Previous Article in Journal
Nano-Biofungicides and Bio-Nanofungicides: State of the Art of Innovative Tools for Controlling Resistant Phytopathogens
Previous Article in Special Issue
In Silico Design of Novel Piperazine-Based mTORC1 Inhibitors Through DFT, QSAR and ADME Investigations
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Biophysica
Volume 5
Issue 2
10.3390/biophysica5020016
biophysica-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Molecular Dynamics Simulation Study of Stabilizer Association with the Val122Ile Transthyretin Variant

by
Kevin Morris
1,
John DeSalvo
1,
Iman Deanparvar
1,
Lucus Schneider
1,
Kaleigh Leach
1,
Matthew George, Jr.
2 and
Yayin Fang
2,*
1
Department of Chemistry, Carthage College, 2001 Alford Park Drive, Kenosha, WI 53140, USA
2
Department of Biochemistry and Molecular Biology, Howard University College of Medicine, Howard University, 520 W Street NW, Washington, DC 20059, USA
*
Author to whom correspondence should be addressed.
Biophysica 2025, 5(2), 16; https://doi.org/10.3390/biophysica5020016
Submission received: 4 March 2025 / Revised: 12 April 2025 / Accepted: 18 April 2025 / Published: 23 April 2025
(This article belongs to the Special Issue Molecular Structure and Simulation in Biological System 3.0)
Browse Figures
Versions Notes

Abstract

:
The tetrameric protein transthyretin (TTR) transports the hormone thyroxine in plasma and cerebrospinal fluid. Certain point mutations of TTR, including the Val122Ile mutation investigated here, destabilize the tetramer leading to its dissociation, misfolding, aggregation, and the eventual buildup of amyloid fibrils in the myocardium. Cioffi et al. reported the design and synthesis of a novel TTR kinetic stabilizing ligand, referred to here as TKS14, that inhibited TTR dissociation and amyloid fibril formation. In this study, molecular dynamics simulations were used to investigate the binding of TKS14 and eight TSK14 derivatives to the Val122Ile TTR mutant. For each complex, the ligand’s solvent accessible surface area (SASA), ligand–receptor hydrogen-bonding interactions, and the free energy of ligand-binding to TTR were investigated. The goal of this study was to identify the TSK14 functional groups that contributed to TTR stabilization. TKS14 was found to form a stable, two-point interaction with TTR by hydrogen bonding to Ser-117 residues in the inner receptor binding pocket and interacting through hydrogen bonds and electrostatically with Lys-15 residues near the receptor’s surface. The free energy of TKS14-TTR binding was −18.0 kcal mol−1 and the ligand’s average SASA value decreased by over 80% upon binding to the receptor. The thermodynamic favorability of TTR binding decreased when TKS14 derivatives contained either methyl ester, amide, tetrazole, or N-methyl functional groups that disrupted the above two-point interaction. One derivative in which a tetrazole ring was added to TKS14 was found to form hydrogen bonds with Thr-106, Thr-119, Ser-117, and Lys-15 residues. This derivative had a free energy of TTR binding of −21.4 kcal mol−1. Overall, the molecular dynamics simulations showed that the functional groups within the TKS14 structural template can be tuned to optimize the thermodynamic favorability of ligand binding.

1. Introduction

The protein transthyretin (TTR) is produced in the liver, the choroid plexus of brain, and the retina of the eye. TTR transports the thyroid hormone thyroxine and the retinol-binding protein bound to retinol in both plasma and cerebrospinal fluid [1]. TTR also plays roles in cognition, neurogenesis, nerve regeneration, and axonal growth [2]. Wild-type TTR monomers are composed of 127 residues arranged into eight β-sheet strands and an alpha-helix. The monomers are further connected by non-covalent interactions to form the tetramer shown in Figure 1a. Thyroxine binds to TTR in one of two identical halogen binding pockets (HBP) denoted in Figure 1a as AB and A’B’ [3]. Figure 1b shows a TTR-stabilizing ligand in pocket AB. The residues comprising the pocket on monomer A are highlighted as well.
The TTR protein is associated with rare amyloidosis disorders where amyloid fibrils build up in the myocardium and in the nervous system [4]. These amyloid fibrils form when the TTR tetramer first dissociates into monomers. The monomers then undergo a conformational change causing them to eventually aggregate into the amyloid fibrils. The dissociation of the TTR tetramer to form monomers is the rate-limiting step in the process, therefore many drug therapies target kinetic stabilization of the intact tetramer [5].
Amyloidosis diseases are associated with both wild-type and point-mutated TTR. In familial amyloid cardiomyopathy (FAC), point mutations such as Leu55Pro and Val122Ile destabilize the tetramer, making it more susceptible to dissociation and aggregation [6]. Over one hundred single-point mutations of TTR have been identified, some of which are amyloidogenic, while others are not. For example, the mutation Leu55Pro increases the tetramer dissociation rate nine-fold, while the mutation Thr119Met provides kinetic stabilization [7]. The pathogenic TTR mutant investigated here is Val122Ile or a substitution of residue Val-122 with an Isoleucine. Buxbaum et al. reported that the Val122Ile mutation is present in 3.4% of African Americans. Amyloidosis caused by the Val122Ile mutation is likely underdiagnosed, with patients often at an advanced disease stage when diagnosis occurs [8,9]. Compared to wt TTR, the Val122Ile mutation shifts the tetramer-monomer equilibrium toward the monomer and causes a three-fold increase in the rate of tetramer to folded monomer dissociation [10].
FAC treatments include liver transplantation to remove the source of the mutant protein and gene editing using CRISPR-Cas9 technology [11,12]. Another treatment approach uses TTR-targeted drug therapies to stabilize the tetramer and reduce the rate of tetramer dissociation. Current drug therapies include tafamidis, which has been approved for treatment of FAC by the FDA and European Medicines Agency; the Parkinson’s drug tolcapone; and the anti-inflammatory drug diflunisal [1,13]. In addition, natural flavonoids, polyphenols, crown ethers, natural products like γ-Mangostin, and the bivalent compound Mds84 have all been identified as TTR amyloidosis inhibitors [7,14,15].
The drug tafamidis was first identified by Bulawa et al. in 2012 using structure-based design methods [16]. Tafamidis’ structure is shown in Figure 2a. Kinetic stabilization by tafamidis has been reported with both wt TTR and mutants Val30Met, Val122Ile, and Leu111Met. X-ray crystal structures show that tafamidis places its chlorine-containing ring in the inner part of the HBP, referred to as P3, and containing residues Ala-108, Leu-110, Ser-117, and Thr-119 [16]. The tafamidis benzoazole ring is located near the center of the HBP, often denoted P2 and containing residues Lys-15, Leu-17, Ala-109, and Leu-110 [7]. A subsequent molecular dynamics (MD) simulation study in 2020 of the mechanism of TTR amyloidosis inhibition by tafamidis reported that the ligand remained in this orientation within the HBP throughout a 100 ns MD simulation and that the ligand experienced electrostatic and hydrogen-bonding interactions between its carboxylate function group and residue Lys-15. Hydrogen bonds also formed between the tafamidis nitrogen atom and residue Thr-118 [17].
The molecular structure of the drug AG10 or acoramidis is shown in Figure 2b. AG10 was identified in a high throughput study of TTR amyloidosis inhibitors. It has also been reported that AG10 outperformed tafamidis in tetramer stabilization studies with both wt and Val122Ile mutant TTR [18]. Recently, results from Phase III FDA trials showed that patients administered AG10 performed significantly better than a placebo group with respect to morbidity and the other factors investigated [19]. X-ray structures showed that when TTR-bound, AG10’s pyrazole ring was inserted into the inner P3 section of the HBP. Hydrogen bonds also formed between AG10 nitrogen atoms and TTR Ser-117 residues [18]. The stabilization provided by these H-bonds was found to mimic tetramer stabilization provided by the mutation Thr119Met [20]. Electrostatic interactions between the AG10 carboxylate and Lys-15 residues were present in the X-ray structure as well [18]. Subsequent MD simulation studies of the AG10-Val122Ile complex showed that the AG10 ligand maintained its stable two-point interaction with residues Ser-117 and Lys-15 throughout the 100 ns MD simulation. The AG10 atoms’ solvent accessibility also decreased considerably upon TTR binding [17].
Zhou et al. used MD and metadynamics simulations to compare the mechanisms of AG10’s and tafamidis’ association and dissociation with wt TTR [21]. Binding free energy studies showed the energy barrier for tafamidis-TTR dissociation was lower than the corresponding AG10 barrier. Hydrogen-bonding interactions were the main factor governing AG10-TTR dissociation, while the tafamidis dissociation mechanism was more complex and featured multiple steps. Overall, the authors concluded that the structural features shared by both molecules, namely an aromatic ring containing H-bond donors/acceptors and an anionic carboxylate at the opposite end of the molecule, both played important roles in governing TTR binding and dissociation [21].
In 2021, Cioffi et al. reported a structure-based drug design study of TTR tetramer kinetic stabilizers [22]. By utilizing a TTR computational docking model derived from X-ray crystallographic data and a Structure–Activity Relationship (SAR) study, Cioffi et al. designed and synthesized the compound shown in Figure 2c, referred to here as TTR kinetic stabilizer 14 or TKS14 [22]. The authors’ hypothesis was that replacing the methylene chain of AG10 with a more rigid piperazine ring would reduce the conformational flexibility of the TTR-bound ligand and lead to further stabilization of the TTR tetramer. Cioffi et al. found that the TKS14 stabilizer exhibited excellent potency at TTR (TTR fluorescence polarization (FP) IC50 = 220 nM), which is approximately two-fold better than tafamidis (TTR FP IC50 = 410 nM) and comparable to AG10 (TTR IC50 = 160 nM). The study also showed that TSK14 prevented TTR aggregation in a gel-based assay and possessed desirable pharmacokinetics in mice. Furthermore, molecular modeling experiments showed that like AG10, TKS14 experienced a two-point interaction with the TTR receptor by forming H-bonds with Ser-117 residues in the inner HBP and with Lys-15 residues in the outer part of the pocket [22]. These H-bonds are also present in the Figure 1b structure, which shows the TKS14 ligand bound to the TTR HBP.
Here, MD simulations were used to investigate the binding of the TSK14 molecule and eight of its derivatives to the Val122Ile TTR mutant. These derivatives are designated D1–D8 and their structures are shown in Figure 2d–k. Previous MD simulation studies of TTR and TTR inhibitor complexes have compared the stabilities of the wt tetramer to pathogenic variants [23]. The mechanism of tetramer dissociation and conformational changes preceding the formation of amyloid fibrils have also been investigated with MD simulation experiments [24,25]. Finally, MD simulations and other in silico approaches have also been used to identify novel TTR amyloidosis inhibitors [26,27].
The goal of the project was to study how changes in the TKS14 functional groups affected the ligand’s binding to the Val122Ile variant. Solvent accessible surface area calculations and ligand superpositions at different simulations times were used to verify that the MD simulation analyses yielded physically reasonable results. Hydrogen bond formation between the TKS14 derivatives and Val122Ile were also investigated and free energies of ligand binding, ΔGbinding, were calculated for TKS14 and compounds D1–D8 (Figure 2). These experiments were conducted to address how modifying functional groups within the TKS14 structural template enhanced or disrupted hydrogen-bonding interactions with TTR residues and how these modifications changed the ligands’ binding free energies.

