Chiral pyrrolo[2,3-d]pyrimidine and pyrimido[4,5-b]indole derivatives: Structure-activity relationships of potent, highly stereoselective A1- adenosine receptor antagonists

Article abstract of DOI:10.1021/jm960011w

A series of 33 novel, mostly chiral pyrrolo[2,3-d]pyrimidine and pyrimido[4,5-b]indole derivatives has been synthesized and investigated in radioligand binding assays at the high-affinity adenosine receptor (AR) subtypes A₁ and A(2a). The compounds can be envisaged as adenine and hypoxanthine analogs lacking the nitrogen in the 7-position (7-deazaadenines and 7-deazahypoxanthines). 7-Deazaadenines were much more potent than 7- deazahypoxanthines at AR with A₁AR affinities in the low-nanomolar range, extraordinarily high selectivity for the rat brain A₁AR versus the A(2a)AR (several thousandfold), and high stereoselectivity (up to 96-fold). Pyrimido[4,5-b]indoles were more potent A₁AR antagonists compared to pyrrolo[2,3-d]pyrimidines. Compound 34a (APEPI) is one of the most potent and most selective non-xanthine A₁AR antagonists known to date (K(i) = 2.8 nM, >2000-fold A₁-selective). A new class of very potent A₁AR antagonists has been identified, namely, 2-phenyl-7-deazaadenines bearing a substituent at the exocyclic amino group (N⁴-substituted pyrrolo[2,3-d]pyrimidines). (R)- N(1-Phenylethyl)-4-amino-5,6-dimethyl-2-phenyl-7H-pyrrolo[2,3-d]pyrimidine (DPEAP, 17a) showed a K(i) value of 6.7 nM at A₁AR and >4000-fold A₁ selectivity. Different binding modes are postulated for the N⁴-substituted 4-aminopyrrolo[2,3-d]pyrimidines (e.g., 17a) and the 7-substituted derivatives (e.g., 1a), based on a comparison of steric, electronic, and hydrophobic properties of the two classes of compounds. Water solubility and lipophilicity have been determined for selected compounds. 4-Amino-5,6- dimethyl-2-(3-chlorophenyl)-7H-pyrrolo[2,3-d]pyrimidine (4a) showed the highest water solubility/A₁AR affinity ratio of 368 in the present series, over 2000-fold A₁ selectivity, and 64-fold stereoselectivity (R > S). Therefore, 4a should be an interesting compound for in vivo evaluation.

Full text of DOI:10.1021/jm960011w

2482
J . Med. Chem. 1996, 39, 2482-2491
Ch ir a l P yr r olo[2,3-d ]p yr im id in e a n d P yr im id o[4,5-b]in d ole Der iva tives:
Str u ctu r e-Activity Rela tion sh ip s of P oten t, High ly Ster eoselective
A1-Ad en osin e Recep tor An ta gon ists†,‡
Christa E. Mu¨ller,*,∇ Uli Geis,∇ Bettina Grahner,∇ Wolfgang Lanzner,§ and Kurt Eger§
J ulius-Maximilians-Universita¨t Wu¨rzburg, Institut fu¨r Pharmazie und Lebensmittelchemie, Pharmazeutische Chemie,
Am Hubland, D-97074 Wu¨rzburg, and Universita¨t Leipzig, Institut fu¨r Pharmazie, Pharmazeutische Chemie,
Bru¨derstrasse 34, D-04103 Leipzig, Germany
Received J anuary 3, 1996X
A series of 33 novel, mostly chiral pyrrolo[2,3-d]pyrimidine and pyrimido[4,5-b]indole derivatives
has been synthesized and investigated in radioligand binding assays at the high-affinity
adenosine receptor (AR) subtypes A1 and A2a. The compounds can be envisaged as adenine
and hypoxanthine analogs lacking the nitrogen in the 7-position (7-deazaadenines and
7-deazahypoxanthines). 7-Deazaadenines were much more potent than 7-deazahypoxanthines
at AR with A1AR affinities in the low-nanomolar range, extraordinarily high selectivity for
the rat brain A1AR versus the A2aAR (several thousandfold), and high stereoselectivity (up to
96-fold). Pyrimido[4,5-b]indoles were more potent A1AR antagonists compared to pyrrolo[2,3-
d]pyrimidines. Compound 34a (APEPI) is one of the most potent and most selective non-
xanthine A1AR antagonists known to date (Ki ) 2.8 nM, >2000-fold A1-selective). A new class
