Synthesis, biological evaluation and molecular modeling studies of the PPARβ/δ antagonist CC618

Article abstract of DOI:10.1038/415813a

Herein, we describe the synthesis, biological evaluation and molecular docking of the selective PPARβ/δ antagonist (4-methyl-2-(4-(trifluoromethyl)phenyl)-N-(2-(5-(trifluoromethyl)-pyridin-2-ylsulfonyl)ethyl)thiazole-5-carboxamide)), CC618. Results from in vitro luciferase reporter gene assays against the three known human PPAR subtypes revealed that CC618 selectively antagonizes agonist-induced PPARβ/δ activity with an IC₅₀ Combining double low line 10.0 μM. As observed by LC-MS/MS analysis of tryptic digests, the treatment of PPARβ/δ with CC618 leads to a covalent modification of Cys249, located centrally in the PPARβ/δ ligand binding pocket, corresponding to the conversion of its thiol moiety to a 5-trifluoromethyl-2-pyridylthioether. Finally, molecular docking is employed to shed light on the mode of action of the antagonist and its structural consequences for the PPARβ/δ ligand binding pocket.

Full text of DOI:10.1038/415813a

letters to nature
15. Wroe, S. F. et al. An imprinted transcript, antisense to Nesp, adds complexity to the cluster of
imprinted genes at the mouse Gnas locus. Proc. Natl Acad. Sci. USA 97, 3342–3346 (2000).
16. Lee, Y. J. et al. Mit1/Lb9 and Copg2, new members of mouse imprinted genes closely linked to Peg1/
Mest1. FEBS Lett. 472, 230–234 (2000).
affected only a small region of the X chromosome that contains a
.
dosage-sensitive gene24,25
A
Methods
Generation of the Air-T allele
17. Rougeulle, C., Cardoso, C., Fontes, M., Colleaux, L. & Lalande, M. An imprinted antisense RNA
overlaps UBE3A and a second maternally expressed transcript. Nature Genet. 19, 15–16 (1998).
18. Smilinich, N. J. et al. A maternally methylated CpG island in KvLQT1 is associated with an antisense
paternal transcript and loss of imprinting in Beckwith-Wiedemann syndrome. Proc. Natl Acad. Sci.
USA 96, 8064–8069 (1999).
The targeting vector was constructed from a 9.6-kb BglII fragment (bp 117,878–127,575;
AJ249895). A selection cassette containing neomycin and the thymidine kinase gene, each
driven by a PGK promoter flanked by loxP sites plus a 1.2-kb fragment from the rabbit b-
globin gene (bp 31,392–32,590; M18818) containing part of exon 2, complete intron 3
and complete exon 3 with the polyadenylation signal in the correct orientation for the Air
promoter, was inserted at the BamHI site in this fragment. The selection cassette was
deleted by electroporation of a plasmid encoding Cre recombinase26 and transient
puromycin selection. We generated chimaeric mice by injecting paternally targeted Air-T
embryonic stem (ES) cells into C57/Bl6 blastocysts27. Mice carrying the Air-T allele were
maintained on an FVB/N background.
19. Avner, P. & Heard, E. X-chromosome inactivation: counting, choice and initiation. Nature Rev. Genet.
2, 59–67 (2001).
20. Clemson, C. M., McNeil, J. A., Willard, H. F. & Lawrence, J. B. XIST RNA paints the inactive X
chromosome at interphase: evidence for a novel RNA involved in nuclear/chromosome structure.
J. Cell Biol. 132, 259–275 (1996).
21. Lee, J. T., Strauss, W. M., Dausman, J. A. & Jaenisch, R. A 450 kb transgene displays properties of the
mammalian X-inactivation center. Cell 86, 83–94 (1996).
22. Sheardown, S. A. et al. Stabilization of Xist RNA mediates initiation of X chromosome inactivation.
Cell 91, 99–107 (1997).
Methylation and expression analyses
23. Wutz, A. & Jaenisch, R. A shift from reversible to irreversible X inactivation is triggered during ES cell
differentiation. Mol. Cell 5, 695–705 (2000).
Digestion of methyl-sensitive enzymes was monitored by hybridization to mitochondrial
DNA. We carried out RPA using an RPAIII kit (Ambion). Signals were quantified with a
Phosphorimager (Fujix). For RT-PCR, 5 mg heart RNA was reverse transcribed with
Superscript (Gibco) using the linker-poly(dT) primer RT, 50 -CTGGGAAACAGCTATGA
CCATGATCGATTTTTTTTTTTTTTTTTN-3 0 , and amplified with PX6R, 50 -GAAGCAC
AGCACCGCCAGTTAC-30 , for 30 cycles (94 8C, 30 s; 59 8C, 30 s; 72 8C, 120 s).
24. Lyon, M. F. Imprinting and X-chromosome inactivation. Results Probl. Cell Differ. 25, 73–90 (1999).
25. Graves, J. A. Mammals that break the rules: genetics of marsupials and monotremes. Annu. Rev. Genet.
30, 233–260 (1996).
26. O’Gorman, S., Dagenais, N. A., Qian, M. & Marchuk, Y. Protamine-Cre recombinase transgenes
efficiently recombine target sequences in the male germ line of mice, but not in embryonic stem cells.
