Uracils as potent antagonists of the human gonadotropin-releasing hormone receptor without atropisomers

Article abstract of DOI:10.1016/j.bmcl.2005.03.057

Uracil derivatives were designed and synthesized to avoid atropisomers observed in the 6-methyluracils as antagonists of the human GnRH receptor. Optimization at the 1- and 5-positions of the uracil resulted in potent compounds such as 24 (K_i = 0.45 nM).

Full text of DOI:10.1016/j.bmcl.2005.03.057

Bioorganic & Medicinal Chemistry Letters 15 (2005) 2519–2522
Uracils as potent antagonists of the human
gonadotropin-releasing hormone receptor without atropisomers
Zhiqiang Guo,a Yongsheng Chen,a Charles Q. Huang,a Timothy D. Gross,a
Joseph Pontillo,a Martin W. Rowbottom,a John Saunders,a Scott Struthers,b
Fabio C. Tucci,a Qiu Xie,b Warren Wade,a Yun-Fei Zhu,a Dongpei Wua and Chen Chena,*
aDepartment of Medicinal Chemistry, Neurocrine Biosciences, Inc. 12790 El Camino Real, San Diego, CA 92130, USA
bDepartment of Exploratory Research, Neurocrine Biosciences, Inc. 12790 El Camino Real, San Diego, CA 92130, USA
Received 27 January 2005; revised 14 March 2005; accepted 16 March 2005
Available online 12 April 2005
Abstract—Uracil derivatives were designed and synthesized to avoid atropisomers observed in the 6-methyluracils as antagonists
of the human GnRH receptor. Optimization at the 1- and 5-positions of the uracil resulted in potent compounds such as 24
(Ki = 0.45 nM).
Ó 2005 Elsevier Ltd. All rights reserved.
Gonadotropin-releasing hormone (GnRH), also known
as luteinizing hormone-releasing hormone (LH-RH), is
a linear decapeptide amide, pGlu-His-Trp-Ser-Tyr-
Gly-Leu-Arg-Pro-Gly-NH2. GnRH exerts its action by
activating its cell surface receptor, a member of the class
A G-protein-coupled receptor superfamily, in the
pituitary to stimulate the secretion of the gonadotro-
pins-luteinizing hormone (LH) and follicle-stimulating
hormone (FSH).1 These two hormones then act on the
reproductive organs where they participate in the regu-
lation of steroid production, gametogenesis, and ovul-
ation.2 Several disease conditions, such as endometriosis,
uterine fibroids, and prostate cancer, can be treated by
suppression of the pituitary–gonadal axis. GnRH pep-
tide super-agonists, represented by leuprorelin,3 based
on a receptor down-regulation mechanism, are currently
used clinically in the treatment of these conditions.4 Re-
cently, clinical evidence has shown that peptidic GnRH
antagonists could directly lower gonadal sex hormone
levels to alleviate these disease symptoms without the
concomitant flare effect caused by super-agonists.5 An
orally bioavailable small molecule GnRH antagonist
will have the advantage of flexible dosing and titration
of drug concentrations, which may provide novel clini-
cal management options.6
Previously, we reported that 6-methyluracils such as 1
(NBI 42902, Fig. 1) are potent and orally active antago-
nists of the human gonadotropin-releasing hormone
receptor.7 Compound 1 has been developed into clinical
evaluation for potential treatment of GnRH-related dis-
eases such as endometriosis and uterine fibroids. While
the methyl group at the 6-position of uracil 1 provides
the environment for an orthogonal relationship between
the 5-aryl group and the uracil core, which is an impor-
tant feature of pharmacophore for this class of com-
pounds, it causes the carbon–carbon bond connecting
the two aromatic rings to rotate slowly, as evidenced
by both NMR and HPLC analyses for compound 1
and its close analogues. The interchange half-life for
the two atropisomers (rotational diastereoisomers) of 1
is estimated to be 35 min at 25 °C.8 While these inter-
OMe
OMe
F
F
O
1 N
F
O
N
F
3
5
N
N
6
H2N
H2N
O
R
O
X2
F
S-2: X2 = F
S-9: X2 = CF3
1: R = Me
2: R = H
Keywords: GnRH; antagonist; Uracil; Atropisomer.
