Corticotropin-releasing hormone receptor antagonists: Framework design and synthesis guided by ligand conformational studies

Article abstract of DOI:10.1021/jm980223o

As described in the preceding paper (Arvanitis et al. J. Med. Chem. 1999, 42), anilinopyrimidines I were identified as potent antagonists of corticotropin-releasing hormone-1 receptor (CRH₁-R, also referred to as corticotropin-releasing factor, CRF₁-R). Our next goal was to understand the receptor-bound conformation of the antagonists and to use this information to help guide preclinical optimization of the series and to develop new leads. Since receptor structural information was not available, we assumed that these small, high-affinity antagonists would tend to bind in conformations at or energetically close to their global minima and that rigid analogues that maintained the important stereoelectronic features of the bound anilinopyrimidine would also bind tightly. Conformational preferences and barriers to rotation of the anilinopyrimidines were determined by semiempirical methods, and X-ray and variable-temperature NMR spectroscopy provided experimental results that correlated well with calculated structures. Using these data, a key dihedral angle was constrained to design fused-ring analogues, substituted N-arylpyrrolopyridines II, synthesis of which provided CRH₁ receptor antagonists with potency equal to that of the initial congeneric leads (K(i) = 1 nM) and which closely matched the conformation held by the original compound, as determined by crystallography. In addition to providing a useful template for further analogue synthesis, the study unequivocally determined the active conformation of the anilinopyrimidines. Theoretical and spectroscopic studies, synthesis, and receptor binding data are presented.

Full text of DOI:10.1021/jm980223o

J . Med. Chem. 1999, 42, 819-832
819
Cor ticotr op in -Relea sin g Hor m on e Recep tor An ta gon ists: F r a m ew or k Design
a n d Syn th esis Gu id ed by Liga n d Con for m a tion a l Stu d ies
C. Nicholas Hodge,* Paul E. Aldrich, Zelda R. Wasserman, Christina H. Fernandez, Gregory A. Nemeth,
Argyrios Arvanitis, Robert S. Cheeseman, Robert J . Chorvat, Engelbert Ciganek, Thomas E. Christos,
Paul J . Gilligan, Paul Krenitsky,† Everett Scholfield, and Philip Strucely‡
Department of Chemical and Physical Sciences, DuPont Pharmaceuticals Company, Experimental Station,
P.O. Box 80500, Wilmington, Delaware 19880-0500
Received October 1, 1998
As described in the preceding paper (Arvanitis et al. J . Med. Chem. 1999, 42), anilinopyrimidines
I were identified as potent antagonists of corticotropin-releasing hormone-1 receptor (CRH1-
R, also referred to as corticotropin-releasing factor, CRF1-R). Our next goal was to understand
the receptor-bound conformation of the antagonists and to use this information to help guide
preclinical optimization of the series and to develop new leads. Since receptor structural
information was not available, we assumed that these small, high-affinity antagonists would
tend to bind in conformations at or energetically close to their global minima and that rigid
analogues that maintained the important stereoelectronic features of the bound anilinopyri-
midine would also bind tightly. Conformational preferences and barriers to rotation of the
anilinopyrimidines were determined by semiempirical methods, and X-ray and variable-
temperature NMR spectroscopy provided experimental results that correlated well with
calculated structures. Using these data, a key dihedral angle was constrained to design fused-
ring analogues, substituted N-arylpyrrolopyridines II, synthesis of which provided CRH1
receptor antagonists with potency equal to that of the initial congeneric leads (Ki ) 1 nM) and
which closely matched the conformation held by the original compound, as determined by
crystallography. In addition to providing a useful template for further analogue synthesis, the
study unequivocally determined the active conformation of the anilinopyrimidines. Theoretical
and spectroscopic studies, synthesis, and receptor binding data are presented.
In tr od u ction
series. Additionally, whether the new series was ulti-
As described in the first paper in this series,1 there
is strong evidence that potent, bioavailable antagonists
of the corticotropin-releasing hormone (CRH1) receptor
(CRH1-R) will have clinically beneficial properties in a
number of indications, particularly anxiety and depres-
sion. Having discovered potent anilinopyrimidine an-
tagonists I (Chart 1), the next task was to broaden the
scope of the lead into new structural classes. For the
project to be considered successful, we determined that
new antagonists should have: (1) potency equal to or
greater than that of the original lead; (2) a novel core
structure that would provide insight into the bound
conformation and allow us to rapidly and systematically
explore the effect of changes in side chain functionality;
(3) a favorable oral bioavailability potential (that is, the
size, functional groups, and polarity of potent com-
pounds in the new series should be within the ranges
associated with favorable absorption, distribution, me-
tabolism, and elimination profiles2); (4) a favorable
potential for distribution into the CNS;3 (5) a stable,
formulatable, nontoxic core structure.
mately as attractive as the anilinopyrimidines in one
or more of the above areas, the process of design, if
carefully executed, would provide insight into the bound
shape of the original lead and thereby assist in optimi-
zation of the anilinopyrimidines. Finally, different core
structures would broaden the scope of functionality that
could be introduced into the molecules, by virtue of
different chemistry.
In designing the new structural series, we were
guided by an active ligand rather than its receptor. The
CRH1 receptor is a seven-helical bundle that spans the
cell membrane, and although its tertiary structure is
qualitatively well-understood, it was clear that current
receptor models possessed insufficient accuracy for
design at the atomic level. Additionally, unlike the
catecholamine neurotransmitter receptors, which are
known to interact with a critical aspartic acid on
transmembrane helix 3, significant uncertainty existed
even as to where ligands bind the receptor, i.e., inside
or outside the seven-helix bundle, inserted into the
membrane, or extra- or even intracellularly.4 Kinetic
data suggested that unlike peptide CRH-based antago-
nists,5 the anilinopyrimidines do not compete with
peptide agonists (S. Culp, unpublished results) in recep-
tor-binding assays, which obviated the use of models of
the bound natural agonist for mimicking the effect of
anilinopyrimidines.
Achieving this goal would provide several advantages
to our program: first, it would expand our knowledge
of active structures in addition to the anilinopyrimidine
* Corresponding author. Current address: Caliper Technologies
Corp., 605 Fairchild Dr., Mountain View, CA 94043-2234. E-mail:
nick.hodge@calipertech.com.
† Current address: ArQule, Inc., 200 Boston Ave, Medford, MA
The most direct and reliable method for new ligand
design was therefore to understand the bound shape of
02155.
‡ Deceased.
10.1021/jm980223o CCC: $18.00 © 1999 American Chemical Society
Published on Web 02/17/1999

820 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 5
Hodge et al.
Ch a r t 1. Optimum Anilinopyrimidine Substituents for CRH1-R Potency
the already-identified lead series and to design mol-
ecules that present these features rigidly and unam-
biguously in space. This process is conceptually similar
to other successful structure-based designs;6,7 here our
emphasis is on the rigorous determination of the bound
ligand conformation using information derived only from
free ligand spectra and calculations.
The method we employed to design the new series
involved the following steps: (1) Determine low-energy
conformations of potent anilinopyrimidines by NMR,
X-ray, and semiempirical methods. (2) Design confor-
mationally restricted analogues, starting with the most
favorable conformation. (3) Synthesize the compounds
and characterize their three-dimensional shape. (4) Test
for biological activity, and if necessary redesign based
on the next-lowest-lying conformation.
This approach depends on the hypothesis that a low-
energy free conformation is predictive of the bound
conformation. Explicitly stating this assumption high-
lights the importance of examining low-energy confor-
mations under several conditions: in crystals, in non-
polar and polar solvents, and calculated in the gas
phase. Clearly this assumption will not always hold.
However, in the absence of a known bound geometry
for these molecules, the lowest energy is a good place
to start. These are small, potent molecules, and only
limited interactions are available for binding to the
receptor. It is reasonable to suggest, therefore, that in
order to avoid an internal strain energy penalty on
binding, the internal strain energy of the bound con-
formation will be close to the lowest energy conforma-
tion available in the test medium (in this case pH 7.2
buffer). This hypothesis obviously does not imply that
the bound conformer and the low-energy free conformer
are geometrically close, so a careful analysis requires
examination of the full potential energy surface of the
molecule in question. It is also possible that a high-
energy ligand conformation is stabilized by a high-
energy receptor conformation, and the resulting complex
is lower in energy than the two free species. This
occurrence is uncommon with small ligands, in our
experience. In any case, if the conformational locking
is done carefully, a negative result is also valuable since
it rules out one of a few possible bound conformations.
It should also be clear that this approach becomes
experimentally untenable as the number of equi-
energetic minima occupied by the reference ligand
increases. The use of structural and computational
chemistry to obtain a detailed understanding of the
conformational free energy surface of the reference
ligand is therefore a useful initial step in deciding
whether this approach is appropriate.
In this report we detail the successful application of
this method to the design, synthesis, and characteriza-
tion of CRH1-R antagonists (Table 1) that meet the
criteria described above. The anilinopyrimidines that
were characterized spectrally and computationally were
synthesized as described in the first paper in this series
and are summarized in Table 2. A limited structure-
activity study of these potent bicyclic antagonists is
presented here as well. The third paper in this series8
reports expanded SAR studies on bicyclic analogues, as
well as a summary of reports on related antagonists
from other laboratories.
Meth od s
Sin gle-Cr ysta l X-r a y An a lysis. Structural charac-
terization of compounds 2, 3, 9, 25, and 29 (Tables 1
and 2) was carried out using standard X-ray crystal-
lographic techniques. Complete reports including crys-
tallization conditions, atomic coordinates, thermal pa-
rameters, and interatomic distances and angles are
available as Supporting Information.
NMR Solu tion Str u ctu r e. Variable-temperature 1H
NMR experiments were carried out on a Varian XL-400
spectrometer at 5-10 mg of sample/mL of solvent.