Proc. Natl Acad. Sci. USA 94, 14602–14607 (1997).
27. Hogan, B. L. M., Beddington, R. S. P., Costantini, F. & Lacy, E. Manipulating the Mouse Embryo (Cold
Spring Harbor Laboratory Press, Cold Spring Harbor, 1994).
Probes for DNA analyses
We used the following probes: probe A, a 325-bp PCR fragment (bp 102,813–103,137;
AJ249895); probe C, a 0.9-kb EcoRI/HinCII fragment from EST AA592338 (ref. 6)
collinear with genomic DNA; probe D, (bp 121,834–122,862; AJ249895); and probe E
(bp 124,992–126,086; AJ249895).
Supplementary Information accompanies the paper on Nature’s website
(http://www.nature.com).
Probes for RNA analyses
Probe B, described as probe GFPAIR14, is 322 bp and protects 185 bp of Air RNA at the
Igf2r promoter. Probes C and D are the same as the DNA analyses probes. Probe F is 174 bp
and protects 47 bp (bp 126,181–126,227; AJ249895) immediately downstream from the
principal Air transcription start site6. Probe G (bp 126,086–126,293; AJ249895), described
as probe MlMs1 (refs 6,14), detects unspliced (207, 171 and 148 bp) Air RNA fragments.
This relatively large probe did not produce a clear signal for the spliced Air-T RNA
(predicted fragments of 112, 76 and 53 bp), which should have been recognized (see probe
K below). Probe H, described as probe RPA1 (ref. 6), protects 221 bp of Air RNA
(bp 85,250–85,029; AJ249895). Probe I, described as probe MS1B8 (ref. 6) protects
196 bp. Probe J is 173 bp and protects 155 bp of Air RNA (bp 123,233-123,387; AJ249895).
Probe K is 364 bp, contains the RT-PCR fragment made with probes RT and PX6R
(Supplementary Information) and protects 287 bp of spliced and 240 and 47 bp of
unspliced polyadenylated RNA. Probe L spans 558 bp (bp 32,032–32,590; M18818) of the
b-globin gene (including the polyadenylation signal) and protects 173 bp of
polyadenylated RNA (bp 32,032–32,204; M18818). Probe L extends 385 bp downstream
from the polyadenylation signal and the absence of protected fragments longer than
173 bp indicates use of the polyadenylation signal. The Aprt template is a 252-bp XhoI/
XbaI fragment (bp 2,165–2,417; M11310) and protects Aprt exon 3 (134 bp). For RNA
blots, we used the following probes: for Igf2r, complementary DNA exons 3–6; for
Slc22a2, bp 989–1,605 (AJ006036); for Slc22a3 bp 1–2,766 (AF078750); for Gapd,
complete cDNA.
Acknowledgements
We thank K. van Veen, K. van het Wout, P. Krimpenfort for help in generating mice;
S. Greven, T. Maidment and N. Bosnie for care of the mice; A. Berns, H. te Riele,
M. van Lohuizen, R. Beijersbergen, P. Borst and A. Frischauf for comments; and A. Berns
for help and encouragement. This research was supported by the Dutch Cancer Society
(KWF).
Correspondence and requests for materials should be addressed to D.P.B.
(e-mail: dbarlow@imb.oeaw.ac.at).
..............................................................
Structural basis for antagonist-
mediated recruitment of nuclear
co-repressors by PPARa
Received 27 September; accepted 4 December 2001.
1. Reik, W. & Walter, J. Genomic imprinting: parental influence on the genome. Nature Rev. Genet. 2,
21–32 (2001).
H. Eric Xu, Thomas B. Stanley, Valerie G. Montana, Millard H. Lambert,
Barry G. Shearer, Jeffery E. Cobb, David D. McKee, Cristin M. Galardi,
Kelli D. Plunket, Robert T. Nolte, Derek J. Parks, John T. Moore,
Steven A. Kliewer, Timothy M. Willson & Julie B. Stimmel
2. Sleutels, F. & Barlow, D. P. in Homology Effects (eds Wu, C.-t. & Dunlap, C.) (Academic, San Diego, in
the press).
3. Wutz, A. et al. Imprinted expression of the Igf2r gene depends on an intronic CpG island. Nature 389,
745–749 (1997).
4. Wutz, A. et al. Non-imprinted Igf2r expression decreases growth and rescues the Tme mutation in
mice. Development 128, 1881–1887 (2001).
Nuclear Receptor Discovery Research, GlaxoSmithKline, Research Triangle Park,
North Carolina 27709, USA
.............................................................................................................................................................................
5. Zwart, R., Sleutels, F., Wutz, A., Schinkel, A. H. & Barlow, D. P. Bidirectional action of the Igf2r imprint
control element on upstream and downstream imprinted genes. Genes Dev. 15, 2361–2366 (2001).
6. Lyle, R. et al. The imprinted antisense RNA at the Igf2r locus overlaps but does not imprint Mas1.
Nature Genet. 25, 19–21 (2000).
