Pyrazolo[1,5-a]pyrimidines as estrogen receptor ligands: Defining the orientation of a novel heterocyclic core

Article abstract of DOI:10.1016/j.bmcl.2004.08.046

We investigated the pyrazolo[1,5-a]pyrimidine system as a novel heterocyclic scaffold for the development of estrogen receptor (ER) ligands. By altering the pattern of hydroxyl substitution, we established the orientation that is most favorable for ER bin

Full text of DOI:10.1016/j.bmcl.2004.08.046

Bioorganic & Medicinal Chemistry Letters 14 (2004) 5681–5684
Pyrazolo[1,5-a]pyrimidines as estrogen receptor ligands:
defining the orientation of a novel heterocyclic core
Dennis R. Compton, Kathryn E. Carlson and John A. Katzenellenbogen*
Department of Chemistry, University of Illinois, 600 S. Matthews Avenue, Urbana, IL 61801-3731, USA
Received 26 April 2004; accepted 18 August 2004
Available online 16 September 2004
Abstract—We have examined the pyrazolo[1,5-a]pyrimidine scaffold as a novel core structure for estrogen receptor ligands. Attach-
ment of various substituents has helped to define the orientation of this heterocycle in the ligand-binding pocket as one in which a
pendant phenol rather than the hydroxylpyrimidine serves as a mimic of the A-ring of estradiol.
Ó 2004 Elsevier Ltd. All rights reserved.
The estrogen receptor (ER) is a nuclear hormone recep-
tor of pharmaceutical interest as a target for the treat-
ment of osteoporosis, breast cancer, and other
endocrine disorders.1 Although its natural ligand is the
steroid 17b-estradiol (E2), the ER is capable of binding
a wide variety of steroidal and nonsteroidal ligands.2
Many ER ligands are built upon a heterocyclic core as
a scaffold onto which are attached additional substitu-
ents. Among these cores are 5,6-fused heterocycles
(Fig. 1A), such as the benzothiophenes3 and indoles.4
Both of these systems, as well as others, incorporate a
hydroxyl group into the six-membered ring that serves
as a mimic of the A-ring phenol of estradiol. As part
of our larger effort to diversify the core structures of
ER ligands,5–7 we considered the pyrazolo[1,5-a]pyrim-
idine system (Fig. 1B): This structure is commensurate
with other 5,6-fused heterocycles, is easily synthesized,
and can readily be adorned with substituents. The versa-
tility of the core has allowed it to be studied as a scaffold
in a wide variety of pharmaceutical targets. Thus, the
pyrazolo[1,5-a]pyrimidine scaffold has recently be exam-
ined as a CRF antagonist8 and COX-2 inhibitor.9
In our initial consideration of the pyrazolo[1,5-a]pyrim-
idine core as an ER ligand building block, we incorpo-
rated a hydroxyl at the 6-position, in a manner similar
to the benzothiophene and indole systems, and we ap-
pended aromatic rings at the 2- and 3-positions, which
Figure 1. (A) Benzothiophene and indole heterocycles, (B) pyr-
azolo[1,5-a]pyrimidine core and numbering system. 6-Hydroxypyr-
azolo[1,5-a]pyrimidines a potential heterocyclic estrogen receptor ligand.
*
Corresponding author. Tel.: +1 217 333 6310; fax: +1 217 333
7325; e-mail: jkatzene@uiuc.edu
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.08.046