2. Materials and Methods

The structure of the AG10 ligand complexed with Val122Ile was taken from the Protein Data Bank (PDB code 4HIQ) [28]. To generate the complexes containing TKS14 and the derivatives shown in Figure 2, the AG10 complex was edited with the software package MOE (Molecular Operating Environment, Chemical Computing Group Inc., Montreal, QC H3A 2R7, Canada) [29]. The AG10 methylene chain was replaced with a piperazine ring to give TKS14. The other functional groups were modified in a similar manner. In receptor–ligand complexes containing TKS14 and derivatives D4–D8, the ligand charge was −1. In complexes D1–D3, the ligand charges were neutral. Atomic charges for the heavy atoms in TKS14 and derivatives D1–D8 are given in Supplemental Figure S1. The MOE software program was also used to add missing residues to the proteins’ N and C termini.
AMBER 16 and the ff14SB force field were used to carry out all MD simulations after sodium counter-ions and approximately 18,000 TIP3P water molecules were added to each complex [16,30]. All MD simulations began with an energy minimization step followed by a twenty picosecond MD simulation to bring the system to 300 K. A 1 ns MD simulation was also carried out to equilibrate the system. The total simulation time for each production run was 100 ns. The production run used cubic periodic boundary conditions, and the MD simulation time step was 2 fs. Complex structures were stored every 0.2 ps. The system was at physiological pH in each MD simulation. Supplemental Information Figure S2 shows RMSD versus simulation time plots for each receptor–ligand complex. These relatively flat RMSD plots suggest that for each complex, the system is equilibrated, and that the MD simulations ran stably.
The mm-PBSA method was used to calculate the binding free energies of each Val122Ile–ligand complex [31]. Equation (1) shows that these free energy values represent the difference between the mm-PBSA free energy of the Val122Ile–ligand complex and the sum of the individual free energies of the ligand and the receptor.
ΔGbinding = Gcomplex − Greceptor − Gligand
In Equation (1), G = Gsolute + Gsolvent. The Gsolute term equals E − T·S, where S represents the entropy contributions to ligand binding and E is the MM energy averaged over the MD simulation. T is Kelvin temperature. The energy term included contributions from both electrostatic and van der Waals interactions for the complex, receptor, and ligand.
The AMBER 16 cpptraj utility was used in all trajectory analyses. When analyzing hydrogen bonds, the distance cutoff was 3.5 Å and the angle cutoff between the donor and acceptor atoms was ±30°. The hydrogen bond percent occupancies reported below represent the percent of the total MD simulation time that each hydrogen bond was present. Solvent accessible surface areas were also calculated with the AMBER16 cpptraj utility using the linear combination of pairwise overlap (LCPO) method developed by Weiser et al. [32].