*
Corresponding author. Tel.: +1 858 617 7600; fax: +1 858 617
7967; e-mail: cchen@neurocrine.com
Figure 1. Structure of uracil analogues.
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.03.057

2520
Z. Guo et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2519–2522
changeable atropisomers with relative short half-lives
equilibrate in solution and seem to be not an issue for
development,9 the atropisomerism of these compounds
could cause some difficulties in term of manufacturing,
such as batch to batch reproduciblility.
class of compounds, we conducted a study to modify
the 1- and 5-substituents on the uracil ring in order to
improve the potency. Here we report the results of this
des-methyl series of compounds as GnRH receptor
antagonists without atropisomers.
In our efforts to avoid the atropisomerism caused by
6-methyluracils, we synthesized a close analogue of 1 by
deleting the 6-methyl group (Fig. 1). Thus, compound
2 exhibited a Ki value of 5.3 nM, which was about 10-
fold less potent than 1 (Ki = 0.56 nM). However, 2 does
not possess the atropisomeric property, as demonstrated
by a single set of proton and carbon NMR signals.
Compound 2 was also much easier to be crystallized
than 1. Interestingly, like the 6-methyluracil 1, the crys-
tal structure of 2 also displayed an orthogonal relation-
ship between the 5-aryl group and the uracil core in solid
state (Fig. 2). Since the 6-methyl group on the uracil also
dictates the conformation of the 1-benzyl group, which
is another important pharmacophoric element in this
The synthesis of these uracils started from the commer-
cially available 5-bromouracil 27, which was alkylated
with different substituted benzyl bromides, promoted
by N,O-bis(trimethylsilyl)acetamide (BSA), to provide
intermediates 28. A Mitsunobu coupling reaction of
28 with (R)-N-Boc-phenylglycinol (DEAD/PPh3/THF)
afforded the 1,3-dialkylated uracils 29. Suzuki coupling
reactions of 29 with various arylboronic acids
(Pd[PPh3]4/Na2CO3/H2O–dioxane/reflux) afforded the
final products 2–4, 7–9, and 16–26, after Boc-deprotec-
tion with trifluoroacetic acid (Scheme 1).10
Nucleophilic displacement of the ortho-fluorine of inter-
mediates 30 with alkylthiols promoted by a base such
as sodium hydride afforded the sulfides 31. The sulfide
analogue 10 was obtained from Boc-deprotection
with trifluoroacetic acid. Oxidation of 31 provided the
corresponding sulfones 5, 6, and 11–13 in very good
yields after deprotection (Scheme 2). Sulfonamides 14
and 15 were synthesized from 31 by oxidation with
N-chlorosuccinamide in dichloromethane, followed by
a coupling reaction with either dimethylamine or mor-
pholine under basic conditions, and then TFA deprotec-
tion (Scheme 3).
The synthesized compounds were tested for their ability
to displace a radiolabeled GnRH peptide in the GnRH
receptor binding assay as previously reported,7 and the
results are listed in Table 1. Initial structure–activity
relationship study of the non-methyl uracil analogues
2–4 and 7 suggests that substitutions on the 5-phenyl
ring are very important for high affinity. Thus, in com-
parison with 2 (Ki = 5.3 nM), the 2-fluoro-3-hydroxy-
phenyl derivative 3 had a dramatic reduction in
binding affinity (Ki = 660 nM). On the other hand, the
2-chloro-3-methoxyphenyl analogue 7 increased the
potency about 5-fold (Ki = 1.0 nM). As expected,
removal of the 3-methoxy group of 7 resulted in about
4-fold loss in affinity (4, Ki = 6.0 nM).
The substituted benzyl group at the 1-position of the ura-
cil is also crucial for high affinity binding of these com-
pounds to the GnRH receptor.11 Based on our receptor
Figure 2. ORTEP drawing of the X-ray structure of compound 2.16
Y
O
O
N
O
O
N
Br
Br
X2
Br
X2
b
a
c
HN
N
HN
N
BocHN
HN
P
O
N
O
X2
O
N
H
O
27
X1
X1
X1
29
30: P = Boc
2-4,7-9,16-26: P = H
28
Scheme 1. Reagents and conditions: (a) BSA/ArCH2Br/ACN/80 °C, 85%; (b) (R)-PhCH(NHBoc)CH2OH/DEAD/Ph3P/THF, 91%; (c) ArB(OH)2/
Pd(0)/Na2CO3; then TFA/CH2Cl2, 60–95%.