CRH Receptor Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 5 821
Ta ble 1.
no.
R1
R2
R3
CN
R4
yielda (%)
mp (°C)
anal.
X-rayg
Ki (nM)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Me
Me
Ph
Me
Me
Ph
Me
Me
Me
H
Me
Me
Me
Ph
Ph
Me
Me
Me
H
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4,6-(OMe)2
2-Br, 4,6-(OMe)2
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4-isopropyl (7-N-oxide)
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4-isopropyl
4-isopropyl
78
66
77
2b
108
76
130
198
115
96
186
202
125
176
oil
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
HRMS
C,H,Nf
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
HRMS
HRMS
363
47
H
x
x
CN
CN
H
H
CN
H
CN
CN
H
>104
34
92
82
92
20c
10
10d
57
85
45
56
79
71
41
73
62
73
100
89
65
36
>104
29
1
x
1277
>104
222
>104
>104
446
79
Me
H
Me
H
Me
Me
Me
Me
H
H
oil
OHe
Cl
CN
CN
H
CN
CN
H
CN
H
H
202
124
75
179
123
oil
192
101
oil
237
242
Cl
Me
Me
Me
Me
Me
Me
Me
Me
>104
722
67
112
28
2462
>104
>104
Cl
Cl
1-morpholino
1-morpholino
Me
Me
Me
CONH2
CO2H
2-Br, 4-isopropyl
2-Br, 4-isopropyl
a
b
Last step. Side product from preparation of 4. c Yields suffered from O-demethylation; cited yield is after remethylation. See
d
Experimental Section.
Coproduct from preparation of 9; total yield of 9 + 10 was 21%. e Or keto tautomer. f Microanalysis not
g
accuratessee Experimental Section. Check (x) indicates these data were obtained and appear in the Supporting Information.
Ta ble 2.
no.
R5
R6
R7
R8
Tc1 (°C)a
∆Gact-1 (kcal)a
Tc2 (°C)b
∆Gact-2 (kcal)b
X-rayc
Ki (nM)
24
25
26
27
28
29
30
31
Me
Me
Me
Me
Me
Me
Me
Me
morpholino
Et
Et
Me
Me
Et
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-Br, 4-isopropyl
2-I, 4-isopropyl
2-I, 4-isopropyl
2,4-dimethoxy, 6-Br
2-I, 4-isopropyl
46
22
96
48
38
5
x
Me
Me
Me
Me
Cl
0
0
100
20
13.9
14.7
20.3
16.9
22.5
<-70
90
>150
120
[<9]f
17.7
Et
Et
x
19.2
d
140
>150
43e
32
>150
[>23]f
100
18.2
>1000
a
b
Coalescence temperature and calculated activation energy for pyrimidine methyl group(s). Coalescence temperature and calculated
activation energy for N-CH2 group, where applicable. c Check (x) indicates these data were obtained and appear in the Supporting
d
Information. Ki values of chloropyrimidines were not considered reliable due to potential reactivity under assay conditions. The potency
is approximately equal to that of the dimethylpyrimidine. e Methanesulfonate salt. f Assumes the same ∆ν value as 28. Compounds listed
in this table were synthesized as described in the first paper in this series.1
Acquisition and delay times were 4.07 and 2 s, respec-
tively, and pulse width was 4.62 µs. Normally 32
transients were collected at each temperature. Data
were collected at 20-40 °C increments, from -70 to 30
°C (CD2Cl2), -50 to 30 °C (CDCl3), or 30 to 150 °C
(DMSO-d6).

822 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 5
Hodge et al.
F igu r e 1. A. Key dihedrals in X-ray structure of 29: φ1 ) 108°, φ2 ) 4°, φ3 ) 169°. B. Stereoview of X-ray structure of 29. C.
Space-filling model of X-ray structure of 29. D. Stereoview of X-ray structure of 25.
Con for m a tion a l An a lysis. For reasons presented in
the Results and Discussion section, preliminary model-
ing indicated that molecular mechanics force fields
would not provide accurate low-energy conformations
of the anilinopyrimidines. As a result, exploration of the
potential surface and calculation of rotational barriers
were done with the program MOPAC using the AM1
Hamiltonian.9,10
ortho-substituted phenyl ring but under similar condi-
tions (ethanol/concentrated HCl at reflux), the 4-methyl-
6-phenyl isomer 3 is in fact the dominant product, as
shown by X-ray crystallography (see Supporting Infor-
mation). The 4-phenyl isomer could be isolated in very
small amounts from this reaction mixture. Alterna-
tively, the yield of 4 could be improved at the expense
of overall yield by running the reaction under conditions
that favored initial attack by nitrogen onto the less
hindered carbonyl (refluxing xylene with removal of
water; see Experimental Section). The 3-carboxamido-
and 3-carboxypyrrolopyrimidines 22 and 23 (Table 1)
were also prepared by procedures described in this
reference.
To calculate the potential surface, an initial model of
compound 29 was built and minimized in Discover using
the cff91 force field.11 Starting with this geometry, the
key dihedrals (φ1 and φ2, Figure 1A) were rotated in 30°
increments, and each of the 144 single-point energies
was determined in MOPAC. Minima were located, and
each of them was fully minimized.
The 3-unsubstituted derivatives, 2, 5, 6, 8, 11, and
12 were prepared in a single step by treatment of the
appropriate nitrile with 65% sulfuric acid at reflux
(approximately 140 °C) for 1 h. The ring system is
remarkably stable to this treatment; yields are sum-
marized in Table 1.
To calculate the barrier to rotation about selected
bonds, we started from the global minimum and changed
the dihedral angle by 20° increments. Smaller (4°)
increments were used near the transition maxima. This
dihedral was held fixed, and the structure was mini-
mized in MOPAC. The force-constant matrix was di-
agonalized to verify that these were transition states.
The 4- and 6-chloro derivatives 14, 15, 17, and 18
(Table 1 and Scheme 2) were examined as intermediates
for the exploration of non-carbon-linked substituents at
these positions. The 6-chloro derivative 14 was prepared
from pyridone 13 in low yield, and elemental analyses
were poor. However, the cyano group was removed
without difficulty to form 15, which was characterized
adequately. The 4-chloro derivative was prepared via
oxidation of the pyridyl nitrogen of 4-unsubstituted
derivative 9 to N-oxide 16, followed by displacement in
the 4-position with phosphorus oxychloride. (Mixtures
of products were reported under similar Meisenheimer
conditions on naphthyridine N-oxide,16 but since the
6-position was blocked in our case, the reaction went
smoothly.) Unfortunately, displacing chloride failed
under all conditions attempted. Under forcing condi-
tions, reaction with the aryl bromide occurred. After
considerable experimentation the N-oxide was treated
with triflic anhydride followed directly by morpholine,
Ch em istr y. The target antagonists, as discussed
below, contained the 1-phenyl-7-azaindole or 1H-pyr-
rolo[2,3-b]pyridine core (Table 1; hereafter referred to
as pyrrolopyridines). The chemistry of pyrrolopyridines
has been studied in some depth by Yakhontov12 and was
reviewed by Willette.13 Derivatives with different sub-
stitution patterns than those described here have been
reported as CNS depressants and antihypertensives.14
Brodrick and Wibberley15 described a synthesis that
provided entry into a substitution pattern that was ideal
for our purposes (Scheme 1). The syntheses of cyanopy-
rrolopyridines 1, 3, 7, 9, 10, and 13 (Table 1) were
carried out following the basic procedure published by
these authors, with modifications described in the
Experimental Section. In this reference the authors
assign the major cyclization product of N-phenylpyrrole
IV (Scheme 1) with 1-phenylbutane-1,3-dione as the
6-methyl-4-phenylpyrrolopyridine isomer, based on NMR
shifts in analogous compounds. In our hands, with an

CRH Receptor Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 5 823
Sch em e 1
Sch em e 2
29 was obtained, shown as a stereo ball-and-stick model
in Figure 1B. The important geometric features for
designing a conformationally locked mimetic are the
following: the phenyl and pyrimidine rings form a plane
angle of 56°, the central nitrogen has a pronounced
pucker of 17°, and the N-C-N-C dihedral angle, φ2,
between the pyrimidine ring and the central nitrogen,
is almost planar, at 4°.
A space-filling model (Figure 1C) clearly shows the
steric hindrance around the two bonds joining the
central nitrogen. Only one molecule lies in the unit cell,
and intermolecular contacts do not appear to contribute
to the observed geometry (see Supporting Information).
Note there are no H-bond donor-acceptor pairs avail-
able in the molecule, and no solvent appears in the unit
cell.
A crystal structure of asymmetrically substituted
pyrimidine 25 was also obtained (Figure 1D). Here the
plane angle between phenyl and pyrimidine rings is 84°,
the pucker angle of the anilino nitrogen is only 5°, and
the N-C-N-C dihedral, φ2, is also 4°. Interestingly,
the rotamer which places the morpholino group close
to the phenyl ring is observed in the crystalline state.
In the following discussions, the pyrimidine substituent
that lies closest to the phenyl ring (e.g., the morpholine
group in the case of 25) is referred to as the endo group,
and the more distant group is referred to as exo. The
same nomenclature is used for the pyrimidine ring
nitrogens.
forming 19 in moderate yields. The nitrile was removed
under standard conditions to form 4-morpholino deriva-
tive 20.
The anilinopyrimidine antagonists shown in Table 2
were synthesized as described in the first paper in this
series.1 During the course of this work,17,18 the group
at Pfizer disclosed the CRH affinity of related bicyclic
systems.19 The full scope of the chemistry of the bicyclic
antagonists is discussed in more detail in the following
paper in this series.8
NMR Solu tion Str u ctu r e. Since the crystal envi-
ronment may not accurately represent conformational
minima in solution, and to obtain an understanding of
the dynamic conformational behavior of these molecules,
1
we examined variable-temperature H NMR spectra. It
seemed likely that the steric hindrance observed in the
X-ray structure around the anilino nitrogen would result
in discrete observable conformations at experimentally
accessible temperatures. Spectra were recorded (see
Methods section) at various increments from -70 to 30
Resu lts a n d Discu ssion
X-r a y Str u ctu r e An a lysis. As a starting point for
analyzing the low-energy conformations of anilinopyri-
midines, a single-crystal X-ray structure of compound

824 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 5
Hodge et al.
F igu r e 2. Variable-temperature NMR of 27. a. Spectrum at
-70 °C in CD2Cl2. b. Spectra at indicated temperatures in CD2-
Cl2, showing pyrimidine methyl resonances. c. Spectra at
indicated temperatures in DMSO-d6, showing pyrimidine
methyl resonances. Peak height is normalized such that
highest peak is full scale.
F igu r e 3. Variable-temperature NMR of 28. a. Spectrum at
-70 °C in CD2Cl2. b. Spectra at indicated temperatures in CD2-
Cl2, showing pyrimidine methyl and N-CH2 resonances. c.
Spectra at indicated temperatures in DMSO-d6, showing
pyrimidine methyl and N-CH2 resonances. Peak height is
normalized such that highest peak is full scale.
°C (CD2Cl2) and from 30 to 150 °C (DMSO-d6). A
representative experiment with 2-(N-methyl-N-(2-iodo-
4-isopropylphenyl)amino)-4,6-dimethylpyrimidine (27)
is shown in Figure 2. The pyrimidine methyl groups at
2.1 and 2.3 ppm are in distinct chemical environments
at -70 °C in CD2Cl2 (Figure 2a), with reasonably small
line widths. By -30 °C, noticeable broadening has
occurred, and the peaks coalesce somewhere between
-10 and 10 °C (Figure 2b). When DMSO-d6 is solvent
(Figure 2c), the peak is still broad at 30 °C but sharpens
to a minimum width by 90 °C, indicating free rotational
averaging of the methyl groups. The N-methyl hydro-
gens are equivalent even at the lowest temperature
(-70 °C).
Spectra of the N-ethyl analogue 28 under the same
conditions are shown in Figure 3. The coalescence
temperature (Tc1) of the pyrimidine methyls remains the
same (∼0 °C). The behavior of the two methylene
protons next to the central nitrogen is informative in
this compound. The protons are well-separated at low
temperature (dt, 3.35 ppm; dt, 4.25 ppm; Figure 3a).
Some broadening is visible at 30 °C in CD2Cl2 (Figure
3b), and the coalescence temperature (Tc2) in DMSO-d6
is about 90 °C (Figure 3c). The peak has not fully
sharpened by 140 °C, suggesting some hindered rotation
even at this temperature. Over this entire range, the
unhindered isopropyl methyl and methylene resonances
show virtually no change in chemical shift or coupling
constants.
Spectra of 29-32 were obtained similarly. Com-
pounds 30 and 31 were particularly useful for analyzing
the relative distribution and Tc of rotamers of asym-
metrically substituted pyrimidines. Compound 30, a
4-chloro-6-methylpyrimidine, shows two pyrimidine meth-
yl peaks at low temperature, which we assign to the exo
(downfield, 2.17 ppm, 60%) and endo (upfield, 2.07 ppm,
40%) rotamers. These peaks coalesce at about 20 °C and
form a single sharp peak above 90 °C. Asymmetric
4-morpholino-6-methyltriazine 31, on the other hand,
also shows two methyl peaks but with the reverse
distribution (upfield, 1.99 ppm, 65%, and downfield, 2.17
ppm, 35%). The peaks begin to broaden at 70 °C and
have not fully coalesced by 130 °C. The increase in Tc
values for the 2-chloro-4-methylpyrimidine derivative