3. Results and Discussion

Solvent accessible surface area (SASA) analyses were carried out to compare the ligands’ solvent exposure in the two HBPs to exposure in free solution. The results of these analyses verified that all ligands remained bound to the Val122Ile receptor throughout the MD simulations and that the ligand behaviors observed in the two HBP’s were similar to one another. Figure 3 shows SASA results for TKS14 and one of the derivatives (D8). For TKS14 (Figure 3a), the average free solution SASA value was 413 ± 3 Å2. The average SASA of TKS14 ligands in the AB and A’B’ pockets were 63 ± 9 Å2 and 70 ± 11 Å2, respectively. A t-test showed these two values were not different at the 95% confidence level. Furthermore, the TKS14 SASA decreased by 84.6% moving from free solution into pocket AB and by 83.0% moving into pocket A’B’. Analogous behavior was seen with ligand D8 as shown in Figure 3b. In both cases, the SASA plots were stable throughout the MD simulation and the %SASA decreases moving from free solution into each pocket were very similar. Corresponding percent SASA decrease values for all ligands studied were in the 80–90% range, with numerical values given in Supplemental Table S1. SASA plots for the other ligands shown in Figure 2 are given in Supplemental Figure S3.
Figure 3c,d show superimposed ligand structures extracted from the MD simulations at different time steps for TKS14 and ligand D8, respectively. The structures used for these superpositions were chosen by calculating the average structure for each MD simulation and then choosing time steps throughout that had low RMSD values with respect to the average. For TKS14 and D8, the superpositions show that the ligands’ positions in both HBPs were relatively stable, and their positions and conformations were relatively unchanged throughout the MD simulations. Superpositions extracted in an analogous manner for the other ligands in Figure 2 are shown as inserts in Supplemental Figure S3. Taken together, the SASA plots and superpositions serve to validate the stability of the MD simulations and confirm that all ligands remained TTR-bound throughout each MD simulation.
Free energies of ligand binding—ΔGbinding—values calculated for all ligands investigated are shown in Figure 4. The colors in Figure 4 indicate the ligand ring where functional groups were modified. ΔGbinding values represent the free energy difference between the ligand–Val122Ile complex and the sum of the free energies of the free ligand and receptor. Supplemental Table S2 reports ΔGbinding values for TKS14 and ligands D1–D8 along with the standard error of the mean calculated using AMBER16 software (San Francisco, CA, USA) for each ligand. The later values were then used to estimate the 95% interval for each free energy of binding [33]. These calculations yielded a confidence interval of [−18.2, −17.8] for TKS14. Confidence intervals for ligands D1–D8 are also reported in Supplemental Table S2 and were found to be similar to the TKS14 interval.
Table 1 reports the results of hydrogen bond analyses for ligands D1–D5. In Table 1, the hydrogen bond donor and acceptor atoms as well as the percent occupancy of each hydrogen bond are given. In the hydrogen bond and ΔGbinding analyses, we first considered results from an MD simulation with a complex containing Val122Ile and the TKS14 ligand. Analysis of the MD simulation showed ligand binding was thermodynamically favorable, with a ΔGbinding of −18.0 kcal mol−1. The hydrogen bonds formed between the ligand and receptor were also investigated and the two-point ligand–receptor interaction reported previously was confirmed [16,18]. Two high-occupancy hydrogen bonds formed in the AB HBP between Ser-117 residues and the TKS14 atoms to form the first connection of the two-point interaction. For example, a hydrogen bond of 56.6% occupancy formed between the ligand N atom and the Ser-117 hydroxyl atom of monomer A. Additionally, another bond of 50.7% occupancy was formed between the NH atom of the ligand’s pyrazole ring and a Ser-117 hydroxyl atom of monomer B. The outer part of the AB halogen-binding pocket also formed hydrogen bonds with the ligand, thus yielding the second connection in the two-point interaction. These hydrogen bonds occurred at 46.0% and 30.0% occupancy between the Lys-15 residue of monomer B and the carboxylate oxygens of the ligand. Moreover, the carboxylate oxygens of the ligand also hydrogen bound with the Lys-15 residue of monomer A with a 29.2% occupancy.
Similar hydrogen-bonding interactions occurred in the receptor’s A’B’ pocket with the hydrogen bonds generally having a greater percent occupancy than in the AB pocket described above. The carboxylate-Lys-15 hydrogen bonds, for example, had percent occupancies of 68.5% and 17.9% in the A’B’ pocket, compared to 46.0% and 30.0% in the AB pocket. The same trend was observed with the hydrogen bonds between the pyrazole ring and Ser-117 residues. However, when the hydrogen bonding in both pockets is considered, despite small differences in the percent occupancies, the values were similar, and the hydrogen bonds formed between the same residues.
The hydrogen-bonding interactions between Val122Ile and the TKS14 ligand are also very similar to the interactions between the protein and AG10, in that both exhibit two-point ligand–receptor interactions between the same TTR residues and ligand atoms [16,18]. While the hydrogen-bonding interactions between the TKS14 and AG10 pyrazole rings and receptor Ser-117 residues were comparable, in general, more hydrogen bonds were observed between the receptor Lys-15 residues and the TKS14 carboxylate than between the same receptor residues and the carboxylate of AG10. For example, the AG10 pyrazole ring hydrogen bound with Ser-117 at 59.4% and 64.4% occupancy in the AB and A’B’ pockets, respectively [16]. These hydrogen bonds closely resemble Ser-117’s hydrogen-bonding interactions with the pyrazole ring in TKS14. However, while the AG10 carboxylate hydrogen bound with Lys-15, as does the carboxylate of TKS14, the hydrogen-bonding interactions between the AG10 and receptor residues have a significantly lower percent occupancy than observed for the TKS14 ligand. Specifically, the highest percent occupancy of the AG10/Lys-15 hydrogen bond was 18.1%, which is significantly less than the maximum value of 68.5% observed from the TKS14/Lys-15 interaction [16]. Finally, it should be noted that the ΔGbinding of the AG10 was −10.3 kcal mol−1, compared to −18.0 kcal mol−1 for TKS14 [16]. Therefore, it is likely that the binding of the TKS14 ligand to the receptor is more thermodynamically favored, in part because of the higher occupancy hydrogen-bonding interactions between the ligand and receptor. Furthermore, the TKS14 ligand contains a piperazine ring that connects the ligand moieties experiencing the two-point interaction, whereas the AG10 ligand connects these moieties with an aliphatic chain. The central piperazine ring may lead to TKS14 having a more rigid and less conformationally flexible structure than AG10, allowing for TKS14 to hydrogen bond with the receptor Lys-15 residues more frequently than AG10.
Finally, Figure 5a further illustrates the H-bonds formed between TKS14 and Val122Ile. In the figure, H-bonds between ligand and receptor residues in the AB pocket are colored blue and H-bonds in the A’B’ pocket are colored orange. Note that the ligand pyrazole ring atoms form hydrogen bonds to both the A and B monomers in pocket AB and to the A’ and B’ monomers in pocket A’B’. Similar patters can be seen with H-bonds between the ligand carboxylate and Lys-15 residues. Therefore, when TKS14 is in the HBP, it forms H-bonds with both the AA’ and BB’ Val122Ile dimers. By hydrogen bonding to both dimers, TKS14 likely stabilizes the Val122Ile tetramer and as a result inhibits tetramer dissociation [16,18,20].
Next, derivatives of the TKS14 ligand were studied to investigate the roles that its different functional groups played in receptor binding. In the first derivative depicted as D1 in Figure 2d, the TKS14 carboxylate functional group was replaced with a methyl ester, thus removing the ligand’s negative charge. It was hypothesized that the binding of the D1 ligand to Val122Ile would be less thermodynamically favored than TKS14, as there would be fewer favorable interactions between the ligand and receptor residues in the methyl ester. Hydrogen bonds formed by Val122Ile and the D1 ligand are shown in Table 1. These data show that while the methyl ester ligand still had a two-point interaction with the receptor, the methyl ester functional group only formed hydrogen bonds with its carbonyl oxygen. Additionally, the hydrogen bonds formed between the carbonyl oxygen and Lys-15 (A/A’) residues were of 26.1% occupancy in the AB pocket, and 4.7% occupancy in the A’B’ pocket compared to 29.2% and 23.71% occupancy in the same pockets, respectively, with the TKS14 ligand. Therefore, it is likely that removing the TKS14 charge in forming the methyl ester eliminated the electrostatic attraction between the ligand and Lys-15 residues. These residues, therefore, spent less time close to the methyl ester carbonyl oxygen and thus formed fewer hydrogen bonds. Finally, it should be noted that the percent occupancies of the H-bonds formed by the D1 ligand’s methyl ester carbonyl oxygens were unexpectedly different in the AB and A’B’ pockets. For TKS14 and other derivatives though, H-bonding behavior in the two pockets was more similar. This unexpected result may have come from the ligand having a neutral charge and the Lys-15 residue being near the TTR N-terminus where the backbone was likely more flexible. At some time during the MD simulation, the Lys-15 residue may have moved away from the D1 ligand. Since there was no electrostatic attraction between the methyl ester and the Lys-15 residue, the two may have remained separated during the remainer of the simulation time, thus preventing the hydrogen bond from reforming.
The DGbinding for the methyl ester ligand was found to be −9.1 kcal mol−1, compared to −18.0 kcal mol−1 for TKS14 as reported above. The less thermodynamically favorable ligand–receptor binding observed for ligand D1 likely resulted from both reduced hydrogen bonding and elimination of the electrostatic interaction between Lys-15 residues and the ligand carboxylate. However, in MD simulations with the D1 ligand, high-occupancy hydrogen-bonding interactions similar to those in TKS14 still formed between the ligand’s pyrazole ring and the receptor’s Ser-117 residues. These H-bonds had occupancies of 26.1% and 54.3% in the AB pocket and 57.9% and 57.7% occupancy in the A’B’ pocket, compared 56.6%, 50.7%, 63.4%, and 42.1% occupancy between the same residues with the TKS14 ligand, respectively.
