![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Departments of Biochemistry (S.V., C.C.) and Chemistry (M.P., J.I.K.), University of Kuopio, Kuopio, Finland; and BioXell, Nutley, New Jersey (M.R.U.)
Received November 18, 2002; accepted February 14, 2003
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
,25-dihydroxyvitamin D3
[1,25(OH)2D3]. From a mechanistic point of view,
the most interesting analog of 1,25(OH)2D3 is the
one that carries two side chains, referred to as Gemini. In this study,
molecular dynamics (MD) simulations of the Gemini-VDR complex were performed
that demonstrated that the binding of a ligand with a 25% increased volume
does not disturb the overall structure of the ligand-binding domain (LBD). It
was found that one of the two side chains takes exactly the same position as
the single side chain of the natural ligand, which suggests that the molecular
mechanism of the agonism of Gemini is identical to that of
1,25(OH)2D3. VDR single and double point mutants
at L227, A303, I313, and L397 and in vitro and ex vivo assessment of their
agonistic action confirmed the predictions of the MD simulations. Moreover, it
was found that the second side chain of Gemini can choose between two binding
positions within the ligand-binding pocket of the VDR. These two newly
identified "corners" were characterized most specifically by the
amino acids pairs L227/A303 and I313/L397. Therefore, Gemini is an important
model compound that allows further insight into the molecular actions of the
VDR but is, in parallel, also a promising precursor for the design of even
more potent 1,25(OH)2D3 analogs.
).
NRs are characterized by a highly conserved DNA-binding domain and a less
conserved ligand-binding domain (LBD). The LBD is formed by 11 to 13
-helical structures, and its overall architecture seems to be very
similar for all NRs, because it is folded into a three-layered antiparallel
-helical sandwich that comprises a central core layer of the helices 5,
9, and 11 that is sandwiched between two additional layers of helices creating
a ligand-binding pocket (Wurtz et al.,
1996). Binding of an agonistic ligand to the LBD initiates the
dissociation of corepressor proteins and an enhanced interaction with
dimerization partners, which consequently increases the amount of complex
formation with DNA. However, the major consequence of an agonist-induced
conformational change of the LBD is a repositioning of the most
carboxy-terminal -helix (helix 12), which contains a short
transactivation function 2 (AF-2) domain
(Bourguet et al., 2000). A
glutamate residue of the AF-2 domain forms, together with a lysine residue in
helix 3, a charge clamp that positions the NR interaction domain of
coactivator proteins to a hydrophobic cleft on the surface of the LBD
(Feng et al., 1998). The
correct positioning of helix 12 is critical for the distance of the glutamate
to the lysine residue and influences in this way the efficacy of the
NR-coactivator interaction. Coactivator proteins in turn contact components of
the basal transcriptional machinery resulting in an enhanced transcription of
NR target genes (Freedman,
1999).
Classic endocrine NRs, such as the receptors for steroid hormones, vitamin
D, thyroid hormone, and retinoic acid, bind their ligands with a
Kd value on the order of 1 nM or lower
(Beato et al., 1995). In
contrast, adopted orphan NRs that are sensors for lipids, such as certain
polyunsaturated fatty acids, oxysterols, bile acids, and other cholesterol
derivatives, interact with their ligands in a concentration range of 1 µM
or higher (Chawla et al.,
2001). The clear difference in the Kd values
of endocrine and adopted orphan NRs is related to the much larger
ligand-binding pocket of the latter. On the average, the ligand-binding pocket
of an adopted orphan NR is filled to only 30 to 40% with ligand, whereas
nuclear hormones occupy the ligand-binding pockets of their specific receptors
to 60% or more. This tight adaptation of hormones to their receptors seems to
restrict the possibilities to increase the ligand volume during the creation
of effective synthetic analogs.
The vitamin D receptor (VDR) belongs to the endocrine NRs
(Carlberg, 1995), because it
binds the biological active form of the secosteroid vitamin D,
1,25-dihydroxyvitamin D3
[1,25(OH)2D3], with a Kd
value of 0.1 nM. However, in contrast to other endocrine NRs, the ligand
binding pocket of VDR is, with 697 Å3, rather large and only
56% filled by the natural ligand (Rochel
et al., 2000). 1,25(OH)2D3 plays an
important role in calcium resorption and bone formation
(DeLuca et al., 1990) and was
also shown to act as a regulator of cellular proliferation, differentiation,
and apoptosis (Walters, 1992).
VDR acts preferentially as a heterodimer with the retinoid X receptor (RXR) on
specific DNA sequences in promoter regions of
1,25(OH)2D3 target genes, referred to as
1,25(OH)2D3 response elements (VDREs)
(Carlberg and Polly, 1998).