of very potent A1AR antagonists has been identified, namely, 2-phenyl-7-deazaadenines bearing
a substituent at the exocyclic amino group (N4-substituted pyrrolo[2,3-d]pyrimidines). (R)-N-
(1-Phenylethyl)-4-amino-5,6-dimethyl-2-phenyl-7H-pyrrolo[2,3-d]pyrimidine (DPEAP, 17a) showed
a Ki value of 6.7 nM at A1AR and >4000-fold A1 selectivity. Different binding modes are
postulated for the N4-substituted 4-aminopyrrolo[2,3-d]pyrimidines (e.g., 17a ) and the 7-sub-
stituted derivatives (e.g., 1a ), based on a comparison of steric, electronic, and hydrophobic
properties of the two classes of compounds. Water solubility and lipophilicity have been
determined for selected compounds. 4-Amino-5,6-dimethyl-2-(3-chlorophenyl)-7H-pyrrolo[2,3-
d]pyrimidine (4a ) showed the highest water solubility/A1AR affinity ratio of 368 in the present
series, over 2000-fold A1 selectivity, and 64-fold stereoselectivity (R > S). Therefore, 4a should
be an interesting compound for in vivo evaluation.
In tr od u ction
(AR) activation can lead to an inhibition of adenylate
cyclase activity, while A2a- and A2bAR activation causes
a stimulation of adenylate cyclase. Other second-
messenger systems have been described for A1AR,
including coupling to calcium and potassium channels.6
A3ARs have been reported to be coupled to phosphoi-
nositide turnover in some systems.5
All known AR agonists are derivatives of the physi-
ological agonist adenosine. Adenine derivatives and
analogs lacking the ribose moiety, however, have been
shown to act as antagonists at AR.7 We have been
engaged in the search for and development of novel
subtype-selective AR antagonists, especially for the
high-affinity A1- and A2aAR subtypes.8 Such AR an-
tagonists are needed as pharmacological tools and are
of considerable interest as drugs. A1-antagonists are
currently developed for the treatment of cognitive
diseases, renal failure, and cardiac arrhythmias, while
A2a-antagonists may be beneficial for patients suffering
from Morbus Parkinson.9
During our search for novel AR antagonists, we had
found that 7-deazaadenines and 7-deazahypoxanthines
with pyrrolo[2,3-d]pyrimidine or pyrimido[4,5-b]indole
structure (see Chart 1) were potent and selective ad-
enosine receptor antagonists.10,11 4-Aminopyrrolo[2,3-
d]pyrimidines bearing a 2-phenyl ring had turned out
to be particularly active at A1AR and very selective for
that receptor subtype.11 The most potent compound was
1a (ADPEP) with a Ki value at A1AR of rat brain of 4.7
nM and ca. 789-fold selectivity for that receptor subtype
Adenosine receptors belong to the superfamily of
purine receptors which are currently subdivided into P1
(adenosine) and P2 (ATP, ADP, and other nucleotide)
receptors.1 There is increasing evidence, however, that
it is not the purine part which is essential for activity
of natural and synthetic agonists at these receptors2 but
rather the ribose moiety. Therefore we have suggested
to name this receptor family “riboside receptors” 3 in
accordance with the originally proposed nomenclature
for adenosine receptors by Londos and co-workers.4 The
riboside receptors could be subdivided into two sub-
classes, one activated by nucleosides, the other subclass
activated by nucleotides. Four receptor subtypes for the
nucleoside adenosine have been cloned so far from
various species including humans.5,6 Two receptor sub-
types (A1 and A2a) exhibit high affinity for adenosine
(generally in the low-nanomolar range), while the two
other known subtypes (A2b and A3) are low-affinity
receptors, typically showing affinity for adenosine in the
low-micromolar range.6 A1- and A3-adenosine receptor
† Dedicated to Prof. Dr. G. Seitz, Marburg, on the occasion of his
60th birthday.