5682
D. R. Compton et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5681–5684
are typically sites for aryl or alkyl group substitution on
related systems. By incorporating a hydroxyl group into
the 6-position of the core scaffold, we believed that the
hydroxylpyrimidine portion would be poised to act as
a mimic of the aromatic A-ring of estradiol. The aro-
matic ring substituents in such ligands can either be
phenyl or 4-hydroxyphenyl groups, the latter depicted
in Figure 1B.
orientation to the ER was most likely to occur, that is,
which ring would be the mimic of the A-ring of estra-
diol. Once the core orientation was established, we felt
that we would be in a better position to explore addi-
tional systems to optimize ligand binding.
Pyrazolo[1,5-a]pyrimidines have historically been syn-
thesized by a condensation of a dicarbonyl compound
with a 3-amino-pyrazole under acidic conditions.11,12
For the systems of interest to us, this involves the
condensation of a substituted 1,3-propanedial and a 3-
amino-4,5-diaryl-pyrazole (Scheme 1). The 3-amino-
pyrazole derivatives can be obtained in two steps, by
Claisen condensation of a phenylacetonitrile with a
methyl benzoate, followed by cyclization with hydra-
zine. The desired pyrazolo[1,5-a]pyrimidine can be then
prepared in a few steps.
Although we envisioned that the 6-hydroxyl group
would act as the A-ring mimic of estradiol and hydrogen
bond with Glu353 and Arg394 in the ligand-binding
pocket of ERa, when multiple hydroxyl groups are pre-
sent (i.e., with 4-hydroxyphenyl substituents), the possi-
bility exists that one of the other rings could act as the
A-ring mimic. To examine whether the 6-hydroxy-
pyrimidine system could act as the A-ring mimic of E2,
and whether the compounds would fit into the binding
pocket of the estrogen receptor, we docked the trihydr-
oxy pyrazolopyrimidine system into the crystal structure
of ERa ligand-binding domain complex with E2 (1ERE)
(Fig. 2).10
The synthesis of the desired 3-aminopyrazoles is illus-
trated in Scheme 2. A Claisen condensation between a
phenyl acetonitrile 1a–b and a methyl benzoate 2a–b
produces the 3-ketopropionitriles 3a–d in good yields.13
Condensation of compounds 3a–d with hydrazine in the
presence of acid produces the desired 3-aminopyrazoles
4a–d.
Using an approach we have reported previously,6 we
overlaid the 6-hydroxyl substituent with the hydroxyl
of E2 and allowed the structure to minimize in the recep-
tor after E2 was deleted from the pocket. In this manner,
we found that the binding pocket would accommodate
this system, with the 6-hydroxy-pyrazolo[1,5-a]pyrim-
idine acting as an A-ring mimic of the phenol of E2.
In addition to forming the desired hydrogen bonds with
Glu353 and Arg394, the modeling also indicated poten-
tial hydrogen bonding between Thr347 and hydroxyl of
the phenyl at the 2-position. In further modeling studies,
we examined the possibilities of either of the additional
4-hydroxyphenyl rings acting as the A-ring mimic of E2
(results not shown). We found that 2,3-bis-(4-hydroxy-
phenyl)-pyrazolo[1,5-a]pyrimidines could also be
accommodated in an alternate orientation, with the pen-
dant C(2)-phenol in the A-ring binding site (not shown).
Considering these alternative possibilities and recogniz-
ing that molecular modeling would not provide a defin-
itive means for selecting among them, we set about
synthesizing a set of molecules to explore which binding
The required 2-hydroxy-1,3-propanedial is a highly
functionalized system that is not likely to be stable.
Scheme 1.
Figure 2. Docking of 2,3-bis-(4-hydroxyphenyl)-6-hydroxypyrazolo-
[1,5-a]pyrimidine into the ERa binding pocket (1ERE), using SYBYL
V. 6.6. Shown is the model with the 6-hydroxyl group mimicking the
phenol of estradiol.
Scheme 2. Reagents and conditions: (a) THF, NaH, 65°C, 2days; (b)
H2NNH2, EtOH, HCl, 80°C, 8h.

