5-aryl-1,2-dihydrochromeno[3,4-f]quinolines: A novel class of nonsteroidal human progesterone receptor agonists

Article abstract of DOI:10.1021/jm9705768

The development of a novel class of nonsteroidal human progesterone receptor (hPR) agonists, 5-aryl-1,2-dihydro-5H-chromeno[3,4-f]quinolines 2, is described. The introduction of a 5-aryl group into the 1,2- dihydrocoumarino[3,4-f]quinoline core 1 is the k

Full text of DOI:10.1021/jm9705768

J . Med. Chem. 1998, 41, 291-302
291
5-Ar yl-1,2-d ih yd r och r om en o[3,4-f]qu in olin es: A Novel Cla ss of Non ster oid a l
Hu m a n P r ogester on e Recep tor Agon ists
Lin Zhi,*,† Christopher M. Tegley,† E. Adam Kallel,† Keith B. Marschke,‡ Dale E. Mais,§
Marco M. Gottardis,§ and Todd K. J ones†
Departments of Medicinal Chemistry, New Leads Discovery, and Endocrine Research, Ligand Pharmaceuticals, Inc.,
10275 Science Center Drive, San Diego, California 92121
Received August 28, 1997
The development of a novel class of nonsteroidal human progesterone receptor (hPR) agonists,
5-aryl-1,2-dihydro-5H-chromeno[3,4-f]quinolines 2, is described. The introduction of a 5-aryl
group into the 1,2-dihydrocoumarino[3,4-f]quinoline core 1 is the key for progestational
activities. The structure-activity relationship (SAR) studies of the 5-aryl substituents
generated a series of potent hPR agonists, which exhibited similar biological activity (EC50
)
8-30 nM) to the natural hormone progesterone (EC50 ) 2.9 nM) in cell-based assays with
efficacies ranging from 28% to 96%. Most of the analogues displayed similar or greater binding
affinity (Ki ) 0.41-3.6 nM) than progesterone (Ki ) 3.5 nM). Three representative analogues
(13, 15, and 24) demonstrated in vivo activities in mammary gland morphology/uterine wet
weight assay in ovariectomized rats.
Ch a r t 1
In tr od u ction
Progesterone (Chart 1) is a unique reproductive
hormone, and it plays a trophic role for tissues of female
reproduction. Its principal target organs are uterus,
breast, and brain.1 Since the isolation of progesterone
in 1934,1a development of new progestins and under-
standing their action have been part of one of the
greatest chemical and biological research efforts on a
single group of substances (steroids). In the late 1930s
and early 1940s, on the basis of animal experiments,
several groups reported the idea that progesterone could
be used to prevent ovulation in women, but the concept
was not realized until the 1950s.2 Progesterone was
originally obtained by an inefficient and expensive
extraction from animal ovaries and is rapidly metabo-
lized in vivo,1b which prevented the natural hormone
from becoming a widely used medicine. In the early
1950s, Djerassi and Rosenkranz3 of Syntex and Colton4
of G. D. Searle and Co. reported the synthesis of
norethindrone and norethynodrel, respectively. The
discovery that these 19-nortestosterones have proges-
tational activity made synthetically modified progestins
of tremendous therapeutic importance. It resulted in
the successful development of oral contraceptive (OC)
agentss“the pill” in the 1960s.1b In addition to the
primary use as birth control for women, progestins,
combined with estrogen, are widely used in hormone
replacement therapy (HRT) and have been increasingly
prescribed during the last two decades due to the
increase of life expectancy for women. It has been
reported that addition of a progestin in HRT signifi-
cantly reduces the risk of endometrial cancer with
postmenopausal women who take estrogen.5,6 Progestins
are also used to treat several gynecological disorders:
dysmenorrhea, endometriosis, and dysfunctional uterine
bleeding caused by hormonal deficiency or imbalance.1b
Like most steroidal drugs, steroidal progestins can
have undesirable side effects due to varying degrees of
cross-reactivities with the closely related androgen
receptor, glucocorticoid receptor, estrogen receptor, and
mineralocorticoid receptor and the interactions with
GABA (γ-aminobutyric acid) receptor through a nonge-
nomic pathway. Some cross-reactivity with other recep-
tors may occasionally be desirable,7 requiring inten-
* Address correspondence to this author. Phone: (619) 550-7663.
Fax: (619) 550-7249. E-mail: lzhi@ligand.com.
† Department of Medicinal Chemistry.
‡ Department of New Leads Discovery.
§ Department of Endocrine Research.
S0022-2623(97)00576-1 CCC: $15.00 © 1998 American Chemical Society
Published on Web 01/13/1998

292 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Zhi et al.
Sch em e 1a
tional control of the selectivity profile. Development of
new generations of progestins to improve the selectivity
profile of progestins has been a great challenge.8 Ad-
ditionally, exploration of therapeutic applications such
as oncology demands progestins with new profiles.9
The number of steroid drugs has not substantially
increased during the last two decades, but the knowl-
edge of their mechanisms of action has. Since the
cloning of the first intracellular receptor (IR) in 1985,
more than fifty members of the IR superfamily have
been identified.10 The understanding of the IRs at
molecular level makes it possible to devise small-
molecule pharmacological agents to modulate gene
expression for therapeutic benefit, which opens new
opportunities to discover more efficient and more selec-
tive orally available small molecules to mimic or block
the action of natural hormones.11 A group of cell-based
high-throughput assays, termed cotransfection assays,
have been developed12 and are utilized to detect the
interaction of novel small molecules with the IRs to
identify individual IR agonists, antagonists, or partial
agonists by using cloned human IR cDNAs. These
powerful assays can be used to identify novel lead
structures13 and also to profile an existing drug for the
purpose of optimization.14,15 Here we report the results
of our endeavor to develop a novel human progesterone
receptor (hPR)16 agonist pharmacophore, 5-aryl-1,2-
dihydro-5H-chromeno[3,4-f]quinoline 2, and the opti-
mization results of the 5-aryl substituents by using the
hPR cotransfection assay as a guide and other IR assays
to characterize their selectivity profiles, including hu-
man androgen receptor (hAR),17 human glucocorticoid
receptor (hGR),18 human estrogen receptor (hER),19 and
human mineralocorticoid receptor (hMR).20
a
Reagents: (a) acetone, I2, 100 °C; (b) PhLi, THF, -30 °C; (c)
TFA, Et3SiH, CH2Cl2.
acid directly to the reaction mixture, requiring a large
excess of fuming nitric acid to reach the desired acid
concentration. Alternatively, isolation of the 2:1 mix-
ture of the mononitrated products followed by nitration
with fuming nitric acid provided a 2:1 mixture of 4,2′-
dinitro-2-biphenylcarboxylic acid (9a ) and the 4,4′-
dinitro isomer (9b). Formation of the lactone 10 from
the mixture of the dinitrated products was completed
through an unusual cyclization of the carboxylate onto
the nitro phenyl ring at 120 °C in N,N-dimethylaceta-
mide (DMA).22 Direct precipitation of the lactone from
the water-diluted reaction mixture separated 10 from
the undesired 4,4′-dinitro-2-biphenylcarboxylic acid (9b).
Nitro lactone 10 was reduced by a palladium-catalyzed
hydrogenation to afford 8-amino-6H-dibenzo[b,d]pyran-
6-one in high yield. The Skraup reaction of the 8-amino
compound required 120-130 °C and 15 h in a sealed
tube with 40% iodine. However, it regioselectively
afforded quinoline 1 in over 50% yield.23 This practical
synthetic route to intermediate 1 provided easy access
to a large number of 5-aryl compounds, quickly estab-
lishing the structure-activity relationship (SAR) around
the 5-aryl group.
The 5-aryl group was introduced by nucleophilic
addition of an aryllithium or aryl Grignard reagent to
the 1,2-dihydrocoumarino[3,4-f]quinoline 1 followed by
reduction with triethylsilane in the presence of a Lewis
acid such as TFA or BF3-OEt2.24 Most of the aryl-
lithium reagents were prepared by treatment of the
corresponding aryl bromides with n-BuLi at -78 °C in
THF. It was found that the disubstituted aryllithium
species could not be generated by n-BuLi in THF but
could be made in diethyl ether25 (e.g., 3,4-dichlorophe-
nyllithium, 3,5-difluorophenyllithium, and 4-chloro-3-
fluorophenyllithium).
Ch em istr y
The 1,2-dihydro-2,2,4-trimethylquinoline moiety of
the core structure is synthesized by the classic Skraup
reaction.21 Treatment of 3-amino-6H-dibenzo[b,d]py-
ran-6-one (3) with refluxing acetone in the presence of
a catalytic amount of iodine provided a 2:1 mixture of
two regioisomers of isocoumarinoquinoline (4 and 5) in
90% yield. Elevation of reaction temperature in a sealed
tube accelerated the Skraup reaction. Scheme 1 shows
the synthesis of analogues 6 and 7. The aryl group was
introduced by a nucleophilic addition of phenyllithium
to isocoumarinoquinolines 4 and 5 followed by a reduc-
tion with triethylsilane in the presence of TFA.
The synthesis of 5-aryl-1,2-dihydro-5H-chromeno[3,4-
f]quinolines 2 is outlined in Scheme 2, which includes
the synthesis of the 1,2-dihydrocoumarino[3,4-f]quino-
line core 1 and the introduction of the 5-aryl groups.
8-Nitro-6H-dibenzo[b,d]pyran-6-one (10) was prepared
by a modified literature procedure,22 starting from the
controlled double nitration of commercially available
2-biphenylcarboxylic acid (8). The first nitration with
70% nitric acid was carried out at room temperature,
generating a 2:1 mixture of two compounds favoring 2′-
nitration over 4′-nitration. The selectivity, which may
arise from the acid coordination with the active nitration
species, was not optimized. The second nitration at the
4-position required a higher nitric acid concentration
and can be run by two different procedures. The one-
pot procedure was performed by adding fuming nitric

Novel Class of hPR Agonists
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 293
Sch em e 2a
Ch a r t 2
Ta ble 1. Cotransfection and Competitive Binding Data for the
Quinoline Analoguesa
hPR-A
hPR agonist,b
hPR antagonist,b
binding
Ki (nM),
mean
mean ( SEM
mean ( SEM
a
Reagents: (a) 70% HNO3, rt; (b) fuming HNO3, rt; (c) DMA,
pot
pot
130 °C; (d) DMF, H2, 10% Pd/C, rt; (e) acetone, I2, 130 °C; (f) PhLi
or PhMgX, THF or ether, -78 °C to rt; (g) TFA or BF3-OEt2,
Et3SiH, CH2Cl2, rt.
compd
eff (%) (EC50, nM)c eff (%) (IC50, nM)c
( SEM
d
progesterone 100
2.9 ( 0.9
-
-
3.5 ( 0.2
669 ( 71
119 ( 24
274 ( 121
1
4
5
6
7
13
-
-
-
-
-
-
-
-
-
-
95
305
83 ( 6 194 ( 23
73 ( 7 170 ( 51
81 ( 4 933 ( 462 80 ( 12
>1000
75 ( 4 290 ( 121 3.6 ( 0.7
The aryl adducts presumably are a mixture of two
isomers: the cyclic hemiacetal 11 and the open hydroxy
ketone 12. Determination of the ratio of 11 and 12 by
NMR spectra was unsuccessful. Peak broadening was
observed due to the rapid interconversion between 11
and 12. In the case of p-N,N-dimethylphenyl adduct,
the open form 12 was the only isomer observed, and it
was not converted by TFA-Et3SiH conditions to the
5-aryl product 2 (R′ ) H, R ) NMe2). Usually, BF3-
OEt2 mediated reduction was faster and cleaner than
TFA-mediated reduction. However, when the 5-aryl
group contains heteroatoms, such as the 5-pyridyl
compounds, TFA was superior. In the synthesis of
compound 21, BF3-OEt2/Et3SiH treatment of the cor-
responding hemiacetal 11 gave the overreduced 5-(4-
ethylphenyl) product in high yield. Compounds 13-34
in Chart 2 were synthesized from 1 by the method
outlined in Scheme 2.
-
-
42 ( 10 156 ( 71
a
Values with standard errors (SEM) represent the mean value
of at least three separate experiments with triplicate determina-
tions, and values without standard errors represent a single
experiment. Agonist efficacies were compared to that of proges-
b
terone (100%), and antagonist efficacies were determined as a
function (%) of maximal inhibition of progesterone at EC50
concentration. c All EC50 and IC50 values were determined from
full dose-response curves ranging from 10-12 to 10-5 M in CV-1
d
cells. Stands for the efficacy < 20% and potency > 10 000 nM.
compound 4, based on the observation that RU48626 and
ZK98,29927 (Chart 1), the well-known PR antagonists,
were developed by introduction of the 11â-aryl substitu-
ent on the relatively planar steroid skeleton. To mimic
the known steroidal PR modulators, we synthesized
quinoline analogues 6 by introducing an aromatic sub-
stituent to lead compound 4. The binding affinity of
compound 6 is stronger than that of the lead compound
4, but 6 has less hPR antagonist activity in cotransfec-
tion assay. A Skraup reaction regioisomer (5) of 4 and
its 10-phenyl analogue (7) were also tested in the similar
assays, and 7 is totally inactive. Switching lactone 4
into lactone 1 generated a platform which was used to
prepare 5-aryl analogues having more similarity with
steroidal antiprogestins. The assay results for 5-phenyl
compound 13 were much more promising. Its binding
Resu lts a n d Discu ssion
Our initial hPR modulator lead, isocoumarinoquino-
line 4, was prepared to mimic the steroid skeleton and
showed moderate hPR antagonist activity and moderate
binding affinity to PR-A (human A-subtype)16 (see Table
1). In an effort to develop a novel class of potent PR
modulators from this lead, we introduced an aromatic
moiety which projects out of the planar structure of lead