‡ Preliminary results were presented in Philadelphia at the 5th
International Symposium on Adenosine and Adenine Nucleotides,
1994; abstract published in Drug Dev. Res. 1994, 31, 301.
* Corresponding author. Phone: +49-(0)931-888-5440. E-mail:
mueller@pharmazie.uni-wuerzburg.de.
∇ Universita¨t Wu¨rzburg.
§ Universita¨t Leipzig.
X Abstract published in Advance ACS Abstracts, May 15, 1996.
S0022-2623(96)00011-8 CCC: $12.00 © 1996 American Chemical Society

Pyrrolopyrimidines and Pyrimidoindoles as AR Antagonists
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13 2483
Sch em e 1. Synthesis of Pyrrolo[2,3-d]pyrimidinesa
Ch a r t 1. Basic Structure of Investigated
7-Deazaadenines and 7-Deazahypoxanthines
a
For R, see Table 1 or 2. (a) Benzonitrile derivative (see Table
1), sodium methylate; (b) HCOOH; (c) POCl3; (d) (R)- or (S)-1-
phenylethylamine; (e) benzoyl chloride derivative (see Table 1);
(f) P2O5, dimethylcyclohexylamine, H2O; (g) polyphosphoric acid.
compared to the high-affinity A2AR of rat striatum.11
The compound exhibited stereoselectivity, the R-con-
figurated stereoisomer 1a being more potent than its
S-enantiomer.
Pyrrolo[2,3-d]pyrimidin-4-one derivatives 13 and 14
were prepared by condensation of 2 with benzoyl
chloride or p-chlorobenzoyl chloride, respectively. N7,N4-
Disubstituted pyrrolo[2,3-d]pyrimidines 10a ,b were syn-
thesized from 2a ,b in three steps: Ring closure of 2 with
formic acid yielded 9 which was treated with POCl3 to
afford the corresponding 4-chloro derivative. Substitu-
tion of the chlorine atom by (R)- or (S)-1-phenyleth-
ylamine could be achieved by heating the chloro deriva-
tive with a 10-fold excess of the amine for 96 h to yield
10. 7-Unsubstituted N4-substituted pyrrolo[2,3-d]pyri-
midine 17, an isomer of 1a , was prepared from com-
pound 13. Dealkylation in the 7-position by heating
with polyphosphoric acid13 to 15 and subsequent reac-
tion with POCl3 yielded 16. Substitution with a 10-fold
excess of (R)- or (S)-1-phenylethylamine afforded 17,
requiring a reaction time of several days.
N-Substituted 2-aminotetrahydroindole-3-carboni-
triles 19-21 (Scheme 2) were obtained in a one-pot
reaction developed by Eger et al.14 Hydroxycyclohex-
anone was condensed with an arylalkylamine in the
presence of p-toluenesulfonic acid. Subsequent base-
catalyzed cyclization with malononitrile yielded the
desired products. The arylalkylamines used were chiral,
and both enantiomers of each compound (19-21) were
prepared starting from the R- or S-configurated amines,
respectively. Reaction of 19-21 with benzonitrile and
sodium methylate in 2-propanol yielded the 2-phe-
nyltetrahydropyrimido[4,5-b]indolamine derivatives 22-
24 in analogy to described procedures.11,15 The 4-oxo
analog 29 of the 4-aminotetrahydroindole 22 was pre-
pared by condensation of 19 with benzoyl chloride to
yield 28 and subsequent cyclization with phosphorus
pentoxide in the presence of a mixture of dimethylcy-
clohexylamine and water at 190 °C.16 Despite such
drastic reaction conditions, the phenylethyl residue at
N9 was stable, while treatment of similar compounds
In the present study a new series of 7-deazaadenine
and 7-deazahypoxanthine derivatives has been synthe-
sized and investigated in A1- and A2AR binding assays.
Our goal was to get more insight into structure-activity
relationships of this potent class of non-xanthine AR
antagonists and eventually to identify more potent and
more selective compounds. In an earlier study most
investigated compounds had been either achiral or
racemic. We have now synthesized the enantiomers of
a new series of chiral compounds to investigate the
stereochemical requirements of this class of compounds
for binding to AR. We were particularly interested in
obtaining information about the binding mode of 7-dea-
zaadenines to AR with respect to adenosine derivatives
(AR agonists). Since potent AR antagonists often show
high lipophilicity, low water solubility, and hence low
bioavailibility, we also investigated physicochemical
properties, especially water solubility of the compounds.