Repression of gene transcription by nuclear receptors is mediated
by interactions with co-repressor proteins such as SMRTand N-
CoR1,2, which in turn recruit histone deacetylases to the chro-
matin3–5. Aberrant interactions between nuclear receptors and
co-repressors contribute towards acute promyelocytic leukaemia
and thyroid hormone resistance syndrome6–8. The binding of co-
repressors to nuclear receptors occurs in the unliganded state,
and can be stabilized by antagonists9. Here we report the crystal
structure of a ternary complex containing the peroxisome pro-
liferator-activated receptor-a ligand-binding domain bound to
the antagonist GW6471 and a SMRT co-repressor motif. In this
structure, the co-repressor motif adopts a three-turn a-helix that
prevents the carboxy-terminal activation helix (AF-2) of the
receptor from assuming the active conformation. Binding of
7. Beechey, C. V., Cattanach, B. M. & Selley, R. L. Mouse Imprinting Data and References. MRC
Mammalian Genetics Unit [online] khttp://www.mgu.har.mrc.ac.uk/imprinting/imprinting.htmll
(2000).
8. Schmidt, J. V., Levorse, J. M. & Tilghman, S. M. Enhancer competition between H19 and Igf2 does not
mediate their imprinting. Proc. Natl Acad. Sci. USA 96, 9733–9738 (1999).
9. Hark, A. T. et al. CTCF mediates methylation-sensitive enhancer-blocking activity at the H19/Igf2
locus. Nature 405, 486–489 (2000).
10. Constancia, M. et al. Deletion of a silencer element in Igf2 results in loss of imprinting independent of
H19. Nature Genet. 26, 203–206 (2000).
11. Bell, A. C. & Felsenfeld, G. Methylation of a CTCF-dependent boundary controls imprinted
expression of the Igf2 gene. Nature 405, 482–485 (2000).
12. Reik, W. & Murrell, A. Genomic imprinting. Silence across the border. Nature 405, 408–409 (2000).
13. Wang, Z. Q., Fun, M. R., Barlow, D. P. & Wagner, E. F. Regulation of embryonic growth and lysosomal
targeting by the imprinted Igf2/Mpr gene. Nature 372, 464–467 (1994).
14. Sleutels, F. & Barlow, D. P. Investigation of elements sufficient to imprint the mouse Air promoter.
Mol. Cell. Biol. 21, 5008–5017. (2001).
© 2002 Macmillan Magazines Ltd
NATURE | VOL 415 | 14 FEBRUARY 2002 | www.nature.com
813

letters to nature
the co-repressor motif is further reinforced by the antagonist, such as N-CoR (nuclear co-repressor) and SMRT (silencing
which blocks the AF-2 helix from adopting the active position. mediator for retinoid and thyroid hormone receptors). Coactiva-
Biochemical analyses and structure-based mutagenesis indicate tors contain an LXXLL motif that forms two turns of a-helix and
that this mode of co-repressor binding is highly conserved across binds into a hydrophobic cleft on the surface of the receptor. In the
nuclear receptors.
PPARa/GW409544/SRC-1 structure, the SRC-1 coactivator motif is
The peroxisome proliferator-activated receptor-a (PPARa) regu- secured by a charge clamp between residues on the AF-2 helix and
lates hepatic fatty acid metabolism and mediates the effects of helix 3, similar to coactivator complexes with oestrogen receptor13,
fibrate lipid-lowering drugs10. We recently described the L-tyrosine thyroid receptor14, retinoid X receptor15 and PPARg15,16. As nuclear
analogue GW409544 as a potent PPARa agonist11. The crystal receptor antagonists decrease binding of coactivators and increase
structure of GW409544 bound to PPARa reveals that the carboxylic binding of co-repressors9, we examined the effect of GW6471 on the
acid group of GW409544 makes a direct hydrogen bond with Y464 interaction of PPARa with peptides derived from coactivator and
on the C-terminal AF-2 helix, and this interaction stabilizes the co-repressor motifs17 – 21. GW6471 disrupted the interactions
receptor in the active confirmation11. On the basis of these obser- between PPARa and coactivator motifs derived from SRC-1 or
vations, we designed a PPARa antagonist by substituting an ethyl CBP, but promoted the binding of the co-repressor motifs from
amide group for the acid group of GW409544 to disrupt the SMRTor N-CoR22 (Fig. 1c). The potency of GW6471 for displace-
hydrogen bond with Y464 (Fig. 1a). The resulting compound, ment of coactivators and recruitment of co-repressors matched the
GW6471, is a potent PPARa antagonist. In a cell-based reporter IC50 value for PPARa antagonism in cell-based assays. The binding
assay, GW6471 completely inhibited GW409544-induced activation affinity of the SMRT or the N-CoR motifs to the PPARa ligand-
of PPARa with a median inhibitory concentration (IC50) of 0.24 mM binding domain (LBD) was enhanced fivefold by GW6471 (Fig. 1d).
(Fig. 1b).
GW6471 also enhanced the interactions of PPARa with the N-CoR
The transcriptional activity of nuclear receptors is a function of or the SMRT motifs in a mammalian two-hybrid assay (Fig. 1e).
their interactions with coactivators such as SRC-1 (steroid receptor These results demonstrate that GW6471 induces a PPARa LBD
coactivator-1)12 or CBP (CREB-binding protein), or co-repressors conformation that interacts efficiently with co-repressors.
