Helix mimetics as inhibitors of the interaction of the estrogen receptor with coactivator peptides

Article abstract of DOI:10.1002/anie.200700657

The short and curlies: A new α-helix mimetic based on a pyridylpyridone scaffold has been developed to bind to the estrogen receptor (ER) by mimicking the key leucine side chains of coactivator LXXLL boxes (L= leucine, X = any amino acid). These inhibitor

Full text of DOI:10.1002/anie.200700657

Angewandte
Chemie
DOI: 10.1002/anie.200700657
a Helix Mimetics
Helix Mimetics as Inhibitors of the Interaction of the Estrogen
Receptor with Coactivator Peptides**
Jorge Becerril and Andrew D. Hamilton*
The estrogen receptor (ER) is a transcription factor that
mediates the expression of estrogen-activated genes. The ER
has been associated with a variety of diseases including breast
cancer, osteoporosis, and cardiovascular disease and is there-
fore an important target for therapeutic intervention.[1]
Binding of an estrogen molecule to the ligand binding
domain (LBD) of the ER ultimately leads to interaction
with specific DNA promoters and recruitment of coactivator
proteins. These coactivator proteins mediate the assembly of
the transcriptional machinery and are therefore essential for
Figure 1. a) X-ray structure of ER LBD (blue) bound to GRIP1 coactiva-
tor peptide (red) and an agonist molecule (yellow). b) Magnified view
of the GRIP1 peptide (green) on the surface of the ER (red=hydropho-
bic, blue=hydrophilic).[4]
expression of the ER-regulated genes. Traditionally, inhib-
ition of the ER has been attempted by using antagonist
molecules that bind to the LBD and trigger a conformational
change that prevents the ER from recruiting the coactivator
proteins.[2] An alternative and underexploited approach
involves the small-molecule inhibition of the interaction
between the estrogen-activated ER and the coactivator
proteins.[3a,b] Importantly, it has been shown that an analogous
strategy can be used to target other nuclear receptors.[3c]
The coactivator proteins possess multiple copies of a
conserved LXXLL motif, also known as nuclear receptor box
(where L is leucine and X is any amino acid including
leucine). Extensive studies have shown that this short LXXLL
sequence is necessary and sufficient for binding to the ER.[5]
The X-ray structure of the ligand-bound ER and a fragment
of the coactivator peptide, glutamate receptor interacting
protein 1(GRIP1; Figure 1), shows that the LXXLL peptide
adopts an a-helical conformation where the leucine side
chains in positions i and i + 4 are projected into a hydrophobic
groove on the ER surface while that in the i + 3 position
projects into a hydrophobic pocket (Figure 1b).[4] Addition-
ally, the crystal structure suggests that interactions between
the coactivator peptide backbone and the charged residues
that flank the binding groove on the ER further stabilize the
complex.
However, there have been only two reports of small-molecule
inhibitors with only one of them (with an inhibition constant
(Ki) of 29 mm) designed to bind to this surface region of the
ER and block the coactivatorꢀs approach.[3]
We have previously reported a broad strategy for the
disruption of a-helix–protein interactions that involves the
design of rigid scaffolds from which groups mimicking the
surface functionality of an a helix can be projected.[7] For
example, 2,3’,3’’-trisubstituted terphenyls can mimic the i, i +
4, and i + 7 residues of two turns of an a helix and lead to
potent inhibitors of protein–helix contacts such as those
between Bcl-xL/Bak and MDM2/p53.[8] In the case of the
coactivator LXXLL motif, a modified approach is needed to
incorporate the features of the i + 3 leucine. We and others
have shown that this can be simply achieved by placing a
second ortho substituent on a biaryl scaffold.[9] Separation of
the elements of the i + 3 side chain by a single methylene (as
in Figure 2a) allows the adoption of a relative side chain
conformation on the biaryl that closely mimics the distances
and angular projections of the i, i + 3, and i + 4 groups of an
a helix.
In the search for inhibitors of this interaction, various
short peptide derivatives based on the LXXLL sequence have
been shown to disrupt the ER–coactivator interaction.[6]
We synthesized a series of substituted pyridylpyridone
derivatives as shown in Figure 2b, with the expectation that
the bis-heteroaryl scaffold will have improved water solubility
[*] J. Becerril, Prof. A. D. Hamilton
Department of Chemistry
Yale University
P. O. Box 208107, New Haven, CT 06520-8107 (USA)
Fax: (+1)203-432-6144
E-mail: andrew.hamilton@yale.edu
[**] We thank the National Institutes of Health (GM69850) for financial
support of this work and Prof. John Katzenellenbogen for valuable
conversations in the early stages of this work.
Figure 2. a) Trisubstituted biaryl scaffold mimicking the i, i+3, and
i+4 residues of an a helix. b) Structure of the pyridylpyridone deriva-
tives 1–7.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2007, 46, 4471 –4473
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4471