These data may be useful for selecting compounds for
in vivo studies.
Ch em istr y
Most of the products prepared were chiral, and in
most cases both enantiomers were synthesized. All
compounds with R-configuration were designated a and
those with S-configuration b after the compound num-
ber.
N4- Or 7-(1-phenylethyl)-substituted 5,6-dimethylpyr-
rolo[2,3-d]pyrimidines 3-10 and 13-17 (Scheme 1)
were prepared starting from pyrrole derivative 2 (2a ,
R-enantiomer, or 2b, S-enantiomer). Compound 2 was
obtained in analogy to compounds 19-21 (see below;
Scheme 2) in a one-pot reaction of 3-hydroxy-2-bu-
tanone, 1-phenylethylamine, and malonodinitrile.11-14

2484 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13
Mu¨ller et al.
Sch em e 2. Synthesis of Pyrrolo[4,5-b]indoles and
Spiro[cyclopentane-pyrrolo[2,3-d]pyrimidines]a
tives. Introduction of a 2-phenyl group into 32 resulted
in a tremendous increase in A1AR affinity (compounds
34a ,b).
The effects of substituents on the 2-phenyl ring of 1a
and its enantiomer 1b were investigated. Substitution
of the 2-phenyl group with chloro or methoxy functions
generally reduced AR affinity. Meta- and ortho-chloro
substitutions were better tolerated (ca. 2-fold decrease
in A1AR affinity) than para-chloro substitution, which
led to a 10-fold reduction of A1 affinity. Meta-chloro
substitution, however, was favorable for high A1 selec-
tivity which was increased from 789-fold in 1a to over
2000-fold in compound 4a .
A regioisomer of 1a in which the (R)-1-phenylethyl
residue is attached to the exocyclic amino function
rather than to the pyrrole nitrogen was prepared. The
compound 17a (DPEAP) exhibited high affinity to A1-
AR (Ki ) 6.7 nM) and showed extraordinarily high A1
selectivity (>4000-fold) and stereoselectivity (17a /17b:
96-fold; R > S).
A comparison of the steric and electronic properties
of regioisomers 1a and 17a revealed excellent confor-
mity, if the N4 nitrogen atom of compound 17a was
aligned with the N7 of compound 1a and N1 of one
compound was aligned with N3 of the other one (see
Figure 1, N4/N7-binding model). Both compounds ex-
hibited similarly high A1AR affinity, A1 selectivity, and
stereoselectivity. We therefore postulate different bind-
ing modes for 7-substituted 4-aminopyrrolo[2,3-d]pyri-
midines, such as 1a , and N4-substituted derivatives,
such as 17a . The postulated binding mode was tested
by a molecular modeling study. At first we searched
for low-energy conformations of compounds 1a and 17a .
Then, the structures were aligned according to the
proposed model, with N7 of 1a /N4 of 17a and N1 of 1a /
N3 of 17a superimposed. Figure 2, top, shows a
comparison of the van der Waals volumes of 1a and 17a
which are very similar for both compounds. In Figure
2, bottom, the molecular electrostatic potentials (MEP),
which are also considered to be important for molecular
recognition, of 1a and 17a are compared. Again, both
compounds show a striking similarity when superim-
posed according to the N7/N4 binding model. In addi-
tion, lipophilic properties of 1a and 17a were calculated
and compared. The two compounds are regioisomers,
and therefore, partition coefficients (P values) should
be similar. In fact, calculation of log P values resulted
in very similar numbers for compounds 1a and 14a (see
Table 3).
a
(a) RNH2, CH2(CN)2; (b) benzonitrile, sodium methylate,
2-propanol; (c) Pd/C, 1-methylnaphthalene; (d) NaIO4; (e) benzoyl
chloride; (f) P2O5, dimethylcyclohexylamine, H2O.
with polyphosphoric acid at 70 °C had led to dealkyla-
tion13 (see above, compounds 13/15). Dehydration of
the tetrahydroindoles 23, 24, and 29 to obtain the
desired aromatic indole derivatives 25, 26, and 30 was
achieved catalytically with Pd/C at high temperatures.