Both SMRT and N-CoR contain multiple receptor-interacting
motifs19–21. To understand the molecular basis of co-repressor
a
recruitment, we determined the crystal structure of a ternary
complex containing the PPARa LBD bound to GW6471 and a
peptide derived from the C-terminal receptor-interacting motif of
SMRT (Figs 2a and 3a). The refined structure contains four PPARa/
GW6471/SMRT complexes (see Methods and Table 1), and reveals a
clear electron density for the core co-repressor motif and the
antagonist GW6471 (Fig. 2b, c). The PPARa LBD contains 13
a-helices and 4 small b-strands that fold into a typical nuclear
receptor helical sandwich. The SMRT motif adopts a three-turn
a-helix and docks into a hydrophobic groove formed by helices 3,
30 , 4 and 5 (Fig. 2a). Superposition of the agonist- and antagonist-
bound PPARa (Fig. 2d) reveals that the co-repressor-binding site
partly overlaps with the coactivator-binding site. The additional
helical turn in the co-repressor motif extends into the space that is
left by the repositioning of the AF-2 helix, and prevents this helix
from folding back into its active conformation.
GW6471 adopts a U-shaped configuration and wraps around
C276 of helix 3 (Fig. 2c). The modified amide head group of
GW6471 adopts a configuration that is no longer able to form a
hydrogen bond with Y464 on the AF-2 helix. Instead, this larger
head group extends into the space that is normally occupied by the
side chain of Y464, and pushes the AF-2 helix out of the agonist-
bound position (Fig. 2e). Unlike the previously determined antago-
R
H
O
R = C02H, X = H
GW409544
(agonist)
HN
N
O
O
R = CH2NHC(O)Et
X = CF3
GW6471
(antagonist)
X
b
c
3
80
60
2
1
40
20
0
0
0.001 0.01 0.1 1.0 10
0.001 0.01 0.1 1.0 10
GW6471 (µM)
GW6471 (µM)
e
d
3
2
1
0
1.0
0.8
0.6
0.4
0.2
0
Table 1 Statistics of crystallographic data and structures
Crystals
.............................................................................................................................................................................
PPARa/SMRT complex
PPARa/SRC-1 complex
Space group
P212121
30.0–3.0
29,380
99.0
23.3 (2.0)
6.8
P21212
20.0–1.8
30,147
99.4
39.5 (5.2)
4.5
NC GW
N-CoR
NC GW
SMRT
1
10
20
˚
Resolution (A)
PPARα (µM)
Unique reflections
Completeness (%)
I/j
Figure 1 GW6471 recruits co-repressors to PPARa. a, Chemical structures of
GW409544 and GW6471. b, Effect of GW6471 on activation of PPARa in the presence
(filled circles) or absence (open circles) of 10 nM GW409544. c, Fluorescence energy
transfer assays for the effect of GW6471 on peptide binding to PPARa. SMRT, filled
circles; N-CoR, filled squares; CBP, filled triangles; SRC-1, open squares. d, Binding of
SMRT (squares) or N-CoR (circles) motifs to PPARa in the presence (filled) or absence
(open) of GW6471, as measured by fluorescence polarization. e, Effect of GW6471 on
binding of SMRT or N-coR motifs to PPARa in a mammalian two-hybrid assay. NC, no
compound; GW, 10 mM GW6471. The results are the average of three experiments, with
error bars showing s.d.
Rsym* (%)
Refinement statistics
Rfactor† (%)
Rfree (%)
25.7
29.0
21.9
26.7
˚
r.m.s.d. bond lengths (A)
r.m.s.d. bond angles (deg)
Total non-hydrogen atoms
0.008
1.956
9,615
0.007
1.561
2,659
.............................................................................................................................................................................
r.m.s.d. is the root-mean-square deviation from ideal geometry.
Rsym ¼ SjIavg ÿ Il j=SIl.
Rfactor ¼ SjFO ÿ FCj=SFO, where FO and FC are observed and calculated structure factors, Rfree
was calculated from a randomly chosen 10% of reflections excluded from refinement, and Rfactor
was calculated for the remaining 90% of reflections.
© 2002 Macmillan Magazines Ltd
814
NATURE | VOL 415 | 14 FEBRUARY 2002 | www.nature.com

letters to nature
nist-bound oestrogen-receptor structures13, the AF-2 helix does not prevent the coactivator LXXLL motif binding tightly to the receptor
occupy the coactivator-binding groove, but is loosely packed against in the absence of the AF-2 helix, like the co-repressor motif. This is
helix 3 (Fig. 2a, d). The PPARa/SMRT complex reveals that removal in agreement with previous studies, which found that replacing
of the AF-2 helix from its active position destroys the integrity of the residue þ5 and þ8 with leucines in the co-repressor motif results in
charge clamp but leaves sufficient space to accommodate the a non-functional co-repressor21 The third difference between co-
additional helical turn of the co-repressor motif.
