A Novel synthesis of 2-arylpyrrolo[1,2-a]pyrimid-7-ones and their structure-activity relationships as potent GnRH receptor antagonists

Article abstract of DOI:10.1016/S0960-894X(01)00780-6

In the process of developing GnRH receptor antagonists, a novel base-catalyzed cyclization of compounds 5a-b was discovered, which led to the formation of the 2-aryl pyrrolo[1,2-a]pyrimid-7-one core stuctures 6a-b. These intermediates were further modifie

Full text of DOI:10.1016/S0960-894X(01)00780-6

Bioorganic & Medicinal Chemistry Letters 12 (2002) 403–406
A Novel Synthesis of 2-Arylpyrrolo[1,2-a]pyrimid-7-ones
and Their Structure–Activity Relationships as Potent GnRH
Receptor Antagonists
Yun-Fei Zhu,* Keith Wilcoxen, John Saunders, Zhiqiang Guo, Yinghong Gao,
Patrick J. Connors, Jr., Timothy D. Gross, Fabio C. Tucci, R. Scott Struthers,
Greg J. Reinhart, Qiu Xie and Chen Chen*
Neurocrine Biosciences, Inc., 10555 Science Center Drive, San Diego, CA 92121, USA
Received 4 October 2001; accepted 12 November 2001
Abstract—In the process of developing GnRH receptor antagonists, a novel base-catalyzed cyclization of compounds 5a–b was
discovered, which led to the formation of the 2-aryl pyrrolo[1,2-a]pyrimid-7-one core stuctures 6a–b. These intermediates were
further modified at positions 1, 2, 4 and 6 to afford a series of potent GnRH antagonists with low nanomolar Ki values. # 2002
Elsevier Science Ltd. All rights reserved.
In the previous letter,1 we discussed the initial SAR
study of a novel series of 1-aminomethyl-2-aryl-3-
cyano-pyrrolo[1,2-a]pyrimid-7-ones (1) as human go-
nadotropin-releasing hormone (hGnRH) receptor
antagonists. Here, we report a novel synthesis of the
bicyclic 2-arylpyrrolo[1,2-a]pyrmid-7-one core structure
as well as further SAR studies of its derivatives as
potent hGnRH receptor antagonists.
The novel synthesis of 2-arylpyrrolo[1,2-a]pyrimid-7-
ones, represented by 6a–b, is outlined in Scheme 1.
Amidine 2 was refluxed with diethyl ethoxymethylene
malonate in the presence of EtONa in EtOH2 to afford
the corresponding pyrimidone 3. Compound 3 was then
treated separately with a-bromoacetophones 4a and 4b
in the presence of tetrabutylammonium fluoride
(TBAF) in THF to give pyrimidones 5a and 5b, respec-
tively. The regioselective N1-alkylation was confirmed by
NOE experiments. Catalyzed by NaH in THF, pyrim-
idones 5a–b underwent intramolecular cyclization to
give the bicyclic core structures 6a–b in good yields.
Intermediates 6a–b were then exposed to 2-fluorobenzyl
bromide and 1 M TBAF in THF to form compounds
7a–b. Compound 7b was further reduced by hydro-
genation over Pd/C to yield the corresponding aniline
7c. Subsequent acylation with carboxylic anhydrides
gave the amides 7d (7e–g).
Scheme 1. Reagents and conditions: (a) diethyl ethoxymethylene mal-
onate, EtONa/EtOH reflux; (b) TBAF, THF; (c) NaH, THF; (d)
TBAF/THF, 2-fluorobenzyl bromide; (e) H2, Pd/C, HOAc, 30 psi; (f)
(RCO)2O, Et3N, THF.
*Corresponding author. Tel.: +1-858-658-7745; fax: +1-858-658-
7601; e-mail: fzhu@neurocrine.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00780-6