294 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Zhi et al.
Ta ble 2. Cotransfection and Competitive Binding Data for the Quinoline Analoguesa
hPR agonist,b mean ( SEM
hPR antagonist,b mean ( SEM
hPR-A binding Ki (nM),
compd
eff (%)
pot (EC50, nM)c
eff (%)
pot (IC50, nM)c
mean ( SEM
d
progesterone
100
2.9 ( 0.9
0.15 ( 0.05
156 ( 71
8.0 ( 7.0*
14 ( 2
14 ( 6
8.0 ( 3.2
18 ( 5
8.0 ( 3.5
33 ( 14
-
-
-
-
3.5 ( 0.2
0.34 ( 0.04
3.6 ( 0.7
5.3 ( 1.1
0.70 ( 0.14
0.55 ( 0.21
0.69 ( 0.21
0.43 ( 0.11
1.1 ( 0.6
10.8 ( 1.1
8.8 ( 1.2
2.3 ( 0.5
13.3 ( 1.2
2.6 ( 0.2
3.4 ( 0.7
0.41 ( 0.08
1.2 ( 0.2
6.8 ( 0.5
1.8 ( 0.8
1.7 ( 0.2
2.4 ( 0.4
0.87 ( 0.03
3.9 ( 1.4
13.8 ( 4.2
MPA
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
80 ( 7
42 ( 10
28 ( 4*
77 ( 5
59 ( 7
48 ( 13
76 ( 8
77 ( 11
35 ( 2
-
-
75 ( 4
77 ( 16
31
42 ( 22*
74
37
41 ( 10
49
68 ( 19
38
65 ( 3
-
42
290 ( 121
342 ( 175
600
3138 ( 2272*
150
930
1136 ( 326
150
132 ( 95
250
105 ( 40
-
720
-
-
4600
-
-
1207 ( 269*
1100
54 ( 15*
71 ( 18
59 ( 5
-
37 ( 11
-
96 ( 11
66 ( 7
59 ( 19
68 ( 18
60 ( 13
83 ( 23
57 ( 7
49 ( 16
65 ( 24
24
13 ( 2
18 ( 8
17 ( 16
14 ( 8
25 ( 13
16 ( 10
50 ( 37
27 ( 12
17 ( 8
3900
-
-
56
-
-
65 ( 8*
31
86 ( 8*
86 ( 9
21
1700
a
Values with standard errors (SEM) represent the mean value of at least three separate experiments with triplicate determinations,
values without standard errors represent a single experiment, and values with an asterisk represent the mean value of two experiments
b
with standard deviation. Agonist efficacies were compared to that of progesterone (100%), and antagonist efficacies were determined as
a function (%) of maximal inhibition of progesterone at EC50 value. c All EC50 and IC50 values were determined from full dose-response
d
curves ranging from 10-12 to 10-5 M in CV-1 cells. Stands for efficacy < 20% and potency > 10 000 nM.
affinity (Ki ) 3.6 nM) is the same order of magnitude
as the natural ligand progesterone (Ki ) 3.5 nM).
These preliminary results encouraged us to exten-
sively explore the SAR of the aryl substituent. Prepar-
ing analogues in the 5-aryl series (structure 2) gener-
ated a number of interesting compounds that showed
nanomolar agonist activity on hPR with efficacies rang-
ing from 28% to 96%. The assay results of the repre-
sentative 5-aryl analogues in cotransfection assays in
both agonist and antagonist modes are summarized in
Table 2. The introduction of a chlorine or bromine
group into the 5-aryl moiety of parent compound 13
enhanced the agonist activity by about 10-fold with
similar or higher efficacy, regardless of the substitution
at para or meta positions (compounds 15, 18 and 16,
19). These chloro or bromo 5-aryl compounds represent
the most active analogues in this series. Fluorine
substitution had a relatively smaller effect at the para
position (14) and, at the meta position, generated a
potent partial agonist (17). The m-(trifluoromethyl)-
phenyl compound 24 showed excellent full agonist
activity, while its para analogue 20 only had 35%
efficacy. Compounds containing either the electron-
withdrawing acetyl group (21) or the electron-donating
methoxy group (23) behaved as hPR antagonists. It
appears that the size of the substituents play a greater
role than their electronic properties.
analogues with best binding affinity (Ki ) 0.41-1.8 nM)
are those compounds bearing a halogen in the 5-aryl
ring (compounds 15-19). The analogues which showed
activity in the antagonist mode had slightly lower
binding affinities (compounds 21, 23, 33, and 34).
To study the effect of two substituents on the activi-
ties, disubstituted 5-aryl analogues were synthesized
and tested (Table 2). The compounds were not sub-
stantially more potent than the monosubstituted ana-
logues. Two chlorines were not better than one (com-
pare 25 with 15 and 18). And in the binding assay,
compound 25 was 6-7-fold weaker than the monochloro
compounds 15 and 18. The 3,5-dichlorophenyl analogue
30 also showed weaker activity than the monochloro
compound 18 in both binding and cotransfection assays.
Compound 26 was less potent than both of its mono-
substituted analogues (15 and 17) but equally effica-
cious relative to the parent compounds, although it
exhibited similar binding affinity as its parent com-
pound. The activity of 27 was very similar to that of
parent compound 15, which implies that the methyl
group in the meta position has little influence on the
activity of 15. The combination of p-fluoro and m-
trifluoromethyl (analogue 28) did not improve the
potency of 24 but resulted in a significant decrease of
the efficacy from 96% to 60%. The 3,5-disubstituted
analogues 31 and 32 showed similar binding affinity but
weaker agonist activity than their corresponding parent
compounds. On the basis of the results from the
disubstituted analogues, we did not pursue the trisub-
stituted 5-aryl analogues.
These monosubstituted 5-aryl analogues were also
examined in the hPR-A binding assay. The EC50 value
(2.9 nM) of progesterone in the cotransfection assay is
the same order of magnitude as its Ki value (3.5 nM) in
the binding assay. This is also true for medroxyproges-
terone acetate (MPA). However, the Ki values of most
analogues in Table 2 are 1 order of magnitude better
than their corresponding EC50 or IC50 values in cellular
assays. Similar to the cotransfection assay results, the
The introduction of a heteroatom into the 5-aryl
moiety was accomplished, and the in vitro data of two
representative analogues (33 and 34) are listed in Table
2. Both of the analogues showed that the nitrogen atom
in the 5-aryl ring switched the activity from agonists

Novel Class of hPR Agonists
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 295
Ta ble 3. Cotransfection and Competitive Binding Data for
ER diminished and the selectivity ratio of PR over other
IRs significantly increased. In the competitive binding
assays, the dihydroquinoline PR agonist series have
much less cross-reactivity with AR and GR than proges-
terone and MPA. The results demonstrate that the
cross-reactivities of the dihydroquinoline PR agonist
series can be manipulated and potentially could gener-
ate more selective progestins.
The in vivo experiments were conducted to define the
oral activity of the novel hPR agonist series in two target
organs of the rat: the mammary gland and the uterus.
The approach used in these experiments has the ad-
vantage of simultaneously evaluating the inhibition of
estrone-induced uterine wet weight increases and mam-
mary gland differentiation (lobular bud formation) by
progestins in the same ovariectomized animal.
Optically Active 5-Arylquinoline Analoguesa
hPR agonist,b
mean ( SEM
hPR-A binding
Ki (nM),
mean ( SEM
optical
compd rotation
eff (%)
pot (EC50, nM)c
15
35
36
22
37
38
(
+
-
(
+
-
77 ( 5
32 ( 2*
77 ( 21
59 ( 5
39 ( 5*
73 ( 5
14 ( 2
0.70 ( 0.14
19.1 ( 2.1
0.38 ( 0.04
2.3 ( 0.5
43.1 ( 7.4
1.4 ( 0.4
165 ( 66*
7.0 ( 4.3
37 ( 11
192 ( 18*
22 ( 12
a
Values with standard errors (SEM) represent the mean value
of at least three separate experiments with triplicate determina-
tions, and values with an asterisk represent the mean value of
b
two experiments with standard deviations. Agonist efficacies
were compared to that of progesterone (100%). c All EC50 values
were determined from full dose-response curves ranging from
10-12 to 10-5 M in CV-1 cells.
As expected uterine wet weight increased after 3 days
of treatment by estrone (E) (10 µg/rat) by approximately
4-fold over vehicle-treated animals in all experiments
(data not shown). Addition of MPA to E-treated animals
significantly (p < 0.05, ANOVA) decreased estrogen-
induced uterine wet weight 30% compared to E-treated
alone animals (Figure 1).28 Quinoline derivatives 13,
15, and 24 were as efficacious as MPA in inhibiting
estrogen-induced uterine wet weights in 3.0 mg/animal
dose.
Mammary gland differentiation by progestins was
observed in the quinoline analogue-treated animals in
the ovariectomized rat (Figure 2). MPA and the three
tested analogues (3 mg/rat) all significantly (p < 0.05,
ANOVA) increased the average number of lobular buds
when compared to E-treated alone animals. In contrast
with the results in uterus, compounds 13, 15, and 24
showed different efficacies in stimulating the percentage
of lobular bud observed over E-treated alone animals.
Compound 13 was surprisingly more efficacious than
MPA, while compound 24 was much less efficacious.
These results indicate that compound 24 might be a
tissue-selective progestin, which has full efficacy in
uterus but lower efficacy in breast.
(16 and 19) to antagonists along with about a 10-fold
decrease of the binding affinities.
The 5-arylquinolines 2 contain one stereogenic center.
To address the chirality issue, two racemic compounds,
one agonist (15) and one partial agonist (22), were
resolved by chiral HPLC, and the assay results of the
enantiomers are reported in Table 3. The absolute
stereochemistry of these enantiomers has not been
determined. The (-)-enantiomers, 36 and 38, showed
superior potency or efficacy relative to their racemic
mixtures, 15 and 22. The (+)-enantiomers, 35 and 37,
showed much less activity in the assays. Theoretically,
when one enantiomer contributes to most of the potency,
its potency is about twice that of the racemate. This is
true for enantiomer 36 versus racemate 15.
The cross-reactivity profiles of this series of analogues
with other steroid hormone receptors were determined
in cotransfection assays and receptor binding assays
(Tables 4 and 5). None of the nonsteroidal PR agonist
analogues showed agonist activity in hAR, hGR, hMR,
and hER cotransfection assays, and they showed weak
antagonist activities (micromolar IC50 values) in the
closely related IR assays. Parent compound 13 has
moderate antagonist activities with all IRs tested (PR,
AR, GR, ER, MR); however, when the PR agonist
activity was optimized by introducing substituents at
the 5-phenyl position, the cross-reactivities with MR and
The novel nonsteroidal compounds 2 showed proges-
tational activities in both in vitro and in vivo assays,
which inspires a series of questions of how they mimic
the steroids to interact with the receptors. No crystal
Ta ble 4. Antagonist Cross-Reactivities with hAR, hGR, hER, and hMRa
hAR, mean ( SEM
eff (%)b IC50 (nM)c
hGR, mean ( SEM
eff (%)
IC50 (nM)c
>1000
hMR, mean ( SEM
hER, mean ( SEM
compd
eff (%)
IC50 (nM)c
eff (%)
IC50 (nM)c
d
progesterone
46 ( 7
(159 ( 10
78 ( 5*
85 ( 1*
90 ( 3
88 ( 8
90 ( 7
89 ( 5
90 ( 5
90 ( 4
77 ( 10
88 ( 6
82 ( 4
83 ( 8
84 ( 9
37 ( 2
39 ( 8
157 ( 22
98 ( 2
97 ( 2
98 ( 0*
98 ( 2
99 ( 1
97 ( 2
96 ( 3
99 ( 11
99 ( 1
99 ( 1
98 ( 1
91 ( 4
97
83 ( 6
67 ( 10
90 ( 1
73 ( 3
76 ( 5*
80 ( 6
72 ( 9
70 ( 8
70 ( 16
70 ( 5
70 ( 12
58 ( 7
56 ( 4
-
14 ( 4
1197 ( 852
354 ( 85
1525 ( 504
977 ( 706*
748 ( 580
1068 ( 406
1592 ( 460
2219 ( 644
1062 ( 945
1358 ( 299
2035 ( 1023
1000 ( 390
-
-
-
MPA
13
15
16
18
19
22
24
25
26
27
29
30
32
6.1 ( 1.0
10 ( 1)e
226 ( 55
339 ( 154
115 ( 27*
136 ( 52
266 ( 103
294 ( 161
505 ( 320
80 ( 45
171 ( 39
275 ( 63
185 ( 74
225 ( 84
166
(46 ( 5
89 ( 7
43 ( 23
-
924 ( 203)e
299 ( 85
1803 ( 385
-
745 ( 285*
1550 ( 538*
481 ( 127
253 ( 156
401 ( 13
1239 ( 314
315 ( 42
946 ( 419
2205 ( 925
1426 ( 187
1220 ( 296
1089 ( 351
1236 ( 603
76 ( 10
73 ( 16
53 ( 12*
44 ( 22
52 ( 18
52 ( 14
58 ( 7
54 ( 3
-
1480 ( 691
1095 ( 394
2395 ( 1081*
1537 ( 185
2632 ( 781
1449 ( 382
1220 ( 491
1449 ( 194
-
-
-
49
1037
a
Values with standard errors (SEM) represent the mean value of at least three separate experiments with triplicate determinations,
b
and values with an asterisk represent the mean value of two experiments with standard deviation. Antagonist efficacies were determined
as a function (%) of maximal inhibition of an agonist. c All IC50 values were determined from full dose-response curves ranging from
10-12 to 10-5 M in CV-1 cells. Stands for efficacy < 20% and potency > 10 000 nM. e Agonist efficacy, EC50 (nM).
d