CRH Receptor Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 5 825
30 and the aminotriazine 31 (Table 2) relative to the
dimethylpyrimidine congeners may indicate electronic
effects on the rotational barrier. Other intriguing ob-
servations were made, and some studies on the chirality
of these diastereotopic molecules were carried out but
were not directly relevant to the design process and will
be reported elsewhere.
For the purpose of this qualitative analysis, the
coalescence temperature is reported as the temperature
at which coalescence is actually observed; that is, a very
low, broad baseline hump is observed, as in Figure 3c
(90 °C), or Tc is estimated by extrapolating between the
two measurements at which it occurred, as in Figure
3b between -10 and 10 °C. The uncertainty is therefore
about 5 °C; this corresponds to an error of about 0.3 kcal
in the free energy of activation, ∆Gact, as calculated from
the Eyring equation.20 The values for coalescence of the
pyrimidine methyl(s) (Tc1 and ∆Gact-1) and for coales-
cence of the methylene group on nitrogen (Tc2 and
∆Gact-2), when observed, are listed in Table 2.
It has been reported21 that alkoxydiarylboronates
with substituents at the aryl ortho positions show
similar temperature-dependent rotational averaging of
protons on both the aryl and alkoxy groups, with values
of ∆Gact about 4-6 kcal lower than we observe in the
anilinopyrimidines.
To determine if a protic environment would change
the preferred low-energy conformations in solution, D2O
(up to 40%, 0.8 mg/mL sample in DMSO-d6) was added
to the NMR solvent. No significant change in the room-
temperature shift positions or coupling constants of
compound 27 or 28 was observed under these conditions.
An NOE between the pyrimidine methyl group and
ortho ring proton of compound 28 was observed (not
shown). The pyrimidine methyls are nonequivalent at
room temperature, as discussed above, and the upfield-
shifted methyl shows the NOE. No NOE is observed
between an N-CH3 proton and the ortho proton. The
only conformation that is consistent with these data is
one similar to the X-ray structure, in which one pyri-
midine methyl is tucked in close proximity to the
shielding cone of the aromatic ring. In a model of 28
based on the low-energy conformations calculated below,
the distance between the centroid of the three methyl
protons and the ortho aromatic proton is 4.5 Å, consis-
tent with the observed NOE.
F igu r e 4. A. Contour plot of the calculated energy (1 kcal/
contour line) versus dihedrals φ1 and φ2 of N-methylanilinopy-
rimidine 27. The dashed lines show the lowest energy path
for rotation of the pyrimidine ring to a symmetry-equivalent
structure. B. Conformation of 27 (yellow carbons) at the
minimum energy values of φ1 and φ2 overlaid with the X-ray
structure of 25 (green carbons) for comparison.
formation, and the use of several methods to determine
other close-lying minima is important for design.
The need for caution in applying X-ray conformations
to ligand design is underscored by the observation that
unsymmetrically substituted triazine derivative 31 has
both triazine rotamers present in solution under the
NMR conditions, whereas the crystal structure of 25,
containing an analogously substituted pyrimidine, shows
only the rotamer with the morpholine group endo to the
phenyl ring (vide supra). There are a number of possible
reasons for this difference: the endo-morpholine could
have crystallized preferentially; the lowest energy con-
formations could differ in the crystal versus solution
state; and/or the pyrimidine and triazine could have
different low-energy conformations. But since the SAR
of the designed pyrrolopyridine series (see below and
next paper in this series8) demonstrates unequivocally
that a large pyrimidine/triazine substituent is only
tolerated at the exo position, the relevant point is that
the crystal structure of 25 predicted an inactive con-
From these experimental conformational studies it
appears that a single, fairly deep-welled minimum
energy structure dominates at room temperature in a
variety of solvents, with values of φ1-3 close to those
observed in the X-ray structure. (For nonsymmetrically
substituted heterocycles, two nondegenerate conforma-
tions are present.) We also now possessed experimen-
tally determined values of ∆Gact that could be used to
assess the accuracy of calculated conformations.
Ca lcu la ted Low -En er gy Con for m a tion s. In addi-
tion to the insights we were obtaining into the experi-
mentally preferred low-energy geometry of anilinopy-
rimidines, we also wanted to be sure that we could
model these compounds accurately for purposes of
design, and we wanted to determine if there were other
low-lying minima not detected by spectral techniques
but that might represent the receptor-bound conforma-

826 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 5
Ch a r t 2. Effect of N- and 6-Phenyl Substituents on Affinity
Hodge et al.
tion. We initially conceived a molecular field analysis22
as a rapid means of understanding the bound shape of
the anilinopyrimidines. We found, however, that the
experimentally observed structures were not well re-
produced by force fields available at the time, since the
ones we examined forced the anilino nitrogen into
planarity even at the expense of steric strain, which was
counter to intuition and the experimental details we
were accumulating. Since the limited degrees of internal
freedom in the leads allowed us to undertake confor-
mational analysis by semiempirical methods at reason-
able computational cost, we performed a full analysis
using MOPAC with the AM1 Hamiltonian9,10 (see
Methods section).
A contour plot of the calculated energy (1 kcal/contour
line) versus dihedrals φ1 and φ2 is shown in Figure 4A
for N-methylanilinopyrimidine 27. The conformation at
the minimum energy values of φ1 and φ2 is shown in
Figure 4B, overlaid with the X-ray structure of 25 for
comparison. This conformation is also consistent with
the NOE observed in NMR, since a pyrimidine methyl
is within 5 Å of an ortho proton and the methyl protons
are in the deshielding region of the ring.
correlate crudely with potency: that is, the more rigidly
locked conformations appear to be better antagonists.
This same point was noted by Chen et al. in their SAR
studies on anilinopyrimidines and -triazines.23 The most
striking example is the series of unsubstituted nitrogens
and their N-methyl and N-ethyl congeners (previous
paper), where N-H analogues are inactive unless large
groups are present in the ortho positions (Chart 2). This
result was encouraging since it suggested that the
design of conformationally locked mimetics should be
straightforward, provided that the unbound low-energy
conformation we had deduced was similar to the recep-
tor-bound conformation. The simplest way to mimic the
calculated shape of, for example, 24, is to restrict the
rotation of the dihedral angle closest to 0°, as shown in
Figure 5; since all of the atoms lie within 4° of a plane,
they can be approximated by a planar aromatic ring.
The literature indicated that the N-phenylpyrrolopyri-
dine nucleus was accessible synthetically without dif-
ficulty (see Methods section). This core was initially
chosen to test the proposed locked conformation. How-
ever, the key requirements of the core were only that
A
As shown in Figure 4A, the transition state for
rotation around φ2 in 27 is approximately 7.5 kcal, about
6 kcal lower than the value obtained from variable-
temperature experiments using the Eyring equation
(Table 1).
In summary, we were able to show experimentally
that the calculated conformations were in fact observed
in solution and in crystals (with the exception of the
asymmetric pyrimidine rotamer discussed above) and
that low-energy conformations were well-separated
energetically from other conformers. In addition, the
studies suggest which functionality on the anilinopyri-
midine is important for direct interaction with the
receptor and which are important for presenting the
correct conformation to the receptor (see below). This
information was critical for the next step.
B
Design of Con for m a tion a lly Lock ed Mim etics.
Rotational restriction around the central nitrogen of
anilinopyrimidines is observed spectrally and is sup-
ported by calculations. Looking over the anilinopyrimi-
dine SAR, steric hindrance of this rotation appears to
F igu r e 5. A. Locking of key dihedral φ2 to form pyrrolopyri-
dine. B. Postulated active (endo) and inactive conformations
of isopyrimidine 32. The absence of receptor affinity of 32 is
suggested to be due to the predominance of the exo form in
solution (see text for discussion).

