity. Chem. Res. methods to estimate ligand-binding affinities. Expert Opin. Drug Toxicol. 2003, 16, 123−128. Discov. 2015, 10, 449−461. (76) Kim, J. H.; Oroz, J.; Zweckstetter, M. Structure of Monomeric (59) Hammarstrom, P.; Wiseman, R. L.; Powers, E. T.; Kelly, J. W. Transthyretin Carrying the Clinically Important T119M Mutation. Prevention of transthyretin amyloid disease by changing protein Angew. Chem. Int. Ed. Engl. 2016, 55, 16168−16171. misfolding energetics. Science 2003, 299, 713−716. (77) Sebastião, M. P.; Lamzin, V.; Saraiva, M. J.; Damas, A. M. (60) Purkey, H. E.; Palaninathan, S. K.; Kent, K. C.; Smith, C.; Safe, Transthyretin stability as a key factor in amyloidogenesis: X-ray S. H.; Sacchettini, J. C.; Kelly, J. W. Hydroxylated polychlorinated analysis at atomic resolution. J. Mol. Biol. 2001, 306, 733−744. biphenyls selectively bind transthyretin in blood and inhibit (78) Judge, D. P.; Heitner, S. B.; Falk, R. H.; Maurer, M. S.; Shah, S. amyloidogenesis: rationalizing rodent PCB toxicity. Chem. Biol. 2004, 11, 1719−1728. J.; Witteles, R. M.; Grogan, M.; Selby, V. N.; Jacoby, D.; Hanna, M.; (61) McCammon, M. G.; Scott, D. J.; Keetch, C. A.; Greene, L. H.; Nativi-Nicolau, J.; Patel, J.; Rao, S.; Sinha, U.; Turtle, C. W.; Fox, J. C. Purkey, H. E.; Petrassi, H. M.; Kelly, J. W.; Robinson, C. V. Screening Transthyretin Stabilization by AG10 in Symptomatic Transthyretin transthyretin amyloid fibril inhibitors: characterization of novel Amyloid Cardiomyopathy. J. Am. Coll. Cardiol. 2019, 74, 285−295. multiprotein, multiligand complexes by mass spectrometry. Structure (79) Case, D. A.; Betz, R. M.; Cerutti, D. S.; Cheatham, T. E. I.; 2002, 10, 851−863. Darden, T. A.; Duke, R. E.; Giese, T. J.; Gohlke, H.; Goetz, A. W.; (62) Green, N. S.; Palaninathan, S. K.; Sacchettini, J. C.; Kelly, J. W. Homeyer, N. AMBER 2016. University of California: San Francisco, Synthesis and characterization of potent bivalent amyloidosis 2016. inhibitors that bind prior to transthyretin tetramerization. J. Am. (80) Maier, J. A.; Martinez, C.; Kasavajhala, K.; Wickstrom, L.; Chem. Soc. 2003, 125, 13404−13414. Hauser, K. E.; Simmerling, C. ff14SB: Improving the Accuracy of (63) Ferguson, R. N.; Edelhoch, H.; Saroff, H. A.; Robbins, J.; Protein Side Chain and Backbone Parameters from ff99SB. J. Chem. Cahnmann, H. J. Negative cooperativity in the binding of thyroxine to Theory Comput. 2015, 11, 3696−3713. human serum prealbumin. Preparation of tritium-labeled 8-anilino-1- (81) Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. naphthalenesulfonic acid. Biochemistry 1975, 14, 282−289. A. Development and testing of a general amber force field. J. Comput. (64) Jiang, X.; Buxbaum, J. N.; Kelly, J. W. The V122I Chem. 2004, 25, 1157−1174. cardiomyopathy variant of transthyretin increases the velocity of (82) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. Universal rate-limiting tetramer dissociation, resulting in accelerated amyloi- solvation model based on solute electron density and on a continuum dosis. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 14943−14948. model of the solvent defined by the bulk dielectric constant and (65) Hurshman, A. R.; White, J. T.; Powers, E. T.; Kelly, J. W. atomic surface tensions. J. Phys. Chem. B 2009, 113, 6378−6396. Transthyretin aggregation under partially denaturing conditions is a (83) Frisch, M. J.; Trucks, G. W.; Petersson, G. A.; Nakatsuji, H. downhill polymerization. Biochemistry 2004, 43, 7365−7381. Gaussian 09, Revision D.01, Gaussian, Inc.: Wallingford CT, 2016. (66) Dolado, I.; Nieto, J.; Saraiva, M. J.; Arsequell, G.; Valencia, G.; (84) Humphrey, W.; Dalke, A.; Schulten, K. VMD: visual molecular Planas, A. Kinetic assay for high-throughput screening of in vitro dynamics. J. Mol. Graph. 1996, 14, 33−38. transthyretin amyloid fibrillogenesis inhibitors. J. Comb. Chem. 2005, (85) McGibbon, R. T.; Beauchamp, K. A.; Harrigan, M. P.; Klein, 7, 246−252. C.; Swails, J. M.; Hernandez, C. X.; Schwantes, C. R.; Wang, L. P.; (67) Yokoyama, T.; Mizuguchi, M. Transthyretin Amyloidogenesis Lane, T. J.; Pande, V. S. MDTraj: A Modern Open Library for the Inhibitors: From Discovery to Current Developments. J. Med. Chem. Analysis of Molecular Dynamics Trajectories. Biophys. J. 2015, 109, 2020, 63, 14228−14242. 