In the next TKS14 derivative, the carboxylate functional group was replaced with an amide. The ligand is labeled D2 in Figure 2e. The amide derivative was also studied by Cioffi et al. and was found to have a TTR FP IC50 value of 670 nM, which is significantly higher than the 220 nM value measured for TKS14 [22]. Like the D1 derivative, the D2 ligand is electrically neutral; however, unlike D1, the amide functional group contains NH2 atoms that can form hydrogen bonds with Val122Ile. Despite the presence of these potential H-bond donor atoms, ΔGbinding for ligand D2 was only −8.3 kcal mol−1 compared to −9.1 kcal mol−1 for D1 and −18.0 kcal mol−1 for TKS14. The hydrogen bond analysis also showed that ligand D2 hydrogen bound much more strongly to the inner part of both the AB and A’B’ HBP than to residues in the outer parts of both pockets. For example, the atoms of the D2 pyrazole ring hydrogen bound to Ser-117 residues with 63.3% and 46.5% occupancy in the AB pocket, and 37.0% and 55.8% occupancy in the A’B’ pocket, respectively. On the other hand, the ligand’s amide functional group only formed a 6.2% occupancy hydrogen bond with Thr-119 of monomer B. All other hydrogen bonds observed between the amide functional group and receptor residues were below 5% occupancy. Therefore, as with ligand D1, it is likely that in ligand D2, removal of the −1 charge eliminated the electrostatic attraction between the ligand and receptor Lys-15 residues, so even though the D2 amide NH2 atoms were potential H-bond donors, these atoms did not experience favorable H-bonding interactions with Val122Ile. These effects in turn likely contributed to the ligand’s less favorable ΔGbinding compared to TKS14.
The next TKS14 derivative (D3 in Figure 2f) replaced the ligand’s carboxylate functional group with a tetrazole ring. Ligand D3 was also studied by Cioffi et al. and was found to have a TTR FP IC50 value of 260 nM, compared to 220 nM for TKS14 [22]. The D3 ligand is neutral, but the tetrazole ring introduces new potential hydrogen bond donor and acceptor atoms that are not found in TKS14. The hydrogen bond analysis showed the expected hydrogen bonding between the ligand’s pyrazole ring and the receptor residues within both the AB and A’B’ HBP, as well as a 55.6% occupancy H-bond between the NH atom of the ligand’s tetrazole ring and the Glu-7 residue of monomer B. Additionally, Table 1 shows that the 55.6% occupancy H-bond is the only H-bond formed between the D3 ligand and the outer part of the pocket. In other words, no H-bonding interactions were observed between the ligand’s tetrazole ring and the Val122Ile Lys-15 residues. ΔGbinding of ligand D3 was −14.4 kcal mol−1, meaning its interaction with the receptor is more thermodynamically favored than either D1 or D2, but still less favorable than TKS14. Ligand D3 likely had a more negative binding free energy than ligands D1 or D2 because of the H-bond formed between the D3 tetrazole NH atom and the receptor Glu-7 residue. This D3 H-bond had a 55.6% occupancy, which is considerably larger than the occupancies of similar H-bonds between the D1 methyl ester carbonyl oxygen and Lys-15 (26.1%) or the D2 amide oxygen and Thr-119 (6.2%). In other words, more favorable hydrogen-bonding interactions with residues in the outer part of the HBP likely contributed to ligand D4 interacting with the receptor more favorable than ligands D1 or D2.
Since DGbinding decreased whenever the TSK14’s −1 charge was removed, another ligand D4 was studied in which the TKS14 carboxylate was replaced with a tetrazole ring and the TKS14 fluorine atom was replaced with a carboxylate (Figure 2g). Now, the ligand aromatic ring, compared to TKS14, had additional H-bond donor/acceptor atoms, but the overall compound retained its −1 charge. It can be hypothesized that the binding of the D4 ligand to Val122Ile would be more thermodynamically favored than TKS14 because the former compound’s aromatic ring would experience both hydrogen bonding and electrostatic interactions with the receptor. ΔGbinding for D4 was in fact −21.4 kcal mol−1, meaning ligand binding was more thermodynamically favored than ligands D1, D2, D3, and TKS14. These ΔGbinding results suggest that D4 may be a lead compound warranting further experimental investigation. An analysis of the D4 ligand’s hydrogen-bonding interactions are shown in both Table 1 and Figure 5b. The H-bond analysis showed high-occupancy hydrogen bonds between the ligand’s carboxylate and Lys-15 residues (40.0% (B), 35.0% (B), 13.9% (A), 18.8% (A), 56.7% (B’), 33.1% (B’) 44.3% (A’), and 28.9% (A’)), as well as high-occupancy H-bonds between the ligand’s tetrazole ring and Thr-106 residues of monomer B/B’ (34.3% (B), 75.8% (B’)). The hydrogen bonds to Thr-106 residues were not observed in any of the previous MD simulations, suggesting that the addition of the carboxylate electrostatically attracts positive Lys-15 residues to the ligand’s carboxylate where hydrogen bonding and electrostatic attraction occurs. As a result, the nearby Glu-7 residue of monomer B, which formed a hydrogen bond with the tetrazole ring in the D3 ligand, was no longer near the D4 ligand’s tetrazole ring donor and acceptor atoms. Therefore, those atoms instead formed high-occupancy hydrogen bonds with the Thr-106 (B) residue.
Figure 6 shows two alternative depictions of the ligand–receptor interactions experienced by the D4 ligand. In Figure 6a, a structure was extracted from the MD simulation at 63.7 ns and MOE was used to generate a ligand interaction map for the 63.7 ns timestep. The structure at 63.7 ns was chosen because it had a low RMSD with respect to the average structure. The map in Figure 6a shows the Val122Ile residues that are near the D4 ligand in the HBP. H-bonds formed between the ligand and receptor Thr-106, Ser-117, and Lys-15 residues are shown as well. In the map, polar or charged residues are purple and non-polar residues are green. Analogous ligand binding maps extracted at 10.56 ns and 98.45 ns showing the same ligand–receptor interactions are shown in Supplemental Information Figure S4. In the Figure 6a ligand map, a box is drawn around the TTR residues in the P2 part of the HBP (Leu-17, Lys-15, Ala-109, and Leu-110 of monomer A and Lys-15 of monomer B). Likewise, a red circle is drawn around TTR residues in the P3 part of the HBP (Ala-108 and Thr-119 of monomer A). Finally, Figure 6b shows another depiction of a structure also extracted from the MD simulation at 63.7 ns. Here, the ligand as well as Val122Ile residues Thr-106, Ser-117, and Lys-15 are shown. Again, both the ligand–receptor two-point interaction and the H-bond formed between the D4 tetrazole ring and Val122Ile Thr-106 residues are evident in Figure 6b.
In a final substitution to functional groups on the TKS14 aromatic ring, the ligand’s fluorine atom was replaced with a methoxy group (ligand D5 in Figure 2h). The ligand was studied experimentally and its TTR FP IC50 of 300 nM was found to be higher than TKS14 (220 nM) [21]. In the MD simulation with the D5 ligand though, the methoxy functional group did not hydrogen bond with any of the receptor residues. In other words, the methoxy-containing ligand experienced largely the same receptor interactions as the fluorine atom in TKS14. As a result, ΔGbinding for ligand D5 was −17.0 kcal mol−1 or very similar to the corresponding value of 18.0 kcal mol−1 calculated for TKS14.
After several modifications to the TKS14 ligand’s aromatic ring, attention was shifted to the ligand’s central piperazine ring. No hydrogen-bonding interactions were observed thus far between any of the ligand’s piperazine atoms and Val122Ile. Therefore, in ligand D6 (Figure 2i), the N atoms were each moved one position on the six-membered ring to potentially make them more accessible for hydrogen bond formation. Hydrogen bond analyses for ligands D6–D8 are reported in Table 2. Additionally, because changing the N atoms’ position removes one of their bonds to a carbon atom, the N atoms of the piperazine ring become NH functional groups, thus introducing additional potential hydrogen bond donor atoms. However, Table 2 shows that after the nitrogen atom’s positions were shifted, still no hydrogen bonds formed between the receptor and the D6 ligand’s central piperazine ring. Therefore, since the central part of the HBP is largely hydrophobic, it seems unlikely that other similar modifications to the ligand’s piperazine ring would yield new hydrogen-bonding interactions with the receptor [1]. Finally, while ligand D6 maintained its two-point interaction with the receptor pocket via its pyrazole ring and carboxylate, its ΔGbinding of −11.2 kcal mol−1 was less thermodynamically favored than TSK14. Thus, shifting the nitrogen atoms in ligand D6 may create unfavorable interactions between the polar NH atoms and non-polar residues in the central part of the HBP thus leading to less thermodynamically favored binding [1].
TKS14 derivative D7 (Figure 2j) replaced TKS14’s piperazine ring with two five-membered fused rings. The D7 ligand was studied experimentally by Cioffi et al. and had a TTR FP IC50 much higher than TSK14. The primary difference between TKS14 and ligand D7 is the size, steric bulk, and rigidity of the central part of the molecule, which connects the pyrazole and aromatic rings. ΔGbinding of ligand D7 was −9.5 kcal mol−1 or considerably less thermodynamically favorable than TKS14. In both pockets, the hydrogen bond analysis in Table 2 shows high-occupancy H-bonds between the ligand’s inner pyrazole ring and Ser-117 residues, but fewer hydrogen bonds between the D7 carboxylate functional group and Val122Ile Lys-15 residues, when compared to TKS14. This result may be caused by the fused central rings lengthening the ligand, so the carboxylate and Lys-15 residues do not align as optimally in D7 as in TKS14 and other derivatives. The reduced H-bonding with Lys-15 residues in ligand D7 also likely contributed to its relatively less favored ΔGbinding.
The final TKS14 derivative investigated modified the ligand’s pyrazole ring. Here, the NH of the pyrazole ring was replaced with a N-methyl functional group, which removed a hydrogen bond donor atom from the affected ring (ligand D8 in Figure 2k). The purpose of the N-methyl substitution was to quantify the contribution of hydrogen bonds formed by the NH atom of the TKS14 pyrazole ring to the ligand’s ΔGbinding. Ligand D8 had a ΔGbinding of −10.5 kcal mol−1, which was, as expected, less thermodynamically favorable than TKS14, which had a ΔGbinding of −18.0 kcal mol−1. Furthermore, the hydrogen bond analysis in Table 2 showed no hydrogen-bonding interactions between the nitrogen of the D8 N-methyl functional group and the TTR receptor. However, hydrogen-bonding interactions between the N atom of the D8 ligand’s pyrazole ring and Ser-117 residues increased to 73.6% occupancy from monomer A, and 69.6% occupancy from monomer B’, compared to 56.6% and 63.7% occupancy observed between TKS14 and the same receptor residues, respectively. These H-bonds are also shown in Figure 5c. This observation suggests that removing a hydrogen bond donor from the pyrazole ring decreased ΔGbinding but also increased the occupancy of hydrogen bonds from an adjacent H-bond acceptor atom, likely to supplement the loss of hydrogen-bonding interaction experienced by the pyrazole NH atom.