CRH Receptor Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 5 827
The final point to be addressed by the SAR of the
anilinopyrimidines is the loss of the methyl group of the
potent N-ethyl analogues. This methyl group projects
above the plane of the new ring of 2 (Figure 5B), which
would be difficult to mimic with a planar heterocycle.
As in the ortho position of the aromatic ring, the
conformational studies appeared to favor a locking role
for the nitrogen substituent, rather than (or in addition
to) a direct interaction with the receptor. Even if the
extra carbon were important, the loss of the methyl
group was not considered a serious flaw for the series
since the N-methylanilinopyrimidine is also moderately
active. The additional carbon atom completing the five-
membered ring occupies a space not filled by the
anilinopyrimidines, so its effect on binding was uncer-
tain.
Syn th esis, Str u ctu r e, a n d Biologica l Activity of
P yr r olop yr id in es. The initial target 2 was selected for
the reasons described above and also since the desired
starting materials were readily synthesized. The syn-
thetic routes are described in Methods and Experimen-
tal Section. The very first compound tested in the
designed series proved to be active in the r-CRH1
receptor-binding assay (1, 363 nM, Table 1). Removal
of the cyano group improved the activity enough that
we felt certain we had chosen the right conformer to
mimic (2, 47 nM). Several correlations in the SAR with
the anilinopyrimidine SAR also gave us confidence: (a)
removing the ortho substituent from the phenyl ring (21,
2462 nM) decreased activity; (b) substituting both
phenyl ortho positions improved activity (7, 29 nM; 8,
1 nM); and (c) placing a polar group on the central
nitrogen substituent decreased activity (3-cyano ana-
logues and the inactive 3-carboxamido and 3-carboxylate
derivatives 22 and 23).
F igu r e 6. Crystal structure of conformationally locked pyr-
rolopyridine 2 (yellow carbons) overlaid with the calculated
structure of anilinopyrimidine 24 (cyan carbons).
the bicyclic system remain planar, the phenyl ring lie
between 60° and 90° with respect to the bicyclic system,
a lone pair acceptor lie on the inner face of the V-shaped
molecule, and valency for the required substituents be
available. Numerous variations on the position and
number of nitrogens in the planar, bicyclo[4,3,0]-ring
system were envisaged and subsequently shown to be
active, as discussed in the next paper and references
therein.
We next considered what the SARs in the anilinopy-
rimidine series told us about the best groups to place
on the constrained ring system. First, the anilinopyri-
midine maintains good potency with methyl groups in
both the 4- and the 6-positions of the pyrimidine ring,
so the dimethylpyrrolopyridine was a logical first choice
(Figure 5A). Second, although we have constrained one
degree of freedom, the phenyl ring on the pyrrole
nitrogen still has a low barrier to rotation, and the SAR
of anilinopyrimidines indicated that this is an important
factor. The aryl group in anilinopyrimidine 24 was
selected as the best test for the initial pyrrolopyridine,
leading to 2. The calculated low-energy geometries of
these two molecules overlapped very closely (not shown;
see Figure 6).
Third, the pyrrolopyridine replaces one of the pyri-
midine ring nitrogens with the ring fusion carbon,
resulting in the loss of this potential H-bond acceptor
and alteration of the electronic character of the ring.
The complete lack of antagonist potency of isopyrimidine
32 could be explained by the absence of a nitrogen in
this position. However, NMR spectra of 32, which show
only two sharp, well-separated methyl peaks between
-70 and 150 °C, provided an equally valid possibility:
that the preferred solution conformation had the pyri-
midine 2-nitrogen locked exo to the aromatic ring
(Figure 5B), perhaps due to a higher barrier for rotation
of the methine hydrogen past the N-ethyl group. The
presence of only one rotamer of the isopyrimidine in the
variable temperature NMR studies up to 140 °C is also
very strong support for this hypothesis (Table 1). These
data and the SAR seemed to point toward the essential
ring nitrogen being in the endo position; in any event,
this hypothesis would be tested by the synthesis of
pyrrolopyridine 2.
Single-crystal X-ray structures of bicyclic inhibitors
2, 3, and 9 were obtained. An overlay of the X-ray
structure of 2 with the calculated low-energy conforma-
tion of anilinopyrimidine 24 is shown in Figure 6. The
angle of the phenyl ring relative to the pyridine ring,
the only significant degree of freedom remaining, is very
close to the experimental and calculated values of
anilinopyrimidines. In addition, the crystal structure of
3 allowed regioisomeric assignment of the phenyl and
methyl groups in 3 and 4 (see Methods and Supporting
Information).
The high potency and close correlation of the structure
and SAR with the original lead confirmed that we had
developed an attractive new series with a defined
binding mode. We could now use the fixed position of
the pyrimidine ring substituents to confirm unambigu-
ously the bound conformation of anilinopyrimidines that
we proposed from the spectral and SAR studies, as
discussed above. Removing a methyl group from the
4-position resulted in a decrease in activity of about
5-fold (11 vs 2). Removal of the 6-methyl, however,
caused a greater than 500-fold loss of activity (12 vs 2).
This result shows that an endo-methyl is required for
activity. Replacing the 6-methyl with the larger phenyl
group (3 vs 2) also destroys activity, while a slight
improvement is noticed in the 4-phenyl replacement (4
vs 2). Although the 6-morpholino compound was not
prepared, analogous compounds in closely related series
are less active.8 The 4-morpholino derivatives 19 and

828 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 5
Hodge et al.
20 retain potency. Replacing methyl with chloro in the
6-position causes only a 2-fold decrease in activity (15),
suggesting that steric contacts are primarily responsible
for the sensitivity of this site to substitution. These
results clearly establish the endo-methyl conformation
as the active conformation when the other ring sub-
stituent is large; again, this is not observed in the X-ray
structure but appears to be present in solution (vide
supra).
principle required only 10-12 weeks by a single labora-
tory. Third, several methods (e.g., X-ray, variable-
temperature NMR, and semiempirical calculations, in
this case) are important to ensure a full understanding
of the conformational preferences of the molecules to
be mimicked; just an X-ray structure is frequently
misleading. Finally, and most importantly, the design
process depended heavily on the accurate SAR gener-
ated by chemists in the original lead series.
These bicyclic antagonists were subsequently ex-
tended to include a variety of other heterocyclic cores
with promising CRH1 receptor affinity, as described in
the following paper.8 Modification of the heterocycle and
side chains to optimize potency, log P, solubility, oral
availability, and CNS penetration led to the identifica-
tion of compounds with properties that warrant further
evaluation as potential clinically useful therapeutic
agents.
Comparison of the receptor-binding potency of the
acyclic and cyclic series can be made in two ways:
compounds 2 and 24 are identical homologues by
number of carbon atoms and have effectively identical
affinities (46 and 47 nM). On the other hand, the methyl
group of 24 occupies a region that cannot be filled by 2,
and it might be argued that the acyclic N-methyl
analogue 26 (96 nM) is a better comparator. In either
case the effect of cyclization is relatively minor. It is not
surprising that any entropy gain due to preorganiza-
tion24 would be small, given the restricted number of
accessible states available to the acyclic compound.
The best compound, incorporating the potent 2-bromo-
4,6-dimethoxyaryl group, binds to the r-CRH1 receptor
with an affinity of about 1 nM (8, Table 1). The
compounds are selective for the human and rat CRH1
receptor and do not bind with significant affinity to
other common cell surface receptors (not shown).
Exp er im en ta l Section
Gen er a l. All procedures were carried out under inert gas
in oven-dried glassware unless otherwise indicated. Proton
NMR spectra were obtained on VXR or Unity 300- or 400-MHz
instruments (Varian Instruments, Palo Alto, CA) with TMS
as an internal reference standard. Melting points were deter-
mined on a Mettler SP61 apparatus and are uncorrected.
Elemental analyses were performed by Quantitative Technolo-
gies, Inc., Bound Brook, NJ . High-resolution mass spectra were
carried out on a VG 70-VSE instrument with NH3 chemical
ionization. Thin layer and column chromatography were
carried out on plates or silica gel from E. Merck, Darmstadt,
FRG. Separation of optical isomers was performed using
supercritical fluid chromatography with a Chiracel column
(Daicel Chemical Ind. Ltd.) and 20% methanol modified CO2
mobile phase. Optical rotations were obtained on a Perkin-
Elmer 241 polarimeter. Solvents and reagents were obtained
from commercial vendors in the appropriate grade and used
without further purification unless otherwise indicated.
1-(2-Br om o-4-isop r op ylp h en yl)-3-cya n o-4,6-d im eth yl-
7-a za in d ole (1). P a r t A. A solution of 42.80 g (0.200 mol) of
the potassium salt of formylsuccinonitrile25 and 29.20 g (0.200
mol) of 2-bromo-4-isopropylaniline in a mixture of 50 mL of
glacial acetic acid and 120 mL of ethanol was refluxed for 2 h.
The mixture was stripped of most of the acetic acid and
ethanol. The residue was taken up in ethyl acetate, washed
with 10% sodium bicarbonate solution, dried with anhydrous
sodium sulfate, and evaporated to give a dark, oily residue.
This was chromatographed on silica gel with 80:20 hexanes-
ethyl acetate to give 24.23 g (40%) of N-(2-bromo-4-isopropy-
Con clu sion s
This report details a design and synthesis project that
successfully yielded both a new series of antagonists and
insight into the bound conformation of the original
series. The original hypothesissthat free low-energy
conformations were useful starting points for inferring
a bound conformationswas validated for this series by
the results. The first compound synthesized was active,
and we attained equal potency with the lead series as
well as structural novelty. The well-defined shape of the
pharmacophore that resulted from the study also al-
lowed the design of completely different scaffolds that
will be described at a later time.
It is a fair critique to point out that these systems
are simple, possess limited degrees of freedom, and
therefore are ideal and perhaps predisposed to success
in a study of this kind. In defense of taking the time to
obtain an accurate initial understanding of the confor-
mational preferences, however, it is important to reiter-
ate several points that were made in the discussion:
first, the choice of methods is crucial to success. This
method should not be chosen to design mimetics of
ligands with many equi-energetic minima and internal
degrees of freedom unless resources are available to
study them all systematically. The approach becomes
much less efficient with shallow conformational energy
surfaces, and this can be determined computationally
at the outset. Second, even in this simple system there
are many possible ways of conformational locking, and
many chemist-months have been spent in unsuccessful
attempts to mimic molecular shapes with “tied-up” or
“tied-back” analogues without understanding the three-
dimensional shape of the initial lead. In contrast,
obtaining experimental and calculated structural infor-
mation required just a few weeks of effort, and the
synthesis of the initial pyrrolopyrimidine for proof of
lphenyl)aminomethylenesuccinonitrile. MS: (M + NH4)+
321.0; calcd, 321.0.
)
P a r t B. To a solution of 10 mL of 1 M potassium tert-
butoxide in tetrahydrofuran and 10 mL of ethanol was added
1.11 g (3.65 mmol) of N-(2-bromo-4-isopropylphenyl)amino-
methylenesuccinonitrile (part A), and the mixture was stirred
for 16 h. The solvents were removed by evaporation. The
residue was taken up in ethyl acetate and was washed
successively with 1 N hydrochloric acid, 10% sodium bicarbon-
ate solution, and brine. The solution was dried with anhydrous
sodium sulfate and evaporated to give a dark residue. The
residue was dissolved in dichloromethane, 20 g of silica gel
was added, and the mixture was evaporated to dryness. This
mixture was placed on top of a chromatographic column of 150
g of silica gel in hexane. The column was eluted successively
with 10, 15, 20, 25, and 30% ethyl acetate in hexane to give
0.65 g (59% yield) of 1-(2-bromo-4-isopropylphenyl)-2-amino-
4-cyanopyrrole. MS: (M + H)+ ) 304.0; calcd, 304.0. Rf ) 0.22
on silica gel thin-layer chromatography (70:30 hexanes-ethyl
1
acetate). H NMR (300 MHz, CDCl3) δ 7.58 (1H, d, J ) 1.10
Hz), 7.29 (2H, m), 6.84 (1H, d, J ) 1.46 Hz), 5.75 (1H, d, J )
1.56 Hz), 3.22 (2H, br s), 2.97 (6H, d, J ) 6.96).