repressor and coactivator binding is that the three-turn helix of the
Although there is overlap in the co-repressor- and coactivator- co-repressor motif deviates from a regular a-helix, which allows the
binding modes, there are numerous differences in their binding residues of L þ 1, I þ 5 and L þ 9 to align on the same face of the
interfaces. First, the three helical turns in the co-repressor motif helix (Figs 2f and 3b), and to make the core hydrophobic inter-
result in a larger interaction interface with PPARa than the two actions with the receptor. Besides these core interactions, the co-
helical turns of the coactivator motif. The SMRT LXXXIXXXL repressor motif makes additional interactions with the receptor
2
˚
motif covers 736 A of Connolly surface of PPARa, whereas the through peripheral residues composed by I þ 4, A þ 8, E þ 2 and
2
˚
SRC-1 LXXLL motif buries only 478 A . The larger co-repressor R þ 6, which are oriented at angles of ,908 to either side of the
interaction surface may account for the preferential binding of co- central interface (Fig. 3b). On one side, the side chains of I þ 4 and
repressor over coactivator to PPARa in the presence of GW6471. A þ 8 are partially exposed and make additional hydrophobic
Second, the co-repressor helix is anchored to PPARa by three contacts along one ridge of the binding groove. On the opposite
hydrogen bonds between the C-terminal carbonyls and the con- side, residues E þ 2 and R þ 6 form a strong intramolecular H-
served K292 from helix 3, which also caps the coactivator helix11. bond, which positions these residues so as to make a pair of
Compared with the coactivator helix, the co-repressor motif pivots intermolecular hydrogen bonds with K310 and N303 of helix 4.
,15 degrees about its C-terminal end into the void left behind by This hydrogen bond network is further strengthened by additional
displacement of the AF-2 helix, bringing the central isoleucine hydrophobic interactions between these two side chains and L302
˚
residue (I þ 5 in Fig. 3a) 2–3 A closer to helices 3, 4 and 5 (Fig. and V306, which form the ridge on this side of the groove.
2d, e). The unexpected binding mode of the co-repressor motif
We next examined the receptor–co-repressor interface by ala-
provides a structural explanation for the sequence differences nine-scanning mutations of the co-repressor motif (Fig. 3c).
between co-repressor and coactivator motifs. The first and the Replacement of any of the three core interface residues (L þ 1,
second leucines of the coactivator LXXLL motif are equivalent in I þ 5 and L þ 9) with alanine markedly reduced co-repressor
the positions of the central isoleucine (þ5) and the A þ 8 of the co- binding. Mutating the peripheral interface residues (E þ 2, I þ 4,
repressor motif in the superposition of their respective structures. R þ 6 and M þ 10) also decreased co-repressor binding. The
However, the larger leucine side chains at positions þ5 and þ8 interaction profile of these mutated co-repressor motifs was very
Figure 2 Structure of the PPARa/GW6471/SMRT complex. An overview of the ternary
complex. a, PPARa LBD, yellow; AF-2 helix, red; SMRT motif, white; GW6471, green
spheres. b, c, Electron density maps of the PPARa–SMRT interface and the antagonist
the PPARa/GW409544/SRC-1 complexes. PPARa LBD, yellow in SMRT complex or blue
in SRC-1 complex; SRC-1, purple; SMRT, white. e, Superposition of SMRT peptide and
GW6471 with the PPARa/GW409544/SRC-1 complex. The short arrow indicates
GW6471. The maps are calculated with FO coefficiency and contoured at 1j. The carbon interference between GW6471 and Y464; the long arrow indicates interference between
atoms of SMRT and GW6471 are in green. d, Overlay of the PPARa/GW6471/SMRT and SMRT and the active AF-2 helix. f, Interface between PPARa (yellow) and SMRT (white).
© 2002 Macmillan Magazines Ltd
NATURE | VOL 415 | 14 FEBRUARY 2002 | www.nature.com
815

letters to nature
a
b
123456789
L
I
K310 V306 N303 L302
Co-repressor motif
LXX-IXXX-
I
H
N-CoR1 (2040–2065)
SMRT1 (2124–2149)
N-CoR2 (2251–2275) GHSFADPASNLGLEDIIRKALMGSF
VPRTHRLITLADHICQIITQDFAR
PGVKGHQRVVTLAQHISEVITQDYTR
M
E
R
K
D
N
K292
SMRT2 (2329–2358) QAVQEHASTNMGLEAIIRKALMGKYDQWEE
I
A L
F297
Q305
Y314 V313 V284 T288,L309 V306
L
I
Coactivator motif
LXXLL
T285
V281
SRC1-2 (676–700)
CBP-1 (58–80 )
CPSSHSSLTERHKILHRLLQEGSPS
NLVPDAASKHKQLSELLRGGSGS
c
100
80
60
40
20
0
d
5
4
3
TRβ
PPARα
2
1
N
M L
E
A
I
I R K A L M K Y D Q W
Figure 3 Conservation of co-repressor binding to nuclear receptors. a, Co-repressor and c, Alanine scan of the SMRT2 motif. Binding of peptides to PPARa and TRb was
coactivator motifs. The underlined sequence was used in crystallography. b, Interactions measured by fluorescence polarization. IC50 values for the wild-type SMRT motif are 8.0
between the SMRT motif (green cylinder) and PPARa (boxed residues). Residue colours: ^ 3.4 uM and 6.7 ^ 2.3 uM for PPARa and TRb, respectively. d, Affinities of mutated
H-bond donors, blue; acceptors, red; hydrophobic contacts, grey (SMRT) and black
TRb LBDs for the SMRT2 motif as measured by fluorescence polarization. WT, wild type.