VDR-RXR-VDRE complexes are the molecular cores of DNA-dependent
1,25(OH)2D3 signaling
(Carlberg et al., 2001) and the
stabilization of the agonistic conformation of the LBD of the VDR is the most
critical step in this signaling process. This is achieved by a hydrogen bond
between the C25-hydroxyl group of 1,25(OH)2D3 and
H397 of the receptor (Rochel et al.,
2000) and is supported by an additional, minor important hydrogen
bond with H305 (Väisänen et al.,
2002b). In the presence of agonist, H397 is able to form van der
Waals contacts with F422 of the AF-2 domain, which keeps helix 12 in an
optimal position for the charge clamp between E420 (helix 12) and K246 (helix
3).
More than 2000 analogs of 1,25(OH)2D3 have
been synthesized with the goal of improving the potency and specificity of the
physiological effects of vitamin D
(Bouillon et al., 1995). The
large majority of these analogs have been modified at the side chain, which in
most cases increases their metabolic stability and, in the case of potent
agonists, increases the half-life of the VDR-ligand complex
(van den Bemd et al., 1996;
Bury et al., 2001b). However,
most super-agonists carry only minor modifications compared with the natural
hormones and stabilize the same agonistic VDR conformation via the H397-F422
interaction (Tocchini-Valentini et al.,
2001). A very interesting exception is Gemini, which is the first
1,25(OH)2D3 analog that carries two side chains
(Uskokovic et al., 1997;
Herdick et al., 2000;
Norman et al., 2000); i.e.,
this analog has a 
25% higher volume than the natural hormone. Gemini was
shown to act as a potent agonist and seems to be able to bind the VDR in its
agonistic conformation (Herdick et al.,
2000; Herdick and Carlberg,
2000), but it is not obvious how a ligand with such a drastically
increased volume can fit into the tight ligand-binding pocket of the VDR.
Therefore, in this study, molecular dynamics (MD) simulations of the
Gemini-VDR complex were performed on the basis of the X-ray structure of the
1,25(OH)2D3-bound VDR-LBD
(Rochel et al., 2000). One of
the two side chains of Gemini was found to have the same location as in the
natural hormone and contacts H397 and H305, whereas for the second side chain,
two approximately equal positions were identified. The results of the MD
simulations were confirmed by receptor mutagenesis, coactivator interaction
studies in vitro, and functional assays in living cells. It was indicated that
Gemini uses both possible positions and that the addition of fluor atoms
improves the potential of Gemini even more. Taken together, this study shows
that the ligand-binding pocket of the VDR is flexible enough to accommodate
large ligands, such as Gemini, that act even more efficiently than the natural
hormone.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
,25(OH)2D3. RO4383582 and
RO4383586 are hexafluoro-derivatives of Gemini that differ in their
configuration at carbon 20. The synthesis of Gemini and its derivatives was
described elsewhere (Uskokovic et al.,
1997; Manchand and Uskokovic,
1999;
2000;
Norman et al., 2000), their
structures are shown in Fig. 1.
All compounds were dissolved in 2-propanol; further dilutions were made in
dimethyl sulfoxide (for in vitro experiments) or in ethanol (for cell culture
experiments).
|
|
|
,25(OH)2D3-VDR-LBD complex (Brookhaven Protein
Data Bank code 1DB1
[PDB]
) determined to 1.8-Å resolution
(Rochel et al., 2000). The
missing amino acid residues of the X-ray structure (residues 118, 119,
375–377, and 424–427) were build using the Quanta98 molecular
modeling package (Molecular Simulations Inc., San Diego, CA). The four
residues missing from the C terminus (424–427) were built in an
-helical conformation (
=-57°, 
=-47°).
Crystallographic water molecules were included in simulation systems
neutralized by placing four Na+ ions in the positions of largest
electrostatic potential as determined by the program CION of the AMBER6.0
simulation package (University of California, San Francisco, CA). Gemini was
docked to the ligand-binding pocket of the LBD using the locally enhanced
sampling (LES) method, which is a mean-field technique providing ability to
focus on the interesting part of the system
(Simmerling and Elber, 1995).
In practice, multiple copies are employed on the part of the system in which
conformational sampling may be critical, and the rest of the system is treated
as a single structure. For the LES calculations, the protein-ligand complexes
were solvated by adding a sphere of water molecules with a 25-Å radius
from the mass center of the ligand. This resulted in 764 water molecules for
the complexes. Only residues having atoms within 12 Å of the atoms of
the ligand were allowed to move. The water molecules of the complex were first
energy-minimized for 2000 steps, heated to 300 K in 5 ps, and equilibrated by
30 ps at a constant temperature of 300 K. After that, the moving part of the
protein-ligand complex was minimized for 2000 steps, the temperature of the
system was increased to 300 K in 5 ps, and equilibrated for 75 ps using a time
step of 2 fs. The system was then heated to 400 K in 5 ps and equilibrated for
40 ps using a time step of 1 fs and applying harmonic positional restraint of
2.0 kcal/mol. After that, the positional restraints were removed, velocities
of the atoms were assigned to correspond to 340 K, and the system was cooled
from 340 to 0 K in 160 ps.