296 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Zhi et al.
Ta ble 5. Competitive Binding Data of 5-Arylquinoline
Analogues with hPR, hAR, and hGRa
Ki (nM), mean ( SEM
compd
hPR
hAR
hGR
progesterone
3.5 ( 0.2
0.34 ( 0.04
3.6 ( 0.7
8.5 ( 3.1
2.9 ( 0.2
30.5 ( 1.9
13.2 ( 1.8
154 ( 16
152 ( 16
67 ( 3
MPA
13
15
16
18
19
22
24
25
26
27
29
30
32
1856 ( 251
2685 ( 312
0.70 ( 0.14
0.55 ( 0.21
0.43 ( 0.11
1.1 ( 0.6
>10000
1526 ( 360
2138 ( 341
2330 ( 166
795 ( 95
58 ( 21
34 ( 4
2.3 ( 0.5
80 ( 36
44 ( 12
59 ( 22
39 ( 8
147 ( 52
60 ( 22
100 ( 16
42 ( 7
2.6 ( 0.2
3.4 ( 0.7
5831 ( 4168
1457 ( 611
1860 ( 364
1581 ( 602
2584 ( 22
3499 ( 451
0.41 ( 0.08
1.2 ( 0.2
1.8 ( 0.8
1.7 ( 0.2
0.87 ( 0.03
a
Values with standard errors (SEM) represent the mean value
of at least three separate experiments on receptor expressed in
SF12 cells in a baculovirus expression system.
F igu r e 2. Stimulation of lobular alveolar bud formation in
the ovariectomized rat by quinoline derivatives or MPA (3.0
mg/rat, n ) 4). All values represent the mean percent change
( SEM of lobular alveolar buds from animals treated with E
alone (*p < 0.05 vs E ANOVA).
fields of progesterone were used as a template to fit
alignment.
In attempting to understand the relationship of these
compounds with their steroidal counterparts, it was
decided to use a predictive cross-validated r2 in a
CoMFA analysis of a set of steroidal and nonsteroidal
progesterone agonists as the criteria for judging the
validity of these overlaps. Attempts to manually over-
lap isosteric moieties were performed, and none of these
produced an alignment which led to a favorable cross-
validated r2. Manual attempts were made to overlap
these structures with progesterone and other steroidal
progesterone agonists. None of these produced signifi-
cant correlation with the cotransfection EC50 values in
a CoMFA study. When the alignment generated from
the CoMFA field of progesterone was used, the cross-
validated r2 obtained was predictive and correlative.31
The modeling results can be understood by the align-
ments shown in Figure 3. Progesterone can be seen in
its alignment with 2. The quinoline ring overlays with
the steroid D-ring, and the aromatic D-ring of the
quinoline 2 overlays with the steroid A-ring. By the
modeling alignment, the 5-position of quinoline 2 over-
lays with the 7-position of the steroid skeleton rather
than the 11-position. Figure 3 represents a possible
explanation of the orientation of our nonsteroidal ago-
nists with respect to the native hormone.
F igu r e 1. Inhibition of estrone-induced uterine wet weight
in the ovariectomized rat by quinoline derivatives or MPA (3.0
mg/rat, n ) 4). All values represent the mean percent change
( SEM of uterine wet weight from animals treated with E
alone (*p < 0.05 vs E ANOVA).
structure of the ligand-binding domain exists for PR;
consequently, other less direct methods must be em-
ployed to delineate the physical alignment between the
steroidal and nonsteroidal progesterone agonists. One
method to gain possible insight into the correlation of
molecular structure with the biological response is
CoMFA.29 CoMFA correlates a three-dimensional grid
of steric and electrostatic interaction energies between
a series of overlapped molecules and two probe atoms,
an sp3 carbon atom and a proton. The results of the
CoMFA modeling will be discussed in a subsequent
paper.30 However to force a physically meaningful
alignment of the nonsteroidal PR compounds, a CoMFA
model of six steroidal PR agonists including progester-
one was developed. When the predictive ability of that
CoMFA model was verified, the steric and electrostatic
Con clu sion
A novel series of nonsteroidal hPR agonists, 5-aryl-
1,2-dihydro-2,2,4-trimethyl-5H-chromeno[3,4-f]quino-
lines 2, was developed by using cotransfection and
binding assays as guides. This 5-aryl series exhibits
biological activity (EC50 ) 8-30 nM) similar to that of
the natural hormone progesterone (EC50 ) 2.9 nM) in
cotransfection assays with efficacies ranging from 28%

Novel Class of hPR Agonists
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 297
F igu r e 3. Top and side views of progesterone and 13 showing the alignment rule. All other compounds overlap with the same
rule.
1,2-Dih ydr o-2,2,4-tr im eth yl-10-isocou m ar in o[3,4-f]qu in -
olin e (4) an d 1,2-Dih ydr o-2,2,4-tr im eth yl-10-isocou m ar in o-
[4,3-g]qu in olin e (5). A solution of 3-amino-6H-dibenzo[b,d]-
pyran-6-one (3) (185 mg, 0.87 mmol) and iodine (25 mg, 0.10
mmol, 11 mol %) in acetone (20 mL) was heated at 100 °C in
a sealed tube for 15 h. The mixture was concentrated, and
chromatography (30- × 150-mm column, hexane-EtOAc, 10:1
to 2:1 gradient) afforded 4 (150 mg, 60%) and 5 (75 mg, 30%)
to 96%. Most analogues of the series have the same or
greater binding affinity (Ki ) 0.41-3.6 nM) than
progesterone (Ki ) 3.5 nM), including 5-chloro- and
5-bromophenyl compounds (15, 16, 18, and 19). The
novel pharmacophore of the 5H-chromeno[3,4-f]quino-
line provides a number of opportunities for further
optimization. The results of the D-ring modifications
of the agonist series will be published in a separate
report.32 The novel structure of 5-arylquinoline 2 is
quite distinct from the structure of the steroidal PR
agonists. Our molecular modeling results suggest a
possible overlap of the quinoline A-ring in 2 with the
steroid D-ring, which may be how they mimic steroidal
interaction with the receptors.
1
as yellow solids. Data for 4: mp 197-199 °C; H NMR (400
MHz, acetone-d6) 8.20 (d, J ) 7.6, H7), 8.10 (d, J ) 7.6, H11),
7.83 (d, J ) 8.6, H12), 7.77 (t, J ) 7.6, H9), 7.44 (t, J ) 7.6, H8),
6.64 (d, J ) 8.6, H10), 5.88 (br s, NH), 5.38 (s, H3), 2.39 (s, 3 ×
H
4a), 1.29 (s, 6 × H2a); 13C NMR (100 MHz, CDCl3) 161.5, 149.6,
146.9, 136.6, 134.8, 130.4, 128.3, 126.6, 122.8, 120.8, 119.0,
111.1, 109.6, 108.6, 51.7, 30.4, 23.6; IR (KBr) 3350 br, 2960
m, 1711 s, 1608 s, 1566 m, 1468 m, 1311 m. Anal. (C19H17
-
NO2) C, H, N. Data for 5: mp 246-248 °C; 1H NMR (400 MHz,
CDCl3) 8.30 (d, J ) 7.6, H9), 7.93 (d, J ) 7.6, H6), 7.72 (t, J )
7.6, H7), 7.63 (s, H5), 7.41 (t, J ) 7.6, H8), 6.36 (s, H12), 5.40 (s,
H3), 4.13 (br s, NH), 2.10 (s, 3 × H4a), 1.33 (s, 6 × H2a); 13C
NMR (100 MHz, CDCl3) 162.2, 152.7, 146.1, 136.4, 134.9,
130.7, 129.1, 127.2, 126.5, 120.5, 119.4, 119.0, 117.8, 107.6,
99.8, 52.7, 31.8, 19.0. Anal. (C19H17NO2) C, H, N.
Exp er im en ta l Section
Gen er a l Exp er im en ta l Meth od s. 1H and 13C NMR
spectra were obtained on a Bruker AC400 spectrometer at 400
and 100 MHz, respectively, using CDCl3 as solvent and TMS
(0.00 ppm 1H, 0.00 ppm 13C) or CHCl3 (7.26 ppm 1H, 77.00
ppm 13C) as internal standards, unless otherwise noted.
Chemical shifts (δ) are given in parts per million (ppm), and
coupling constants (J ) are given in hertz (Hz). Selected data
are reported in the following order: chemical shift, coupling
constants, and assignment. Infrared (IR) spectra were re-
corded on a Mattson Galaxy Series 3000 FT infrared spec-
trometer. Liquid samples were measured as neat films on
NaCl plates; solid samples were measured as KBr pellets.
Peaks are reported (cm-1) with the following relative intensi-
ties: s (strong, 70-100%), m (medium, 40-70%), w (weak, 20-
40%), br (broad). Elemental analyses were performed by
Oneida Research Services, Inc., Whitesboro, NY, or Galbraith
Laboratories, Inc., Knoxville, TN. Melting points were taken
on an Electrothermal IA9100 digital apparatus and are
uncorrected. Optical rotations were obtained on a Perkin-
1,2-Dih yd r o-2,2,4-t r im et h yl-5-cou m a r in o[3,4-f]q u in o-
lin e (1). The intermediate 8-nitro-6H-dibenzo[b,d]pyran-6-
one (10) was prepared by a modified literature procedure.23
To a 50-mL round-bottom flask charged with 2-biphenylcar-
boxylic acid (8) (5.0 g, 25 mmol) was added 10 mL of 70% nitric
acid, and the resulting yellow slurry was stirred at room
temperature (rt) for 30 min.34 The second nitration was
performed in two procedures (A and B). For the one-pot
operation, procedure A, 25 mL of fuming nitric acid was
directly introduced to the above reaction mixture dropwise in
20 min, giving rise to a clear yellow solution. The reaction
mixture was stirred at rt for 15 h and was poured into ice
water (100 mL). The crude mixture was extracted with ethyl
acetate (EtOAc) (3 × 60 mL), and the combined extract was
washed with water (2 × 20 mL) and brine (2 × 20 mL).
Removal of solvent under reduced pressure afforded a crude
yellow solid, which was a 2:1 mixture of two regioisomers, 4,2′-
dinitrobiphenyl-2-carboxylic acid (9a ) and 4,4′-dinitrobiphenyl-
2-carboxylic acid (9b). For the alternative procedure B, the
mixture of the first nitration was poured into ice water (50
mL) and the products were precipitated out from the solution.
Filtration of the mixture afforded the mononitrated products
as a white solid, which was then treated with fuming nitric
acid (10 mL) at rt for 6 h followed by the same workup
procedure as described in procedure A to afford the same ratio
mixture of dinitrated products 9a ,b in quantitative yield.
Elmer 241 polarimeter and are reported as follows: [R]temp
(concentration in g/100 mL, solvent). Flash column
-
wavelength
chromatography refers to the method of Still33 using Merck
230-400 mesh silica gel. Analytical thin-layer chromatogra-
phy (TLC) was performed using Merck 60-F-254 0.25-mm
precoated silica gel plates. Preparative thin-layer chroma-
tography (PTLC) was performed using Merck 60-F-254, 0.50-
or 1.00-mm precoated silica gel plates. Analytical HPLC was
performed on a Beckman System Gold HPLC system. Pre-
parative HPLC was performed on a Waters Delta Prep 4000.
The detector wavelength was set to 254 nm. Ethyl ether and
tetrahydrofuran were distilled directly prior to use from
sodium/benzophenone. Dichloromethane (CH2Cl2), benzene,
and toluene were stored under nitrogen over 4-Å molecular
sieves. “Brine” refers to a saturated aqueous solution of NaCl.
Unless otherwise specified, solutions of common inorganic salts
used in workups are aqueous solutions. Reactions were
conducted under a nitrogen atmosphere unless otherwise
noted. Reported yields are not optimized.
The mixture of the dinitrobiphenylcarboxylic acid was
dissolved in DMA (80 mL), and the solution was heated at
reflux for 12 h. The reaction mixture was cooled to rt and
diluted with water (20 mL). The desired product was precipi-
tated out from the solution upon standing at rt overnight.
Filtration of the mixture afforded the product (2.9 g, 50%) as
a brown solid, which was directly used in the next step without