404
Y.-F. Zhu et al. / Bioorg. Med. Chem. Lett. 12 (2002) 403–406
Scheme 2. Reagents and conditions: (a) R2R3NH, CH2O in water,
EtOH; (b) 3-pentanol, BuLi, THF.
Scheme 4. Reagents and conditions: (a) NaH, 2-amino-2-methyl-pro-
panol, THF; (b) SOCl2, THF; (c) 2-(2-methylaminoethyl)pyridine,
CH2O in water; (d) LiOH, THF/MeOH/ H2O; (e) K2CO3, 1-bromo-
pinacolone, DMF; (f) NH4OAc/HOAc, 100 ꢀC.
converted to a variety of carboxamides by reaction0 with
pre-formed triethyl aluminum and amine (R4R4 NH)
complexes in 1,2-dichloroethane at 90 ꢀC, followed by
Mannich reactions to give products 26–35 as depicted in
Scheme 3.
0
Scheme 3. Reagents and conditions: (a) R4R4 NH, Et3Al, DCE,
reflux; (b) R2R3NH, CH2O in water.
As illustrated in Scheme 2, advanced intermediates 7a–c
and 7e–g were diversely modified at position 1 via a
simple Mannich reaction3 with various amines in the
presence of formaldehyde. The structures of a selected
series of final compounds 8–15 are presented in Table 1.
In addition, modifications of the 6-ethoxy-carbonyl
group on 7a–c and 7e–g were also explored (Scheme 2)
and the ethyl esters were trans-esterified to the corre-
sponding 3-pentyl esters by lithium 3-pentoxide, formed
in situ from butyl lithium and 3-pentanol in THF, fol-
lowed by Mannich reactions to afford the amines 16–25
as shown in Table 2. Furthermore, ethyl ester 7g was
Scheme 4 shows the synthesis of two heterocyclic deri-
vatives (36 and 37) of the 6-carboxylic ester. Reaction of
7b with pre-mixed NaH and 2-amino-2-methyl-prop-
anol solution in THF, followed by treatment with thio-
nyl chloride formed an oxazoline,4 which was then
subjected to Mannich conditions to yield 36. Compound
7b was hydrolyzed by LiOH in a mixture of THF, MeOH
and water and the resulting acid was treated with 1-bro-
mopinacolone in the presence of K2CO3, followed by
reflux in HOAc with NH4OAc to produce an oxazole.5
Mannich reaction yielded the desired compound 37.
Table 1. Binding affinities of compounds 8–15 on the hGnRH receptor7
Compd
R1
R2R3NH
Ki (nM) human
8
9
MeO
MeO
NO2
BnNHMe
(2-pyr)-(CH2)2NHMe
BnNHMe
BnNHMe
BnNHMe
BnNHMe
(2-pyr)-(CH2)2NHMe
(2-pyr)-(CH2)2NHMe
400
55
450
190
42
44
14
1.2
10
11
12
13
14
15
NH2
MeCONH
i-PrCONH
i-PrCONH
n-PrCONH
Table 2. Binding affinities of compounds 16–25 on the hGnRH receptor7
Compd
R1
R2R3NH
Ki (nM) human
16
17
18
19
20
21
22
23
24
25
MeO
MeO
NH2
BnNHMe
(2-pyr)-(CH2)2NHMe
BnNHMe
BnNHMe
(2-pyr)-(CH2)2NHMe
BnNHMe
(2-pyr)-(CH2)2NHMe
Me2NH
Et2N(CH2)2NH-Me
PhCH2)2NHMe
62
14
100
4.0
2.8
3.8
2.7
130
5.8
2.1
MeCONH
MeCONH
n-PrCONH
n-PrCONH
n-PrCONH
n-PrCONH
n-PrCONH