For periodic box simulations, the LBD-ligand complexes were solvated by
10,778 water molecules in a periodic box of 69 x 61 x 87 Å.
The minimization and equilibration of water molecules were done as in the
water cap simulations. Then the whole system was again energy-minimized for
2000 steps, heated to 300 K in 5 ps, and equilibrated by 300 ps at 300 K and
pressure of 101.325 Pa. After that, 2-ns production simulations were started.
Bonds involving hydrogen atoms were constrained to their equilibrium lengths
using the SHAKE algorithm. The time step of 1.5 fs was used in periodic box
simulations. The electrostatics were treated using the particle-mesh Ewald
method and continuum model correction was applied for energy and pressure. The
center of mass velocity was removed each 75 ps. All the simulations were done
using the AMBER6 program and the force field described by Cornell et al.
(1995).
DNA Constructs and Point Mutagenesis. The full-length cDNAs for
human VDR (Carlberg et al.,
1993) and human RXR
(Levin et al., 1992) were
subcloned into the simian virus 40 promoter-driven pSG5 expression vector
(Stratagene, LaJolla, CA). These constructs are also suitable for
T7 RNA polymerase-driven in vitro transcription/translation of the
respective cDNAs. The point mutants of VDR were generated using the Quik
Change site-directed mutagenesis kit (Stratagene) and confirmed by sequencing.
The luciferase reporter gene was driven by four copies of the rat atrial
natriuretic factor (ANF) gene promoter DR3-type VDRE (core sequence,
AGAGGTCATGAAGGACA) (Kahlen
and Carlberg, 1996) fused to the tk promoter (bold
characters indicate the hexameric core binding motifs). The NR interaction
domain of the human coactivator transcription intermediary factor 2 (spanning
amino acids 646–926) (Voegel et al.,
1996) was subcloned into the glutathione S-transferase
(GST) fusion vector pGEX (Amersham Biosciences, Uppsala, Sweden).
In Vitro Protein Translation and Bacterial Fusion Protein
Overexpression. In vitro translated VDR and RXR proteins were generated by
transcribing their respective pSG5-based cDNA expression vector with
T7 RNA polymerase and translating these RNAs in vitro using rabbit
reticulocyte lysate as recommended by the supplier (Promega, Madison, WI).
35S-Labeled VDR was generated by translation in the presence of
[35S]methionine. Bacterial overexpression of
GST-TIF2646–926 was facilitated in the Escherichia
coli BL21(DE3)pLysS strain (Stratagene) by induction with
isopropyl-
-D-thio-galactopyranoside (0.25 mM) for 3 h at
37°C.
GST Pull-Down Assay. The GST pull-down assay was performed with 40 µl of a 50% Sepharose bead slurry of GST-TIF2646–926 (pre-blocked with 1 µg/µl bovine serum albumin) and 20 ng of in vitro translated, 35S-labeled VDR (wild-type or mutant) in the presence of indicated ligands. VDR proteins were incubated in immunoprecipitation buffer (20 mM HEPES, pH 7.9, 200 mM KCl, 1 mM EDTA, 4 mM MgCl2, 1 mM dithiothretiol, 0.1% Nonidet P-40 and 10% glycerol) for 20 min at 30°C. In vitro translated proteins that were not bound to GST-fusion proteins were washed away with immunoprecipitation buffer. GST-fusion protein bound 35S-labeled VDRs were resolved by electrophoresis through 10% SDS-polyacrylamide gels and quantified on an FLA3000 reader (Fuji, Tokyo, Japan) using Image Gauge software (Fuji).
Supershift Assay. Heterodimers formed by in vitro translated VDR
(wild type or mutant) and RXR were incubated with saturating concentrations of
1,25(OH)2D3 or its analogs for 15 min at room
temperature in a total volume of 20 µl of binding buffer [10 mM HEPES, pH
7.9, 1 mM dithiothretiol, 0.2 µg/µl poly(dI-C), and 5% glycerol]. The
buffer had been adjusted to 150 mM by addition of KCl. Three micrograms of
bacterially expressed GST-TIF2646–926 fusion protein were
included in the incubation. Approximately 1 ng of the 32P-labeled
DR3-type VDRE from the rat ANF gene was then added to the
protein-ligand mixture and incubation was continued for 20 min. Protein-DNA
complexes were resolved through 8% nondenaturing polyacrylamide gels in
0.5x Tris borate/EDTA (45 mM Tris, 45 mM boric acid, and 1 mM EDTA, pH
8.3) and were quantified on a Fuji FLA3000 reader.