D. R. Compton et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5681–5684
5683
Scheme 3. Reagents and conditions: (a) NaH, rt, 6h; (b) EtOH,
BF3ÆOEt2, Hg(II)O; (c) EtOH, LiBH4, 7h.
Therefore, we prepared it as the tetraalkoxy-bisacetal.
Balsevich14 prepared the 2-keto-1,1,3,3-tetraethoxy-pro-
pane 8 (Scheme 3) by a procedure that involves coupling
of dimethylsulfoxide 6 with ester 5 to generate the b-
ketosulfoxide 7. Pummerer rearrangement of 7 with
mercuric oxide yields ketone 8. Baganz15 prepared 8
via an alternative method; however, they described the
reduction of 8 with lithium borohydride to give 9, the
1,3-propanedial derivative we desired. The preparation
of 9 allows for the synthesis of 6-hydroxypyrazolo[1,5-
a]pyrimidines.
To obtain a more complete picture of the potential of
the core to act as an estrogen receptor ligand and to
study the cyclization conditions, we purchased 1,1,3,3-
tetramethoxypropane (10), a compound that lacks the
hydroxyl group needed to produce the 6-hydroxyl in
the pyrazolo[1,5-a]pyrimidines, and we also reacted it
with compounds 4a–d (Scheme 4). For the in situ con-
version of the bis-acetal to the reactive dialdehyde, a sol-
vent sufficiently polar to dissolve the 3-aminopyrazoles
and acid was required. Thus, in acetic acid the reaction
reached completion in just a few hours, providing com-
pounds 11a–d. The methyl ethers of 11b–d were cleaved
with BF3ÆSMe2 to reveal the free hydroxyls in com-
pounds 12a–c. The lack of the 6-hydroxyl group in these
systems allowed us to study the possibility of the pen-
dant hydroxyl as the A-ring mimic of estradiol.
Scheme 4. Reagents and conditions: (a) HOAc, 110°C, 3h; (b)
CH2Cl2, BF3ÆSMe2, 16–20h.
Having established conditions for pyrazolo[1,5-a]pyrim-
idine formation and phenol deprotection, we prepared
the desired analogs containing the 6-hydroxyl group
(Scheme 4) by reaction of compounds 4a–d with 9 to
yield compounds 13a–d. Deprotection of 13b–d as above
provided the desired products 14a–c, which incorpo-
rated a hydroxyl group at the 6-position.
The first compound to notice is 11a; it has no hydroxyl
groups and shows no binding to either ERa or ERb.
This indicates that the pyrazolo[1,5-a]pyrimidine core
simply acts as a scaffold for the substituents to hydrogen
bond with the ER and does not by itself interact in an
effective manner directly with the receptor. Compound
13a, which only contains the 6-hydroxyl for hydrogen
bonding, shows at most minimal binding to the ER.
To best understand the remaining data, compounds
12a–c should be compared to 14a–c, respectively. From
this comparison, it is evident that with one exception,
compounds containing a 6-hydroxyl substituent (14a–
c) consistently show significantly lower RBA values
than the corresponding systems without the 6-hydroxyl
substituent (12a–c), not at all what we had origi-
nally sought. Also of note is that all compounds having
The two series produced a total of eight compounds
whose binding for both ERa and ERb was determined
using a competitive radiometric assay.16 The affinity of
the compounds are expressed as a percent relative to est-
radiol, which is normalized to 100%. These relative
binding affinity (RBA) values are given in Table 1.
Although the binding affinities are low, the structure–
affinity relationships are informative of the key issue
of ligand core orientation.