4. Conclusions

The binding of TKS14 to the Val122Ile transthyretin variant was found to be more thermodynamically favored than the binding of AG10. Respective ΔGbinding values were −18.0 and −10.3 kcal mol−1. When a tetrazole ring was added to TKS14 to form derivative D4, additional hydrogen-bonding interactions occurred and ΔGbinding decreased from −18.0 to −21.4 kcal mol−1. All other modifications to the TKS14 functional groups disturbed the ligands’ two-point interaction with the receptor and raised ΔGbinding. For example, when the TKS14 carboxylate was changed to a methyl ester or amide, ΔGbinding increased to −9.1 and −8.3 kcal mol−1, respectively. These ΔGbinding values are 9–10 kcal mol−1 higher than the value calculated for TKS14. Furthermore, these ΔGbinding results suggest that electrostatic interactions between Lys-15 residues and the TKS14 carboxylate are important contributors to the stability of the complex and that these interactions contribute ~ 10 kcal mol−1 to the overall binding free energy. Modifications to the central piperazine ring also increased ΔGbinding when compared to TKS14. In ligand D6, this change was attributed to unfavorable interactions likely occurring between the ring’s NH atoms and non-polar residues in the HBP. In D7, the five-membered fused rings disrupted H-bonding interactions between the ligand and receptor Lys-15 residues. Finally, addition of a methyl group to the TKS14 pyrazole ring nitrogen atom disrupted H-bonding interactions with Ser-117 residues and raised ΔGbinding to −10.5 kcal mol−1. This result suggests that H-bonding interactions between the TKS14 pyrazole ring in the HBP contribute ~ 8 kcal mol−1 to the overall binding free energy. Overall, the MD simulations show that the TKS14 structure provides an effective scaffold or template upon which functional groups can be adjusted to tune the thermodynamic favorability of ligand binding.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/biophysica5020016/s1, Table S1: Percent decrease in solvent accessible surface area (SASA) for TKS14 and derivatives D1–D8 moving from free solution into Val122Ile pockets AB and A’B’; Table S2: Free energies of ligand binding, standard error of the mean from the free energy calculations, and 95% confidence intervals for TKS14 and derivatives D1–D8. Figure S1. MD simulation atomic charges for ligands (a) TKS14 and derivatives (b) D1, (c) D2, (d) D3, (e) D4, (f) D5, (g) D6, (h) D7, and (i) D8; Figure S2. RMSD plots for (a) TKS14 and derivatives (b) D1, (c) D2, (d) D3, (e) D4, (f) D5, (g) D6, (h) D7, and (i) D8; Figure S3: Solvent accessible surface area plots and ligand superpositions for TKS14 derivatives (a) D1, (b) D2, (c) D3, (d) D4, (e) D5, (f) D6, and (g) D7; Figure S4: (a) Ligand interaction map generated in MOE for ligand D4 at 10.6 ns; (b) hydrogen bonds formed between ligand D4 and Val122Ile at 10.6 ns; (c) ligand interaction map generated in MOE for ligand D4 at 98.5 ns; (d) hydrogen bonds formed between ligand D4 and Val122Ile at 98.5 ns.