1528−1532. (68) Miller, M.; Pal, A.; Albusairi, W.; Joo, H.; Pappas, B.; Haque (86) Boymond, L.; Rottlander, M.; Cahiez, G.; Knochel, P. Tuhin, M. T.; Liang, D.; Jampala, R.; Liu, F.; Khan, J.; Faaij, M.; Park, Preparation of Highly Functionalized Grignard Reagents by an 14690 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Iodine-Magnesium Exchange Reaction and its Application in Solid- Phase Synthesis. Angew. Chem. Int. Ed. Engl. 1998, 37, 1701−1703. (87) Jorgensen, W. L.; Bollini, M.; Thakur, V. V.; Domaoal, R. A.; Spasov, K. A.; Anderson, K. S. Efficient discovery of potent anti-HIV agents targeting the Tyr181Cys variant of HIV reverse transcriptase. J. Am. Chem. Soc. 2011, 133, 15686−15696. (88) Andrew, R. G.; Raphael, R. A. A new and total synthesis of aaptamine. Tetrahedron 1987, 43, 4803−4816. (89) Bailey, K.; Tan, E. W. Synthesis and evaluation of bifunctional nitrocatechol inhibitors of pig liver catechol-O-methyltransferase. Bioorg. Med. Chem. 2005, 13, 5740−5749. (90) Saiz-Poseu, J.; Alcón, I.; Alibés, R.; Busqué, F.; Faraudo, J.; Ruiz-Molina, D. Self-assembly of alkylcatechols on HOPG inves- tigated by scanning tunneling microscopy and molecular dynamics simulations. CrystEngComm 2012, 14, 264−271. (91) Learmonth, D. A.; Alves, P. C. Improved method for demethylation of nitro-catechol methyl ethers. Synth. Commun. 2002, 32, 641−649. (92) Learmonth, D. A.; Vieira-Coelho, M. A.; Benes, J.; Alves, P. C.; Borges, N.; Freitas, A. P.; da Silva, P. S. Synthesis of 1-(3,4-dihydroxy- 5-nitrophenyl)-2-phenyl-ethanone and derivatives as potent and long- acting peripheral inhibitors of catechol-O-methyltransferase. J. Med. Chem. 2002, 45, 685−695. (93) Vega, S.; Abian, O.; Velazquez-Campoy, A. A unified framework based on the binding polynomial for characterizing biological systems by isothermal titration calorimetry. Methods 2015, 76, 99−115. (94) Freire, E.; Schön, A.; Velazquez-Campoy, A. Isothermal titration calorimetry: general formalism using binding polynomials. Methods Enzymol. 2009, 455, 127−155. (95) Ferreira, N.; Cardoso, I.; Domingues, M. R.; Vitorino, R.; Bastos, M.; Bai, G.; Saraiva, M. J.; Almeida, M. R. Binding of epigallocatechin-3-gallate to transthyretin modulates its amyloidoge- nicity. FEBS Lett. 2009, 583, 3569−3576. (96) Almeida, M. R.; Macedo, B.; Cardoso, I.; Alves, I.; Valencia, G.; Arsequell, G.; Planas, A.; Saraiva, M. J. Selective binding to transthyretin and tetramer stabilization in serum from patients with familial amyloidotic polyneuropathy by an iodinated diflunisal derivative. Biochem. J. 2004, 381, 351−356. (97) Saraiva, M. J.; Costa, P. P.; Goodman, D. S. Transthyretin (prealbumin) in familial amyloidotic polyneuropathy: genetic and functional aspects. Adv. Neurol. 1988, 48, 189−200. (98) Juanhuix, J.; Gil-Ortiz, F.; Cuni, G.; Colldelram, C.; Nicolas, J.; Lidon, J.; Boter, E.; Ruget, C.; Ferrer, S.; Benach, J. Developments in optics and performance at BL13-XALOC, the macromolecular crystallography beamline at the ALBA synchrotron. J. Synchrotron Radiat. 2014, 21, 679−689. (99) Kabsch, W. Integration, scaling, space-group assignment and post-refinement. Acta Crystallogr. D Biol. Crystallogr. 2010, 66, 133− 144. (100) Winn, M. D.; Ballard, C. C.; Cowtan, K. D.; Dodson, E. J.; Emsley, P.; Evans, P. R.; Keegan, R. M.; Krissinel, E. B.; Leslie, A. G.; McCoy, A.; McNicholas, S. J.; Murshudov, G. N.; Pannu, N. S.; Potterton, E. A.; Powell, H. R.; Read, R. J.; Vagin, A.; Wilson, K. S. Overview of the CCP4 suite and current developments. Acta Crystallogr. D Biol. Crystallogr. 2011, 67, 235−242. (101) Adams, P. D.; Afonine, P. V.; Bunkoczi, G.; Chen, V. B.; Davis, I. W.; Echols, N.; Headd, J. J.; Hung, L. W.; Kapral, G. J.; Grosse-Kunstleve, R. W.; McCoy, A. J.; Moriarty, N. W.; Oeffner, R.; Read, R. J.; Richardson, D. C.; Richardson, J. S.; Terwilliger, T. C.; Zwart, P. H. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr. D Biol. Crystallogr. 2010, 66, 213−221. (102) Emsley, P.; Lohkamp, B.; Scott, W. G.; Cowtan, K. Features and development of Coot. Acta Crystallogr. D Biol. Crystallogr. 2010, 66, 486−501. 14691 https://doi.org/10.1021/acs.jmedchem.2c01195 J. Med. Chem. 2022, 65, 14673−14691