(PPARa). Arrows indicate hydrogen bonds; lines indicate van der Waals contacts.
Error bars are the s.d. of three experimemts.
Crystalization and data collection
similar to the one for TRb(thyroid hormone receptor-b) (Fig. 3c),
demonstrating a similar mode of co-repressor binding to these
receptors. This idea is further supported by structure-based
sequence alignment, which reveals that the residues involved in
co-repressor binding are highly conserved across the nuclear recep-
tor family (data not shown). Nine out of the fourteen residues that
contact the co-repressor domain in PPARa were previously
mutated in TRb19 – 21. Each of these mutations decreased the
interaction between TRb and the co-repressor motif. We mutated
the remaining five contact residues in TRb and showed that these
changes also result in decreases in co-repressor interaction (Fig. 3d).
These results suggest that the co-repressor interaction interface
observed in PPARa is conserved in other nuclear receptors.
The PPARa/GW6471/SMRT crystals were grown at room temperature in hanging drops
containing 1 ml of the above protein-ligand solutions and 1 ml of well buffer (5.5–9.6%
polyethylene glycol (PEG) 35K, 50 mM di-ammonium hydrogen citrate, pH 4.9, 50 mM
1,3 bis-tris-propane (BTP), pH 7.0, 10% MPD). Crystals grew to a size of 100–200 mm
within several weeks. Data collections were performed with crystals that were transiently
mixed with the well buffer containing an additional 10% 2-methyl-2,4-pentanediol
(MPD), and then flash frozen in liquid nitrogen.
ꢀ
ꢀ
The crystals formed in the P212121 space group, with a ¼ 103:4 A, b ¼ 112:8 A,
ꢀ
c ¼ 123:9 A. Each asymmetric unit contained four PPARa LBDs, four SMRT co-
repressor peptides and four GW6471 antagonist molecules with a solvent content of 50%.
Crystals were screened with a Rigaku R-Axis II detector and data were collected with a
MAR CCD (charge-coupled device) detector at 17-ID in the facilities of the Industrial
Macromolecular Crystallography Association (IMCA), at the Advanced Photon Source.
We processed the observed reflections with the Denzo and Scalepack programs in the
HKL2000 package24
.
The structure of the PPARa/GW6471/SMRT ternary complex
demonstrates that the co-repressor motif adopts an irregular three-
turn helix that fits tightly into a receptor groove, which also serves
as the coactivator-binding site. Nuclear receptors distinguish co-
repressors from coactivators by the length of their helical inter-
action motifs, which depends on the AF-2 conformation. In the
presence of an antagonist, the AF-2 helix is blocked from assuming
the active conformation, resulting in a larger pocket that can
accommodate the three-turn a-helix of the co-repressor
LXXXIXXXL motif. The high degree of conservation of the co-
repressor and coactivator interaction interfaces suggests that this is a
Structure determination and refinement
The PPARa/GW6471/SMRT structure was determined by molecular replacement
methods with the AmoRe program25 using a 1.8 A structure of a PPARa/SRC-1 complex as
˚
the initial model (Table 1). The best-fit solution generated with AmoRe gave a correlation
coefficiency of 36.1% and an R-factor of 46.9%. The phases from the molecular
replacement were extensively refined with solvent flattening, histogram matching, and
fourfold non-crystallographic symmetry (NCS) averaging as implemented in the DM
program26. The electron density maps were further improved with multiple crystal
averaging of three PPARa/SMRT data sets by the DMMULTI program. We applied a
negative b-factor to the data to enhance the feature of the electron density for the protein
side chains27. Multiple cycles of manual model building were performed with QUANTA
(Molecular Simulations). Structure refinements were proceeded with CNX28, using the
maximum likelihood target and NCS constraints, which was relaxed in the final stages of
refinement. The statistics of the structure and data sets are summarized in Table 1.
model that applies to many of the nuclear receptors.
A
Functional assays
Cell-based assays for GW6471 as an antagonist were performed in CV-1 cells using the
(UAS)5-tk-SPAP reporter and the human PPARa–GAL4 chimaeras as described
previously11. We carried out transfections using lipofectamine (Life Technologies) with a
b-galactosidase vector as a normalization control. The mammalian two-hybrid assays were
performed according the published procedures29. Transfection mixes included 8 ng
(UAS)5-tk-CAT reporter plasmid, 8 ng VP16-human PPARa expression plasmid, 2 ng of
either GAL4–N-CoR or GAL4–SMRT expression plasmid, 25 ng b-galactosidase
expression plasmid as internal control, and 35 ng carrier plasmid. The synthesis of
GW6471 and GW409544 will be described elsewhere.
Methods
Protein preparation
The human PPARa LBD (amino acids 198–468; GenBank accession number S74349)
tagged with MKKHHHHHHLVPRG (thrombin cleavage site underlined) was expressed
from the T7 promoter of the pRSETA vector (Invitrogen). The wild-type and mutated
LBDs of TRb were also expressed from pRSETA, but were tagged with
MKKGHHHHHHG. The proteins were purified as described23, and the tag was removed
from the PPARa LBD by incubation with thrombin (200:1 molar ratio) overnight at 4 8C.