Y.-F. Zhu et al. / Bioorg. Med. Chem. Lett. 12 (2002) 403–406
405
All of the synthesized compounds were evaluated for
their ability to compete for des-Gly10[125I-Tyr,5DLeu,6
NMeLeu,7Pro9-NEt]GnRH radioligand binding to the
cloned human receptor.6 The binding assay results on
varying the R1 group with either N-methyl-benzyl-
Table 2 lists the binding affinity data of compounds 16–
25, in which all molecules contain the 3-pentyl carboxy-
late instead of the ethyl carboxylate at position 6. A
direct comparison between compounds 16, 17, 18 and
compounds 8, 9, 11 reveals that the 3-pentyl carboxylate
significantly enhanced the binding affinity. Consequently,
combination of 3-pentyl carboxylate and acetamido
groups in compound 19 gave a Ki value of 4 nM, while
its corresponding ethyl carboxylate 12 was 10-fold less
potent. Interestingly, the binding enhancement by
incorporation of the N-methyl-N-2-(2-pyridyl)ethyl-
amine at the 1-methyl position in the ethyl carboxylates
was no longer observed here, as compound 20 was only
slightly more potent than compound 19. This trend was
also observed between compounds 21 and 22. Further
evaluation of different substituents on the basic amine
revealed that the simple dimethylamino analogue 23
had respectable potency (Ki =130 nM). A non-aromatic
side chain containing a basic amine (24) was only 2-fold
less potent than compound 22. The binding affinity of
25 provided more evidence that the N-methyl-N-2-(2-
pyridyl)ethylamine was no longer crucial for high
potency, since phenethyl replacement of the 2-pyr-
idylethyl group as the substituent on the basic amine side
chain gave very similar potency. It was worth noting that
such replacement led to complete loss of binding affinity
in the previous SAR study on compound 1.1
amine
or
N-methyl-N-(2-pyridyl)ethylamine
as
R2R3NH are presented in Table 1.7 Compounds 8 and
9 were substantially more potent than their corre-
sponding 3-cyano analogues.1 These data suggest that
position 3 of the bicyclic core prefers not to be sub-
stituted. Replacement of the electron donating methoxy
group of R1 with a strong electron withdrawing nitro
group had no substantial impact on potency (10, 450
nM). However, reduction of the nitro compound to
the corresponding amino analogue 11 yielded a 2-fold
improvement in the binding affinity. A hydrogen
bond acceptor together with a lipophilic group seems
to be the preferred R1 group as the acylated analo-
gues 12 and 13 had 5-fold improvement in potency,
in comparison with the anilino compound 11. Fur-
thermore,
a 3-fold enhancement of potency was
obtained by use of N-methyl-N-2-(2-pyridyl)ethylamine
as R2R3NH (14, Ki=14 nM). Surprisingly, a slight
change from the branched iso-butyrylamino group (i-
PrCONH) to
a linear n-butyrylamino group (n-
PrCONH) of R1 provided a 10-fold increase in the
binding affinity (15 vs 14).
Table 3. Binding affinities of compounds 26–35 on the hGnRH receptor7
0
Compd
R2R3NH
R4R4 NH
Ki (nM) human
26
(2-pyr)-(CH2)2NHMe
21
27
28
29
30
31
32
33
34
35
(2-pyr)-(CH2)2NHMe
(2-pyr)-(CH2)2NHMe
(2-pyr)-(CH2)2NHMe
(2-pyr)-(CH2)2NHMe
(2-pyr)-(CH2)2NHMe
(2-pyr)-(CH2)2NHMe
BnNHMe
9.2
10
32
3.3
630
1.1
17
BnNHMe
154
440
Et2N(CH2)2NH–Me