Communications
and bioavailability while allowing the ready introduction of
coactivator interaction with low micromolar activity
(Table 1). As expected, all the compounds were soluble
under assay conditions. Compound 2 most closely mimics the
LXXLL sequence and showed good binding with Ki = 16 mm.
substituents into the 2-pyridyl and 1,5-pyridone positions.
Compounds 1–7 were synthesized through radical monoalky-
lation of p-cyanopyridine followed by the addition of a
Grignard reagent to obtain the alkylpyridylketone deriva-
tive.[10] Imine formation followed by reaction with benzyloxy-
acetyl chloride and subsequent cyclization yielded the appro-
priately substituted pyridylpyridone.[11]
Table 1: Results of the fluorescence polarization assay.[a]
Compound
Ki [mm]
Compound
Ki [mm]
SRC-1 NR II
1.0 (0.3)
34 (3)
16 (3)
4
5
6
7
>50
>50
The crystal structure of 5 shows a nonplanar conformation
(Figure 3a) with an aryl–aryl dihedral angle of 828 and
distances of 5.6, 5.4, and 5.7 Š between the atoms that mimic
1
2
3
6.5 (0.5)
4.2 (0.5)
9.4 (2.0)
[a] Active compounds were tested in triplicate in at least three
independent experiments. The values in brackets are the corresponding
standard deviations.
Shortening of the i + 3-mimicking N-alkyl side chain by one
methylene, as in 1, led to a twofold decrease in affinity,
supporting the conformational requirements for optimal i + 3
and i + 4 mimicry. Compounds 3 and 6 containing one and two
benzyl groups, respectively, in the i and i + 3 positions were
based on previously reported peptidomimetics and gave Ki
values of 9.4 and 6.5 mm, respectively.[6d] Introduction of a
naphthyl group in 7 led to a further improvement with a Ki
value of 4.2 mm. Finally, steric constrain of the hydrophobic
pockets on the ER could explain the weak affinity of 4 and 5
in which one or both of the i and the i + 3 groups were
converted to bulkier tert-butyl and neopentyl groups.
Inhibition of coactivator binding by a small molecule
could involve either direct competition with the coactivator
peptide for the ER surface or binding of the small molecule to
the ligand binding site as an antagonist, deactivating the ER
and preventing coactivator recruitment.
To rule out the latter mechanism, a competitive radio-
ligand assay with [3H]estradiol was carried out to specifically
measure the affinity of 3 for the estradiol binding site.
According to the radioligand assay, 3 binds to this site with an
affinity less than 0.001% of the affinity of 17 b-estradiol. As
the concentration of estradiol in the FP assays is kept constant
at 2 mm, we estimated that compound 3 should have a Ki value
of 300 mm, which is six orders of magnitude higher than the Ki
value found in the FP assay (9.4 mm). From this, we conclude
that our helix mimetics prevent the coactivator association
through direct competition for its binding site on the ER
surface.
Figure 3. a) X-ray structure of 5. b) Stereoview of the X-ray structure of
5 superimposed on the a-helical LXXLL motif of the GRIP1 peptide.
Hydrogen atoms and nonrelevant amino acid side chains have been
omitted for clarity.
the b carbons of the key amino acid residues. These distances
are similar to those found in a-helical LXXLL motifs. The b-
carbon atoms of the i, i + 3, and i + 4 leucine residues of the
GRIP1peptide superimposed with the corresponding carbon
atoms of 5 showed good matching with a root-mean-square
(rms) deviation of 0.36 Š (Figure 3b).
To test the effectiveness of these derivatives in blocking
the ER–coactivator interaction, we used a fluorescent polar-
ization (FP) assay based on rhodamine-labeled peptide D22
(Rho-LPYEGSLLLKLLRAPVEEV-COOH), which con-
tains a single LXXLL motif.[5b] Displacement of this fluores-
cent peptide from the surface of estradiol-activated ER leads
to a decrease in the polarization value. A control peptide that
mirrors the second NR box of the coactivator protein SRC-1
(SRC-1NR II) gave a Ki of 0.95 mm (Figure 4), which is
comparable with the reported value of 1 mm.[3,12]
Evaluation of the binding affinities of the pyridylpyridone
compounds showed that most were able to inhibit the ER–
In conclusion, a new a-helix mimetic based on a pyr-
idylpyridone scaffold was designed to mimic the surface
functionality of an a-helical LXXLL motif. Results from FP
indicate that most compounds bind with Ki values in the low
micromolar range. The most potent inhibitors had compara-
ble affinity to that of the control SRC-1NR II; a peptide that
mirrors the second LXXLL motif of the natural SRC-1
coactivator. Extensive structure–activity relationship studies
to improve the affinity of this scaffold as well as other
structurally related molecules are currently underway.
Figure 4. Titration curves for compounds 1 (red), 5 (blue), 6 (green),
and control peptide SRC-1 NR II (pink).
Received: February 12, 2007
Published online: May 9, 2007
4472 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 4471 –4473