Transfection and Luciferase Reporter Gene Assay. MCF-7 human breast cancer cells were seeded into six-well plates (105 cells/ml) and grown overnight in phenol red-free DMEM supplemented with 5% charcoal-treated fetal bovine serum. Liposomes were formed by incubating 1 µg of the reporter plasmid and 1 µg each of pSG5-based receptor expression vectors for VDR (wild-type or mutant) and RXR with 10 µg of N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methylsulfate (Roth, Karlsruhe, Germany) for 15 min at room temperature in a total volume of 100 µl. After dilution with 900 µl of phenol red-free DMEM, the liposomes were added to the cells. Phenol red-free DMEM supplemented with 15% charcoal-treated fetal bovine serum (500 µl) was added 4 h after transfection. At this time, VDR ligands were also added. The cells were lysed 16 h after onset of stimulation using the reporter gene lysis buffer (Roche Diagnostics, Mannheim, Germany), and the constant light signal luciferase reporter gene assay was performed as recommended by the supplier (Canberra-Packard, Groningen, The Netherlands). The luciferase activities were normalized with respect to protein concentration and induction factors were calculated as the ratio of luciferase activity of ligand-stimulated cells to that of solvent controls.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
), which has been completed by molecular modeling with the
amino acid residues 118, 119, 375–377, and 424–427. Twelve
different LBD-Gemini conformations were simulated with the LES method by
systematically rotating both side chains of Gemini in steps of 20° around
the C17–C20 bond, whereas the first side chain stays at the same
position as within the 1,25(OH)2D3-VDR-LBD
structure. This conformational analysis resulted in two possible positions of
the extra side chain of Gemini, which were studied further with long periodic
box MD simulations. The whole VDR-LBD view
(Fig. 2A) shows the positions
of the second side chain of Gemini in relation to its first one.
Interestingly, the first side chain of Gemini keeps the same position than the
single side chain of 1,25(OH)2D3. The detailed
views on Gemini in position 1 (Fig. 2, B
and C) and position 2 (Fig. 2,
D and E) were restricted to those 12 amino acid residues that are
near the second side chain. Interestingly, in position 1, the extra side chain
of Gemini points toward the same direction as the methyl-group (C21) at C20 of
the 1,25(OH)2D3-VDR-LBD crystal structure
(Rochel et al., 2000), whereas
in position 2, the second side chain is rotated by 120° in relation to its
location in position 1. In detail, the dihedral angle of C16-C17-C20-C21 of
1,25(OH)2D3 in the X-ray structure was
-32.6°, the respective angles of Gemini in positions 1 and 2 were
-29.5° and -69.8°, respectively, for the first side chain and
102.7° and 155.6°, respectively for the extra side chain. The
C25-hydroxy group of the extra side chain of Gemini was found to form a
hydrogen bond with Q400 in position 1 and with V300 in position 2.
|
The distance between the 12 amino acids of the binding site and the second
side chain (Tab. 1) suggests
that the amino acid residues M272, L309, I313, L393 and H397 are significantly
closer to the extra side chain when Gemini is in position 1. In contrast,
amino acid residues L227, L230, and A303 are closer to that side chain when
the ligand is in position 2. The crystal structure of the
1,25(OH)2D3-VDR-LBD complex and the simulated
Gemini-VDR-LBD complex were compared via the root-mean-square deviation (RMSD)
value, which was calculated from the last nanosecond of the 2-ns simulation.
The RMSD value of the 12 C-atoms of the binding site of the
extra side chain was found to be 0.62 Å (1.26 Å for all atoms) for
position 1 and 0.67 Å (1.26 Å for all atoms) for position 2.
Interestingly, a corresponding MD simulation of the
1,25(OH)2D3-VDR-LBD complex provided already a
RMSD-value of 0.45 Å for these 12 C-atoms. Moreover,
the RMSD-value of all C-atoms of the VDR-LBD was 2.1 Å
for position 1 and 1.9 Å for position 2; i.e., for both positions, only
rather minor changes in the VDR-LBD structure were necessary to accommodate
the second side chain of Gemini.
|
For an experimental confirmation of the model derived by MD simulations (Fig. 2) and for a possible discrimination of the two predicted binding positions of the extra side chain of Gemini, amino acid residues L227 and A303 were selected to be most characteristic for position 2, whereas the residues I313 and L393 should represent most specifically position 1 (see Table 1). In the full-length VDR expression construct, these four amino acids were mutated individually to phenylalanine to disturb most effectively with this bulky, aromatic side chain the binding of Gemini to the respective position. In three different assay systems, the four VDR mutants were compared with wild-type VDR for the agonistic potential of Gemini in relation to the natural hormone (Fig. 3). Moreover, hexafluoro-derivatives of Gemini, RO4383582, and RO4383586, in which one of the two side chains was constrained by a triple bond and made more electronegative by replacing six hydrogen atoms with fluor atoms (for structures see Fig. 1), were included in the comparison.