Z. Guo et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2519–2522
2521
Y
Y
Y
O
N
F
O
N
S
O
a
b
N
N
N
BocHN
H2N
HN
P
X2
O
X2
O
X2
O
N
S
R1
R1
30 (X2 = F or CF3)
31: P = Boc
10: P = H
O O
5-6, 11-13
Scheme 2. Reagents and conditions: (a) R1SNa/DMF/100 °C, 98% (R1 = Me); (b) oxone/acetone/H2O, 90%; then TFA/CH2Cl2, 90% (R1 = Me).
Table 1. SAR of uracil analogues with different 1-benzyl and 5-aryl
groups
OMe
OMe
F
F
O
N
S
O
N
S
Y
O
N
N
H2N
BocHN
a
O
F
O
F
N
H2N
R3
N
O
N
X2
Bn
R2
14, 15
O O
31
1
2-26
Compd X1
X2
Y
Ki (nM)a
Scheme 3. Reagents and conditions: (a) i. NCS/DCM; ii. R2R3NH; iii.
TFA/CH2Cl2, 40–50% overall yields.
1
0.56
5.3
570
660
6.0
1.3
3100
1.0
1.4
0.64
55
2
F
F
F
F
F
F
F
F
F
2-F–3-OMe
2-F–3-OMe
2-F–3-OH
2-Cl
S-2
3
modeling and mutagenesis studies, this group may bind
in a pocket formed by three tyrosine residues (Tyr-283,
284, and 290) of the transmembrane domain six.7,12
These results suggest an electron-deficient aromatic ring
will be favored at this position for a strong interaction.13
This is further supported by the affinity of 4 and 5. The
methylsulfone 5 had a Ki value of 1.3 nM, which is about
5-fold more potent than its fluorine analogue 4
(Ki = 6.0 nM). The increase in potency of 5 over 4 could
be caused by a stronger electron-withdrawing effect of
the methylsulfonyl group, and hence increasing the
aromatic interaction of the substituted benzyl group
with the tyrosine cluster. It is also possible that the large
sized sulfone alters the benzyl group to a more favored
position. Interestingly, the 2-trifluoromethyl-6-methyl-
sulfonylbenzyl analogue 6 (Ki = 3,100 nM) exhibited
a large reduction in binding affinity from 5.
4
5
SO2Me
2-Cl
2-Cl
6
CF3 SO2Me
7
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
Cl
2-Cl–3-OMe
2-F–3-OMe
2-F–3-OMe
2-F–3-OMe
2-F–3-OMe
2-F–3-OMe
2-F–3-OMe
8
9
CF3
CF3
S-9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
MeS
SO2Me
SO2Pr-i
19
0.90
30
SO2CH2CH2OH 2-F–3-OMe
SO2NMe2 2-F–3-OMe
SO2morphlinyl-1 2-F–3-OMe
9.1
120
1400
7.8
8.0
1.2
4.0
18
CF3
CF3
CF3
CF3
CF3
CF3
CF3
CF3
CF3
CF3
CF3
H
2-F
2-Cl
2-CF3
2-F–3-OH
2-F–3-OEt
1.8
To further explore the role of the 1-benzyl group, a ser-
ies of 2-fluorobenzyl with a different substituent at the 6-
position of the phenyl ring were synthesized and tested
2-F–3-OCH2CF3 8.2
2-F–3-OPr-i
2-Cl–3-OMe
2-Cl–4-Me
2-Cl–4-CF3
6.3
0.45
2.6
(compounds 8–15). The chlorinated derivative
8
(Ki = 1.4 nM) was about 4-fold better than the corre-
sponding fluorinated analogue 2. The trifluoromethyl
9 exhibited further improvement in potency (Ki =
0.64 nM). While 10 with a weakly electron-donat-
ing methylsulfide had a Ki of 19 nM, which is about
4-fold less active than the fluoro analogue 2, the corre-
sponding sulfone 11 exhibited over 20-fold increase in
affinity from 10. The binding affinity of the isopropyl-
sulfone 12 decreased 33-fold from the methyl analogue
11. This result demonstrates a size-limitation at this
position, which is further supported by compounds
13–15, which were all less potent than 11.