Author Contributions

Conceptualization, K.M., M.G.J. and Y.F.; methodology, K.M. and Y.F.; software, K.M. and Y.F.; validation, K.M. and Y.F.; formal analysis, K.M., J.D., I.D., L.S., K.L. and Y.F.; investigation, K.M., J.D., I.D., L.S., K.L. and Y.F.; resources, K.M. and Y.F.; data curation, K.M.; writing—original draft preparation, K.M. and I.D.; writing—review and editing, K.M. and Y.F.; visualization, K.M. and Y.F.; supervision, K.M. and Y.F.; project administration, K.M. and Y.F.; funding acquisition, K.M. and Y.F. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by National Science Foundation grants 1924092 to Y.F. and 2203506 to K.M. and by National Institutes of Health grant 2U54MD007597 to the RCMI Program at Howard University.

Data Availability Statement

All data are archived at Carthage College. Contact Kevin Morris (kmorris@carthage.edu) for access.

Acknowledgments

We acknowledge the generosity and support of the Ralph E. Klingenmeyer family.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
TTRtransthyretin
TKS14transthyretin kinetic stabilizer 14
SASAsolvent accessible surface area
FACfamilial amyloid cardiomyopathy
Val122Ilevaline-122 isoleucine point mutation
HBPhalogen-binding pocket
MDmolecular dynamics
ΔGbindingfree energy of ligand binding
MOEMolecular Operating Environment
D1–D8TKS14 derivatives, shown in Figure 2