The GST pull-down assay (Fig.
3A) monitored the ligand-induced interaction of VDR with the
coactivator TIF2 in solution. Compared with solvent control, all four ligands
were able to increase the VDRwt-TIF2 interaction from 5 to
40%. The mutant VDRL227F interacted more efficiently with TIF2
when it was induced by 1,25(OH)2D3 than by
Gemini, whereas both hexafluoro derivatives were comparable with each other
and more effective than Gemini. On the mutant VDRA303F, the
agonistic action of both 1,25(OH)2D3 and Gemini
was nearly completely blocked, whereas both Gemini derivatives still showed
approximately half of their maximal potential. The action of
1,25(OH)2D3 on VDRA303F was affected,
because amino acid 303 is also near the binding site of the first side chain.
Concerning the mutants VDRI313F and VDRL393F, the GST
pull-down assay proved to be insufficiently stringent to allow discrimination
among the four ligands. However, it could be shown that VDRI313F
has only half of the coactivator interaction potential of wild-type VDR,
whereas VDRL393F displayed nearly full activity.
In the supershift assay (Fig.
3B), the ligand-induced interaction of DNA-bound VDR-RXR complexes
with TIF2 was assessed. The results were quite comparable with those from the
GST-pull down assay (Fig. 3A);
i.e., on wild-type VDR, all four ligands showed comparable effects, whereas on
each of the four mutants, the hexafluoro derivatives were more effective than
1,25(OH)2D3 and Gemini. Moreover, in agreement
with the GST pull-down assay, the supershift assay demonstrated that on the
mutant VDRL227F 1,25(OH)2D3 was
clearly more effective than Gemini, whereas both ligands were blocked with
VDRA303F.
Finally, the reporter gene assay (Fig.
3C) analyzed the agonistic potential of the four ligands in the
transiently transfected model cell line MCF-7 (human breast cancer). Also,
this assay showed results similar to those of the two previous ones. The only
remarkable difference is that Gemini and its derivatives showed activities in
living cells slightly higher than those of the natural hormone. In agreement
with the two previous assays, no significant difference between the two
hexafluoro derivatives could be detected, and on all four VDR mutants, they
proved to be more active than 1,25(OH)2D3 and
Gemini. On the mutants VDRL227F, VDRI313F, and
VDRL393F, 1,25(OH)2D3 was found to be
more active than Gemini, whereas with VDRA303F, both showed low
activity.
For a more detailed analysis, double point mutants of VDR were created in
which each two amino acid residues were mutated, so that both are affecting
Gemini only in position 1 (VDRI313F/L393F) or only in position 2
(VDRL227F/A303F) or each is influencing a different binding site
(VDRL227F/I313F and VDRL227F/L393F). Please note that
the two other possible double point mutants (VDRA303F/I313F and
VDRA303F/L393F) were not considered useful, because
VDRA303F on its own already inhibited the action of
1,25(OH)2D3 and Gemini (see
Fig. 3). With the four double
point mutants (Fig. 4), the
same set of assays were performed as with the single point mutants
(Fig. 3). All three assays
provided comparable results exception of Gemini on mutant
VDRI313F/L393F, which was found to be quite active in the GST
pull-down assay (Fig. 4A) but
not in the supershift (Fig. 4B)
or reporter gene assays (Fig.
4C). This deviation was most probably attributable to the lower
stringency of the GST pull-down assay, which was done in solution without RXR,
whereas the supershift and reporter gene assays were performed on DNA in the
presence of RXR. However, a consistent observation was that on the mutants
that affect only one position at the time (VDRL227F/A303F and
VDRI313F/L393F), the activity of Gemini was still significantly
higher than on the crossover double mutants (VDRL227F/I313F and
VDRL227F/L393F). Similarly, both hexafluoro derivatives of Gemini
showed still reasonable activity on VDRL227F/A303F and
VDRI313F/L393F, which decreased on VDRL227F/I313F and
VDRL227F/L393F. In contrast,
1,25(OH)2D3 did not differentiate between the
different type of double mutants.