200
a Key compounds were tested for two or more times in independent
experiments. Ki values were highly reproducible with an average
standard deviation of less than 45% for replicate Ki determinations.
Having identified the optimally substituted benzyl group
at the 1-position, we next studied the 5-phenyl group
with different substitutions (compounds 16–27). The
un-substituted phenyl compound 16 exhibited a Ki of
7.8 nM, while the 2-fluoro analogue 17 possessed a sim-
ilar Ki value (8.0 nM). These results are quite interesting,

2522
Z. Guo et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2519–2522
since in the 6-methyluracil series, the 2-fluorine group
causes about 4-fold increase in binding affinity. This
may indicate that the small fluorine atom, without the
6-methyl group at the uracil ring, has little effect on ori-
entating the 5-phenyl ring. The larger chlorine atom
(compound 18, Ki = 1.2 nM), however, increased the
binding affinity about 6-fold over 16. On the other hand,
the trifluoromethyl group (19, Ki = 4.0 nM) had mini-
mal impact. The hydroxy analogue 20 (Ki = 18 nM)
was slightly less active than its parent 17 (Ki = 8.0 nM),
but much less potent than the methoxy compound 9
(Ki = 0.64 nM), suggesting that the methoxy group is
strongly favored. Further increasing the bulk over the
methoxy group led to progressive decrease in binding
affinity (compound 21–23). Although the 2-chloro-3-
methoxylphenyl compound 24 (Ki = 0.45 nM) possessed
similar binding affinity to 9, 24 (IC50 = 0.53 nM) was
much more potent than 9 (IC50 = 7.2 nM) in the func-
tional IP3 turnover assay.14 While adding a methyl
group at the 4-position on the 5-phenyl ring did not im-
prove binding affinity (25, Ki = 2.6 nM), incorporating a
trifluoromethyl group actually dramatically reduced its
potency (26, Ki = 200 nM). These results may imply that
an electron-deficient phenyl ring is unfavored at this
position.
structure–activity relationship studies at the 1- and
5-positions of the uracil core resulted in re-optimized
compounds with subnanomolar potency (i.e., 24,
Ki = 0.45 nM). Importantly, these compounds do not
possess atropisomeric property.
Acknowledgements
This work was supported, in part, by National Institutes
of Health grants 1-R43-HD38625-01 and 2-R44-
HD38625-02.
References and notes
1. (a) Sealfon, S. C.; Weinstein, H.; Millar, R. P. Endocr.
Rev. 1997, 18, 180; (b) Millar, B. P.; Lu, Z.-J.; Pawson, A.
J.; Flanagan, C. A.; Morgan, K.; Maudsley, S. R. Endocr.
Rev. 2004, 25, 235.
2. Cheng, K. W.; Leung, P. C. K. Can. J. Physiol. Pharma-
col. 2000, 78, 1029.
3. Cottreau, C. M.; Ness, R. B.; Modugno, F.; Allen, G. O.;
Goodman, M. T. Clin. Cancer Res. 2003, 9, 5142.
4. Karten, M. J.; Rivier, J. E. Endocr. Rev. 1986, 7, 44.
5. Huirne, J. A.; Lambalk, C. B. Lancet 2001, 358, 1793.
6. For recent review articles on small molecule GnRH
antagonists, see: (a) Tucci, F. C.; Chen, C. Curr. Opin.
Drug Discovery Dev. 2004, 7, 832; (b) Zhu, Y.-F.; Chen,
C.; Struthers, R. S. Non-peptide GnRH antagonists.
Annu. Rep. Med. Chem. 2004, 39, 99.