References

  1. Tschöpe, C.; Elsanhoury, A. Treatment of transthyretin amyloid cardiomyopathy: The current opinions, the future, and the challenges. J. Clin. Med. 2022, 11, 2148. [Google Scholar] [CrossRef]
  2. Vishwakarma, K.; Mittal, S. Probing the role of V30M mutation in human transthyretin aggregation using molecular simulations. Curr. Sci. 2024, 11, 1343–1348. [Google Scholar] [CrossRef]
  3. Williams, M.A.C.; Shankar, B.; Vaishnav, J.; Ranek, M.J. Current and potential therapeutic strategies for transthyretin cardiac amyloidosis. Front. Drug Discov. 2022, 2, 1015545. [Google Scholar] [CrossRef]
  4. Nativi-Nicolau, J.N.; Karam, C.; Khella, S.; Maurer, M.S. Screening for ATTR amyloidosis in the clinic: Overlapping disorders, misdiagnosis, and multiorgan awareness. Heart Fail. Rev. 2022, 27, 785–793. [Google Scholar] [CrossRef]
  5. Mangione, P.P.; Verona, G.; Corazza, A.; Marcoux, L.; Canetti, D.; Giorgetti, S.; Raimondi, S.; Stoppini, M.; Esposito, M.; Relini, A.; et al. Plasminogen activation triggers transthyretin amyloidogenesis in vitro. J. Biol. Chem. 2018, 293, 14192–14199. [Google Scholar] [CrossRef] [PubMed]
  6. Ruberg, F.L.; Berk, J.L. Transthyretin (TTR) cardiac amyloidosis. Circulation 2012, 126, 1286–1300. [Google Scholar] [CrossRef] [PubMed]
  7. Zhong, B.; Huang, X.; Zheng, Y.; Guo, X.; Wu, L. The discovery and development of transthyretin amyloidogenesis inhibitors: What are the lessons. Future Med. Chem. 2021, 13, 2083–2105. [Google Scholar] [CrossRef]
  8. Buxbaum, J.N.; Ruberg, F.L. Transthyretin Val122Ile (pVal122Ile)* cardiac amyloidosis: An age-dependent autosomal dominant cardiomyopathy too common to be overlooked as a cause of significant heart disease in elderly African Americans. Genet. Med. 2017, 19, 733–742. [Google Scholar] [CrossRef] [PubMed]
  9. Kittleson, M.M.; Maurer, M.S.; Ambardekar, A.V.; Bullock-Palmer, R.P.; Chang, P.P.; Eisen, H.J.; Nair, A.P.; Nativi-Nicolau, J.; Ruberg, F.L. American heart association heart failure and transplantation committee of the council on clinical cardiology. Cardiac amyloidosis: Evolving diagnosis and management: A scientific statement from the American heart association. Circulation 2020, 142, e7–e22. [Google Scholar] [CrossRef]
  10. Jiang, X.; Buxbaum, J.N.; Kelly, J.W. The Val122Ile cardiomyopathy variant of transthyretin increases the velocity of rate-limiting tetramer dissociation, resulting in accelerated amyloidosis. Proc. Natl. Acad. Sci. USA 2021, 26, 14943–14948. [Google Scholar] [CrossRef]
  11. Grande-Trillo, A.; Baliellas, C.; Lladó, L.; Casasnovas, C.; Franco-Baux, J.V.; Gracia-Sánchez, L.; Bravo, M.G.; González-Vilatarsana, E.; Caballero-Gullón, L.; Echeverri, E.; et al. Transthyretin amyloidosis with cardiomyopathy after domino liver transplantation: Results of a cross-sectional study. Am. J. Transplant. 2021, 21, 372–381. [Google Scholar] [CrossRef]
  12. Gillmore, J.D.; Gane, E.; Taubel, J.; Kao, J.; Fontana, M.; Maitland, M.L.; Seitzer, J.; O’Connell, D.; Walsh, K.R.; Wood, K.; et al. CRISPR-Cas9 in vivo gene editing for transthyretin amyloidosis. N. Engl. J. Med. 2021, 385, 493–502. [Google Scholar] [CrossRef] [PubMed]
  13. Sant’Anna, R.; Gallego, P.; Robinson, L.Z.; Pereira-Henriques, A.; Ferreira, N.; Pinheiro, F.; Esperante, S.; Pallares, I.; Huertas, O.; Almeida, M.R.; et al. Repositioning tolcapone as a potent inhibitor of transthyretin amyloidogenesis and associated cellular toxicity. Nat. Commun. 2016, 7, 10787. [Google Scholar] [CrossRef] [PubMed]
  14. Corazza, A.; Verona, G.; Waudby, C.A.; Mangione, P.P.; Bingham, R.; Uings, I.; Canetti, D.; Nocerino, P.; Taylor, G.W.; Pepys, M.B. Binding of monovalent and bivalent ligands by transthyretin causes different short- and long-distance conformational changes. J. Med. Chem. 2019, 62, 8274–8283. [Google Scholar] [CrossRef] [PubMed]
  15. Almeida, M.; Gales, L.; Damas, A.M.; Cardoso, I.; Saraiva, M.J. Small transthyretin (TTR) ligands as possible therapeutic agents in TTR amyloidoses. Curr. Drug Targets CNS Neurol. Disord. 2005, 4, 587–596. [Google Scholar] [CrossRef]
  16. Bulawa, C.E.; Connelly, S.; Devit, M.; Wang, L.; Weigel, C.; Fleming, J.A.; Packman, J.; Powers, E.T.; Wiseman, R.L.; Foss, T.R.; et al. Tafamidis, a potent and selective transthyretin kinetic stabilizer that inhibits the amyloid cascade. Proc. Natl. Acad. Sci. USA 2012, 109, 9629–9934. [Google Scholar] [CrossRef]
  17. Morris, K.F.; Geoghegan, R.M.; Palmer, E.E.; George, M., Jr.; Fang, Y. Molecular dynamics simulation study of AG10 and tafamidis binding to the Val122Ile transthyretin variant. Biochem. Biophys. Rep. 2020, 21, 100721–100732. [Google Scholar] [CrossRef]
  18. Penchala, S.C.; Connelly, S.; Wang, Y.; Park, M.S.; Zhao, L.; Baranczak, A.; Rappley, I.; Vogel, H.; Liedtke, M.; Witteles, R.M.; et al. AG10 inhibits amyloidogenesis and cellular toxicity of the familial amyloid cardiomyopathy-associated Val122Ile transthyretin. Proc. Natl. Acad. Sci. USA 2013, 110, 9992–9997. [Google Scholar] [CrossRef]
  19. Gillmore, J.D.; Judge, D.P.; Cappelli, F.; Fontana, M.; Garcia-Pavia, P.; Gibbs, S.; Grogan, M.; Hanna, M.; Hoffman, J.; Masri, A.; et al. Efficacy and safety of acoramidis in transthyretin amyloid cardiomyopathy. N. Engl. J. Med. 2024, 390, 132–142. [Google Scholar] [CrossRef]
  20. Miller, M.; Pal, A.; Albusairi, W.; Joo, H.; Pappas, B.; Tuhin, M.T.H.; Liang, D.; Jampala, R.; Liu, F.; Khan, K.; et al. Enthalpy-driven stabilization of transthyretin by AG10 mimics a naturally occurring genetic variant that protects from transthyretin amyloidosis. J. Med. Chem. 2018, 61, 7862–7876. [Google Scholar] [CrossRef]
  21. Zhou, S.; Ge, S.; Zhang, W.; Zhang, Q.; Lo, G.V.; Dou, Y. Conventional molecular dynamics and metadynamics simulation studies of the binding and unbinding mechanism of TTR stabilizers AG10 and Ttafamidis. ACS Chem. Neurosci. 2020, 11, 3025–3035. [Google Scholar] [CrossRef]
  22. Cioffi, C.L.; Raja, A.; Muthuraman, P.; Jayaraman, A.; Jayakumar, S.; Varadi, A.; Racz, B.; Petrukhin, K. Identification of transthyretin tetramer kinetic stabilizers that are capable of inhibiting the retinol-dependent retinol binding protein 4-transthyretin interaction: Potential novel therapeutics for macular degeneration, transthyretin amyloidosis, and their common age-related comorbidities. J. Med. Chem. 2021, 64, 9010–9041. [Google Scholar] [CrossRef]
  23. Lei, M.; Yang, M.; Huo, S. Intrinsic versus mutation dependent instability/flexibility: A comparative analysis of the structure and dynamics of wild-type transthyretin and its pathogenic variants. J. Struct. Biol. 2004, 148, 153–168. [Google Scholar] [CrossRef] [PubMed]
  24. Zhou, S.; Zou, H.; Wang, Y.; Lo, G.V.; Yuan, S. Atomic mechanisms of transthyretin tetramer dissociation studied by molecular dynamics simulations. J. Chem. Inf. Model. 2022, 62, 6667–6678. [Google Scholar] [CrossRef] [PubMed]
  25. Das, J.K.; Mall, S.S.; Bej, A.; Mukherjee, S. Conformational flexibility tunes the propensity of transthyretin to form fibrils through non-native intermediate states. Angew. Chem. Int. Ed. 2014, 53, 12781–12783. [Google Scholar] [CrossRef]
  26. Zhou, Z.; Liu, H.; Wu, J.W.; Tsao, C.; Chen, W.; Liu, K.; Ho, Y. Computer-aided discovery of potential inhibitors for transthyretin-related amyloidosis. J. Chin. Chem. Soc. 2014, 61, 263–273. [Google Scholar] [CrossRef]
  27. Pinheiro, F.; Pallarès, I.; Peccati, F.; Sánchez-Morales, A.; Varejão, N.; Bezerra, F.; Ortega-Alarcon, D.; Gonzalez, D.; Osorio, M.; Navarro, S.; et al. Development of a highly potent transthyretin amyloidogenesis inhibitor: Design, synthesis, and evaluation. Med. Chem. 2022, 21, 14673–14691. [Google Scholar] [CrossRef] [PubMed]
  28. RCSB Protein Data Bank. Available online: https://www.rcsb.org/ (accessed on 21 April 2025).
  29. Molecular Operating Environment (MOE). Chemical Computing Group. 2020. Available online: https://www.chemcomp.com (accessed on 21 April 2025).