406
Y.-F. Zhu et al. / Bioorg. Med. Chem. Lett. 12 (2002) 403–406
Since position 6 was well tolerated for modification, a
limited study was undertaken to explore the use of more
stable heterocyclic groups to replace the esters and
amides. The resulting compounds 36 and 37 (Table 4)
had Ki values of 270 and 64 nM, respectively, which
were comparable to the potency of the corresponding
esters and amides. These results promoted us to perform
more modifications at position 6 using aryl groups and
the results will be presented elsewhere in the near future.
In order to demonstrate functional antagonism, selected
compounds were evaluated for their ability to inhibit
GnRH stimulated calcium flux.8 As shown in Figure 1,
compounds 20 and 22 at a concentration of 1 mM were
able to completely block Ca++ flux stimulated by 10
nM GnRH. No indication of stimulatory activity for
these or other compounds tested was observed.
Figure 1. Inhibition of GnRH stimulated Ca++ flux by compounds
20 and 22.
In conclusion, we have developed a novel and efficient
two-step synthesis for 2-arylpyrrolo[1,2-a]pyrimid-7-
ones. Further modifications of these structures led to
the discovery of a series of highly potent hGnRH
receptor antagonists. SAR study of these antagonists
indicated that hydrogen is more preferred substituent
than a cyano group at position 3 of this bicyclic core
structure. Position 6 was amenable to substitution by a
variety of groups without compromising the binding
affinity to the hGnRH receptor.
Table 4. Binding affinities of compounds 36 and 37 on the hGnRH
receptor7
Compd
Ki (nM) human
36
37
270
64
With these results, subsequent SAR studies were
focused on replacing the 6-carboxylates with carbox-
amides. As shown in Table 3, 3-pentyl carboxamide 26
was substantially less potent than its 3-pentyl ester ana-
logue 22. However, cyclopentyl analogue 27 was 2-fold
more potent than 26. While expanding the ring from
cyclopentyl to cyclohexyl (29) caused a decrease in
potency, a linear butyl carboxamide 28 was equally
potent as the cyclopentyl analogue 27. Further exten-
sion of the butyl to hexyl (30) provided a 3-fold increase
in binding affinity. However, insertion of an oxygen
atom in the butyl chain for reducing lipophilicity also
reduced potency (31). The preferred side chain from this
limited optimization study was a 3-phenylpropyl car-
boxamide group (32) which yielded another 3-fold
increase in potency in comparison with its hexyl analo-
gue 30. Unlike the 3-pentyl ester, incorporation of car-
boxamide at position-6 required the presence of the N-
methyl-N-[2-(2-pyridyl)]ethylamine on 1-methyl position
for high potency again. For example, 33 decreased
almost 17-fold in potency only due to switch of side
chain of the basic amine from a 2-(2-pyridyl)ethyl group
to a benzyl group. Similar results were observed in 34
and 35. Compared to 28, the potencies of these two
compounds were dramatically reduced simply because
benzyl and 2-diethylaminoethyl instead of 2-(2-pyr-
idyl)ethyl were substituted on the basic amine.
References and Notes
1. Zhu, Y.-F.; Struthers, R. S.; Connors, P. J., Jr.; Gao, Y.;
Gross, T. D.; Saunders, J.; Wilcoxen, K.; Reinhart, G. J.; Ling
N.; Chen, C. Bioorg. Med. Chem. Lett. 2002, 12, 399
2. Baxter, R. L.; Hanley, A. B.; Henry, W. S. J. Chem. Soc.,
Perkin Trans. 1 1990, 11, 2963.
3. Hanck, A.; Kutscher, W. Z. Physiol. Chem. 1964, 338, 272.
4. Jones, W. D., Jr.; Ciske, F. L. J. Org. Chem. 1996, 61, 3920.
5. Chihiro, M.; Nagamoto, H.; Takemura, I.; Kitano, K.;
Komatsu, H.; Sekiguchi, K.; Tabusa, F.; Mori, T.; Tominaga,
M.; Yabuuchi, Y. J. Med. Chem. 1995, 38, 353.
6. Human GnRH receptor was stably expressed in HEK293
cells and a 96-well filtration assay was used.
7. On each assay plate, a standard antagonist of comparable
affinity to those being tested was included as a control for
plate-to-plate variability. Overall, Ki values were highly
reproducible with an average standard deviation of 45% for
replicate Ki determinations. Most of compounds reported here
were assayed 2–8 times.
8. HEK293 cells stably expressing the hGnRH receptor were
loaded with the calcium sensitive dye Indo-1 then pre-incu-
bated with compound 20, 22 or vehicle control for 1 min prior
to stimulation with 10 nM GnRH. Calcium mobilization was
measured by the change in fluorescence intensity ratio (490
nm/405 nm) following excitation at 350 nm.