1-Methylnaphthalene proved to be a suitable solvent for
this reaction due to its high boiling point (244 °C).
Solvents with lower boiling point such as decaline or
mesitylene gave unsatisfactory results.15 Reactions of
tetrahydroindoles 22 and 29 with sodium periodate led
to the corresponding spiro[cyclopentane-pyrrolo[2,3-d]-
pyrimidine] derivatives in analogy to described reac-
tions.15
Yields and analytical data, including optical rotation
of enantiomers, are given in Table 1. 1H and 13C NMR
spectral data were consistent with the proposed struc-
tures. Selected NMR data are given in the Experimen-
tal Section. Further H and 13C NMR data are available
1
as Supporting Information.
Biologica l Eva lu a tion
The compounds were tested in radioligand binding
assays for affinity at A1- and A2a-adenosine receptors
in rat cortical membrane and rat striatal membrane
preparations, respectively. [3H]-N6-(R)-(Phenylisopro-
pyl)adenosine ((R)-PIA) was used as A1 ligand and [3H]-
5′-(N-ethylcarbamoyl)adenosine (NECA) as A2a ligand
in the presence of 50 nM N6-cyclopentyladenosine, the
latter to block A1-receptors present in the striatal tissue.
Disubstitution with 1-phenylethyl at both nitrogen
atoms, N4 and N7, in a compound lacking the 2-phenyl
group resulted in compounds with low AR affinity
(10a ,b) and loss of stereoselectivity. A similar effect had
been observed in a series of analogous N6,9-disubsti-
tuted adenine derivatives, another class of adenosine
receptor antagonists.17 At present it is not clear whether
the lacking of the 2-phenyl group, the disubstitution,
or both is responsible for the low activity of compounds
10a ,b.
Resu lts a n d Discu ssion
Str u ctu r e-Activity Rela tion sh ip s. The 2-phenyl
substituent of 7-deaza-2-phenyladenines is known to
contribute to a large extent to the high potency of the
compounds at A1AR as had been shown in a series of
pyrrolo[2,3-d]pyrimidines.11 This result could now be
confirmed in a series of pyrimido[4,5-b]indole deriva-
7-Deazaadenines with pyrimido[4,5-b]indole structure
were more potent than those with 5,6-dimethylpyrrolo-
[2,3-d]pyrimidine structure, such as 1a . The most
potent compound of the whole series of 7-deazapurines
was an analog of 1a , 34a (APEPI), with a Ki value at
A1AR of 2.6 nM and >2000-fold selectivity for A1- versus

Pyrrolopyrimidines and Pyrimidoindoles as AR Antagonists
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13 2485
Ta ble 1. Reagents, Yields, and Analytical Data of New Compounds Synthesized
b
[R]20
D
EI-MS
configuration
R S
amount
[g (mmol)]
yield
[g (%)]
(70 eV)
compda
reagent
formula
anal.