CRH Receptor Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 5 829
P a r t C. Parts A and B were scaled up for part C. A mixture
of 18.51 g (0.0609 mol) of 1-(2-bromo-4-isopropylphenyl)-2-
amino-4-cyanopyrrole, 300 mL of ethanol, 0.6 mL of concen-
trated hydrochloric acid, and 10 mL (9.75 g, 0.0974 mol) of
2,4-pentanedione was refluxed with stirring under a nitrogen
atmosphere for 4 h. The mixture was allowed to cool, and the
solvent was removed under reduced pressure. The residue was
dissolved in ethyl acetate. The solution was washed with 10%
sodium bicarbonate solution and then with brine. The solution
was dried with anhydrous sodium sulfate and evaporated to
give 21.76 g of dark, tarry residue. The residue was chromato-
graphed on silica gel by eluting in step gradients of 0, 10, 15,
20, 25, and 30% ethyl acetate in hexane. The initial fraction
was 17.6 g (78%) of 1-(2-bromo-4-isopropylphenyl)-3-cyano-4,6-
dimethyl-7-azaindole, mp 108.2 °C. High-resolution MS: (M
+ H)+ ) 368.0749; calcd, 368.0762 (79Br). Rf ) 0.45 on silica
gel thin-layer chromatography with 70:30 hexanes-ethyl
separator for 2 h. The solvent was removed by evaporation,
and the residue was chromatographed on silica gel as described
above to yield 138 mg (16%) of 1-(2-bromo-4-isopropylphenyl)-
3-cyano-6-methyl-4-phenyl-7-azaindole (4), along with 350 mg
of the 4-methyl-6-phenyl isomer (41%) and a substantial
amount of unreacted pyrrole.
1-(2-Br om o-4-isop r op ylp h en yl)-6-m et h yl-4-p h en yl-7-
a za in d ole (5). A mixture of 130 mg (0.302 mmol) of 1-(2-
bromo-4-isopropylphenyl)-3-cyano-6-methyl-4-phenyl-7-azain-
dole (vide supra) and 10 mL of 65% sulfuric acid was refluxed
for 1 h. The mixture was poured onto ice, and concentrated
ammonium hydroxide was added until the mixture was basic
to pH paper. The mixture was extracted with ethyl acetate.
The extract was evaporated and chromatographed on silica
gel with 70:30 hexanes-ethyl acetate. There was obtained 112
mg (92% yield) of 1-(2-bromo-4-isopropylphenyl)-6-methyl-4-
phenyl-7-azaindole, mp 115 °C. High-resolution MS: (M + H)+
) 405.0986; calcd, 405.0966. 1H NMR (300 MHz, CDCl3) δ 7.71
(1H, d, J ) 1.83 Hz), 7.68 (1H, d, J ) 0.73 Hz), 7.54 (1H, d, J
) 0.73 Hz), 7.35-7.48 (4H, m), 7.26 (1H, d, J ) 1.83 Hz), 7.23
(1H, d, J ) 3.56 Hz), 7.02 (1H, s), 6.68 (1H, d, J ) 3.61 Hz),
2.91 (1H, sept, J ) 6.96 Hz), 2.56 (3H, s), 1.25 (6H, d, J )
6.96 Hz). Anal. Calcd for C23H21BrN2‚0.1CH3COOC2H5: C,
67.86; H, 5.31; N, 6.76. Found: C, 67.73; H, 5.53; N, 6.38.
1-(2-Br om o-4-isop r op ylp h en yl)-4-m et h yl-6-p h en yl-7-
a za in d ole (6). In the same way, 1-(2-bromo-4-isopropylphen-
yl)-4-methyl-6-phenyl-7-azaindole was obtained, mp 95.8 °C
(82% yield). 1H NMR (300 MHz, CDCl3) δ 8.01 (1H, d, J )
1.46 Hz), 7.98 (1H, s), 7.62 (1H, d, J ) 1.73 Hz), 7.51 (1H, d,
J ) 8.05 Hz), 7.29-7.43 (6H, m), 6.66 (1H, d, J ) 3.66 Hz),
2.99 (1H, sept, J ) 6.96 Hz), 2.66 (3H, s), 1.32 (6H, d, J )
6.96 Hz). High-resolution MS: (M + H)+ ) 405.0972; calcd,
405.0966. Anal. Calcd for C23H21BrN2: C, 68.15; H, 5.22; N,
6.91. Found: C, 68.22; H, 5.46; N, 6.54.
1
acetate. H NMR (300 MHz, CDCl3) δ 7.73 (1H, s), 7.61 (1H,
d, J ) 1.47 Hz), 7.40 (1H, d, J ) 8.42 Hz), 7.33 (1H, dd, J )
1.84, 8.43 Hz), 6.93 (1H, s), 2.99 (1H, sept, J ) 6.96 Hz), 2.77
(3H, s), 2.53 (3H, s), 1.32 (6H, d, J ) 6.96 Hz). Anal. Calcd for
C
19H18BrN3: C, 61.97; H, 4.93; N, 11.41. Found: C, 61.78; H,
4.91; N, 11.42.
1-(2-Br om o-4-isop r op ylp h en yl)-4,6-d im et h yl-7-a za in -
d ole (2). A mixture of 4.00 g of 1-(2-bromo-4-isopropyl)-3-
cyano-4,6-dimethyl-7-azaindole and 40 mL of 65% sulfuric acid
was refluxed for 1 h. The solution was cooled and poured onto
ice. Concentrated ammonium hydroxide was added until the
mixture was alkaline to pH paper. The mixture was extracted
with ethyl acetate. The solution was dissolved in 60:40
hexanes-ethyl acetate and was passed through a short column
of silica gel. The eluate was evaporated, and the residue was
crystallized from 20 mL of hexane to give 2.45 g (66% yield)
of 1-(2-bromo-4-isopropylphenyl)-4,6-dimethyl-7-azaindole, mp
76 °C. High-resolution MS: (M + H)+ ) 343.0818; calcd,
343.0810. Rf ) 0.57 on silica gel (70:30 hexanes-ethyl acetate).
1H NMR (300 MHz, CDCl3) δ 7.58 (1H, d, J ) 1.83 Hz), 7.43
(1H, d, J ) 8.06 Hz), 7.29 (1H, dd, J ) 1.83, 8.06 Hz), 7.23
(1H, d, J ) 3.66 Hz), 6.82 (1H, s), 6.59 (1H, d, J ) 3.29 Hz),
2.90 (1H, sept, J ) 6.95 Hz), 2.55 (3H, s), 2,54 (3H, s), 1.30
(6H, d, J ) 6.96 Hz). Anal. Calcd for C18H19BrN2: C, 62.98;
H, 5.58; N, 8.16. Found: C, 62.92; H, 5.60; N, 8.15.
1-(2-Br om o-4-isop r op ylp h en yl)-3-cya n o-4-m et h yl-6-
p h en yl-7-a za in d ole (3) a n d (2-Br om o-4-isop r op ylp h en -
yl)-3-cya n o-6-m eth yl-4-p h en yl-7-a za in d ole (4). A mixture
of 737 mg (2.00 mmol) of 1-(2-bromo-4-isopropylphenyl)-2-
amino-4-cyanopyrrole (vide supra), 324 mg (2.00 mmol) of
benzoylacetone, 10 mL of ethanol, and 0.025 mL of concen-
trated HCl was refluxed together for 48 h. The mixture was
dissolved in dichloromethane-ethyl acetate and chromato-
graphed on silica gel eluting in step gradients with 0, 5, 10,
and 15% ethyl acetate in hexane to yield 1-(2-bromo-4-
isopropylphenyl)-3-cyano-4-methyl-6-phenyl-7-azaindole, mp
130 °C (0.66 g, 77% yield) and 1-(2-bromo-4-isopropylphenyl)-
3-cyano-6-methyl-4-phenyl-7-azaindole, mp 198 °C (0.02 g, 2%
yield). The Rf values were respectively 0.38 and 0.28 (silica
gel with 80:20 hexanes-ethyl acetate).
1-(2-Br om o-4,6-dim eth oxyph en yl)-3-cyan o-4,6-dim eth yl-
7-a za in d ole (7). P a r t A. N-(2-Bromo-4,6-dimethoxyphenyl)-
aminomethylenesuccinonitrile was prepared from 2-bromo-4,6-
dimethoxyaniline by the method described for N-(2-bromo-4-
isopropylphenyl)aminomethylenesuccinonitrile (vide supra).
MS: (M + H)+ ) 322.0; calcd, 322.16 (79Br). Rf ) 0.19 (silica
gel with 60:40 hexanes-ethyl acetate).
P a r t B. The product from part A was cyclized with potas-
sium tert-butoxide as described above for the cyclization of
N-(2-bromo-4-isopropylphenyl)aminomethylenesuccinonitrile
to give 1-(2-bromo-4,6-dimethoxyphenyl)-2-amino-4-cyanopy-
rrole, mp 161 °C, in 79% yield. Rf ) 0.19 (silica gel with 60:40
hexanes-ethyl acetate). High-resolution MS: (M + H)+
)
1
321.0095; calcd, 321.0113. H NMR (300 MHz, CDCl3) δ 6.82
(1H, d, J ) 2.56 Hz), 6.76 (1H, d, J ) 2.83 Hz), 6.53 (1H, d, J
) 2.56 Hz), 5.76 (1H, d, J ) 2.83 Hz), 3.85 (3H, s), 3.79 (3H,
s), 3.13 (2H, br s).
P a r t C. The product from part B was treated with 2,4-
pentanedione as described above for the preparation 1-(2-
bromo-4-isopropylphenyl)-3-cyano-4,6-dimethyl-7-azaindole to
give 1-(2-bromo-4,6-dimethoxyphenyl)-3-cyano-4,6-dimethyl-
7-azaindole, mp 186 °C, in 92% yield. MS: (M + H)+ ) 388.0;
calcd, 388.0. Rf ) 0.44 (silica gel with 60:40 hexanes-ethyl
acetate). 1H NMR (300 MHz, DMSO-d6) δ 8.37 (1H, s), 7.03
(1H, s), 7.01 (1H, d, J ) 2.56 Hz), 6.84 (1H, d, J ) 2.20 Hz),
3.89 (3H, s), 3.69 (3H, s), 2.67 (3H, s), 2.41 (3H, s). Anal. Calcd
for C18H16BrN3O2: C, 55.97; H, 4.19; N, 10.88. Found: C, 55.79;
H, 4.09; N, 10.65.
1-(2-Br om o-4,6-d im eth oxyp h en yl)-4,6-d im eth yl-7-a za -
in d ole (8). A mixture of 200 mg of 1-(2-bromo-4,6-dimethoxy-
phenyl)-3-cyano-4,6-dimethyl-7-azaindole and 10 mL of 65%
sulfuric acid was refluxed for 1 h. The mixture was worked
up as described above for other decyanylations in this series
to give 185 mg of crude product. MS of this crude material
showed (M + H)+ ) 361.0 (major), 347.0, and 330.0. These
numbers suggest that some mono- and bis-O-demethylation
accompanied the decyanylation procedure. A 40-mg portion
was purified by preparative HPLC on a nitrile column using
95:5 1-chlorobutane-acetonitrile to give 11 mg of 1-(2-bromo-
4,6-dimethoxyphenyl)-4,6-dimethyl-7-azaindole. High-resolu-
For the 4-methyl-6-phenyl compound: 1H NMR (300 MHz,
CDCl3) δ 7.98 (1H, d, J ) 1.46 Hz), 7.95 (1H, s), 7.85 (1H, s),
7.64 (1H, d, J ) 1.47 Hz), 7.53 (1H, s), 7.34-7.48 (5H, m),
3.01 (1H, sept, J ) 6.98 Hz), 2.88 (3H, s), 1.34 (6H, d, J )
6.96 Hz). High-resolution MS: (M + H)+ ) 430.0901; calcd,
430.0919.
For the isomeric 6-methyl-4-phenyl compound: 1H NMR
(300 MHz, CDCl3) δ 7.82 (1H, s), 7.64 (2H, 2s), 7,34-7.59 (6H,
m), 7.14 (1H, s), 3.01 (1H, sept, J ) 6.96 Hz), 2.62 (3H, s),
1.33 (6H, d, J ) 6.96 Hz). High-resolution MS: (M + H)+
)
430.0911; calcd, 430.0919. Anal. Calcd for
C24H20BrN3‚
0.1C6H14: C, 67.31; H, 4.91; N, 9.57. Found: 67.20; H, 4.83;
N, 9.53.
To improve the yield of the 4-phenyl isomer, 25 mL of
xylene, 737 mg (2.00 mmol) of 1-(2-bromo-4-isopropylphenyl)-
2-amino-4-cyanopyrrole (vide supra), and 324 mg (2.00 mmol)
of benzoylacetone were heated in a flask equipped with a water