298 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Zhi et al.
further purification. In a 1-L round-bottom flask, a solution
of 8-nitro-6H-dibenzo[b,d]pyran-6-one (10) (2.9 g, 12 mmol) in
EtOAc (600 mL) was treated with 10% Pd/C (1.0 g, 0.94 mmol,
7.8 mol %) under hydrogen balloon for 24 h. Filtration from
the catalyst and removal of solvent afforded 2.2 g of 8-amino-
6H-dibenzo[b,d]pyran-6-one (86%) as a yellowish solid. A 210-
mL Ace-Thred pressure tube charged with the amino com-
pound (2.2 g, 10.4 mmol), iodine (1.1 g, 4.3 mmol, 41 mol %),
and acetone (150 mL) was sealed with the thred cap at rt. The
tube was heated in an oil bath at 110-130 °C for 24 h. The
dark reaction mixture was concentrated, and purification by
flash chromatography (40- × 200-mm column, hexane-EtOAc,
9:1 to 2:1 gradient) afforded 1.5 g of the title compound 1 (50%)
as a yellow solid: mp 190-191 °C; 1H NMR (400 MHz, CDCl3)
7.90 (d, J ) 7.8, H10), 7.78 (d, J ) 8.4, H11), 7.38-7.22 (m, H7,
H8, H9), 7.01 (d, J ) 8.4, H12), 5.58 (s, H3), 4.31 (br s, NH),
2.12 (s, 3 × H4a), 1.33 (s, 6 × H2a); 13C NMR (100 MHz, CDCl3)
160.3, 150.5, 145.7, 132.4, 131.6, 128.4, 124.2, 122.0, 121.4,
121.2, 119.3, 118.4, 117.2, 50.8, 29.9, 28.6; IR (KBr) 3352 br,
2966 s, 2924 m, 1712 s, 1626 m, 1450 m, 1356 m, 1251 m,
1205 m, 740 m. Anal. (C19H17NO2) C, H, N.
CDCl3) 7.64 (d, J ) 7.8, H9), 7.42 (s, H5), 7.39-7.31 (m, H8,
-C6H5), 7.10 (t, J ) 7.8, H7), 6.79 (d, J ) 7.8, H6), 6.09 (s, H10),
6.08 (s, H12), 5.26 (s, H3), 3.81 (br s, NH), 2.06 (s, 3 × H4a),
1.28 (s, 3 × H2ab), 1.27 (s, 3 × H2aa); 13C NMR (100 MHz, CDCl3)
154.7, 145.4, 140.2, 132.4, 131.2, 128.7, 128.5, 128.4, 128.3,
128.2, 127.0, 126.3, 125.7, 120.7, 118.6, 116.6, 112.2, 101.2,
80.0, 52.3, 31.6, 31.5, 19.0. Anal. (C25H23NO) C, H, N.
(R,S)-1,2-Dih yd r o-2,2,4-t r im et h yl-5-p h en yl-5H -ch r o-
m en o[3,4-f]qu in olin e (13). This compound was prepared by
the general procedure from bromobenzene (0.15 mL, 1.4 mmol,
2.8 equiv) and compound 1 (150 mg, 0.50 mmol) in 68% yield
as a white solid (120 mg): mp 149-150 °C; 1H NMR (400 MHz,
CDCl3) 7.53 (d, J ) 7.8, H10), 7.50 (d, J ) 8.2, H11), 7.22-7.12
(m, -C6H5), 7.00 (t, J ) 7.8, H8), 6.92 (s, H5), 6.88 (t, J ) 7.8,
H9), 6.83 (d, J ) 7.8, H7), 6.69 (d, J ) 8.2, H12), 5.46 (s, H3),
3.92 (br s, NH), 1.99 (s, 3 × H4a), 1.29 (s, 3 × H2ab), 1.26 (s, 3
× H2aa); 13C NMR (100 MHz, acetone-d6) 151.5, 147.0, 141.1,
134.4, 134.3, 130.3, 129.1, 129.0, 128.8, 128.5, 127.9, 125.7,
124.3, 122.6, 122.5, 120.7, 119.6, 118.1, 115.8, 115.7, 76.1, 51.0,
29.1, 29.0, 24.0; IR (neat) 3364 br, 2962 m, 1705 m, 1635 m,
1593 m, 1470 s, 1435 m, 1167 m. Anal. (C25H23NO‚1/8H2O)
C, H, N.
Gen er a l P r oced u r e for P r ep a r in g 5-Ar yl Com p ou n d s
fr om th e La cton e 1. This transformation was a two-step
sequence: an addition of a nucleophile either from a com-
mercial reagent or prepared in situ from a metal-halogen-
exchange reaction and a reduction of the resulting cyclic
hemiacetal. To a solution of a bromo compound in THF or
diethyl ether (0.1-0.3 M) at - 78 °C was added n-BuLi (1.6
M solution in hexane, 1.1 equiv) slowly, and the resulting
reaction mixture was allowed to stir at - 78 °C until the anion
was formed. A yellow solution (0.2-0.5 M) of 1 in THF was
added via cannula to the above solution, and the resulting
dark-red mixture was slowly warmed. As soon as the red color
faded (around - 30 °C), the reaction was quenched im-
mediately with water to give a light-yellow solution (with aryl
Grignard reagents the addition reaction took hours at room
temperature). The reaction mixture was extracted with EtOAc
(2×), and the combined extracts were washed with brine and
concentrated. Purification by flash chromatography (hexane-
EtOAc, 10:1 to 2:1 gradient) afforded the hemiacetal interme-
diate as a yellow oil. To a solution of the intermediate in
dichloromethane (0.10 M) at - 78 °C were added triethylsilane
(5.0 equiv) and either trifluoroacetic acid (1.1 equiv) or BF3-
OEt2 (0.20 equiv), and the resulting slurry was warmed to rt,
giving rise to a dark-green solution. The mixture was allowed
to stir at rt or reflux in some cases until the reaction went to
completion. The reaction then was quenched with 5% NaOH
aqueous solution and extracted with ethyl acetate. The
combined extracts were washed with brine and concentrated.
Purification by flash chromatography (hexane-EtOAc, 10:1
to 2:1 gradient) afforded the 5-aryl-1,2-dihydro-2,2,4-trimethyl-
(R,S)-5-(4-F lu or op h en yl)-1,2-d ih yd r o-2,2,4-t r im et h yl-
5H-ch r om en o[3,4-f]qu in olin e (14). This compound was
prepared by the general procedure from 4-fluorobromobenzene
(175 mg, 1.0 mmol, 14 equiv) and compound 1 (20 mg, 0.068
mmol) in 59% yield as a white solid (15 mg): mp 85-87 °C;
1H NMR (400 MHz, acetone-d6) 7.60 (d, J ) 7.8, H10), 7.56 (d,
J ) 8.3, H11), 7.26 (dd, J ) 8.7, 5.7, H2′, H6′), 6.98 (t, J ) 8.7,
H3′, H5′), 6.97 (t, J ) 7.8, H8), 6.92 (s, H5), 6.87 (t, J ) 7.8, H9),
6.83 (d, J ) 8.3, H12), 6.76 (d, J ) 7.8, H7), 5.54 (br s, NH),
5.47 (s, H3), 1.99 (s, 3 × H4a), 1.26 (s, 3 × H2ab), 1.24 (s, 3 ×
H
2aa); 13C NMR (100 MHz, CDCl3) 163.0 (d, J ) 245, C4′), 151.3,
147.1, 137.2, 134.4, 131.3, 131.2, 130.2, 129.0, 128.0, 125.7,
124.4, 122.7, 120.6, 119.5, 118.2, 115.9, 115.5 (d, J ) 22, C3′,5′),
75.5, 51.0, 29.7, 29.1, 24.0; IR (KBr) 3360 br, 2962 m, 1707 m,
1601 m, 1506 s, 1469 s, 1221 m, 1157 m. Anal. (C25H22FNO‚1/
4H2O) C, H, N.
(R,S)-5-(4-Ch lor op h en yl)-1,2-d ih yd r o-2,2,4-t r im et h yl-
5H-ch r om en o[3,4-f]qu in olin e (15). This compound was
prepared by the general procedure from 4-bromochlorobenzene
(1.4 g, 7 mmol, 4.1 equiv) and compound 1 (0.50 g, 1.7 mmol)
1
in 40% yield as a white solid (264 mg): mp 139-140 °C; H
NMR (400 MHz, acetone-d6) 7.59 (d, J ) 7.8, H10), 7.56 (d, J
) 8.4, H11), 7.24 (d, J ) 9.1, H3′, H5′), 7.21 (d, J ) 9.1, H2′, H6′),
6.98 (t, J ) 7.8, H8), 6.92 (s, H5), 6.86 (t, J ) 7.8, H9), 6.83 (d,
J ) 8.4, H7), 6.77 (d, J ) 7.8, H12), 5.54 (br s, NH), 5.48 (s,
H3), 1.99 (s, 3 × H4a), 1.26 (s, 3 × H2ab), 1.24 (s, 3 × H2aa); 13
C
NMR (100 MHz, CDCl3) 137.8, 134.5, 131.0, 129.8, 128.2,
124.5, 75.6, 61.1, 27.9, 22.5, 14.2; IR (KBr) 3371 br, 2964 m,
1593 m, 1469 m, 1435 m. Anal. (C25H22ClNO) C, H, N.
5H-chromeno[3,4-f]quinoline in moderate to good yield.
A
(R,S)-5-(4-Br om op h en yl)-1,2-d ih yd r o-2,2,4-t r im et h yl-
5H-ch r om en o[3,4-f]qu in olin e (16). This compound was
prepared by the general procedure from 1,4-dibromobenzene
(250 mg, 1.0 mmol, 14 equiv) and compound 1 (20 mg, 0.069
mmol) in 54% yield as a white solid (16 mg): mp 97-99 °C;
1H NMR (400 MHz, acetone-d6) 7.58 (d, J ) 7.8, H10), 7.55 (d,
J ) 8.4, H11), 7.39 (d, J ) 8.5, H3′, H5′), 7.16 (d, J ) 8.5, H2′,
H6′), 6.98 (t, J ) 7.8, H8), 6.90 (s, H5), 6.86 (t, J ) 7.8, H9),
6.83 (d, J ) 8.4, H12), 6.77 (d, J ) 7.8, H7), 5.54 (br s, NH),
5.47 (s, H3), 1.99 (s, 3 × H4a), 1.26 (s, 3 × H2ab), 1.23 (s, 3 ×
second flash chromatography might be needed to remove the
silane oxide and the byproduct using a 1:2 mixture of dichlo-
romethane and hexane as eluents.
(R,S)-1,2-Dih yd r o-2,2,4-t r im et h yl-6-p h en yl-6H -isoch -
r om en o[3,4-f]qu in olin e (6). This compound was prepared
by a similar addition-reduction sequence as described in the
general procedure from bromobenzene (157 mg, 1.0 mmol, 4.7
equiv) and lactone 4 (60 mg, 0.21 mmol) in a 85% two-step
yield as a colorless oil (60 mg): 1H NMR (400 MHz, CDCl3)
7.60 (d, J ) 7.7, H7), 7.46-7.30 (m, -C6H5, H8, H11), 7.09 (t, J
) 7.7, H9), 6.73 (d, J ) 7.7, H12), 6.26 (d, J ) 8.2, H10), 6.04 (s,
H6), 5.22 (s, H3), 3.86 (br s, NH), 2.11 (s, 3 × H4a), 1.26 (s, 3 ×
H2ab), 1.23 (s, 3 × H2aa); 13C NMR (100 MHz, CDCl3) 152.0,
146.3, 145.2, 139.6, 132.9, 131.6, 129.3, 128.5, 128.4, 125.9,
125.6, 123.5, 121.2, 113.9, 111.2, 108.5, 80.1, 51.4, 30.2, 30.1,
23.5. Anal. (C25H23NO) C, H, N.
(R,S)-1,2-Dih yd r o-2,2,4-t r im et h yl-10-p h en yl-10H -iso-
cou m a r in o[4,3-g]qu in olin e (7). This compound was pre-
pared by a similar addition-reduction sequence as described
in the general procedure from bromobenzene (157 mg, 1.0
mmol) and quinoline 5 (100 mg, 0.34 mmol) in 83% yield as a
white solid (110 mg): mp 70-72 °C; 1H NMR (400 MHz,
H
2aa); 13C NMR (100 MHz, acetone-d6) 151.4, 147.2, 140.7,
134.6, 132.0, 131.9, 131.4, 129.9, 129.0, 128.1, 125.7, 124.5,
122.8, 122.3, 120.6, 119.6, 118.2, 116.1, 75.6, 51.1, 29.8, 29.2,
24.1; IR (neat) 3364 br, 2962 m, 1699 m, 1593 m, 1574 m, 1483
s, 1469 m, 1435 m, 1249 m, 1213 m, 1167 m, 1010 m. Anal.
(C25H22BrNO) C, H, N.
(R,S)-5-(3-F lu or op h en yl)-1,2-d ih yd r o-2,2,4-t r im et h yl-
5H-ch r om en o[3,4-f]qu in olin e (17). This compound was
prepared by the general procedure from 1-bromo-3-fluoroben-
zene (175 mg, 1.0 mmol, 14 equiv) and compound 1 (20 mg,
0.069 mmol) in 47% yield as a white solid (12 mg): mp 78-80
°C; 1H NMR (400 MHz, acetone-d6) 7.60 (d, J ) 7.9, H10), 7.57
(d, J ) 8.4, H11), 7.26 (td, J ) 7.6, 5.9, H5′), 7.06 (d, J ) 7.6,