The PPARa/GW6471/SMRT complex was prepared by addition of a twofold molar excess
of the SMRT2 peptide (sequence referred to in Fig. 3a) and fivefold molar excess of the
antagonist, concentrated to ,9.0 mg ml-1, and then aliquoted and stored at ,80 8C.
Binding assays
The effects of GW6471 on the interaction of coactivator and co-repressor peptides with
© 2002 Macmillan Magazines Ltd
816
NATURE | VOL 415 | 14 FEBRUARY 2002 | www.nature.com

letters to nature
recognition: structure of a six-finger transcription factor IIIA complex. Proc. Natl Acad. Sci. USA 95,
2938–2943 (1998).
PPARa were determined by chemical-mediated fluorescence energy transfer assays using
the AlphaScreen Technology from Packard BioScience30. The experiments were conducted
with 5 nM PPARa LBD of biotinylated peptide containing individual motifs (Fig. 3a),
following the manufacturer’s instructions for the hexahistidine detection kit in a buffer
containing 50 mM MOPS, pH 7.4, 50 mM NaF, 0.05 mM CHAPS, 0.1 mg ml-1 bovine
serum albumin, and 10 mM dithiothreitol (DTT). The binding signals were detected with
the increasing concentrations of GW6471, and the results from four repeated experiments
were normalized as a percentage of the binding in the absence of GW6471.
28. Brunger, A. T. et al. Crystallography & NMR system: A new software suite for macromolecular
structure determination. Acta Crystallogr. D 54, 905–921 (1998).
29. Oberfield, J. L. et al. A peroxisome proliferator-activated receptor gamma ligand inhibits adipocyte
differentiation. Proc. Natl Acad. Sci. USA 96, 6102–6106 (1999).
30. Ullman, E. F. et al. Luminescent oxygen channeling immunoassay: measurement of particle binding
kinetics by chemiluminescence. Proc. Natl Acad. Sci. USA 91, 5426–5430 (1994).
The effects of GW6471 on the affinity of the SMRT or N-CoR peptides with purified
PPARa LBD were determined by fluorescence polarization in a buffer containing 10 mM
HEPES, pH 7.4, 0.15 M NaCl, 3 mM EDTA, 0.005% polysorbate-20, 5 mM DTTand 2.5%
DMSO. Varied concentration of PPARa LBD in the presence or absence of 40 mM GW6471
were incubated at room temperature with 10 nM of a fluorescein-labelled peptide of N-
CoR2 or SMRT2 (Fig. 3a). The fluorescence polarization values for each concentration of
receptor were determined using a BMG PolarStar Galaxy fluorescence reader with 485 nm
excitation and 520 nm emission filters. The apparent dissociation constant (Kd) values
were determined by the binding curves derived from a nonlinear least-squares-fit of the
data for a simple 1:1 interaction.
Acknowledgements
We thank B. Wisely and R. Bledsoe for making co-repressor expression constructs in early
crystallization studies; W. Burkart and M. Moyer for protein sequencing; M. Iannone for
compound characterizations; G. Waitt and C. Wagner for mass spectroscopy and amino-
acid content analysis; and J. Chrzas and A. Howard for assistance with data collections at
17-ID. Use of the Advanced Photon Source was supported by the US Department of
Energy, Basic Energy Sciences, and Office of Science. We also thank L. Kuyper and
D. Eggleston for support and critical reading of the manuscript.
Mutational analysis of the SMRT co-repressor motif interaction with the PPARa and
TRb LBDs was also performed by fluorescence polarization. To determine the importance
of each amino acid in the SMRT motif for binding to nuclear receptors, SMRT peptides
with alanine substitution at each position were added to inhibit the binding of 1 mM TRb
LBD or 2 mM PPARa to the fluorescent N-CoR2 peptide. For the PPARa experiments we
added 10 mM GW6471. The inhibition curves were constructed and IC50 values were
determined by nonlinear least-squares-fit of the data to a simple 1:1 interaction.
Competing interests statement
The authors declare that they have no competing financial interests.
Correspondence and requests for materials should be addressed to H.E.X.
Received 8 November; accepted 12 December 2001.
(e-mail: ex11957@gsk.com). The Protein Data Bank code for the PPARa/GW6471/SMRT
complex and the PPARa/GW409544/SRC-1 complex is 1KQQ and 1K7L, respectively.
1. Horlein, A. J. et al. Ligand-independent repression by the thyroid hormone receptor mediated by a
nuclear receptor co-repressor. Nature 377, 397–404 (1995).
2. Chen, J. D. & Evans, R. M. A transcriptional co-repressor that interacts with nuclear hormone
receptors. Nature 377, 454–457 (1995).
3. Nagy, L. et al. Nuclear receptor repression mediated by a complex containing SMRT, mSin3A, and
histone deacetylase. Cell 89, 373–380 (1997).
..............................................................
4. Hassig, C. A., Fleischer, T. C., Billin, A. N., Schreiber, S. L. & Ayer, D. E. Histone deacetylase activity is
required for full transcriptional repression by mSin3A. Cell 89, 341–347 (1997).