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
;
Herdick et al., 2000;
Norman et al., 2000) the
analog has been accepted as a potent VDR agonist, but its mechanism of action
and method of binding to the VDR-LBD was not understood. Therefore, the first
important finding of the MD simulations was that one of the two side chains of
Gemini takes exactly the same position as the single side chain of the natural
ligand. Because the side chains of Gemini themselves are not modified, this
observation suggests that the molecular mechanism of the agonism of Gemini is
identical to that of 1,25(OH)2D3. This means that
the hydroxyl group at C25 of one of the two side chains contacts H397 and H305
just as it does in 1,25(OH)2D3. Amino acid H397
interacts with F422, which in turn stabilizes helix 12, in particular E420, in
a position that is optimal for interaction with coactivator proteins. However,
it could be imagined that the extra side chain of Gemini either directly
interferes with this optimal positioning of helix 12 or that it disturbs the
overall structure of the LBD, so that it harms indirectly the stabilization of
agonistic conformation of the VDR. Therefore, it is important that the MD
simulation could indicate that the ligand-binding pocket of VDR has sufficient
extra space to accommodate a second side chain. Compared with the
1,25(OH)2D3-VDR-LBD crystal structure, only minor
shifts of the C-atoms were necessary to fit Gemini in. This
means that the binding of Gemini does not disturb the overall structure of the
VDR-LBD.
Another interesting finding was that there are two different binding sites
for Gemini within the VDR-LBD. In position 1, the extra side chain of Gemini
points in the same direction as the C21-methyl group of
1,25(OH)2D3, whereas in position 2, it is rotated
by 120° in relation to the first position (see
Fig. 2A). Mutated into bulky,
nonpolar phenylalanines, the amino acids at positions 227, 303, 313, and 393
(being by pairs specific for one of the two binding positions of the extra
side chain) were shown to disturb the action of Gemini significantly more than
that of 1,25(OH)2D3. This suggests that Gemini is
using both binding sites. This conclusion is further supported by the analysis
of double point mutants. It could be demonstrated that filling both binding
sites of the extra side chain with one bulky phenylalanine is more severe than
placing two phenylalanines together into one or the other binding site.
Precise interactions of the second side chain of Gemini with the amino acids
of the two new corners of the ligand-binding pocket await further
cocrystallographic analysis.
Gemini derivatives with asymmetric side chains should be able to discriminate between unequal binding sites. However, neither on wild-type VDR nor on any of the VDR mutants could a significant difference in the activity of the two stereoisomers of hexafluoro-Gemini be detected. Interestingly, the hexafluoro derivatives bound so efficiently to the VDR-LBD that, in contrast to Gemini, even two bulky phenylalanines could not completely diminish their activity. The superactivity of RO4383582 and RO4383586 relates to the high electronegativity of the fluor atoms, which provide the Gemini derivatives with higher affinity to the LBD than Gemini or the natural hormone. Therefore, the mutations of the LBD are less severe for RO4383582 and RO4383586 than for Gemini. Taken together, in contrast to all other known VDR ligands, Gemini and its derivatives have the choice between two positions to bind the LBD.
Gemini has been shown previously to stabilize the VDR also in a
nonagonistic conformation (Herdick et al.,
2000; Herdick and Carlberg,
2000; Bury et al.,
2001a). The nonagonistic conformation of the VDR resembles its
apo-form, which is characterized primarily by a different orientation of helix
12. In this conformation VDR is not able to interact with coactivator proteins
but forms complexes with corepressors. A ligand that stabilizes the VDR-LBD in
this conformation has no agonistic effect. However, in contrast to the
antagonistic conformation of the VDR
(Väisänen et al.,
2002a), in which the receptor is permanently blocked to induce
gene activity, the nonagonistic conformation can be converted into the
agonistic conformation, when the binding partners of the VDR are exchanged.
This study focused on the agonistic conformation of the Gemini-bound VDR-LBD,
because as a starting point for the MD simulations, coordinates of the VDR
were available only in its agonistic form, not its apo-form. However, it can
be speculated that with a displaced helix 12, H397 will no longer have a
dominant attraction for the C25-hydroxyl group of the side chains of Gemini.
Then it may be possible that both side chains bind to the two newly identified
corners of the ligand-binding pocket. In this way, Gemini may be able to bind
to the LBD without inducing any agonistic action of the receptor.
In conclusion, this study could demonstrate that endocrine NRs, such as the
VDR, could also interact efficiently with derivatives of their natural ligands
that have a 25% increased volume. Gemini is an outstanding
1,25(OH)2D3 analog, because it has the unique
possibility to choose between two binding positions within the ligand-binding
pocket of the VDR.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
ABBREVIATIONS: NR, nuclear receptor;
1,25(OH)2D3, 1,25-dihydroxyvitamin
D3; AF-2, (trans)activation function-2; ANF, atrial natriuretic
factor; DR3, direct repeat spaced by 3 nucleotides; GST, glutathione
S-transferase; MD, molecular dynamics; RXR, retinoid X receptor; LES,
locally enhanced sampling; LBD, ligand-binding domain; RMSD, root-mean-square
deviation; TIF2, transcriptional intermediary factor 2; VDR,
1,25(OH)2D3 receptor; VDRE,
1,25(OH)2D3 response element; RO272310,
1,25-dihydroxy-21(3-hydroxy-3-methylbutyl)-cholecalciferol; RO4383582,
1,25-dihydroxy-20S,
21(3-hydroxy-3-methylbutyl)-23-yne-26,27-hexafluoro-cholecalciferol;
RO4383586,
1,25-dihydroxy-20R,21(3-hydroxy-3-methylbutyl)-23-yne-26,27-hexafluoro-cholecalciferol;
DMEM, Dulbecco's modified Eagle's medium.