Novel Class of hPR Agonists
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 299
H6′), 7.01-6.81 (m, H7, H8, H9, H12, H2′, H4′), 6.95 (s, H5), 5.58
(br s, H3), 5.49 (s, H3), 2.02 (s, 3 × H4a), 1.27 (s, 3 × H2ab), 1.25
(s, 3 × H2aa); 13C NMR (100 MHz, acetone-d6) 163.5 (d, J )
245, C3′), 151.4, 147.2, 144.3, 134.7, 130.7, 129.9, 128.9, 128.2,
125.6, 125.2, 124.5, 122.8, 120.6, 119.6, 118.3, 116.0, 115.8,
115.6, 115.3, 75.5, 51.1, 29.6, 24.0; IR (neat) 3369 br, 2964 m,
1699 m, 1633 s, 1581 s, 1498 s, 1342 m, 1253 m, 1167 m. Anal.
(C25H22FNO‚1/8H2O) C, H, N.
138.0, 134.3, 130.6, 129.4, 129.2, 127.9, 125.7, 124.2, 122.6,
122.4, 120.7, 119.5, 118.1, 115.8, 115.7, 76.1, 50.9, 24.0, 20.9;
IR (KBr) 3362 br, 2964 m, 1707 m, 1593 m, 1469 s, 1437 m,
1259 m, 1169 m. Anal. (C26H25NO) C, H, N.
(R,S)-1,2-Dih ydr o-5-(4-m eth oxyph en yl)-2,2,4-tr im eth yl-
5H-ch r om en o[3,4-f]qu in olin e (23). This compound was
prepared by the general procedure from 4-bromoanisole (187
mg, 1.0 mmol, 14 equiv) and compound 1 (20 mg, 0.069 mmol)
in 48% yield as a colorless oil (13 mg); 1H NMR (400 MHz,
acetone-d6) 7.59 (d, J ) 7.7, H10), 7.54 (d, J ) 8.4, H11), 7.13
(d, J ) 8.7, H2′, H6′), 6.95 (t, J ) 7.7, H8), 6.87 (s, H5), 6.85 (t,
J ) 7.7, H9), 6.81 (d, J ) 8.4, H12), 6.75 (d, J ) 8.7, H3′, H5′),
6.74 (d, J ) 7.7, H7), 5.47 (br s, NH), 5.45 (s, H3), 3.69 (s,
OCH3), 1.99 (s, 3 × H4a), 1.25 (s, 3 × H2ab), 1.23 (s, 3 × H2aa);
13C NMR (100 MHz, acetone-d6) 160.1, 151.6, 147.0, 134.2,
132.9, 130.5, 127.9, 125.8, 124.2, 122.6, 122.4, 120.7, 119.4,
118.1, 115.7, 114.3, 114.1, 76.0, 55.3, 50.9, 29.7, 24.0; IR (neat)
3366 br, 2960 m, 1608 m, 1510 m, 1469 s, 1249 s, 1168 m.
Anal. (C26H25NO2) C, H, N.
(R,S)-1,2-Dih yd r o-2,2,4-tr im eth yl-5-[3-(tr iflu or om eth -
yl)p h en yl]-5H-ch r om en o[3,4-f]qu in olin e (24). This com-
pound was prepared by the general procedure from 1-bromo-
3-(trifluoromethyl)benzene (450 mg, 2.0 mmol, 4 equiv) and
compound 1 (146 mg, 0.50 mmol) in 57% yield as a pale-white
solid (120 mg): mp 75-77 °C; 1H NMR (400 MHz, acetone-d6)
7.61 (d, J ) 7.6, H10), 7.60 (d, J ) 9.0, H11), 7.56-7.45 (m, H2′,
H4′, H5′, H6′), 7.04 (s, H5), 6.98 (t, J ) 7.6, H8), 6.89-6.83 (m,
H7, H9, H12), 5.60 (br s, NH), 5.55 (s, H3), 2.02 (s, 3 × H4a),
1.27 (s, 6 × H2a); 13C NMR (100 MHz, CDCl3) 150.5, 145.3,
141.3, 134.3, 131.7, 129.1, 128.7, 128.6, 128.0, 125.2, 124.9,
124.8, 124.5, 124.0, 122.3, 121.3, 120.0, 117.8, 115.7, 75.1, 50.9,
29.5, 29.2, 24.0; IR (neat) 2966 m, 1593 m, 1469 m, 1435 m,
1330 s, 1165 s, 1126 s. Anal. (C26H22F3NO‚1/8H2O) C, H, N.
(R,S)-5-(3-Ch lor op h en yl)-1,2-d ih yd r o-2,2,4-t r im et h yl-
5H-ch r om en o[3,4-f]qu in olin e (18). This compound was
prepared by the general procedure from 3-bromochlorobenzene
(195 mg, 1.0 mmol, 14 equiv) and compound 1 (20 mg, 0.069
mmol) in 52% yield as a white solid (14 mg): mp 122-124 °C;
1H NMR (400 MHz, acetone-d6) 7.61 (d, J ) 7.8, H10), 7.57 (d,
J ) 8.4, H11), 7.28-7.18 (m, H8, H2′, H4′, H5′), 7.00 (t, J ) 7.8,
H9), 6.95 (s, H5), 6.89 (d, J ) 7.8, H7), 6.84 (d, J ) 8.4, H12),
6.82 (d, J ) 8.1, H6′), 5.58 (br s, NH), 5.49 (s, H3), 2.01 (s, 3 ×
H
4a), 1.27 (s, 3 × H2ab), 1.25 (s, 3 × H2aa); 13C NMR (100 MHz,
acetone-d6) 151.3, 143.7, 134.7, 134.4, 130.6, 129.7, 129.0,
128.9, 128.7, 128.3, 128.1, 127.7, 124.4, 122.8, 122.7, 120.5,
119.6, 118.2, 116.1, 75.4, 51.0, 30.6, 24.0; IR (neat) 3366 m,
2968 m, 1591 m, 1572 m, 1468 s, 1433 s, 1259 m, 1207 m,
1167 m, 754 m. Anal. (C25H22ClNO) C, H, N.
(R,S)-5-(3-Br om op h en yl)-1,2-d ih yd r o-2,2,4-t r im et h yl-
5H-ch r om en o[3,4-f]qu in olin e (19). This compound was
prepared by the general procedure from 1,3-dibromobenzene
(250 mg, 1.0 mmol, 14 equiv) and compound 1 (20 mg, 0.069
mmol) in 83% yield as a white solid (25 mg): mp 87-89 °C;
1H NMR (400 MHz, acetone-d6) 7.61 (d, J ) 7.8, H10), 7.57 (d,
J ) 8.4, H11), 7.38 (s, H2′), 7.36 (d, J ) 7.8, H4′), 7.26 (d, J )
7.8, H6′), 7.19 (t, J ) 7.8, H5′), 7.00 (t, J ) 8.3, H8), 6.98 (s, H5),
6.81-6.90 (m, H7, H9, H12), 5.60 (br s, NH), 5.50 (s, H3), 2.01
(s, 3 × H4a), 1.27 (s, 3 × H2ab), 1.25 (s, 3 × H2aa); 13C NMR
(100 MHz, acetone-d6) 151.2, 147.1, 144.0, 134.7, 131.9, 131.6,
130.9, 129.6, 128.8, 128.1, 128.0, 125.4, 124.4, 122.8, 122.7,
122.6, 120.4, 119.6, 118.2, 116.1, 75.3, 51.1, 29.5, 24.0; IR (neat)
3364 br, 2962 m, 1699 m, 1591 m, 1469 s, 1437 m. Anal.
(C25H22BrNO) C, H, N.
(R,S)-1,2-Dih yd r o-2,2,4-tr im eth yl-5-[4-(tr iflu or om eth -
yl)p h en yl]-5H-ch r om en o[3,4-f]qu in olin e (20). This com-
pound was prepared by the general procedure from 4-bromo-
(trifluoromethyl)benzene (130 mg, 1.0 mmol, 14 equiv) and
compound 1 (20 mg, 0.069 mmol) in 35% yield as a white solid
(10 mg): mp 79-81 °C; 1H NMR (400 MHz, acetone-d6) 7.61-
7.56 (m, H10, H11, H3′, H5′), 7.45 (d, J ) 8.3, H2′, H6′), 7.01 (s,
H5), 6.97 (t, J ) 7.7, H8), 6.86 (t, J ) 7.7, H9), 6.85 (d, J ) 8.4,
(R,S)-5-(3,4-Dich lor oph en yl)-1,2-dih ydr o-2,2,4-tr im eth yl-
5H-ch r om en o[3,4-f]qu in olin e (25). This compound was
prepared by the general procedure from 1-bromo-3,4-dichlo-
robenzene (226 mg, 1.0 mmol, 7.3 equiv) and compound 1 (40
mg, 0.13 mmol) in 56% yield as a pale-white solid (31 mg):
mp 86-88 °C; 1H NMR (400 MHz, CDCl3) 7.53 (d, J ) 7.8,
H
10), 7.50 (d, J ) 8.3, H11), 7.29 (s, H2′), 7.23 (d, J ) 8.0, H5′),
7.09 (d, J ) 8.0, H6′), 7.04 (t, J ) 7.5, H8), 6.92 (t, J ) 7.5, H9),
6.85 (d, J ) 8.2, H7), 6.83 (s, H5), 6.71 (d, J ) 8.4, H12), 5.48 (s,
H3), 4.0 (br s, NH), 1.97 (s, 3 × H4a), 1.30 (s, 3 × H2ab), 1.26 (s,
3 × H2aa); 13C NMR (100 MHz, CDCl3) 150.4, 145.3, 140.7,
134.3, 132.6, 132.1, 130.5, 130.3, 128.8, 128.6, 128.0, 127.9,
124.5, 124.0, 122.5, 122.4, 121.2, 119.9, 117.8, 115.8, 74.6, 50.9,
29.8, 29.0, 24.0; IR (KBr) 3377 br, 2962 m, 1467 s, 1435 m,
1257, 816 m, 755 m. Anal. (C25H21Cl2NO‚1/4H2O) C, H, N.
H
12), 6.81 (d, J ) 7.7, H7), 5.57 (br s, NH), 5.49 (s, H3), 1.99 (s,
3 × H4a), 1.27 (s, 3 × H2ab), 1.25 (s, 3 × H2aa); 13C NMR (100
MHz, acetone-d6) 151.2, 147.1, 134.7, 129.8, 129.5, 128.8, 128.1,
125.7, 125.5, 124.5, 122.9, 122.8, 120.5, 119.6, 118.1, 116.1,
116.0, 75.5, 51.0, 29.1, 24.0; IR (neat) 2964 br, 1469 m, 1325
s, 1165 m, 1124 m, 1056 m. Anal. (C26H22F3NO) C, H, N.
(R,S)-5-(4-Ch lor o-3-flu or op h en yl)-1,2-d ih yd r o-2,2,4-tr i-
m eth yl-5H-ch r om en o[3,4-f]qu in olin e (26). This compound
was prepared by the general procedure from 2-chloro-5-
bromofluorobenzene (209 mg, 1.0 mmol, 19 equiv) and com-
pound 1 (15 mg, 0.051 mmol) in 64% yield as a colorless oil
(13 mg): 1H NMR (400 MHz, acetone-d6) 7.61 (dd, J ) 7.7,
1.4, H10), 7.57 (d, J ) 8.3, H11), 7.38 (t, J ) 7.9, H5′), 7.13 (dd,
J ) 10.3, 1.8, H2′), 7.05 (td, J ) 7.7, 1.4, H8), 7.00 (dd, J ) 7.9,
1.8, H6′), 6.93 (s, H5), 6.89 (td, J ) 7.7, 1.4, H9), 6.86 (d, J )
8.3, H12), 6.83 (dd, J ) 7.7, 1.4, H7), 5.62 (br s, NH), 5.50 (s,
(R,S)-5-(4-Acet ylp h en yl)-1,2-d ih yd r o-2,2,4-t r im et h yl-
5H-ch r om en o[3,4-f]qu in olin e (21). This compound was
prepared by the general procedure from 2-(4-bromophenyl)-
2-methyl-1,3-dioxane (219 mg, 1.0 mmol, 9.7 equiv) and
compound 1 (30 mg, 0.10 mmol) in 30% yield as a colorless oil
(12 mg): 1H NMR (400 MHz, acetone-d6) 7.83 (d, J ) 8.3, H2′,
H6′), 7.60 (d, J ) 7.6, H10), 7.57 (d, J ) 8.4, H11), 7.36 (d, J )
8.3, H3′, H5′), 6.99 (s, H5), 6.98 (t, J ) 7.6, H9), 6.89-6.79 (m,
H7, H8, H12), 5.56 (br s, NH), 5.50 (s, H3), 2.49 (s, COCH3),
2.00 (s, 3 × H4a), 1.28 (s, 3 × H2aa), 1.25 (s, 3 × H2ab). Anal.
(C27H25NO2) C, H, N.
H3), 2.02 (s, 3 × H4a), 1.27 (s, 3 × H2ab), 1.25 (s, 3 × H2aa); 13
C
NMR (100 MHz, acetone-d6) 157.1 (d, J ) 247.2, C3′), 150.3,
146.4, 142.3 (d, J ) 6.5), 133.9, 130.3, 128.6, 128.0, 127.4,
125.4, 124.6, 123.7, 122.2, 122.0, 119.6, 118.8, 117.4, 116.5 (d,
J ) 22.4), 115.4, 74.2, 50.2, 28.3, 23.2; IR (neat) 3371 br, 2964
m, 1581 m, 1469 s, 1242 m. Anal. (C25H21ClFNO‚1/4 H2O) C,
H, N.
(R,S)-1,2-Dih yd r o-2,2,4-tr im eth yl-5-(4-m eth ylp h en yl)-
5H-ch r om en o[3,4-f]qu in olin e (22). This compound was
prepared by the general procedure from 4-bromotoluene (171
mg, 1.0 mmol, 14 equiv) and compound 1 (20 mg, 0.069 mmol)
in 58% yield as a white solid (15 mg): mp 83-85 °C; 1H NMR
(400 MHz, acetone-d6) 7.58 (d, J ) 7.9, H10), 7.54 (d, J ) 8.5,
(R,S)-5-(4-Ch lor o-3-m et h ylp h en yl)-1,2-d ih yd r o-2,2,4-
tr im eth yl-5H-ch r om en o[3,4-f]qu in olin e (27). This com-
pound was prepared by the general procedure from 5-bromo-
2-chlorotoluene (206 mg, 1.0 mmol, 14 equiv) and 1 (20 mg,
0.069 mmol) in 82% yield as a pale-white solid (22 mg): mp
89-91 °C; 1H NMR (400 MHz, acetone-d6) 7.60 (d, J ) 7.7,
H
11), 7.10 (d, J ) 8.0, H2′, H6′), 7.00 (d, J ) 8.0, H3′, H5′), 6.97
(t, J ) 7.9, H8), 6.89 (s, H5), 6.84 (d, J ) 7.9, H9), 6.81 (d, J )
8.5, H12), 6.75 (d, J ) 7.9, H7), 5.47 (br s, NH), 5.45 (s, H3),
2.19 (s, 3 × H4′a), 1.99 (s, 3 × H4a), 1.25 (s, 3 × H2ab), 1.23 (s,
3 × H2aa); 13C NMR (100 MHz, acetone-d6) 151.6, 147.0, 138.1,
H
10), 7.55 (d, J ) 8.4, H11), 7.21 (d, J ) 7.9, H5′), 7.20 (s, H2′),
7.01 (d, J ) 7.9, H6′), 6.98 (t, J ) 7.7, H8), 6.89 (s, H5), 6.87 (t,
J ) 7.7, H9), 6.83 (d, J ) 8.4, H12), 6.78 (d, J ) 7.7, H7), 5.55