5. Laherty, C. D. et al. Histone deacetylases associated with the mSin3 corepressor mediate mad
transcriptional repression. Cell 89, 349–356 (1997).
correction
6. Hong, S. H., David, G., Wong, C. W., Dejean, A. & Privalsky, M. L. SMRT corepressor interacts with
PLZF and with the PML-retinoic acid receptor alpha (RARa) and PLZF-RARa oncoproteins
associated with acute promyelocytic leukemia. Proc. Natl Acad. Sci. USA 94, 9028–9033 (1997).
7. Grignani, F. et al. Fusion proteins of the retinoic acid receptor-a recruit histone deacetylase in
promyelocytic leukaemia. Nature 391, 815–881 (1998).
Autonomic healing of polymer
composites
8. Yoh, S. M., Chatterjee, V. K. & Privalsky, M. L. Thyroid hormone resistance syndrome manifests as an
aberrant interaction between mutant T3 receptors and transcriptional corepressors. Mol. Endocrinol.
11, 470–480 (1997).
S. R. White, N. R. Sottos, P. H. Geubelle, J. S. Moore, M. R. Kessler,
S. R. Sriram, E. N. Brown & S. Viswanathan
9. Jackson, T. A. et al. The partial agonist activity of antagonist-occupied steroid receptors is controlled
by a novel hinge domain-binding coactivator L7/SPA and the corepressors N-CoR or SMRT. Mol.
Endocrinol. 11, 693–705 (1997).
Nature 409, 794–797 (2001).
.............................................................................................................................................................................
10. Issemann, I. & Green, S. Activation of a member of the steroid hormone receptor superfamily by
peroxisome proliferators. Nature 347, 645–650 (1990).
In this Letter, the middle infrared spectrum in Fig. 3b, correspond-
ing to an authentic sample of poly(DCPD) prepared with Grubbs’
catalyst and DCPD monomer, was a duplicate of the top spectrum
owing to a formatting error. The corrected spectra are shown
11. Xu, H. E. et al. Structural determinants of ligand binding selectivity between the peroxisome
proliferator-activated receptors. Proc. Natl Acad. Sci. USA 98, 13919–13924 (2001).
12. Onate, S. A., Tsai, S. Y., Tsai, M. J. & O’Malley, B. W. Sequence and characterization of a coactivator for
the steroid hormone receptor superfamily. Science 270, 1354–1357 (1995).
13. Shiau, A. K. et al. The structural basis of estrogen receptor/coactivator recognition and the
antagonism of this interaction by tamoxifen. Cell 95, 927–937 (1998).
below.
A
14. Darimont, B. D. et al. Structure and specificity of nuclear receptor–coactivator interactions. Genes
Dev. 12, 3343–3356 (1998).
15. Gampe, R. T.Jr et al. Asymmetry in the PPARg/RXRa crystal structure reveals the molecular basis of
heterodimerization among nuclear receptors. Mol. Cell 5, 545–555 (2000).
16. Nolte, R. T. et al. Ligand binding and co-activator assembly of the peroxisome proliferator-activated
receptor-g. Nature 395, 137–143 (1998).
17. Heery, D. M., Kalkhoven, E., Hoare, S. & Parker, M. G. A signature motif in transcriptional co-
activators mediates binding to nuclear receptors. Nature 387, 733–736 (1997).
18. Yang, W., Rachez, C. & Freedman, L. P. Discrete roles for peroxisome proliferator-activated receptor
gamma and retinoid X receptor in recruiting nuclear receptor coactivators. Mol. Cell Biol. 20,
8008–8017 (2000).
19. Nagy, L. et al. Mechanism of corepressor binding and release from nuclear hormone receptors. Genes
Dev. 13, 3209–3216 (1999).
20. Hu, X. & Lazar, M. A. The CoRNR motif controls the recruitment of corepressors by nuclear hormone
receptors. Nature 402, 93–96 (1999).
21. Perissi, V. et al. Molecular determinants of nuclear receptor–corepressor interaction. Genes Dev. 13,
3198–3208 (1999).
22. Zhou, G. et al. Nuclear receptors have distinct affinities for coactivators: characterization by
fluorescence resonance energy transfer. Mol. Endocrinol. 12, 1594–1604 (1998).
23. Xu, H. E. et al. Molecular recognition of fatty acids by peroxisome proliferator-activated receptors.
Mol. Cell 3, 397–403 (1999).
24. Otwinowski, Z. & Minor, W. in Macromolecular Crystallography (eds Carter, J. C. W. & Sweet, R. M.)
307–326 (Academic, New York, 1997).
25. Navaza, J., Gover, S. & Wolf, W. in Molecular Replacement: Proceedings of the CCP4 Study Weekend (ed.
Dodson, E. J.) 87–90 (SERC, Daresbury, 1992).
1,400 1,200 1,000 800 600
Wavenumber (cm–1
)
26. Cowtan, K. in Joint CCP4 and ESF-EACBM Newsletter on Protein Crystallography 31, 34–38 (1994).
27. Nolte, R. T., Conlin, R. M., Harrison, S. C. & Brown, R. S. Differing roles for zinc fingers in DNA
© 2002 Macmillan Magazines Ltd
NATURE | VOL 415 | 14 FEBRUARY 2002 | www.nature.com
817