Address correspondence to: Prof. Carsten Carlberg, Department of Biochemistry, University of Kuopio, P.O. Box 1627, FIN-70211 Kuopio, Finland. E-mail: carlberg{at}messi.uku.fi
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Bouillon R, Okamura WH, and Norman AW (1995) Structure-function relationships in the vitamin D endocrine system. Endocr Rev 16: 200–257.[CrossRef][Medline]
Bourguet W, Germain P, and Gronemeyer H (2000) Nuclear receptor ligand-binding domains: three-dimensional structures, molecular interactions and pharmacological implications. Trends Pharmacol Sci 21: 381–388.[CrossRef][Medline]
Bury Y, Herdick M, Uskokovic MR, and Carlberg C
(2001a) Gene regulatory potential of 1,
25-dihydroxyvitamin D3 analogues with two side chains. J
Cell Biochem 81:
179–190.[CrossRef][Medline]
Bury Y, Ruf D, Hansen CM, Kissmeyer AM, Binderup L, and Carlberg C (2001b) Molecular evaluation of vitamin D3 receptor agonists designed for topical treatment of skin diseases. J Invest Dermatol 116: 785–792.[CrossRef][Medline]
Carlberg C (1995) Mechanisms of nuclear signalling by vitamin D3. Interplay with retinoid and thyroid hormone signalling. Eur J Biochem 231: 517–527.[Medline]
Carlberg C, Bendik I, Wyss A, Meier E, Sturzenbecker LJ, Grippo JF, and Hunziker W (1993) Two nuclear signalling pathways for vitamin D. Nature (Lond) 361: 657–660.[CrossRef][Medline]
Carlberg C and Polly P (1998) Gene regulation by vitamin D3. Crit Rev Eukaryot Gene Expr 8: 19–42.[Medline]
Carlberg C, Quack M, Herdick M, Bury Y, Polly P, and Toell A (2001) Central role of VDR conformations for understanding selective actions of vitamin D3 analogues. Steroids 66: 213–221.[CrossRef][Medline]
Chawla A, Repa JJ, Evans RM, and Mangelsdorf DJ (2001)
Nuclear receptors and lipid physiology: opening the X-files.
Science (Wash DC) 294:
1866–1870.
Cornell WD, Cieplak P, Bayly CI, Gould I, Merz KM, Ferguson D, Spellmeyer DC, Fox T, Caldwell JW, and Kollman PA (1995) A second generation force field for the simulation of proteins, nucleic acids and organic molecules. J Am Chem Soc 117: 5179–5197.[CrossRef]
DeLuca HF, Krisinger J, and Darwish H (1990) The vitamin D system. Kidney Int 38: S2–S8.
Feng W, Ribeiro RCJ, Wagner RL, Nguyen H, Apriletti JA, Fletterick
RJ, Baxter JD, Kushner PJ, and West BL (1998) Hormone-dependent
coactivator binding to a hydrophobic cleft on nuclear receptors.
Science (Wash DC) 280:
1747–1749.
Freedman LP (1999) Increasing the complexity of coactivation in nuclear receptor signaling. Cell 97: 5–8.[CrossRef][Medline]
Herdick M, Bury Y, Quack M, Uskokovic M, Polly P, and Carlberg C
(2000) Response element- and coactivator-mediated conformational
change of the vitamin D3 receptor permits sensitive interaction
with agonists. Mol Pharmacol
57:
1206–1217.
Herdick M and Carlberg C (2000) Agonist-triggered modulation of the activated and silent state of the vitamin D3 receptor by interaction with co-repressors and co-activators. J Mol Biol 304: 793–801.[CrossRef][Medline]
Kahlen JP and Carlberg C (1996) Functional characterization of a 1,25-dihydroxyvitamin D3 receptor binding site found in the rat atrial natriuretic factor promoter. Biochem Biophys Res Commun 218: 882–886.[CrossRef][Medline]
Levin AA, Sturzenbecker LJ, Kazmer S, Bosakowski T, Huselton C,
Allenby G, Speck J, Kratzeisen C, Rosenberger M, Lovey A, et al.