300 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Zhi et al.
(br s, NH), 5.48 (s, H3), 2.22 (s, 3 × H3′a), 2.00 (s, 3 × H4a),
1.26 (s, 3 × H2ab), 1.24 (s, 3 × H2aa); 13C NMR (100 MHz,
acetone-d6) 151.5, 147.2, 140.3, 136.2, 134.6, 134.2, 132.0,
130.1, 129.4, 129.1, 128.4, 128.1, 125.7, 124.5, 122.8, 120.6,
119.7, 118.3, 116.1, 116.0, 75.6, 51.1, 29.6, 29.2, 24.1, 20.1; IR
(KBr) 3369 br, 2960 m, 1593 m, 1467 s, 1435 s, 1323 m, 1257
m, 1047 m, 762 m. Anal. (C26H24ClNO) C, H, N.
(R,S)-5-[4-F lu or o-3-(tr iflu or om eth yl)p h en yl]-1,2-d ih y-
dr o-2,2,4-tr im eth yl-5H-ch r om en o[3,4-f]qu in olin e (28). This
compound was prepared by the general procedure from
5-bromo-2-fluorobenzotrifluoride (243 mg, 1.0 mmol, 14 equiv)
and compound 1 (20 mg, 0.069 mmol) in 61% yield as a pale-
white solid (18 mg): mp 74-76 °C; 1H NMR (400 MHz,
acetone-d6) 7.62 (d, J ) 7.7, H10), 7.61 (d, J ) 8.3, H11), 7.61-
7.53 (m, H2′, H6′), 7.27 (t, J ) 7.7, H5′), 7.04 (s, H5), 7.02 (t, J
) 7.7, H8), 6.89 (t, J ) 7.7, H9), 6.86 (d, J ) 8.3, H12), 6.83 (d,
J ) 7.7, H7), 5.62 (br s, NH), 5.51 (s, H3), 2.02 (s, 3 × H4a),
1.26 (s, 6 × H2a); 13C NMR (100 MHz, acetone-d6) 159.7 (d, J
) 255), 151.0, 147.3, 138.3 (d, J ) 4.0), 135.5 (d, J ) 8.9), 134.9
(d, J ) 4.0), 129.3, 128.8, 128.3, 127.7 (d, J ) 4.0), 125.4, 124.6,
123.1, 122.9, 122.8 (q, J ) 272), 120.4, 119.6, 118.2, 117.8,
117.6, 116.3 (d, J ) 5.0), 74.8, 51.1, 29.4, 29.3, 24.0; IR (KBr)
3371 br, 2966 m, 1595 m, 1502 s, 1467 s, 1323 m, 1163 m.
Anal. (C26H21F4NO) C, H, N.
7.7, H10), 7.56 (d, J ) 8.4, H11), 7.17 (s, H2′), 7.14 (s, H4′), 7.10
(s, H6′), 7.01 (t, J ) 7.7, H8), 6.91 (s, H5), 6.90-6.82 (m, H7,
H9, H12), 5.58 (br s, NH), 5.50 (s, H3), 2.21 (s, 3 × H5′a), 2.02 (s,
3 × H4a), 1.27 (s, 3 × H2ab), 1.26 (s, 3 × H2aa); 13C NMR (100
MHz, acetone-d6) 151.4, 147.2, 143.9, 141.1, 134.8, 132.2, 129.8,
129.2, 129.1, 128.8, 128.2, 125.5, 124.5, 122.9, 122.8, 122.6,
120.6, 119.8, 118.3, 116.2, 75.4, 51.1, 24.0, 21.1; IR (neat) 3367
br, 2960 m, 1699 m, 1469 s, 1253 m. Anal. (C26H24BrNO) C,
H, N.
(R,S)-5-(4-Br om o-3-pyr idyl)-1,2-dih ydr o-2,2,4-tr im eth yl-
5H-ch r om en o[3,4-f]qu in olin e (33). This compound was
prepared by the general procedure from 2,5-dibromopyridine
(237 mg, 1.0 mmol, 14 equiv) and compound 1 (20 mg, 0.069
mmol) in 46% yield as a colorless oil (14 mg): 1H NMR (400
MHz, acetone-d6) 8.24 (d, J ) 5.2, H5′), 7.62 (dd, J ) 7.7, 1.3,
H
10), 7.57 (d, J ) 8.4, H11), 7.34 (s, H2′), 7.27 (d, J ) 6.5, H6′),
7.06 (td, J ) 7.7, 1.3, H8), 6.97 (s, H5), 6.94-6.88 (m, H7, H9),
6.86 (d, J ) 8.4, H12), 5.68 (br s, NH), 5.55 (s, H3), 2.06 (s, 3 ×
H
4a), 1.29 (s, 3 × H2ab), 1.28 (s, 3 × H2aa); IR (neat) 3373 br,
2962 m, 1593 s, 1574 s, 1467 s, 1429 s, 1120 m. Anal. (C24H21
-
BrN2O‚1/8H2O) C, H, N.
(R,S)-5-(3-Br om o-2-pyr idyl)-1,2-dih ydr o-2,2,4-tr im eth yl-
5H-ch r om en o[3,4-f]qu in olin e (34). This compound was
prepared by the general procedure from 2,6-dibromopyridine
(237 mg, 1.0 mmol, 14 equiv) and compound 1 (20 mg, 0.069
mmol) in 67% yield as a white solid (20 mg): mp 89-91 °C;
1H NMR (400 MHz, acetone-d6) 7.63 (dd, J ) 7.8, 1.5, H10),
7.54 (d, J ) 8.5, H11), 7.53 (t, J ) 7.8, H5′), 7.39 (d, J ) 7.9,
H6′), 7.13 (d, J ) 7.6, H4′), 7.03 (t, J ) 7.6, H8), 6.91 (t, J ) 7.6,
H9), 6.90 (s, H5), 6.83 (d, J ) 7.6, H7), 6.81 (d, J ) 8.5, H12),
5.52 (br s, NH), 5.48 (s, H3), 2.03 (s, 3 × H4a), 1.25 (s, 3 × H2ab),
1.24 (s, 3 × H2aa); 13C NMR (100 MHz, acetone-d6) 161.7, 151.3,
147.4, 142.2, 140.1, 134.4, 129.4, 129.2, 128.7, 128.2, 128.0,
125.4, 124.2, 123.0, 122.9, 122.8, 120.3, 118.1, 116.2, 77.0, 51.1,
29.2, 24.3; IR (neat) 3360 br, 2964 m, 1710 s, 1574 m, 1469 m,
1429 m, 1221 m, 1120 m. Anal. (C24H21BrN2O) C, H, N.
(R,S)-5-(3-F lu or o-4-m et h ylp h en yl)-1,2-d ih yd r o-2,2,4-
tr im eth yl-5H-ch r om en o[3,4-f]qu in olin e (29). This com-
pound was prepared by the general procedure from 4-bromo-
2-fluorotoluene (189 mg, 1.0 mmol, 14 equiv) and compound 1
(20 mg, 0.069 mmol) in 56% yield as a colorless oil (15 mg):
1H NMR (400 MHz, acetone-d6) 7.60 (d, J ) 7.8, H10), 7.56 (d,
J ) 8.4, H11), 7.08 (t, J ) 7.9, H5′), 6.98 (t, J ) 7.8, H8), 6.94
(d, J ) 8.0, H2′), 6.91 (s, H5), 6.90-6.80 (m, H7, H9, H12, H6′),
5.55 (br s, NH), 5.48 (s, H3), 2.12 (s, 3 × H4′a), 2.01 (s, 3 ×
H
4a), 1.26 (s, 3 × H2ab), 1.24 (s, 3 × H2aa); 13C NMR (100 MHz,
acetone-d6) 161.9 (d, J ) 235), 151.4, 147.1, 141.3, 134.5, 132.0,
131.9, 130.0, 129.0, 128.0, 125.6, 125.0, 124.9, 124.4, 122.7,
120.5, 119.5, 118.2, 115.9, 115.5 (d, J ) 23), 75.4, 50.9, 29.5,
24.0, 14.1; IR (KBr) 3369 br, 2962 m, 1593 m, 1575 m, 1467 s,
1259 m, 815 m, 754 m. Anal. (C26H24FNO) C, H, N.
(+)-5-(4-Ch lor op h en yl)-1,2-d ih yd r o-2,2,4-tr im eth yl-5H-
ch r om en o[3,4-f]qu in olin e (35) a n d (-)-5-(4-Ch lor op h e-
n yl)-1,2-d ih yd r o-2,2,4-tr im eth yl-5H-ch r om en o[3,4-f]qu in -
olin e (36). These compounds were prepared by a HPLC
separation of racemic compound 15 by a chiral column,
Chiracel OD-R, using a 9:1 mixture of methanol and water as
mobile phase. The optical purity of 35 was determined by
(R,S)-5-(3,5-Dich lor oph en yl)-1,2-dih ydr o-2,2,4-tr im eth yl-
5H-ch r om en o[3,4-f]qu in olin e (30). This compound was
prepared by the general procedure from 1-bromo-3,5-dichlo-
robenzene (226 mg, 1.0 mmol, 19 equiv) and 1 (15 mg, 0.051
mmol) in 95% yield as a white solid (20 mg): mp 100-102 °C;
1H NMR (400 MHz, acetone-d6) 7.63 (d, J ) 7.7, H10), 7.58 (d,
J ) 8.4, H11), 7.29 (t, J ) 1.9, H4′), 7.20 (d, J ) 1.9, H2′, H6′),
7.03 (t, J ) 7.7, H8), 6.97 (s, H5), 6.93-6.85 (m, H7, H9, H12),
5.63 (br s, NH), 5.53 (s, H3), 2.04 (s, 3 × H4a), 1.28 (s, 3 × H2ab),
1.27 (s, 3 × H2aa); 13C NMR (100 MHz, acetone-d6) 151.0, 147.3,
145.7, 135.5, 135.1, 135.0, 129.0, 128.8, 128.6, 128.4, 127.8,
125.3, 124.6, 123.2, 123.0, 120.3, 119.7, 118.3, 116.4, 116.3,
74.9, 51.2, 24.0; IR (neat) 3350 br, 2940 m, 1690 m, 1590 m,
1480 s, 1070 m. Anal. (C25H21Cl2NO‚1/2H2O) C, H, N.
(R,S)-5-(3-Br om o-5-flu or op h en yl)-1,2-d ih yd r o-2,2,4-tr i-
m eth yl-5H-ch r om en o[3,4-f]qu in olin e (31). This compound
was prepared by the general procedure from 1,3-dibromo-5-
fluorobenzene (254 mg, 1.0 mmol, 4.7 equiv) and compound 1
(60 mg, 0.21 mmol) in 63% yield as a white powder (60 mg):
mp 82-84 °C; 1H NMR (400 MHz, acetone-d6) 7.63 (d, J )
7.7, H10), 7.58 (d, J ) 8.4, H11), 7.23 (d, J ) 5.2, H4′), 7.20 (s,
H2′), 7.08-7.02 (m, H8, H6′), 6.97 (s, H5), 6.94-6.85 (m, H7, H9,
HPLC: >97% ee; [R]20 ) +277 (MeOH). The optical purity
D
of 36 was determined by HPLC: >99% ee; [R]20 ) -254
D
(MeOH).
(+)-1,2-Dih ydr o-2,2,4-tr im eth yl-5-(4-m eth ylp h en yl)-5H-
ch r om en o[3,4-f]qu in olin e (37) a n d (-)-1,2-Dih yd r o-2,2,4-
t r im et h yl-5-(4-m et h ylp h en yl)-5H -ch r om en o[3,4-f]q u in -
olin e (38). These compounds were prepared by a HPLC
separation of racemic compound 22 by a chiral column,
Chiracel OD-R, using a 9:1 mixture of methanol and water as
mobile phase. The optical purity of 37 was determined by
HPLC: >93% ee; [R]20 ) +235 (MeOH). The optical purity
D
of 38 was determined by HPLC: >99% ee; [R]20 ) -246
D
(MeOH).
Cotr a n sfection Assa ys. The function and detailed prepa-
ration procedure of the cotransfection assays have been
described previously.13 Briefly, the cotransfection assays were
carried out in CV-1 cells (African green monkey kidney
fibroblasts), which were transiently transfected, by the stan-
dard calcium phosphate coprecipitation procedure,12 with the
plasmid-containing receptor, MTV-LUC reporter, pRS-â-Gal,
and filler DNA (Rous sarcoma virus chloramphenical acetyl-
transferase). The agonist activity was determined by examin-
ing the LUC expression (normalized response), and the efficacy
readout was a relative value to the maximal LUC expression
produced by a reference agonist, e.g., progesterone for hPR,
dihydrotestosterone (DHT) for hAR, dexamethasone for hGR,
aldosterone for hMR, estradiol for hER. All the cotransfection
experiments were carried out in 96-well plates by automation
(Beckman Biomomek automated workstation).
H
12), 5.64 (br s, NH), 5.53 (s, H3), 2.04 (s, 3 × H4a), 1.28 (s, 3
× H2ab), 1.27 (s, 3 × H2aa); 13C NMR (100 MHz, acetone-d6)
163.4 (d, J ) 250), 151.1, 147.3, 146.4 (d, J ) 7.0), 135.0, 129.1,
128.8, 128.4, 128.3, 125.3, 124.6, 123.2, 123.0, 122.9, 120.4,
119.7, 119.2 (d, J ) 24.8), 118.3, 116.4, 115.2 (d, J ) 22.2),
74.9, 51.2, 29.4, 24.0; IR (neat) 3367 br, 1699 m, 1595 s, 1581
s, 1469 s, 1435 s, 1251 s. Anal. (C25H21BrFNO) C, H, N.
(R,S)-5-(3-Br om o-5-m et h ylp h en yl)-1,2-d ih yd r o-2,2,4-
tr im eth yl-5H-ch r om en o[3,4-f]qu in olin e (32). This com-
pound was prepared by the general procedure from 3,5-
dibromotoluene (250 mg, 1.0 mmol, 14 equiv) and compound
1 (20 mg, 0.069 mmol) in 52% yield as a white solid (16 mg):
mp 87-89 °C; 1H NMR (400 MHz, acetone-d6) 7.61 (d, J )
Recep tor Bin d in g Assa ys. The preparation of receptor
binding assays for hPR-A, hGR, and hAR was described