7. Tucci, F. C.; Zhu, Y.-F.; Struthers, R. S.; Guo, Z.; Gross,
T. D.; Rowbottom, M. R.; Acevedo, O.; Gao, G.;
Saunders, J.; Xie, Q.; Reinhart, G. J.; Liu, X.-J.; Ling,
N.; Bonneville, A. K. L.; Chen, T.-K.; Bozigian, H.; Chen,
C. J. Med. Chem. 2005, 48, 1169.
Similar to the 6-methyluracil series, compounds derived
from R-phenylalaninol were much more potent than
their S-isomers (S-2 and S-9, Ki = 570 and 55 nM,
respectively). This strong stereopreference suggests that
both the amino and the phenyl groups of this side-chain
have major contributions to the binding energy.
These non-methyl uracil GnRH antagonists do not pos-
sess atropisomers as evidenced by a single set of NMR
signals in different solvents at room temperature. No
such atropisomeric property was observed even for
derivatives of the 5-phenyl group substituted with a
larger ortho group such as chlorine.
8. Tucci, F. C.; Hu, T.; Mesleh, M. F.; Bokser, A.; Allsapp,
E.; Gross, T. D.; Guo, Z.; Zhu, Y.-F.; Struthers, R. S.;
Ling, N.; Chen, C. Chirality, submitted for publication.
9. Daniels, J. M.; Nestmann, E. R.; Kerr, A. Drug Info. J.
1997, 31, 639.
10. For experimental details, see: Zhu, Y.; Chen, C.; Tucci,
F. C.; Guo, Z.; Gross, T. D.; Rowbottom, M.; Struthers,
R. S. WO 01/55119.
11. Rowbottom, M. W.; Tucci, F. C.; Zhu, Y.-F.; Guo, Z.;
Gross, T. D.; Reinhart, G. J.; Xie, Q.; Struthers, R. S.;
Saunders, J.; Chen, C. Bioorg. Med. Chem. Lett. 2004, 14,
2269.
12. Betz, S. F.; Reinhart, G. J.; Lio, F. M.; Chen, C.;
Struthers, S. J. Biol. Chem., submitted for publication.
13. For recent review articles on aromatic interactions, see: (a)
Hunter, C. A.; Lawson, K. R.; Perkins, J.; Urch, C. J.
J. Chem. Soc., Perkin. Trans. 2 2001, 651; (b) Janiak, C.
J. Chem. Soc. Dalton Trans. 2000, 3885.
14. Zhou, W.; Rodic, V.; Kitanovic, S.; Flanagan, C. A.; Chi,
L.; Weinstein, H.; Maayani, S.; Millar, R. P.; Sealfon, S.
C. J. Biol. Chem. 1995, 270, 18853.
On the bases of computational modeling of the 3-D hu-
man GnRH receptor,15 the binding site for this 5-phenyl
group could be located in the proximity at the top part
between helices 4 and 5, where two residues, Tyr-211
and Asn-212, are identified to project into the binding
pocket. While the tyrosine aromatic ring might interact
with this 5-phenyl group through p–p stacking, the
asparagine could form hydrogen-bond with the 3-meth-
oxyl group on the phenyl ring. Asn-212 has been known
to be important for the architecture of the ligand-bind-
ing pocket based on mutagenesis and computational
modeling studies, and it is proposed that Asn-212 inter-
acts with pGlu1 of GnRH via hydrogen-bonding.1b In
addition, alanine replacement of Tyr-211 results in a
receptor, which is neither capable of ligand binding
nor signal transduction by GnRH peptide.16
15. Hoffmann, S. H.; ter Laak, T.; Kuhne, R.; Reilander, H.;
Beckers, T. Mol. Endocrinol. 2000, 14, 1099.
16. CCDC 265975 contains the supplementary crystallo-
graphic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif,
by emailing data_request@ccdc.cam.ac.uk, or by contact-
ing The Cambridge Crystallographic Data Centre, 12,
Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223
336033.
In conclusion, a series of uracils with no 6-methyl group
was synthesized to address the atropisomerism of the 6-
methyl analogues such as 1. While initial des-methyl-
ation of 6-methyluracil 1 (Ki = 0.56 nM) caused about
10-fold reduction in binding affinity (2, Ki = 5.3 nM),