830 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 5
Hodge et al.
tion MS: (M + H)+ ) 360.0479; calcd, 360.0473. A better
solution to the attendant O-demethylation was found: the
crude reaction mixture (175 mg, 0.5 mmol) was remethylated
by stirring the mixture with lithium bis(trimethylsilylamide)
(2.28 g, 16.1 mmol) and methyl iodide (2.5 mmol, 1 M in THF)
in tetrahydrofuran (10 mL) for 16 h. The resulting product
could then be purified relatively easily by preparative thin-
layer chromatography on silica gel with 95:5 methylene
chloride-methanol (Rf 0.31). Finally, the product was crystal-
lized from methanol to give crystals (32 mg) of the dimethoxy
compound, mp 202 °C. 1H NMR (300 MHz, CDCl3) δ 7.02 (1H,
d, J ) 3.66 Hz), 6.85 (1H, d, J ) 2.56 Hz), 6.78 (1H, s), 6.60
(1H, d, J ) 3.66 Hz), 6.56 (1H, d, J ) 2.56 Hz), 3.86 (3H, s),
1-(2-Br om o-4-isop r op ylp h en yl)-6-h yd r oxy-3-cya n o-4-
m eth yl-7-azain dole (13) an d 1-(2-Br om o-4-isopr opylph en -
yl)-6-ch lor o-3-cya n o-4-m eth yl-7-a za in d ole (14). P a r t A. A
solution of 3.04 g of 1-(2-bromo-4-isopropylphenyl)-2-amino-
4-cyanopyrrole, 1.9 mL (1.94 g, 14.9 mmol) of ethyl aceto-
acetate, and 0.1 mL of concentrated hydrochloric acid in 30
mL of ethanol was refluxed for 16 h. A precipitate formed upon
cooling. The precipitate was removed by filtration to give 1.68
g of crystals, mp 202 °C, of 1-(2-bromo-4-isopropylphenyl)-4-
methyl-7-azaindol-6-one. TLC on silica gel with 70:30 hex-
anes-ethyl acetate showed a single spot, Rf ) 0.29. MS: (M
+ H)+ ) 370.0; calcd, 370.05 (79Br). 1H NMR (300 MHz, CDCl3)
δ 7.60 (1H, d, J ) 0.73 Hz), 7.48, (1H, s), 7.30 (2H, 2s), 6.21
(1H, d, J ) 0.73 Hz), 3.03 (1H, sept, J ) 6.96 Hz), 2.70 (3H,
s), 1.35 (6H, d, J ) 6.96 Hz).
3.68 (3H, s), 2.54 (3H, s), 2.51 (3H, s). Anal. Calcd for C17H17
-
BrN2O2: C, 56.52; H, 4.78; N, 7.75. Found: C, 56.69; H, 4.38;
N, 7.58.
P a r t B. A mixture of 185 mg of the 7-azaindol-6-one (part
A) and 50 mL of phosphorus oxychloride was heated in an
autoclave at 180 °C for 10 h. The excess phosphorus oxychlo-
ride was removed by distillation at reduced pressure. The
residue was distributed between ethyl acetate and water. The
ethyl acetate layer was separated and washed with 10%
sodium bicarbonate solution and then with brine. The solution
was dried (Na2SO4) and evaporated. TLC of the residue on
silica gel with 70:30 hexanes-ethyl acetate showed a major
new product, Rf ) 0.52 with minor spots at Rf 0.45 and 0.29.
Chromatography on silica gel with step gradients of 5, 10, 15,
and 20% ethyl acetate in hexane gave 109 mg (56%) of the Rf
0.52 product, 1-(2-bromo-4-isopropylphenyl)-6-chloro-3-cyano-
1-(2-Br om o-4-isopr opylph en yl)-3-cyan o-6-m eth yl-7-aza-
in d ole (9) a n d 1-(2-Br om o-4-isop r op ylp h en yl)-3-cya n o-
4-m eth yl-7-a za in d ole (10). To a solution of 1.085 g (5.07
mmol) of 1-(2-bromo-4-isopropylphenyl)-2-amino-4-cyanopyr-
role and 0.80 mL (0.797 g, 6.03 mmol) of acetoacetaldehyde
dimethyl acetal in 20 mL of ethanol was added 0.10 mL of
concentrated hydrochloric acid. The mixture was refluxed for
16 h. The mixture was cooled and evaporated to give a dark,
thick oil. TLC on silica gel with 70:30 hexanes-ethyl acetate
showed two major spots at Rf 0.47 and 0.41. The oil was
dissolved in ethyl acetate, 20 mL silica gel powder was added,
and the mixture was evaporated to dryness. The powdery
residue was loaded on top of a column of 60 mL of silica gel in
hexane. The column was eluted in step gradients of 0, 5, 10,
15, 20, and 25% ethyl acetate in hexane. The first fraction to
elute was 0.32 g of 1-(2-bromo-4-isopropylphenyl)-3-cyano-6-
methyl-7-azaindole, Rf 0.47 (silica gel with 70:30 hexanes-
ethyl acetate). The material was crystallized from hexane to
give 176 mg (10%) of crystals, mp 125 °C. High-resolution
MS: (M + H)+ ) 354.0603; calcd, 354.0606. 1H NMR (300
MHz, CDCl3) δ 8.01 (1H, d, J ) 8.06 Hz), 7.76 (1H, s), 7.62
(1H, d, J ) 1.83 Hz), 7.41 (1H, d, J ) 8,06 Hz), 7.34 (1H, dd,
J ) 1.83, 8,42 Hz), 7.17 (1H, d, J ) 8.06 Hz), 3.00 (1H, sept,
J ) 6.96 Hz), 2.60 (3H, s), 1.32 (6H, d, J ) 6.96 Hz). Anal.
Calcd for C18H16BrN3: C, 61.03; H, 4.55; N, 11.86. Found: C,
60.97; H, 4.48; N, 11.84.
1
4-methyl-7-azaindole, mp 124 °C. H NMR (300 MHz, CDCl3)
δ 7.79 (1H, d, J ) 5.13 Hz), 7.60 (1H, d, J ) 1.83 Hz), 7.28-
7.44 (2H, m), 7.11 (1H, s), 2.99 (1H, sept, J ) 6.96 Hz), 2.81
(3H, s), 1.31 (6H, d, J ) 6.59 Hz). Anal. Calcd for C18H15
-
BrClN3: C, 55.62; H, 3.90; N, 10.81; Cl, 9.12. Found: C, 57.80;
H, 4.10; N, 10.89; Cl, 12.17.
1-(2-Br om o-4-isop r op ylp h en yl)-6-ch lor o-4-m et h yl-7-
a za in d ole (15). A mixture of 52 mg of 1-(2-bromo-4-isopro-
pylphenyl)-6-chloro-3-cyano-4-methyl-7-azaindole and 10 mL
of 65% sulfuric acid was refluxed for 1 h. The cooled solution
was poured onto ice, and 17 mL of concentrated ammonium
hydroxide was added. The alkaline mixture was extracted with
ethyl acetate. The extract was washed (brine), dried (Na2SO4),
and evaporated. TLC of the residue on silica gel with 70:30
hexanes-ethyl acetate showed a major new spot, Rf ) 0.58,
with a trace of unchanged starting material (Rf 0.52). The
crude product was purified by preparative TLC on silica gel
with 70:30 hexanes-ethyl acetate to give 39 mg (79% yield)
of product, which slowly crystallized on standing, mp 75 °C.
High-resolution MS: (M + H)+ ) 363.0247; calcd, 363.0264
(79Br, 35Cl). Anal. Calcd for C17H16BrClN2‚0.14‚C6H14: C, 57.03;
H, 4.82; N, 7.46. Found: C, 57.42; H, 4.54; N, 7.52.
The second fraction to elute was 180 mg (after recrystalli-
zation from hexane) (10%) of 1-(2-bromo-4-isopropylphenyl)-
3-cyano-4-methyl-7-azaindole, mp 176.0 °C, Rf 0.4. MS: (M +
1
H)+ ) 354.0595; calcd, 354.0606. H NMR (300 MHz, CDCl3)
δ 8.31 (1H, d, J ) 4.76 Hz), 7.82 (1H, s), 7.63 (1H, d, J ) 1.83
Hz), 7.40 (1H, d, J ) 8.06 Hz), 7.36 (1H, dd, J ) 1.83, 8.42
Hz), 7.07 (1H, dd, J ) 0.73, 4.76 Hz), 2.99 (1H, sept, J ) 6.96
Hz), 2.84 (3H, s), 1.31 (6H, d, J ) 6.96). Anal. Calcd for C18H16
-
BrN3: C, 61.03; H, 4.55; N, 11.86. Found: C, 61.00; H, 4.39;
N, 11.86.
1-(2-Br om o-4-isop r op ylp h en yl)- 3-cya n o-6-m et h yl-7-
a za in d ole 7-N-Oxid e (16) a n d 1-(2-Br om o-4-isop r op yl-
p h en yl)-4-ch lor o-3-cya n o-6-m eth yl-7-a za in d ole (17). P a r t
A. A solution of 1.24 g of 1-(2-bromo-4-isopropylphenyl)-3-
cyano-6-methyl-7-azaindole (vide supra) and 1.42 g of 85%
3-chloroperoxybenzoic acid in 20 mL of chloroform was refluxed
for 6 h. The mixture was cooled and washed first with 10%
sodium bicarbonate solution and then with brine. The solution
was dried (Na2SO4) and evaporated. TLC of the residue on
silica gel with 95:5 dichloromethane-methanol showed a trace
spot at Rf 0.88 and a major spot at Rf 0.34. The material was
purified by chromatography on silica gel with dichloromethane,
followed by 1% methanol in dichloromethane, to give a trace
of unchanged 1-(2-bromo-4-isopropylphenyl)-3-cyano-6-methyl-
7-azaindole (Rf 0.88) and 0.92 g (71%) of 1-(2-bromo-4-isopro-
pylphenyl)-3-cyano-6-methyl-7-azaindole 7-oxide (Rf 0.34), mp
179 °C. High-resolution MS: (M + H)+ ) 370.0559; calcd,
370.0555 (79Br). 1H NMR (300 MHz, CDCl3) δ 7.64 (1H, d, J )
8.43 Hz), 7.54 (2H, m), 7.42 (1H, d, J ) 8.05 Hz), 7.29 (1H,
dd, J ) 1.83, 8.05 Hz), 7.24 (1H, d, J ) 8.06 Hz), 2.98 (1H,
sept, J ) 6.96 Hz), 2.58 (3H, s), 1.30 (6H, d, J ) 6.59 Hz).
Anal. Calcd for C18H16BrN3O: C, 58.39; H, 4.36; N, 11.35.
Found: C, 58.15; H, 4.17; N, 11.22.
1-(2-Br om o-4-isop r op ylp h en yl)-6-m et h yl-7-a za in d ole
(11). Compound 9 was decyanylated with 65% sulfuric acid
as described above to give the desired product as a viscous
oil. TLC on silica gel with 70:30 hexanes-ethyl acetate showed
Rf ) 0.57. High-resolution MS: (M + H)+ ) 329.0641; calcd,
329.0653 (79Br). 1H NMR (300 MHz, CDCl3) δ 7.84 (1H, d, J )
8.06 Hz), 7.58 (1H, d, J ) 1.83 Hz), 7.44 (1H, d, J ) 8.06 Hz),
7.29 (1H, dd, 1.83, J ) 8.06 Hz), 7.24 (1H, d, J ) 3.66 Hz),
6.99 (1H, d, J ) 8.06 Hz), 6.56 (1H, d, J ) 3.30 Hz), 2.96 (1H,
sept, J ) 6.96 Hz), 2.58 (3H, s), 1.30 (6H, d, J ) 6.96 Hz).
Anal. Calcd C17H17BrN2‚H2O: C, 58.80; H, 5.52; N, 8.07.
Found: C, 58.50; H, 4.93; N, 8.07.
1-(2-Br om o-4-isop r op ylp h en yl)-4-m et h yl-7-a za in d ole
(12). Compound 10 was decyanylated with 65% sulfuric acid
as described above to give the desired product as a viscous
oil. TLC on silica gel with 70:30 hexanes-ethyl acetate showed
Rf ) 0.48. High-resolution MS: (M + H)+ ) 329.063; calcd,
1
329.0653 (79Br). H NMR (300 MHz, CDCl3): δ 8.22 (1H, d, J
) 4.76 Hz), 7.60 (1H, d, J ) 1.83 Hz), 7.41 (1H, d, J ) 8.06
Hz), 7.29-7.33 (2H, m), 6.93 (1H, d, J ) 4.76 Hz), 6.65 (1H, d,
J ) 3.60 Hz), 2.97 (1H, sept, J ) 6.96 Hz), 2.61 (3H, s), 1.30
(6H, d, J ) 6.96 Hz).