Novel Class of hPR Agonists
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 301
previously,13 and the Ki values for the analogues were deter-
mined by application of the Cheng-Prusoff equation.35 The
radioligands used in the competitive binding assays were
progesterone for hPR-A, DHT for hAR, and dexamethasone
for hGR.
Align m en t Ru le Gen er a tion . The CoMFA fields from
which the overlap was derived was developed by minimizing
all structures to a minimum gradient of 10-6 using the Tripos
force field and the BFGS minimizer internal to SYBYL.27
Charges were Mullikan charges taken from MOPAC 5.0 single
points computed with the MNDO model Hamiltonian.36 The
steric and electrostatic fields were projected for progesterone
and used to rms fit our molecules. All alignments were as
depicted in Figure 3. The steric and electrostatic fields of
ketodesgestrel and ICI 182780 were projected and used to
generate alignment rules as well. The alignments were
identical with those developed from progesterone.
Ma ter ia l a n d Meth od s for Ma m m a r y Gla n d Mor p h ol-
ogy/Uter in e Wet Weigh t Assa y in th e Ova r iectom ized
Ra t. Inhibitory effects of progestins on estrogen-induced
uterine wet weight and stimulation of mammary alveolar bud
formation were evaluated using a modification of previously
described methods.37-39 Four- to five-week-old ovariectomized
Sprague-Dawley rats were obtained 1 week after surgery and
allowed to acclimate for an additional week after shipment.
The tested compounds, MPA and estrone, were dissolved in
purified sesame oil. Animals were randomized into treatment
groups (4 rats/group) and administered either MPA or quino-
line derivatives 3.0 mg/rat, 0.5-mL volumes, per os (po), QD
for 3 days in the presence of 10 µg/day, subcutaneous (sc)
estrone. MPA (po, QD) was used as the progestin standard
in all assays. Additional groups of rats, administered estrone
(10 µg/day, sc) or vehicle alone for 3 days, were included as
both positive and negative controls, respectively. Animals
were sacrificed on the fourth day of the experiment. Upon
necropsy, mammary glands were excised, fixed in acetone, and
stained in Gill’s #2 for 6-8 h. Mammary glands were mounted
in Permount and quantified by counting buds in four 1-mm2
microscopic fields, per gland. Uteri were also excised, blotted
on filter paper, and weighed. Statistical analyses of the
resulting data were performed using ANOVA followed by a
modified Fisher’s test.
M.; Gunnet, J . W.; Cryan, E. V.; Granger, K. S.; J ordan, J . J .;
Demarest, K. T. Nonsteroidal Progesterone Receptor Ligands.
2. High-Affinity Ligands with Selectivity for Bone Cell Proges-
terone Receptors. J . Med. Chem. 1995, 38, 4880-4884.
(9) (a) Colletta, A. A.; Wakefield, L. M.; Howell, F. V.; Danielpour,
D.; Baum, M.; Sporn, M. B. The Growth Inhibition of Human
Breast Cancer Cells by a Novel Synthetic Progestin Involves the
Induction of Transforming Growth Factor Beta. J . Clin. Invest.
1991, 87, 277-283. (b) Megase (BMS oncology brand name for
megestrol acetate) for breast cancer.
(10) See recent reviews: (a) Evans, R. M. The Steroid and Thyroid
Receptor Superfamily. Science 1988, 240, 889-895. (b) O’Malley,
B. The Steroid Receptor Superfamily: More Excitement Pre-
dicted for the Future. Mol. Endocrinol. 1990, 4 (3), 363-369.
(c) Miesfeld, R. L. The Structure and Function of Steroid
Receptor Proteins. Critical Reviews in Biochemistry and Molec-
ular Biology. Mol. Biol. 1989, 24, 101-117. (d) Beato, M. Gene
Regulation by Steroid Hormones. Cell 1989, 56, 335-344. (e)
Luisi, B. F.; Schwabe, J . W. R.; Freedman, L. P. The Steroid/
Nuclear Receptors: From Three-dimensional Structure to Com-
plex Function. Vitamins and Hormones; Academic Press: New
York, 1994; Vol. 49, pp 1-47.
(11) Rosen, J .; Day, A.; J ones, T. K.; J ones, E. T. T.; Nadzan, A. M.;
Stein, R. B. Intracellular Receptors and Signal Transducers and
Activators of Transcription Superfamilies: Novel Target for
Small-molecule Drug Discovery. J . Med. Chem. 1995, 38, 4855-
4874.
(12) Berger, T. S.; Parandoosh, Z.; Perry, B. W.; Stein, R. B.
Interaction of Glucocorticoid Analogues with the Human Glu-
cocorticoid Receptor. J . Steroid Biochem. Mol. Biol. 1992, 41,
733-738.
(13) Pathirana, C.; Stein, R. B.; Berger, T. S.; Fenical, W.; Ianiro,
T.; Mais, D. E.; Torres, A.; Glodman, M. E. Nonsteroidal Human
Progesterone Receptor Modulators from the Marine Alga Cy-
mopolia barbata. Mol. Pharm. 1995, 47, 630-635.
(14) J ones, T. K.; Pathirana, C.; Goldman, M. E.; Hamann, L. G.;
Farmer, L. J .; Ianiro, T.; J ohnson, M. G.; Bender, S. L.; Mais,
D. E.; Stein, R. B. Discovery of Novel Intracellular Receptor
Modulating Drugs. J . Steroid Biochem. Mol. Biol. 1996, 56, 61-
66.
(15) Hamann, L. G.; Farmer, L. J .; J ohnson, M. G.; Bender, S. L.;
Mais, D. E.; Wang, M. W.; Chrombie, D.; Goldman, M. E.; J ones,
T. K. Synthesis and Biological Activity of Novel Nonsteroidal
Progesterone Receptor Antagonists Based on Cyclocymopol
Monomethyl Ether. J . Med. Chem. 1996, 39, 1778-1789.
(16) (a) Misrahi, M.; Atger, M.; d′Auriol, L.; Loosfelt, H.; Meriel, C.;
Fridlansky, F.; Guiochon-Mantel, A.; Galibert, F.; Milgrom, E.
Complete Amino Acid Sequence of the Human Progesterone
Receptor Deduced from Cloned cDNA. Biochem. Biophys. Res.
Commun. 1987, 143, 740-748. (b) Sartorius, C. A.; Melville, M.
Y.; Hovland, A. R.; Tung, L.; Takimoto, G. S.; Horwitz, K. B. A
Third Transactivation Function (AF3) of Human Progesterone
Receptors Located in the Unique N-Terminal Segment of the
B-Isoform. Mol. Endocrinol. 1994, 8 (10), 1347-1360. (c) Clemm,
D. L.; Macy, B. L.; Santiso-Mere, D.; McDonnell, D. P. Definition
of the Critical Cellular Components which Distinguish between
Hormone and Antihormone Activated Progesterone Receptor. J .
Steroid Biochem. Mol. Biol. 1995, 53, 487-495.
(17) (a) Chang, C.; Kokontis, J .; Liao, S. Molecular Cloning of Human
and Rat Complementary DNA Encoding Androgen Receptors.
Science 1988, 240, 324-326. (b) Lubahn, D. B.; J oseph, D. R.;
Sullivan, P. M.; Willard, H. F.; French, F. S.; Wilson, E. M.
Cloning of Human Androgen Receptor Complementary DNA and
Localization to the X Chromosome. Science 1988, 240, 327-330.
(18) Hollenberg, S. M.; Weinberger, C.; Ong, E. S.; Cerelli, G.; Oro,
A.; Lebo, R.; Thompson, E. B.; Rosenfeld, M. G.; Evans, R. M.
Primary Structure and Expression of a Functional Human
Glucocorticoid Receptor cDNA. Nature 1985, 318, 635-641.
(19) Green, G. L.; Gilna, P.; Waterfeield, M.; Baker, A.; Hort, Y.;
Shine, J . Sequence and Expression of Human Estrogen Receptor
Complementary DNA. Science 1986, 231, 1150-1154.
(20) Arriza, J . L.; Weinberger, C.; Cerelli, G.; Glaser, T. M.; Handelin,
G. L.; Housman, D. E.; Evans, R. M. Cloning of the Human
Mineralocorticoid Receptor Complementary DNA: Structure and
Functional Kinship with the Glucocorticoid Receptor. Science
1987, 237, 268-275.
Refer en ces
(1) Many general reviews on steroid research, including the
progestins, have been published; see: (a) Witzmann, R. F.
SteroidssKeys to Life; Van Norstrand Reinhold: New York, 1981.
(b) Fullerton, D. S. Steroids and Therapeutically Related Com-
pounds. In Wilson and Gisvold’s Textbook of Organic Medicinal
and Pharmaceutical Chemistry, 9th ed.; Delgado, J . N.; Remers,
W. A., Eds.; J . B. Lippincott Co.: Philadelphia, 1991; Chapter
18. (c) Deghenghi, R.; Givner, M. L. Female Sex Hormones. In
Burger’s Medicinal Chemistry; Wolff, M. E., Ed.; J ohn Wiley &
Sons: New York, 1979-1980; Chapter 29, Vol. 2. (d) Murad, F.;
Kuret, J . Estrogens and Progestins. In The Pharmaceutical Basis
of Therapeutics, 8th ed.; Gilman, A. G., Rall, T. W., Nies, A. S.,
Taylor, P., Eds.; Macmillan: New York, 1990; Chapter 58.
(2) Pincus, G. The Control of Fertility; Academic Press: New York,
1965; references cited therein.
(3) Djerassi, C.; Miramontes, L.; Rosenkranz, G.; Sondheimer, F.
Synthesis of 19-Nor-17R-ethynyltestosterone and 19-Nor-17R-
methyltestosterone. J . Am. Chem. Soc. 1954, 76, 4092-4094.
(4) Colton, F. B. U.S. Patent 2,691,028, 1954 (applied May, 1953).
(5) Breckwoldt, M.; Keck, C.; Karck, U. Benefits and Risks of
Hormone Replacement Therapy (HRT). J . Steroid Biochem. Mol.
Biol. 1995, 53, 205-208.
(6) Mishell, D. R., J r. Breast Cancer Risk with Oral Contraceptives
and Estrogen Replacement Therapy. Aust. N. Z. J . Obstet.
Gynaecol. 1994, 34, 316-320.
(7) Raynaud, J . P.; Azadian-Boulanger, G.; Ojasoo, T. Expansion
Scientifique Francaise. In Pharmacologie Clinique. Bases de la
Therapeutique; Giroud, J . P., Mathe, G., Meyniel, G., Eds.; 1988;
p 2093.
(8) (a) Collins, D. C. Sex Hormone Receptor Binding, Progestin
Selectivity, and the New Oral Contraceptives. Am. J . Obstet.
Gynecol. 1994, 170, 1508-1513. (b) Combs, D. W.; Reese, K.;
Phillips, A. Nonsteroidal Progesterone Receptor Ligands. 1.
3-Aryl-1-benzoyl-1,4,5,6-tetrahydropyridazines. J . Med. Chem.
1995, 38, 4878-4879. Combs, D. W.; Reese, K.; Cornelius, L. A.
(21) (a) Skraup, Z. H. Eine Synthese des Chinolins. Ber. Dtsch. Chem.
Ges. 1880, 13, 2086-2087. (b) Eisch, J . J .; Dluzniewski, T.
Mechanism of the Skraup and Doebner-von Miller Quinoline
Syntheses: Cyclization of R,â-Unsaturated N-Aryliminium Salts
via 1,3-Diazetidinium Ion Intermediates. J . Org. Chem. 1989,
54, 1269-1274. (c) Walter, H.; Sauter, H.; Winkler, T. Eine Neue
Einfache Synthese Spirocyclischer 1H-Chinolin-derivate. Helv.
Chim. Acta 1992, 75, 1274-1280.
(22) Migachev, G. I. Investigations in the Series of Orthosubstituted
Biphenyls 1. Nitration of 2-Biphenylcarboxylic Acid and the
Chemical Properties of Its Nitro Derivatives. J . Org. Chem.,
U.S.S.R. 1979, 15 (3), 503-509.

