Lignans are important biologically active dietary polyphenolic compounds. Consumption of foods that are rich in lignans is associated with positive health effects. Using modeling tools to probe the ligand-binding pockets of molecular receptors, we found that lignans have high docking affinity for the human thyroid hormone receptor β. Follow-up experimental results show that lignans (-) arctigenin and (+) pinoresinol are antagonists of the human thyroid hormone receptor β. The modeled complexes show key plausible interactions between the two ligands and important amino acid residues of the receptor.
Lignans are important biologically active dietary polyphenolic compounds. Consumption of foods that are rich in lignans is associated with positive health effects. Using modeling tools to probe the ligand-binding pockets of molecular receptors, we found that lignans have high docking affinity for the human thyroid hormone receptor β. Follow-up experimental results show that lignans (-) arctigenin and (+) pinoresinol are antagonists of the human thyroid hormone receptor β. The modeled complexes show key plausible interactions between the two ligands and important amino acid residues of the receptor.
Compounds that make up
the noncaloric components of the human diet have profound influence
on the expression of genes and homeostatic regulations in biological
systems although most molecular mechanisms involved in such regulations
remain unknown. Phenolic and polyphenolic molecules constitute a major
group of such compounds. There are over 500 structurally different
dietary phenolic/polyphenol-like compounds. These include anthocyanins,
chalcones, flavanols, flavones, isoflavones, phenolic acids, stilbenes,
lignans, phenolic terpenes, hydroxycoumarins, etc. They are found
in appreciable quantities in plant-derived edibles, such as fruits,
vegetables, nuts, and seeds, as well as in many popular beverages.[1] Over the past two decades, epidemiological studies
have shown that polyphenols promote vascular function, reduce hypertension, and lower the risk of cardiovascular diseases, neurodegenerative diseases,
cancer, and stroke.[2,3] It is well-documented that the
metabolic effects of these compounds are pleiotropic in nature.[4−6] The pleiotropy associated with these compounds seems to stem from
their promiscuity toward numerous molecular targets, for example,
multiple receptors or enzymes. It is becoming increasingly clear,
however, that these compounds may not have therapeutic effects during
pathological states but do have modulatory or hormetic effects that
are largely beneficial in biological systems. These nontherapeutic
effects are due, perhaps, to their relatively weak binding affinities
to cognate receptors/molecular targets in vivo and
to their susceptibility to phase II metabolic alterations.The
molecular targets of most polyphenols with reported biological activity
remain unknown, but many are suspected to either activate or deactivate
membrane-bound or cytosolic receptors. The isoflavones found in leguminous
plants, for example, are known to have moderate binding affinities for
the estrogen receptors. Isoflavones have been shown to have estrogenic
effects which may or may not be advantageous, depending on the exposure
levels and on the developmental or physiological state of the human
subject.[7,8] Also, it was reported recently that some
dietary phytochemicals perturb cell membranes and promiscuously alter
protein function.[9] Human exposure to lignans
occurs predominantly through consumption of flaxseeds and sesame seeds.
Lignans are also present in lower amounts in broccoli, curvy kale,
and apricots. It has been reported that enterolignans, such as enterodiol
and enterolactone, have weak estrogenic activity.[1,10−12]We report in this article that (−) arctigenin
and (+) pinoresinol, lignans present in sesame seeds and olive oil,
respectively, are antagonists of the human thyroid hormone receptor
β (hTRβ), and we describe the molecular features that
define the interactions between the receptor and the two lignans.
Structurally, the hTRβ consists of an N-terminal domain (NTD),
a DNA binding domain (DBD) which serves as the nuclear localization
signal, and a C-terminal ligand binding domain (LBD). The LBD of hTRβ
is made up of 12 alpha-helices. The binding cavity in the LBD is mainly
hydrophobic but also contains a hydrophilic cavity. The hydrophobic
portion is known to interact with the iodinated rings of thyroid hormone.