5684
D. R. Compton et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5681–5684
Table 1. ERa and ERb relative binding affinitiesa
ands. In contrast to our initial design expectations, how-
ever, we found that introducing a 6-hydroxyl into the
core resulted in reduced binding affinity, which sug-
gested that the ligand core is binding to the ER in a re-
versed fashion, with a pendant phenol, rather than the
hydroxyl pyrimidine group, serving as the mimic of
the A-ring of estradiol. Further work based on the
new lead compound, 12c, has led to ER ligands with
higher binding affinity.17
Compound
ERa
ERb
11a
12a
12b
12c
13a
14a
14b
14c
—
—
0.010
0.007
0.004
—
0.050
0.023
0.100
0.005
0.010
0.006
0.034
0.003
—
0.004
a Relative binding affinities (RBA) reported as a percent of E2 activity
where E2 = 100%. These values are the average of replicate experi-
ments, where the range has a CV below 0.3. Where values are not
given, they are below detection, <0.003. The Kd of estradiol is 0.2nM
(ERa) and 0.5nM (ERb).
Acknowledgements
We are grateful for support of this research through a
grant from the National Institutes of Health [PHS
5R37 DK15556].
measurable RBA values are selective for ERb. Com-
pound 12c, which had the highest ERb binding, also
showed the greatest selectivity (25-fold). Compound
14c, of the original design, did show good ERb selectiv-
ity, but had only a third the binding affinity of 12c.
References and notes
1. Katzenellenbogen, B. S.; Katzenellenbogen, J. A. Breast
Cancer Res. 2000, 2, 335–344.
2. Meegan, M. J.; Lloyd, D. G. Curr. Med. Chem. 2003, 10,
181–210.
Based on the consistently higher affinity of the pyraz-
olo[1,5-a]pyrimidine systems lacking the 6-hydroxyl
group, we believe it unlikely that this heterocyclic system
binds with the 6-hydroxyl group mimicking the A-ring
phenol of estradiol. Thus, in work to be described else-
where, we used the highest affinity analog lacking the 6-
hydroxyl group (12c) as a lead to explore how effectively
systems that use the pendant 4-hydroxyphenyl groups as
the A-ring mimic of estradiol can be developed to bind
to the ERs.17 In exploring these other analogs, we chose
those that include alkyl groups at the 5- and 7-positions
(Fig. 3). These alkyl groups could readily be incorpo-
rated into our synthetic scheme as substituents on the
1,3-dicarbonyl unit used to complete the formation of
the pyrimidine unit, and they add hydrophobic bulk to
fill space within the ligand-binding pocket. As described
elsewhere, some of these redesigned pyrazolo[1,5-
a]pyrimidine systems show significantly improved ER
binding affinities, results which can be rationalized by
molecular modeling which places the C-2 (4-hydroxy-
phenyl) group in the A-ring binding pocket. The new
ER analogs also display interesting patterns of selectiv-
ity for ERa versus ERb in terms of binding affinity and
transcription potency and efficacy.17
3. Grese, T. A.; Pennington, L. D.; Sluka, J. P.; Adrian, M.
D.; Cole, H. W.; Fuson, T. R.; Magee, D. E.; Phillips, D.
L.; Rowley, E. R.; Shetler, P. K.; Short, L. L.; Venugo-
palan, M.; Yang, N. N.; Sato, M.; Glasebrook, A. L.;
Bryant, H. U. J. Med. Chem. 1998, 41, 1272–1283.
4. von Angerer, E.; Knebel, N.; Kager, M.; Ganss, B. J.
Med. Chem. 1990, 33, 2635–2640.
5. Fink, B. E.; Mortensen, D. S.; Stauffer, S. R.; Aron, Z. D.;
Katzenellenbogen, J. A. Chem. Biol. 1999, 6, 205–
219.
6. Stauffer, S. R.; Coletta, C. J.; Tedesco, R.; Nishiguchi, G.;
Carlson, K.; Sun, J.; Katzenellenbogen, B. S.; Katzenel-
lenbogen, J. A. J. Med. Chem. 2000, 43, 4934–4947.
7. Mortensen, D. S.; Rodriguez, A. L.; Carlson, K. E.; Sun,
J.; Katzenellenbogen, B. S.; Katzenellenbogen, J. A. J.
Med. Chem. 2001, 44, 3838–3848.
8. Gilligan, P. J.; Baldauf, C.; Cocuzza, A.; Chidester, D.;
Zaczek, R.; Fitzgerald, L. W.; McElroy, J.; Smith, M. A.;
Shen, H. S. L.; Saye, J. A.; Christ, D.; Trainor, G.;
Robertson, D. W.; Hartig, P. Bioorg. Med. Chem. 2000, 8,
181–189.
9. Almansa, C.; de Arriba, A. F.; Cavalcanti, F. L.; Gomez,
L. A.; Miralles, A.; Merlos, M.; Garcia-Rafanell, J.; Forn,
J. J. Med. Chem. 2001, 44, 350–361.
10. Pike, A. C. W.; Brzozowski, A. M.; Hubbard, R. E.;
Bonn, T.; Thorsell, A.-G.; Engstrom, O.; Ljunggren, J.;
Gustafsson, J.-A.; Carlquist, M. EMBO J. 1999, 18, 4608–
4618.
11. Elnagdi, M. H.; Erian, A. W. Bull. Chem. Soc. Jpn. 1990,
63, 1854–1856.
12. Novinson, T.; Miller, J. P.; Scholten, M.; Robins, R. K.;
Simon, L. N.; OꢀBrien, D. E.; Meyer, R. B., Jr. J. Med.
Chem. 1975, 18, 460–464.
In summary, we have explored a new heterocyclic sys-
tem, pyrazolo[1,5-a]pyrimidine, as a core scaffold for
developing novel ER ligands. This synthetically versatile
bicyclic system allows for the introduction of multiple
substituent variations in our search for good ER lig-
13. Kayaleh, N. E.; Gupta, R. C.; Johnson, F. J. Org. Chem.
2000, 65, 4515–4522.
14. Balsevich, J. Can. J. Chem. 1983, 61, 1053–1059.
15. Baganz, H.; May, H. J. Chem. Ber. 1966, 99, 3771–3777.
16. Carlson, K. E.; Choi, I.; Gee, A.; Katzenellenbogen, B. S.;
Katzenellenbogen, J. A. Biochemistry 1997, 36, 14897–
14905.
17. Compton, D. R.; Sheng, S.; Carlson, K. E.; Rebacz, N. A.;
Katzenellenbogen, J. A.; Katzenellenbogen, B. S. J. Med.
Chem., in press.
Figure 3. Redesigned pyrazolo[1,5-a]pyrimidine systems based on lead
compound 12c.