302 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Zhi et al.
(23) The reason for the high regioselectivity is unclear. It is possible
that the carbonyl group of the lactone participated in the
cyclization process (Skraup reaction) and directed the regiose-
lective process. However, in the case of an open ester instead of
a lactone (4-(2′-methoxyphenyl)-3-(methoxycarbonyl)aniline), the
regioselectivity was totally lost.
(24) Kraus, G. A.; Frazier, K. A.; Roth, B. D.; Taschner, M. J .;
Neuenschwander, K. Conversion of Lactones into Ethers. J . Org.
Chem. 1981, 46, 2417-2419.
(32) Edwards, J . P.; West, S. J .; Marschke, K. B.; Mais, D. E.;
Gottardis, M. M.; J ones, T. K. 5-Aryl-1,2-dihydro-5H-chromeno-
[3,4-f]quinolines as Potent, Orally Active, Nonsteroidal Proges-
terone Receptor Agonists: The Effect of D-Ring Substituents.
J . Med. Chem. 1998, 41, 303-310.
(33) Still, W. C.; Kahn, M.; Mitra, A. J . Rapid Chromatographic
Technique for Preparation Separation with Moderate Resolution.
J . Org. Chem. 1978, 43, 2923-2924.
(34) A minimum amount of 70% nitric acid is used in the first
nitration, which minimizes the dilution of the fuming nitric acid
during the second nitration in the one-pot procedure.
(35) Cheng, Y.-C.; Prusoff, W. H. Relationship between the Inhibition
Constant (Ki) and the Concentration of Inhibitor which Causes
50% Inhibition (I50) of an Enzymatic Reaction. Biochem. Phar-
macol. 1973, 22, 3099-3108.
(25) There were a number of observations in the literature that THF
and diethyl ether could play different roles as the lithiation
solvent; see: Bolm, C.; Ewald, M.; Felder, M.; Schlingloff, G.
Enantioselective Synthesis of Optically Active Pyridine Deriva-
tives and C2-Symmetric 2,2′-Bipyridines. Chem. Ber. 1992, 125,
1169-1190.
(26) Teutsch, G. 11â-Substituted 19-Norsteroids: At the Cross Roads
between Hormone Agonists and Antagonists. In Adrenal Steroid
Antagonism; Agarwal, M. K., Ed.; Walter de Gruyter: Berlin,
1984; pp 43-75.
(27) Neef, G.; Beier, S.; Elger, W.; Henderson, D.; Wiechert, R. New
Steroids with Antiprogestational and Antiglucocorticoid Activi-
ties. Steroids 1984, 44, 349-372.
(28) The dose-response curves of MPA and compounds 13, 15, and
24 will be published separately. The 3.0 mg/rat dose level results
in Figures 1 and 2 are presented to simply indicate the efficacy
of the novel series in this in vivo model.
(29) (a) SYBYL, Tripos Associates, St. Louis, MO. (b) Cramer, R. D.;
Patterson, D. E.; Bunce, J . D. Comparative Molecular Field
Analysis (CoMFA). 1. Effect of Shape on Binding of Steroids to
Carrier Proteins. J . Am. Chem. Soc. 1988, 110, 5959-5967.
(30) Kallel, E. A. Manuscript in preparation.
(36) Stewart, J . J . P. MOPAC: A Semiempirical Molecular Orbital
Program. J . Comput.-Aided Mol. Des. 1990, 1-45.
(37) Lerner, L. J .; Hilf, R.; Turkheimer, A. R.; Michel, I.; Engel, S.
L. Effects of Hormone Antagonists on Morphological and Bio-
chemical Changes Induced by Hormonal Steroids in the Im-
mature Rat Uterus. Endocrinology 1966, 78, 111-124.
(38) Sakamoto, S.; Abe, A.; Hideki, K.; Yamada, N.; Seki, K.; Ryohei,
O. Effects of Estrogen and Progesterone on Thymidine Kinase
Activity in the Immature Rat Uterus. Am. J . Obstet. Gynecol.
1983, 145, 711-715.
(39) Michna, H.; Nishino, Y.; Schneider, M. R.; Louton, T.; El Etreby,
M. F. A Bioassay for the Evaluation of Antiproliferative Poten-
cies of Progesterone Antagonists. J . Steroid Biochem. Mol. Biol.
1991, 38, 359-365.
(31) Hellberg, S.; Sjostrom, M.; Skagerberg, B.; Wold, S. Peptide
Quantitative Structure-Activity Relationships, a Multivariate
Approach. J . Med. Chem. 1987, 30, 1126-1135.
J M9705768