m/ z [M+] mp (°C)
3a ,b
4a ,b
5a ,b
6a ,b
7a
2-chlorobenzonitrile
3-chlorobenzonitrile
4-chlorobenzonitrile
3,4-dichlorobenzonitrile
3-methoxybenzonitrile
4-methoxybenzonitrile
2a or 2b
1.4 (10)
1.4 (10)
1.4 (10)
1.72 (10)
1.3 (10)
1.3 (10)
9.57 (40)
5.3 (20)
4.5 (32)
5.6 (32)
2.75 (8)
3.0 (8)
2.0 (53)
1.9 (50)
2.0 (53)
2.4 (58)
2.8 (75)
2.5 (67)
8.5 (79)
4.4 (59)
5.3 (51)
6.0 (53)
1.4 (51)
1.3 (43)
4.3 (90)
2.8 (55)
0.8 (39)
ndc
C22H21ClN4
C, H, N
C, H, N
C, H, N
152
168
150
152
181
156
245
118
215
215
254
286
312-315
177
142
(oil)
128
+24
+3
-23
-3
C
C
C
C
C
C
C
C
C
C
C
C
C
C
22H21ClN4
22H21ClN4
22H20Cl2N4 C, H, N
23H24N4O
23H24N4O
16H17N3O
24H26N4
22H21N3O
22H20N3O
22H21N3O
-6
+6
+2
-2
C, H, N
C, H, N
C, H, N
C, H, N
C, H, N
C, H, N
-21
-10
+56
-110
+169
+163
+82
+79
8a
9a ,b
10a ,b
11a ,b
12a ,b
13a ,b
14a ,b
15
-54
9a or 9b
+108
-164
-161
-84
benzoyl chloride
4-chlorobenzoyl chloride
11a or 11b
12a or 12b
13a or 13b
C, H, N 343.2
22H20ClN3O C, H, N 377.1
-80
6.87 (20)
4.8 (20)
14H13N3O
14H12N3Cl
22H22N4
C, H, N
C, H, N
C, H, N
nd
16
15
17a ,b
19a ,b
20a ,b
(R)- or (S)-1-phenylethylamine 1.55 (6)
(R)- or (S)-1-phenylethylamine 6.06 (20)
(R)- or (S)-(1-(4-methylphenyl) 6.75 (20)
ethyl)amine
+18
nd
+116
-20
nd
-112
9.6 (71)
C
18H21N3
C, H, N
21a ,b
(R)- or (S)-1-methyl-2-
phenylethylamine
20a or 20b
21a or 21b
23a or 23b
24a or 24b
22a
19a or 19b
28a or 28b
29a or 29b
29a
1.35 (10)
nd
nd
(oil)
153
nd
nd
23a ,b
24a ,b
25a ,b
26a ,b
27a
28a ,b
29a ,b
30a ,b
31a
1.65 (6)
1.65 (6)
0.95 (2.5) 0.68 (72)
0.95 (2.5) 0.67 (71)
1.2 (52)
1.5 (65)
C
C
C
C
C
C
C
C
C
25H26N4
25H26N4
25H22N4
25H22N4
24H24N4O
24H23N3O
24H23N3O
24H19N3O
24H23N3O2
C, H, N
C, H, N 382.3
C, H, N
C, H, N
C, H, N 384
C, H, N
C, H, N
C, H, N 365.2
C, H, N 385.3
+19
-19
158-159 -200
150
151
+198
-106
+112
+103
-113
2.2 (6)
0.7 (30)
8.5 (77)
1.3 (44)
0.72 (80)
0.9 (38)
200-201 +4.0
7.95 (30)
2.95 (8)
0.9 (4.2)
2.2 (6)
234
275
257
+164
+14
+114
-160
-16
-110
254-255 +160 (CHCl3)
a
In cases where separate enantiomers were prepared, the R-enantiomer was designated a and the S-enantiomer b after the compound
number. Determined in DMSO unless otherwise noted. c nd ) not determined.
b
pyrrole nitrogen of 7-deazaadenines binds to the same
receptor region as the N6-substituent of adenosine
derivatives.
The exchange of the amino group in 7-deazaadenines
for an oxo function leads to 7-deazahypoxanthines. The
loss of the amino group resulted in a significant decrease
in AR affinity of the compounds investigated (cp. 1a ,b/
13a ,b; 5a ,b/14a ,b; 34a ,b/30a ,b; 27a /31a ). The oxo
analog 13a of 1a , for example, showed a more than 600-
F igu r e 1. Different binding modes for N7-substituted (1a )
and N4-substituted (17a ) pyrrolopyrimidine derivatives: the
N7/N4-binding model.
fold lower affinity than the amino compound 1a . A1AR
selectivity and stereoselectivity, however, was retained.
Stereoselectivity was R > S as observed with the amino
analogs (cp. 13a ,b/1a ,b). Para-chloro substitution of the
2-phenyl ring reduced A1 affinity as in the amino series
(compounds 5a ,b/14a ,b).
A2aAR. The compound exhibited marked stereoselec-
tivity, the R-enantiomer being 25-fold more potent than
the S-enantiomer at A1AR.