Angewandte
Chemie
Geistlinger, R. K. Guy, J. Am. Chem. Soc. 2003, 125, 6852 – 6853;
Keywords: amino acids · drug design · inhibitors ·
protein interactions · proteomimetics
.
c) A. K. Galande, K. S. Bramlett, J. O. Trent, T. P. Burris, J. L.
Wittliff, A. F. Spatola, ChemBioChem 2005, 6, 1991 – 1998;
d) T. R. Geistlinger, A. C. McReynolds, R. K. Guy, Chem.
Biol. 2004, 11, 273 – 281.
[1] J. E. Darnell, Nat. Rev. Cancer 2002, 2, 740 – 749.
[2] B. R. Henke, D. Heyer, Curr. Opin. Drug DiscoveryDev. 2005, 8,
437 – 448.
[7] a) B. P. Orner, J. T. Ernst, A. D. Hamilton, J. Am. Chem. Soc.
2001, 123, 5382 – 5383; b) J. T. Ernst, J. Becerril, H. S. Park, H.
Yin, A. D. Hamilton, Angew. Chem. 2003, 115, 553 – 557; Angew.
Chem. Int. Ed. 2003, 42, 535 – 539; c) H. Yin, A. D. Hamilton,
Bioorg. Med. Chem. Lett. 2004, 14, 1375 – 1379.
[8] a) O. Kutzki, H. S. Park, J. T. Ernst, B. P. Orner, H. Yin, A. D.
Hamilton, J. Am. Chem. Soc. 2002, 124, 11838 – 11839; b) H.
Yin, G. I. Lee, H. S. Park, G. A. Payne, J. M. Rodriguez, S. M.
Sebti, A. D. Hamilton, Angew. Chem. 2005, 117, 2764 – 2767;
Angew. Chem. Int. Ed. 2005, 44, 2704 – 2707.
[9] a) E. Jacoby, Bioorg. Med. Chem. Lett. 2002, 12, 891– 893;
b) I. C. Kim, A. D. Hamilton, Org. Lett. 2006, 8, 1751 – 1754;
c) D. C. Horwell, W. Howson, W. P. Nolan, G. S. Ratcliffe, D. C.
Rees, H. M. G. Willems, Tetrahedron 1995, 51, 203 – 216.
[10] F. Fontana, F. Minisci, M. C. N. Barbosa, E. Vismara, Tetrahe-
dron 1990, 46, 2525 – 2538.
[3] a) A. L. Rodriguez, A. Tamrazi, M. L. Collins, J. A. Katzenel-
lenbogen, J. Med. Chem. 2004, 47, 600 – 611; b) D. L. Shao, T. J.
Berrodin, E. Manas, D. Hauze, R. Powers, A. Bapat, D. I.
Gonder, R. C. Winneker, D. E. Frail, J. Steroid Biochem. Mol.
Biol. 2004, 88, 351– 360; c) L. A. Arnold, E. EstØbanez-Perpiæµ,
M. Togashi, N. Jouravel, A. Shelat, A. C. McReynolds, E. Mar, P.
Nguye, J. D. Baxter, R. J. Fletterick, P. Webb, R. K. Guy, J. Biol.
Chem. 2005, 280, 43048 – 43055.
[4] a) A. K. Shiau, D. Barstad, P. M. Loria, L. Cheng, P. J. Kushner,
D. A. Agard, G. L. Greene, Cell 1998, 95, 927 – 937; Protein Data
Base ID: 3erd; b) Figure 1a was generated by using PyMol
software. Delano, San Carlos, CA, USA.
[5] a) D. M. Heery, E. Kalkhoven, S. Hoare, M. G. Parker, Nature
1997, 387, 733 – 736; b) C. Y. Chang, J. D. Norris, H. Gron, L. A.
Paige, P. T. Hamilton, D. J. Kenan, D. Fowlkes, D. P. McDonnell,
Mol. Cell. Biol. 1999, 19, 8226 – 8239.
[6] a) A. M. Leduc, J. O. Trent, J. L. Wittliff, K. S. Bramlett, S. L.
Briggs, N. Y. Chirgadze, Y. Wang, T. P. Burris, A. F. Spatola,
Proc. Natl. Acad. Sci. USA 2003, 100, 11273 – 11278; b) T. R.
[11] I. Collins et al., J. Med. Chem. 2002, 45, 1887 – 1900. See the
Supporting Information.
[12] A. Tamrazi, K. E. Carlson, J. R. Daniels, K. M. Hurth, J. A.
Katzenellenbogen, Mol. Endocrinol. 2002, 16, 2706 – 2719.
Angew. Chem. Int. Ed. 2007, 46, 4471 –4473
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4473