CRH Receptor Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 5 831
P a r t B. A mixture of 370 mg of the 7-oxide (part A) and 5
mL of phosphorus oxychloride was refluxed for 2 h. The
solution was cooled, poured on ice, and stirred until most of
the phosphorus oxychloride had hydrolyzed. The mixture was
made alkaline with concentrated ammonium hydroxide and
was extracted with ethyl acetate. The extract was dried (Na2-
SO4) and evaporated to give a viscous residue. TLC on silica
gel with 95:5 dichloromethane-methanol showed a major spot
at Rf ) 0.79. The material was purified by preparative TLC
on silica gel with 70:30 hexanes-ethyl acetate to give crystals.
Recrystallization from hexane gave 158 mg (41%) of 1-(2-
bromo-4-isopropylphenyl)-4-chloro-3-cyano-6-methyl-7-azain-
dole, mp 123 °C. High-resolution MS: (M + H)+ ) 388.0197;
calcd, 388.0216 (79Br, 35Cl). 1H NMR (300 MHz, CDCl3) δ 7.79
(1H, s), 7.62 (1H, d, J ) 1.47 Hz), 7.35 (2H, m), 7.16 (1H, s),
3.00 (1H, sept, J ) 6.96 Hz), 2.56 (3H, s), 1.32 (6H, d, J )
6.96 Hz). Anal. Calcd for C18H15ClBrN3: C, 55.62; H, 3.90; N,
10.81. Found: C, 55.67; H, 3.81; N, 10.70.
(1H, d, J ) 3.7 Hz), 6.37 (1H, s), 3.94 (4H, t, J ) 4.8 Hz), 3.46
(4H, t, J ) 4.8 Hz), 2.96 (1H, sept, J ) 6.8 Hz), 2.51 (3H, s),
1.30 (7H, d. J ) 6.6 Hz). High-resolution MS: (M + H)+
)
416.1152; calcd, 416.1161. Anal. Calcd for C21H24BrN3O: C,
60.87; H, 5.84; N, 10.14. Found: C, 60.83; H, 5.96; N, 10.08.
1-(4-Isop r op ylp h en yl)-4,6-d im eth yl-7-a za in d ole (21). A
mixture of 250 mg (0.729 mmol) of 1-(2-bromo-4-isopropyl)-
4,6-dimethyl-7-azaindole, 20 mL of acetic acid, and 50 mg of
10% Pd on carbon was hydrogenated for 16 h at atmospheric
pressure. The catalyst was filtered off, and the filtrate was
evaporated on a rotary evaporator. The residue was dissolved
in ethyl acetate and washed first with 10% aqueous NaHCO3
and then brine. The solution was dried (Na2SO4) and evapo-
rated to give 0.20 g (100%) of 1-(4-isopropylphenyl)-4,6-
dimethyl-7-azaindole, an oil. High-resolution MS: (M + H)+
) 265.1693; calcd, 265.1705. 1H NMR (300 MHz, CDCl3) δ 7.70
(2H, d, J ) 8.8 Hz), 7.38 (1H, d, J ) 3.7 Hz), 7.34 (2H, d, J )
8.4 Hz), 6.82 (1H, s), 6.57 (1H, d, J ) 3.7 Hz), 2.96 (1H, sept,
J ) 6.9 Hz), 2.59 (3H, s), 2.54 (3H, s), 1.29 (6H, d, J ) 7.0
Hz). Anal. Calcd for C18H20N2‚0.5H2O: C, 79.08; H, 7.74; N,
10.25. Found: C. 79.24; H, 7.50; N, 10.03.
1-(2-Br om o-4-isop r op ylp h en yl)-3-ca r boxa m id o-4,6-d i-
m eth yla za in d ole (22). To a stirred solution of 1.10 g of 1-(2-
bromo-4-isopropylphenyl)-3-cyano-4,6-dimethyl-7-azaindole in
2.5 mL of dimethyl sulfoxide were added 1.00 g (7.2 mmol) of
pulverized anhydrous K2CO3 and 5 mL (5.55 g, 1.67 g net, 48.9
mmol) of 30% H2O2 in portions. The mixture frothed. It was
allowed to stir 16 h at room temperature. The mixture was
added to a mixture of ethyl acetate and water. The EtOAc layer
was separated and washed again with water and then with
brine. The solution was dried (Na2SO4) and evaporated to give
1.03 g (89%) of crystals, mp 237 °C. Thin-layer chromatography
on silica gel (95:5 methylene chloride-methanol) showed one
spot, Rf 0.22. MS: (M + H)+ ) 386.1; calcd, 386.1. 1H NMR
(300 MHz, DMSO-d6) δ 7.95 (1H, s), 7.74 (1H, s), 7.54 (1H,
br), 7.46 (2H, 2s), 7.03 (1H, br), 6.92 (1H, s), 3.05 (1H, sept, J
) 6.96 Hz), 2.72 (3H, s), 2.39 (3H, s), 1.28 (6H, d, J ) 6.96
Hz).
1-(2-Br om o-4-isopr opylph en yl)-3-car boxyl-4,6-dim eth yl-
7-a za in d ole (23). A solution of 772 mg (2.00 mmol) of 1-(2-
bromo-4-isopropylphenyl)-3-carboxamido-4,6-dimethyl-7-aza-
indole and 500 mg (425 mg net, 7.57 mmol) of 85% KOH in 10
mL of diethylene glycol was heated at 225 °C for 2 h. The
solution was partially cooled and poured into 100 mL of water.
The solution was filtered to remove a trace of cloudiness.
Addition of 1 N HCl gave a precipitate, which was filtered off,
washed with water, and dried to give 713 mg of light gray solid.
The solid was chromatographed on silica gel by eluting with
1% methanol-methylene chloride. There was obtained 507 mg
(65%) of crystals, mp 242 °C dec. MS: (M + H)+ ) 387.0; calcd,
387.1. 1H NMR (300 MHz, DMSO-d6) δ 8.11 (1H, s), 7.74 (1H,
d, J ) 2.03 Hz), 7.44-7.51 (2H, m), 6.97 (1H, s), 3.03 (1H,
sept, J ) 6.96 Hz), 2.79 (3H, s), 2.39 (3H, s), 1.28 (6H, d, J )
6.96).
1-(2-Br om o-4-isop r op ylp h en yl)-4-ch lor o-6-m et h yl-7-
a za in d ole (18). A mixture of 190 mg of compound 17 and 5
mL of 65% sulfuric acid was refluxed for 30 min. The solution
was poured onto ice and extracted with ethyl acetate. The
extract was washed with brine, dried (Na2SO4), and evapo-
rated. TLC of the residue on silica gel with 60:40 hexanes-
ethyl acetate showed a major spot at Rf ) 0.67. The residue
was purified by preparative TLC to give 130 mg (73%) of
viscous oil, 1-(2-bromo-4-isopropylphenyl)-4-chloro-6-methyl-
7-azaindole. High-resolution MS: (M + H)+ ) 363.0246; calcd,
1
363.0264 (79Br, 35Cl). H NMR (300 MHz, CDCl3) δ 7.59 (1H,
d, J ) 1.83 Hz), 7.40 (1H, d, J ) 8.06 Hz), 7.30 (1H, dd, J )
1.83, 8.06 Hz), 7.27 (1H, d, J ) 4.03 Hz), 7.03 (1H, s), 6.66
(1H, d, J ) 3.66 Hz), 2.98 (1H, sept, J ) 6.96 Hz), 2.55 (3H,
s), 1.31 (6H, d, J ) 6.96 Hz). Anal. Calcd for C17H16ClBrN2‚
0.33H2O: C, 55.34; H, 4.51; N, 7.59. Found: C, 55.36; H, 4.13;
N, 7.38.
1-(2-Br om o-4-isop r op ylp h en yl)-3-cya n o-6-m eth yl-4-(1-
m or p h olin o)-7-a za in d ole (19). A solution of 100 mg of 1-(2-
bromo-4-isopropylphenyl)-3-cyano-6-methyl-7-azaindole 7-ox-
ide in 5 mL of chloroform was treated with 0.10 mL (168 mg)
of trifluoromethanesulfonic anhydride. The mixture was stirred
at room temperature for 20 min. Anhydrous morpholine (0.20
mL, 200 mg) was added, and the mixture was refluxed for 2
h. The mixture was distributed between ethyl acetate and 10%
aqueous NaHCO3. The organic layer was washed with brine,
dried (Na2SO4), and evaporated to give 116 mg (98%) of solid.
Recrystallization from ethanol gave 74 mg of crystalline 1-(2-
bromo-4-isopropylphenyl)-3-cyano-6-methyl-4-(1-morpholino)-
7-azaindole, mp 192 °C. TLC on silica gel with 60:40 hexanes-
ethyl acetate showed one spot, Rf 0.42. High-resolution MS:
(M + H)+ ) 441.1099; calcd, 441.1113 (81Br). 1H NMR (300
MHz, CDCl3) δ 7.69 (1H, s), 7.61 (1H, d, J ) 1.83 Hz), 7.39
(1H, d, J ) 1.83 Hz), 7.33 (1H, dd, J ) 1.83, 8.06 Hz), 6.56
(1H, s), 4.02 (4H, tr, J ) 4.39 Hz), 3.33 (4H, tr, J ) 4.39 Hz),
2.99 (1H, sept, J ) 6.96 Hz), 2.52 (3H, s), 1.31 (6H, d, J )
6.95 Hz). Anal. Calcd for C22H23BrN4O: C, 60.14; H, 5.28; N,
12.75; Br, 18.19. Found: C, 60.14; H, 5.26; N, 12.57; Br, 18.37.
1-(2-Br om o-4-isopr opylph en yl)-6-m eth yl-4-(1-m or ph oli-
n o)-7-a za in d ole (20). A mixture of 48 mg of 1-(2-bromo-4-
isopropylphenyl)-3-cyano-6-methyl-4-(1-morpholino)-7-azain-
dole and 5 mL of 65% sulfuric acid was refluxed for 45 min.
The solution was poured into ice water. The solution was made
alkaline with 10 mL of concentrated ammonium hydroxide,
and a precipitate formed. Ethyl acetate was added, and the
precipitate dissolved upon stirring. The organic extract was
washed with brine, dried (Na2SO4), and evaporated to give 46
mg of crystalline residue. TLC of the residue on silica gel with
60:40 hexanes-ethyl acetate showed a major spot at Rf ) 0.36.
The major component was isolated by preparative TLC of silica
gel using 50:50 hexanes-ethyl acetate to give 33 mg (73%
yield) of product. Recrystallization from hexane gave 19 mg
of crystals of 1-(2-bromo-4-isopropylphenyl)-6-methyl-4-(1-
morpholino)-7-azaindole, mp 100 °C. 1H NMR (300 MHz,
CDCl3) δ 7.58 (1H, d, J ) 1.8 Hz), 7.41 (1H, d, J ) 8.1 Hz),
7.29 (1H, dd, J ) 1.8, 8.1 Hz), 7.12 (1H, d, J ) 3.7 Hz), 6.54
Su p p or tin g In for m a tion Ava ila ble: X-ray crystal data,
descriptions of data collection, treatment, solution, and refine-
ment, and tables of fractional coordinates and isotropic and
anisotropic thermal parameters for compounds 2, 3, 9, 25, and
29. This information is available free of charge via the Internet
at http://pubs.acs.org.
Refer en ces
(1) Arvanitis, A.; Gilligan, P. J .; Chorvat, R. J .; Cheeseman, R. S.;
Christos, T. E.; Bajthavatchalam, R.; Beck, J . P.; Cocuzza, A.
J .; Hobbs, F. W.; Trainor, G. L.; Wilde, R. G.; Arnold, C.;
Chidester, D.; Curry, M.; He, L.; Hollis, A.; Klaczkiewicz, J .;
Krinitsky, P. J .; Rescinito, J . P.; Scholfield, E.; Culp, S.; DeSouza,
E. B.; Fitzgerald, L.; Grigoriadis, D.; Tam, W. S.; Wong, N.;
Huang, S.; Shen, H. Non-Peptide Corticotropin-Releasing Hor-
mone Antagonists: Syntheses and Structure-Activity Relation-
ships of 2-Anilinopyrimidines and -triazines. J . Med. Chem.
1999, 42, 805-818.
(2) Pauletti, G. M.; Gangwar, S.; Siahaan, T. J .; Aube, J .; Borchardt,
R. T. Improvement of Oral Peptide Bioavailability: Peptidomi-
metics and Prodrug Strategies. Adv. Drug Delivery Rev. 1997,
27 (2, 3), 235-256.