Pyrimido[4,5-b]indoles, however, were less active than
the corresponding 5,6-dimethylpyrrolo[2,3-d]pyri-
midines in the 7-deazahypoxanthine series (cp. 13a /
30a ), indicating different binding modes for 7-deazaad-
enines and 7-deazahypoxanthines. Also, a 1-phenyl-
ethyl substituent on the pyrrole nitrogen appeared to
be less favorable for 7-deazahypoxanthines compared
to a phenyl substituent (cp. 37/36), in contrast to results
in the 7-deazaadenine series. In 7-deazaadenines the
introduction of a 2-phenyl ring had led to a tremendous
increase in A1AR affinity.11 2-Phenyl-substituted 7-dea-
zahypoxanthine derivatives had not been investigated
so far. We have now found that a 2-phenyl substituent
in 7-deazahypoxanthine derivatives had a detrimental
effect on AR affinity of the compounds (13a ,b, 14a ,b,
30a ,b, 31a ). These results again indicate different
binding modes for 7-deazahypoxanthines and 7-dea-
zaadenines, at least for those derivatives that bear a
substituent on the pyrrole nitrogen. Oxo analogs of N4-
As had been observed earlier in the pyrrolo[2,3-d]-
pyrimidine series, an (R)-1-phenylethyl substituent on
the pyrrole nitrogen is superior to a phenyl substituent
in that position in pyrimidoindoles as well (cp. 34a /33).
Para-methyl substitution of the phenylethyl substituent
of 34a and its enantiomer 34b led to an about 5-fold
decrease in A1AR affinity and also decreased A2AR
affinity (compounds 25a ,b). It appears that the adenos-
ine receptors cannot accommodate well an additional
methyl group in the phenylethyl-binding receptor do-
main.
Exchange of the (R)-1-phenylethyl group in the 7-po-
sition of 34a by (R)-phenylisopropyl in analogy to the
potent A1AR agonist (R)-PIA18 led to a dramatic (>200-
fold) decrease in A1AR affinity (cp. 34a /26a ). Stereo-
selectivity of the phenylisopropyl derivatives 26a ,b was
very low (only ca. 2-fold) and reversed (S > R) compared
to the 1-phenylethyl-substituted analogs 34a ,b. There-
fore, it appears very unlikely that the substituent at the

2486 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13
Mu¨ller et al.
role in mediating agonistic activity.3,7 Not only is the
ribose part recognized stereoselectively but also sub-
stituents at the exocyclic amino function (N6) of adenos-
ine. A typical example is the classical standard A1AR
agonist (R)-PIA which binds with ca. 50-fold higher
affinity to A1AR than its diastereomer (S)-PIA.18 Among
antagonists stereoselective binding has also been ob-
served. The degree of stereoselectivity, however, was
generally much lower than that observed with agonists.
(R)-1,3-Dipropyl-8-(phenylisopropyl)xanthine, for ex-
ample, was only 9-fold more potent than its S-enanti-
omer.20 Chiral N6-substituted adenines also showed
only a moderate (<10-fold) degree of stereoselectivity.17
7-Deazaadenine derivative 1a had been the first ex-
ample of an AR antagonist exhibiting high (35-fold)
stereoselectivity (R > S).11 In the present study we have
investigated the separate stereoisomers of a new series
of 7-deazaadenines and 7-deazahypoxanthines. We
found that not only is the 7-substituent of pyrrolo[2,3-
d]pyrimidines recognized stereoselectively but also the
N4-substituent, as in 17a (R > S). The highest degree
of stereoselectivity was exhibited by compounds 4a (64-
fold) and 17a (96-fold).
It has been postulated that ARs contain three binding
domains for their ligands, a central aromatic binding
domain, a hydrophobic binding domain, and a ribose
binding domain21 (for agonists) which can be occupied
by lipophilic (often aromatic) substituents of AR an-
tagonists. The structures of potent 7-deazaadenines are
consistent with the proposed model. One important
feature, particularly for A1AR ligands, appears to be a
hydrogen bond donor (N-H) and a hydrogen bond
acceptor (e.g., dN-, dO) in a certain distance (1a ,
N4-H and N3; 17a , N7-H and N1; see Figure 1).
In contrast to AR antagonists, adenosine derivatives
(AR agonists) probably can bind to A1AR only in one
binding mode, determined by the ribose moiety.