Amino acid residues Arg 320, 316, and 282, as well as Asn 331, make
up the hydrophilic pocket. This hydrophilic pocket mainly interacts
with the polar substituent of thyroid hormone. In addition, amino
acid residue His 435 in helix 11 of the ligand binding cavity serves
as a hydrogen bond acceptor.[13,14]
Experimental
Details
Compound and Protein Structure Preparation
The ligands were drawn, and their geometries were optimized using
the molecular mechanics force field (MMFF) algorithm in Spartan ’10
for Windows.[15] Structural information about
the ligands was obtained from the Phenol-Explorer database.[1] The docking studies were carried out using the
crystal structures of the ligand binding domain of hTRβ (PDB
Id: 2pin, 3gws, 2j4a(13,16,17)) from the RCSB Protein Data Bank. The protein
structures were used as rigid model structures. No relaxation was
performed, and assignments of ionic charges on each protein structure
were based on standard protonation states and the default templates
of Molegro Virtual Docker (MVD).[18,19]
Docking Simulation and Scoring
Flexible ligand models
were used for docking and postdocking geometry optimizations. Simulations
were carried out using the ligand binding site of hTRβ. A docking
sphere (15 Å radius) was placed on the binding sites of each
crystal structure in order to allow different orientations of each
ligand to be searched in the binding cavities and for multiple protein–ligand
poses to be returned. The RMSD threshold for multiple cluster poses
was set at <1.00 Å. The docking algorithm was set at maximum
iterations of 1500 with a simplex evolution population size of 50
and a minimum of 30 runs for each ligand. Each binding site of oligomeric
structures was searched, and docking scores of the lowest energy pose
(based on the MVD rerank scores) for each ligand across all protein
structures are presented in Tables S1–S10. The 2D representations of receptor–ligand interactions were
prepared using Molecular Operating Environment (MOE).[20]
Human Thyroid Hormone Receptor
β and Cell Viability Assays
The thyroid hormone receptor
assay was carried out using the hTRβ (NR1A2) luciferase assay
system from Indigo Biosciences (State College, PA) according to the
manufacturer’s instructions. (−) Arctigenin was obtained
from Tocris Bioscience (Bristol, UK), and (+) pinoresinol was obtained
from Sigma-Aldrich (St. Louis, MO). Human TRβ agonist l-triiodothyronine provided with the assay system kit was used as
positive control for receptor activation. The activation/deactivation
of receptor activity was monitored in 8 dose–response experiments
with concentrations ranging from 65 nM to 50 μM for the lignans
and 41 nM to 30 μM for the endogenous agonist. Reporter cell
suspension (100 μL) was dispensed into 96-well assay plates,
and 100 μL of test compounds in compound screening medium was
added to the appropriate wells in triplicate. For antagonist mode
assays, the reporter cell suspension was supplemented with 3.3 μM l-triiodothyronine shortly before test compounds in compound
screening medium were added. The assay plates were placed at 37 °C
in a humidified 5% CO2 incubator for 24 h. After incubating
for 24 h, the Luciferase Detection Reagent (100 μL) was added
to each well and incubated for 15 min at room temperature, and luminescence
was quantified using the Ascent Software on Labsystems Fluoroskan
Ascent FL reader (Helsinki, Finland). EC50/IC50 values were generated using GraphPad Prism 6.00 for Windows (La
Jolla, CA). The effect of (−) arctigenin and (+) pinoresinol
on the viability of the reporter cells was determined using the MTT
assay. They were tested on the reporter cells using the MTT assay
at the following concentrations: 50 μM, 593 nM, and 7 nM.[21] The viability of the treated cells was calculated
based on the mean value of the no treatment control (100% viability).
Results and Discussion
(−)
Arctigenin and (+) Pinoresinol Are Antagonists of the hTRβ
Using molecular modeling tools to explore the structural compatibility
between polyphenolic compounds and a wide range of molecular targets,
we found that lignans have relatively high docking scores for the
ligand binding site of the human TRβ when compared to other
dietary polyphenolic compounds. The docking scores from the simulations
are presented in Figure 1, Tables S1–S10, and Figures S1 and S2. Lignans have
not previously been reported as either agonists or antagonists of
the human TRβ, so we tested the dietarily important lignans (−) arctigenin and (+) pinoresinol for their ability
to activate or deactivate human TRβ.
Figure 1
Docking scores of dietary
polyphenolic compounds with the human thyroid hormone receptor β.
The figure shows that lignans (▲, orange) have relatively high
docking scores for the receptor.
Docking scores of dietary
polyphenolic compounds with the human thyroid hormone receptor β.
The figure shows that lignans (▲, orange) have relatively high
docking scores for the receptor.The results show that (−) arctigenin and (+) pinoresinol
are antagonists of the human TRβ with IC50 values
of 3.8 μM and 8.2 μM, respectively (Figure 2). The lignans were also tested for possible cytotoxicity
on the reporter cells, and the results show that the lignans were
not toxic to the cells at the concentrations tested (Figure 3).
Figure 2
Antagonist activity of (−) arctigenin and (+) pinoresinol
in a cell-based hTRβ reporter assay.
Figure 3
Effect of (−)
arctigenin and (+) pinoresinol on the viability of reporter cells.
Antagonist activity of (−) arctigenin and (+) pinoresinol
in a cell-based hTRβ reporter assay.(−) Arctigenin has been reported as an inhibitor of
cellular metabolism during glucose-deprived conditions. It has also
been shown to inhibit the mitochondria complex 1, in addition to causing
the activation of AMP-activated protein kinase in L6 myotubes and
isolated skeletal muscles.[22,23] Perhaps this previously
reported action of arctigenin may be related to its antagonism of
the TRβ, although this remains to be tested.Effect of (−)
arctigenin and (+) pinoresinol on the viability of reporter cells.