832 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 5
Hodge et al.
(3) Begley, D. J . The Blood-Brain Barrier: Principles for Targeting
Peptides and Drugs to the Central Nervous System. J . Pharm.
Pharmacol. 1996, 48 (2), 136-146.
III. A Novel Synthetic Route from 2-Aminopyrroles. J . Chem.
Soc., Perkins Trans. 1 1975, 19, 1910-1913.
(16) Pokorny, D.; Paudler, W. W. Naphthyridine Chemistry. XIV. The
Meisenheimer Reaction of the I, X.-Naphthyridines.
(4) Strader, C. D.; Fong, T. M.; Tota, M. R.; Underwood, D. J .; Dixon,
R. A. F. Structure and Function of G-protein-coupled Receptors.
Annu. Rev. Biochem. 1994, 63, 101-132.
(17) Hodge, C. N.; Aldrich, P.; Fernandez, C.; Chorvat, R.; Cheese-
man, R.; Christos, T.; Arvanitis, A.; Scholfield, E.; Krenitsky,
P.; et al. Book of Abstracts; 212th ACS National Meeting,
Orlando, FL, Aug 25-29, 1996; American Chemical Society:
Washington, DC; MEDI-094.
(5) Rivier, J .; Rivier, C.; Vale, W. Synthetic Competitive Antagonists
of Corticotropin-Releasing Factor: Effect On ACTH Secretion
In the Rat. Science 1984, 224 (4651), 889-891.
(6) Sugg, E. E. Nonpeptide Agonists for Peptide Receptors. Lessons
from Ligands. Annu. Rep. Med. Chem. 1997, 32, 277-283.
(7) Babine, R. E.; Bender, S. L. Molecular Recognition of Protein-
Ligand Complexes: Applications to Drug Design. Chem. Rev.
1997, 97 (5), 1359-1472.
(8) Chorvat, R. J .; Bakthavatchalam, R.; Beck, J . P.; Gilligan, P.
J .; Wilde, R. G.; Cocuzza, A.; Hobbs, F. W.; Cheeseman, R. S.;
Curry, M.; Rescinito, J . P.; Krenitsky, P.; Chidester, D.; Yarem,
J .; Klackewicz, J . D.; Hodge, C. N.; Aldrich, P. E.; Wasserman,
Z. R.; Fernandez, C. H.; Zaczek, R.; Fitzgerald, L.; Huang, S.-
M.; Shen, H. L.; Wong, Y. N.; Chien, B. M.; Quon, C. Y.;
Arvanitis, A. Synthesis, Corticotropin-Releasing Factor Receptor
Binding Affinity, and Pharmacokinetic Properties of Triazolo-,
Imidazo-, and Pyrrolopyrimidines and -pyridines. J . Med. Chem.
1999, 42, 833-848.
(9) Dewar, M. J . S.; Thiel, W. J . Ground States of Molecules. 38.
The MNDO Methodol. Approximations and Parameters. J . Am.
Chem Soc. 1997, 99, 4899-4907.
(10) Dewar, M. J . S.; Zoebisch, E. F.; Healy, J . J . P.; Stewart, J .
Development and Use of Quantum Mechanical Models. 76.
AM1: a New General Purpose Quantum Mechanical Molecular
Model. J . Am. Chem Soc. 1985, 107, 3902-3909.
(18) Aldrich, P. E.; Arvanitis, A. G.; Cheeseman, R. S.; Chorvat, R.
J .; Christos, T. E.; Gilligan, P. J .; Grigoriadis, D. E.; Hodge, C.
N.; Krenitsky, P. J .; Scholfield, E. L.; Tam, S. W.; Wasserman,
Z. International Publication Number WO Patent 9510506.
(19) Schulz, D. W.; Mansbach, R. S.; Sprouse, J .; et al. CP-154,526:
A potent and selective nonpeptide antagonist of corticotrophin
releasing factor receptors. Proc. Natl. Acad. Sci. U.S.A. 1996,
93, 10477-10482. Chen, Y. L. Pyrrolopyrimidines as CRF
antagonists. WO 9413676-A1, 1994. Chen, Y. L. Pyrrolopyrim-
idines as CRF antagonists. WO 9413677-A1, 1994.
(20) Oki, M. Applications of Dynamic NMR Spectroscopy to Organic
Chemistry, Methods Stereochem. Anal. 1985, 4. Values of ∆ν
obtained at the temperature closest to Tc were used in calculat-
ing ∆G, although ∆ν did not vary significantly over the temper-
atures studied.
(21) Finocchiaro, P.; Gust, D.; Mislow, K. Conformational Dynamics
of Alkoxydiarylboranes. J . Am. Chem. Soc. 1973, 95, 7029-7036.
(22) Cramer, R. D., III; Patterson, D. E.; Bunce, J . D. Recent advances
in comparative molecular field analysis (CoMFA). J . Am. Chem.
Soc. 1988, 110 (18), 5959-67.
(23) Chen, C.; Dagnino, R.; DeSouza, E. B.; Grigoriadis, D. E.; Huang,
C. Q.; Kim, K.; Liu, Z.; Moran, T.; Webb, T. R.; Whitten, J . P.;
Xie, Y. F.; McCarthy, J . R. Design and Synthesis of a Series of
Non-Peptide High-Affinity Human Corticotropin-Releasing Fac-
tor Receptor Antagonists. J . Med. Chem. 1996, 39, 4358-4360.
(24) Lam, P. Y.-S.; Ru, Y.; J adhav, P. K.; Aldrich, P. E.; DeLucca, G.
V.; Eyermann, C. J .; Chang, C. H.; Emmett, G.; Holler, E. R.;
Daneker, W. F.; Li, L.; Confalone, P. N.; McHugh, R. J .; Han,
Q.; Li, R.; Markwalder, J . A.; Seitz, S. P.; Sharpe, T. R.; Bacheler,
L. T.; Rayner, M. R.; Klabe, R. M.; Shum, L.; Winslow, D. L.;
Kornhauser, D. M.; J ackson, D. A.; Erickson-Viitanen, S.; Hodge,
C. N. Cyclic HIV Protease Inhibitors: Synthesis, Conformational
Analysis, P2/P2/Structure-Activity Relationships, and Molec-
ular Recognition of Cyclic Ureas. J . Med. Chem. 1996, 39, 3514-
3525.
(11) Kitson, D. H.; Avbelj, F.; Moult, J .; Nguyen, D. T.; Mertz, J . E.;
Hadzi, D.; Hagler, A. T. Theoretical Study of the Regioselectivity
of Successive 1,3-Butadiene Diels-Alder Cycloadditions to C60
Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 8920-8924. The Biosym
suite of programs is available from MSI, San Diego, CA.
(12) Yakhontov, L. N.; Krasnokutskaya, D. M.; Akalaev, A. N.;
Palant, I. N.; Vainshtein, Yu. I. Khim. Geterotsikl. Soedin. 1971,
7 (6), 789-794.
(13) Willette, R. E. Monoazindoles: The Pyrrolopyridines. Adv.
Heterocycl. Chem. 1968, 9, 27-105.
(14) Yakhontov, L. N.; Liberman, S. S.; Krasnokutskaya, D. M.;
Uritskaya, M. Ya.; Azimov, V. A.; Sokolova, M. S.; Gerchikov,
L. N. Azaindole Derivatives. XLVII. Synthesis and Pharmaco-
logical Study of 3-Aminoalkyl Derivatives of Azaindoles. Khim.-
Farm. Zh. 1974, 8 (11), 5-9.
(25) Gewald, K. Z. Chem. 1961, 1, 349.
(15) Brodrick, A.; Wibberley, D. G. 1H-Pyrrolo-[2,3-b]-pyridines. Part
J M980223O