  30. Case, D.A.; Babin, V.; Berryman, J.T.; Betz, R.M.; Cai, Q.; Cerutti, D.S.; Cheatham, T.E., III; Darden, T.A.; Duke, R.E.; Gohlke, H.; et al. AMBER 16; University of California: San Francisco, CA, USA, 2016. [Google Scholar]
  31. Kollman, P.A.; Massova, I.; Reyes, C.; Kuhn, B.; Huo, S.; Chong, L.; Lee, M.; Lee, T.; Duan, Y.; Wang, W.; et al. Calculating structures and free energies of complex molecules: Combining molecular mechanics and continuum models. Acc. Chem. Res. 2000, 33, 889–897. [Google Scholar] [CrossRef]
  32. Weiser, J.; Shenkin, P.S.; Still, W.C. Approximate Atomic Surfaces from Linear Combinations of Pairwise Overlaps (LCPO). J. Comput. Chem. 1999, 20, 217–230. [Google Scholar] [CrossRef]
  33. McHugh, M. Standard error: Meaning and interpretation. Biochem. Medica 2008, 7–13. [Google Scholar] [CrossRef]
Biophysica 05 00016 g001
Figure 1. (a) Transthyretin (TTR) structure. Monomer chains are designated A, A’, B, B’. The two halogen binding pockets are designated AB and A’B’; (b) TTR-stabilizing ligand in pocket AB. Only residues comprising the pocket on monomer A are labeled for clarity.
Figure 1. (a) Transthyretin (TTR) structure. Monomer chains are designated A, A’, B, B’. The two halogen binding pockets are designated AB and A’B’; (b) TTR-stabilizing ligand in pocket AB. Only residues comprising the pocket on monomer A are labeled for clarity.
Biophysica 05 00016 g001
Biophysica 05 00016 g002
Figure 2. Chemical structures of (a) tafamidis, (b) AG10, (c) TKS14, and (d–k) TKS14 derivatives D1–D8.
Figure 2. Chemical structures of (a) tafamidis, (b) AG10, (c) TKS14, and (d–k) TKS14 derivatives D1–D8.
Biophysica 05 00016 g002
Biophysica 05 00016 g003
Figure 3. Solvent accessible surface area plots versus simulation time for (a) TKS14, (b) D8. Superpositions of ligands at timesteps with a low RMSD with respect to the average structure for (c) TKS14, (d) D8.
Figure 3. Solvent accessible surface area plots versus simulation time for (a) TKS14, (b) D8. Superpositions of ligands at timesteps with a low RMSD with respect to the average structure for (c) TKS14, (d) D8.
Biophysica 05 00016 g003
Biophysica 05 00016 g004
Figure 4. Free energies of ligand binding, ΔGbinding for TKS14 and each derivative investigated. The colors indicate the ligand ring where functional groups were modified.
Figure 4. Free energies of ligand binding, ΔGbinding for TKS14 and each derivative investigated. The colors indicate the ligand ring where functional groups were modified.
Biophysica 05 00016 g004
Biophysica 05 00016 g005
Figure 5. Hydrogen bonds formed between Val122Ile and (a) TKS14, (b) D4, (c) D8. H-bonds between ligand and receptor residues in the AB and A’B’ pockets are colored blue and orange, respectively. Donor and acceptor atoms and the receptor monomer chain are reported for each hydrogen bond. For clarity, percent occupancies of hydrogen bonds to Lysine residues are given in Table 1 and Table 2.
Figure 5. Hydrogen bonds formed between Val122Ile and (a) TKS14, (b) D4, (c) D8. H-bonds between ligand and receptor residues in the AB and A’B’ pockets are colored blue and orange, respectively. Donor and acceptor atoms and the receptor monomer chain are reported for each hydrogen bond. For clarity, percent occupancies of hydrogen bonds to Lysine residues are given in Table 1 and Table 2.
Biophysica 05 00016 g005
Biophysica 05 00016 g006
Figure 6. (a) Ligand interaction map generated in MOE for ligand D4. (b) Hydrogen bonds formed between ligand D4 and Val122Ile. Each structure was extracted from the MD simulation at 63.7 ns.
Figure 6. (a) Ligand interaction map generated in MOE for ligand D4. (b) Hydrogen bonds formed between ligand D4 and Val122Ile. Each structure was extracted from the MD simulation at 63.7 ns.
Biophysica 05 00016 g006
Table 1. Hydrogen bonds formed between TKS14 and derivatives D1–D5, where a functional group on the TKS14 aromatic ring was modified. Donor and acceptor atoms, percent occupancy, and the receptor monomer chain are given for each hydrogen bond.
Table 1. Hydrogen bonds formed between TKS14 and derivatives D1–D5, where a functional group on the TKS14 aromatic ring was modified. Donor and acceptor atoms, percent occupancy, and the receptor monomer chain are given for each hydrogen bond.
TKS14
DonorAcceptor% Occupancy (AB)% Occupancy (A’B’)
Ser-117 -OHTKS14 N (pyrazole)56.6% (A)63.8% (B’)
TKS14 NH (pyrazole)Ser-117 -OH50.7% (B)42.1% (A’)
Lys-15 -NH3+TKS14 COO−46.0%, 30.0% (B) 29.2% (A)68.5%, 17.9% (B’) 23.7%, 26.0% (A’)
D1
DonorAcceptor% Occupancy (AB)% Occupancy (A’B’)
D1 NH (pyrazole)Ser-117 -OH54.3% (B)57.7% (A’)
Ser-117 -OHD1 N (pyrazole)26.1% (A)57.9% (B’)
Lys-15 -NH3+D1 carbonyl oxygen26.1% (A)4.7% (A’)
D2
DonorAcceptor% Occupancy (AB)% Occupancy (A’B’)
Ser-117 -OHD2 N (pyrazole)63.3% (A)55.8% (B’)
D2 NH (pyrazole)Ser-117 -OH46.5% (B)37.0% (A’)
Thr-119 -OHD2 carbonyl oxygen6.2% (B)N/A
D3
DonorAcceptor% Occupancy (AB)% Occupancy (A’B’)
Ser-117 -OHD3 N (pyrazole)68.4% (A)61.3% (B’)
D3 NH (tetrazole)Glu-7 COO− 55.6% (B)N/A
D3 NH (pyrazole)Ser-117 -OH53.5% (B)52.6% (A’)
D4
DonorAcceptor% Occupancy (AB)% Occupancy (A’B’)
Ser-117 -OH D4 N (pyrazole)66.8% (A)N/A
D4 NH (pyrazole)Ser-117 -OH47.2% (B)8.7% (A’)
D4 NH (tetrazole)Thr-106 -OH34.3% (B)75.8% (B’)
Lys-15 -NH3+ D4 COO−40.0%, 35.0% (B)
13.9%, 18.8% (A)		56.7%, 33.1% (B’)
44.3%, 28.9% (A’)
Thr-119 -OHD4 N (tetrazole)N/A13.29% (B’)
D5
DonorAcceptor% Occupancy (AB)% Occupancy (A’B’)
Ser-117 -OHD5 N (pyrazole)61.3% (A)55.7% (B’)
D5 NH (pyrazole)Ser-117 -OH37.9% (B)42.1% (A’)
Lys-15 -NH3+D5 COO−7.8%, 52.7% (B)
56.6%, 19.0% (A)		7.6%, 8.2% (B’)
41.7%, 36.7% (A’)
Table 2. Hydrogen bonds formed between Val122Ile and TKS14 derivatives D6, D7, and D8. Donor and acceptor atoms, percent occupancy, and the receptor monomer chain are given for each hydrogen bond.
Table 2. Hydrogen bonds formed between Val122Ile and TKS14 derivatives D6, D7, and D8. Donor and acceptor atoms, percent occupancy, and the receptor monomer chain are given for each hydrogen bond.
D6
DonorAcceptor% Occupancy (AB)% Occupancy (A’B’)
Ser-117 -OHD6 N (pyrazole)56.1% (A)47.7% (B’)
D6 NH (pyrazole)Ser-117 -OH32.7% (B)47.1% (A’)
Lys-15 -NH3+D6 COO−35.0%, 35.6% (A)
28.6%, 33.2% (B)		49.7%, 4.5% (B’)
29.5%, 36.6% (A’)
D7
DonorAcceptor% Occupancy (AB)% Occupancy (A’B’)
Lys-15 -NH3+D7 COO−77.8% (A)
18.9%, 6.3% (B)		27.7%, 5.2% (B’)
5.4%, 3.9% (A’)
Ser-117 -OH D7 N (pyrazole)70.8% (A)57.6% (B’)
D7 NH (pyrazole)Ser-117 -OH41.5% (B)49.7% (A’)
D8
DonorAcceptor% Occupancy (AB)% Occupancy (A’B’)
Ser-117 -OH D8 N (pyrazole)73.6% (A)69.6% (B’)
Lys-15 -NH3+D8 COO−54.0%, 14.4% (A)
30.1%, 24.9% (B)		18.1%, 42.4% (B’)
39.8%, 26.5% (A’)
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Morris, K.; DeSalvo, J.; Deanparvar, I.; Schneider, L.; Leach, K.; George, M., Jr.; Fang, Y. Molecular Dynamics Simulation Study of Stabilizer Association with the Val122Ile Transthyretin Variant. Biophysica 2025, 5, 16. https://doi.org/10.3390/biophysica5020016

AMA Style

Morris K, DeSalvo J, Deanparvar I, Schneider L, Leach K, George M Jr., Fang Y. Molecular Dynamics Simulation Study of Stabilizer Association with the Val122Ile Transthyretin Variant. Biophysica. 2025; 5(2):16. https://doi.org/10.3390/biophysica5020016

Chicago/Turabian Style

Morris, Kevin, John DeSalvo, Iman Deanparvar, Lucus Schneider, Kaleigh Leach, Matthew George, Jr., and Yayin Fang. 2025. "Molecular Dynamics Simulation Study of Stabilizer Association with the Val122Ile Transthyretin Variant" Biophysica 5, no. 2: 16. https://doi.org/10.3390/biophysica5020016

APA Style

Morris, K., DeSalvo, J., Deanparvar, I., Schneider, L., Leach, K., George, M., Jr., & Fang, Y. (2025). Molecular Dynamics Simulation Study of Stabilizer Association with the Val122Ile Transthyretin Variant. Biophysica, 5(2), 16. https://doi.org/10.3390/biophysica5020016

Article Metrics

Back to TopTop