Structural
Motifs Involved in (−) Arctigenin and (+) Pinoresinol Interaction
with hTRβ
To understand the molecular interactions
that may be responsible for the activity of the two lignans at the
receptor, the modeled complexes of the compounds and hTRβ were
evaluated. Both lignans were predicted to interact with the following
hTRβ amino acid residues: Phe 455, 269, and 272; Ala 234, 279,
and 317; Arg 282 and 316; Asn 233 and 331; Ile 276 and 312; Thr 273
and 329; Val 283; Gly 332, 344, and 345; Leu 330, 341, and 346; Met
310, 313, and 442; Ser 314; and His 435 (Figure 4).
Figure 4
Amino acid residues predicted
to interact with (−) arctigenin (top) and (+) pinoresinol (bottom).
Hydrogen bonding interactions were predicted between the
hydroxyl group of the lignans’ 4-hydroxy-3-methoxyphenyl moiety
and the backbone carbonyl group of hTRβ’s Gly 344. The
guanidinium side-chain of Arg 282 is also predicted to hydrogen bond
with (−) arctigenin’s 3,4-dimethoxyphenyl moiety and
with (+) pinoresinol’s 4-hydroxy-3-methoxyphenyl moiety (Figure 5). The strengths of the predicted hydrogen bonds
are moderate and mostly electrostatic, based on the predicted lengths.[24] The endogenous agonist l-triiodothyronine
is known to hydrogen bond with the guanidinium side-chain of Arg 282,
as well as with His 435.[13,14,25] There are also significant steric interactions between (+) pinoresinol
and Ala 279 of hTRβ and between (−) arctigenin
and Met 313 of hTRβ (Figure 5).
Figure 5
Key hydrogen bonding interactions (blue dash
lines) between human TRβ and endogenous ligand l-triiodothyronine
(top), (+) pinoresinol (middle), and (−) arctigenin (bottom).
Conclusion
From molecular docking simulations, we found that, relative to
other dietary polyphenols, lignans have high structural compatibility
with the ligand binding pocket of the thyroid hormone receptor β.
Our experimental studies revealed that lignans (−) arctigenin
and (+) pinoresinol are antagonists of the human thyroid hormone receptor
β (hTRβ). (−) Arctigenin and (+) pinoresinol have
low micromolar IC50 values and are predicted to interact
with important amino acid residues such as Arg 282, His 435, Ala 279,
and Met 313. Future work on the effects of these compounds on the
multitude of TRβ target genes and on their ability to modulate
the physiological roles of TRβ will be valuable. In addition,
it would be of value to determine the importance, as well as the energetic
contributions, of amino acid residues Gly 344 and Arg 282 to the interactions
between the receptor and the lignans using experimental and computational
mutational studies and molecular dynamics simulations.Amino acid residues predicted
to interact with (−) arctigenin (top) and (+) pinoresinol (bottom).Key hydrogen bonding interactions (blue dash
lines) between human TRβ and endogenous ligand l-triiodothyronine
(top), (+) pinoresinol (middle), and (−) arctigenin (bottom).
Authors: Robert L Dow; Steven R Schneider; Ernest S Paight; Richard F Hank; Phoebe Chiang; Peter Cornelius; Eunsun Lee; William P Newsome; Andrew G Swick; Josephine Spitzer; Diane M Hargrove; Terrell A Patterson; Jayvardhan Pandit; Boris A Chrunyk; Peter K LeMotte; Dennis E Danley; Michele H Rosner; Mark J Ammirati; Samuel P Simons; Gayle K Schulte; Bonnie F Tate; Paul DaSilva-Jardine Journal: Bioorg Med Chem Lett Date: 2003-02-10 Impact factor: 2.823
Authors: G Siasos; D Tousoulis; V Tsigkou; E Kokkou; E Oikonomou; M Vavuranakis; E K Basdra; A G Papavassiliou; C Stefanadis Journal: Curr Med Chem Date: 2013 Impact factor: 4.530
Authors: Helgi I Ingólfsson; Pratima Thakur; Karl F Herold; E Ashley Hobart; Nicole B Ramsey; Xavier Periole; Djurre H de Jong; Martijn Zwama; Duygu Yilmaz; Katherine Hall; Thorsten Maretzky; Hugh C Hemmings; Carl Blobel; Siewert J Marrink; Armağan Koçer; Jon T Sack; Olaf S Andersen Journal: ACS Chem Biol Date: 2014-06-17 Impact factor: 5.100
Authors: Joseph A Rothwell; Jara Perez-Jimenez; Vanessa Neveu; Alexander Medina-Remón; Nouha M'hiri; Paula García-Lobato; Claudine Manach; Craig Knox; Roman Eisner; David S Wishart; Augustin Scalbert Journal: Database (Oxford) Date: 2013-10-07 Impact factor: 3.451