(1992) 9-cis-Retinoic acid stereoisomer binds and
activates the nuclear receptor RXR. Nature
(Lond) 355:
359–361.[CrossRef][Medline]
Manchand PS and Uskokovic MR (1999) inventors, Hoffmann-La Roche, Inc., and Syntex, assignees. Method of using Vitamin D3 analogs with BIS C-20 side chains. U.S. Patent 6,008,209. 1999 Dec 28.
Manchand PS and Uskokovic MR (2000) inventors, and Syntex, assignee. Vitamin D3 analogs with BIS C-20 side chains. U.S. Patent 6,030,962. 2000 Feb 29.
Mangelsdorf DJ, Thummel C, Beato M, Herrlich P, Schütz G, Umesono K, Blumberg B, Kastner P, Mark M, Chambon P, et al. (1995) The nuclear receptor superfamily: the second decade. Cell 83: 835–839.[CrossRef][Medline]
Norman AW, Manchand PS, Uskokovic MR, Okamura WH, Takeuchi JA,
Bishop JE, Hisatake J-I, Koeffler HP, and Peleg S (2000)
Characterization of a novel analog of 1,25(OH)2-vitamin
D3 with two side chains: interaction with its nuclear receptor and
cellular actions. J Med Chem
43:
2719–2730.[CrossRef][Medline]
Rochel N, Wurtz JM, Mitschler A, Klaholz B, and Moras D (2000) Crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5: 173–179.[CrossRef][Medline]
Simmerling CL and Elber R (1995) Computer
determination of peptide conformations in water: different roads to structure.
Proc Natl Acad Sci USA
92:
3190–3193.
Tocchini-Valentini G, Rochel N, Wurtz JM, Mitschler A, and Moras D
(2001) Crystal structures of the vitamin D receptor complexed to
superagonist 20-epi ligands. Proc Natl Acad Sci USA
98:
5491–5496.
Uskokovic MR, Manchand PS, Peleg S, and Norman AW
(1997) Synthesis and preliminary evaluation of the biological
properties of a 1, 25-dihydroxyvitamin D3 analog with two
side-chains, in Vitamin D: Chemistry, Biology and Clinical
Applications of the Steroid Hormone. Proceedings of the 10th International
Vitamin D Workshop; 1997 May 24; Strasbourg, France (Norman AW,
Bouillon R, Thomasset M, eds) pp. 19–21,
University of California, Riverside, CA.
Väisänen S, Peräkylä M, Kärkkäinen JI, Steinmeyer A, and Carlberg C (2002a) Critical role of helix 12 of the vitamin D3 receptor for the partial agonism of carboxylic ester antagonists. J Mol Biol 315: 229–238.[CrossRef][Medline]
Väisänen S, Ryhänen S, Saarela JT, Peräkyla M,
Andersin T, and Maenpää PH (2002b) Structurally and
functionally important amino acids of the agonistic conformation of the human
vitamin D receptor. Mol Pharmacol
62:
788–794.
van den Bemd G-JCM, Pols HAP, Birkenhäger JC, and van Leeuwen
JPTM (1996) Conformational change and enhanced stabilization of
the vitamin D receptor by the 1,25-dihydroxyvitamin D3 analog
KH1060. Proc Natl Acad Sci USA
93:
10685–10690.
Voegel JJ, Heine MJS, Zechel C, Chambon P, and Gronemeyer H (1996) TIF2, a 160 kDa transcriptional mediator for the ligand-dependent activation function AF-2 of nuclear receptors. EMBO (Eur Mol Biol Organ) J 15: 3667–3675.[Medline]
Walters MR (1992) Newly identified actions of the vitamin D endocrine system. Endocr Rev 13: 719–764.[CrossRef][Medline]
Wurtz J-M, Bourguet W, Renaud J-P, Vivat V, Chambon P, Moras D, and Gronemeyer H (1996) A canonical structure for the ligand-binding domain of nuclear receptors. Nat Struct Biol 3: 87–94.[CrossRef][Medline]
This article has been cited by other articles:
|
|
 |
|
  F. Molnar, M. Perakyla, and C. Carlberg Vitamin D Receptor Agonists Specifically Modulate the Volume of the Ligand-binding Pocket J. Biol. Chem., April 14, 2006; 281(15): 10516 - 10526. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Perakyla, M. Malinen, K.-H. Herzig, and C. Carlberg Gene Regulatory Potential of Nonsteroidal Vitamin D Receptor Ligands Mol. Endocrinol., August 1, 2005; 19(8): 2060 - 2073. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. Gonzalez, P. Samenfeld, M. Perakyla, and C. Carlberg Corepressor Excess Shifts the Two-Side Chain Vitamin D Analog Gemini from an Agonist to an Inverse Agonist of the Vitamin D Receptor Mol. Endocrinol., October 1, 2003; 17(10): 2028 - 2038. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