A
comparison of adenosine derivatives and 7-deazaad-
enines (antagonists) would at first sight lead to the
conclusion that adenosines and N4-substituted pyrrolo-
[2,3-d]pyrimidines, such as 17a , bind to AR in the same
binding mode (Figure 1) with the heterocyclic ring
systems and the amino substituents superimposed.
Structure-activity relationships of adenosine deriva-
tives, however, are not consonant with either of the two
binding modes proposed for 7-deazaadenine derivatives
such as 1a and 17a (see Figures 1 and 2) for the
following reasons. (1) 7-Deazaadenosine is nearly inac-
tive at AR,19,22 while 7-deazaadenines, such as 17a , are
very potent. (2) PIA is a potent A1AR agonist exhibiting
a high degree of stereoselectivity (R > S), while (phe-
nylisopropyl)-7-deazadenine 26 shows very low A1 af-
finity and a low degree of stereoselectivity, which is
reversed (S > R). Therefore we conclude that adenosine
derivatives may bind to A1AR in yet another binding
mode, different from those for N7- and N4-substituted
7-deazaadenines.
F igu r e 2. Comparison of (top) van der Waals volumes and
(bottom) molecular electrostatic potentials (MEP) of 1a in cyan
and 17a in magenta. (Bottom) Positive (+4.2 kJ /mol) MEP are
shown in red for 1a , in magenta for 17a ; negative MEP (-4.2
kJ /mol) are shown in blue for 1a , in green for 17a .
substituted pyrrolopyrimidines or pyrimidoindoles with-
out a substituent on the pyrrole nitrogen, e.g., O4-
substituted oxo analogs of 17a , however, have not been
investigated so far. 7-Deazahypoxanthine analogs sub-
stituted at the pyrrole nitrogen are lacking an N-H
function at the appropriate position (such as the N4-H
in 1a or the N7-H in 17a ; see Figure 1) for hydrogen
bonding to the receptors.
Spiro-7-deazahypoxanthines 31a and 35-37 showed
relatively low AR affinity. 7-Deazaadenine analog 27a
of 7-deazahypoxanthine derivative 31a was relatively
potent at A1AR and quite selective but much less potent
than the corresponding 5,6-dimethylpyrrolopyrimidine
1a or the pyrimidoindole 34a .
N6,9-Disubstituted adenine derivatives, in which the
ribose of adenosine is substituted by an alkyl or aryl
function, are moderately potent A1AR antagonists, e.g.,
N6-cyclopentyl-9-methyladenine.7 Structure-activity
relationships indicate a similar binding mode for this
class of compounds as for adenosine derivatives but,
again, a different binding mode from 7-deazaadenines.7,17
Adenosine receptor agonists (adenosine derivatives)
bind to AR in a highly stereoselective manner. Stere-
ochemistry at the ribose moiety is particularly impor-
tant,19 since the ribose is believed to play the essential
Another related class of potent AR antagonists, which

Pyrrolopyrimidines and Pyrimidoindoles as AR Antagonists
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13 2487
has been described recently, is the 8-azaadenines.
Independently from our model,23 Biagi et al. postulated
a similar binding model with two different binding
modes for 2-phenyl-8-azaadenines.24 In contrast to their
hypothesis, however, we propose a third binding mode
for adenosine derivatives different from the two binding
modes postulated for 7-deazaadenines (Figures 1 and
2). The investigation of the third binding mode, that
for adenosines with respect to 7-deazaadenines 1a and
17a , is currently under investigation.
solubility (1.5 µM) and an about 4-fold increase in the
BF index compared to its isomer 1a . With compounds
4a , 17a , and 34a , we have now been successful in
developing 7-deazaadenine derivatives with a BF index
of >100. These compounds should be sufficiently soluble
to exhibit activity in vivo.
Solubilities of the investigated 7-deazaadenines were
in the low-micromolar range, or even below. In the AR
assays, however, we had been able to measure competi-
tion with radioligands at membrane preparations up to
much higher concentrations of test compounds. As
reported in Table 2, concentrations up to 30 µM of most
compounds, including 5a and 17a , could be used without
any apparent problems due to limited solubility. In
order to investigate whether the higher DMSO concen-
tration (2.5% in the binding assays vs 1% for solubility
determination) was the reason for this discrepancy, we
determined solubilities of three selected compounds in
buffer containing 2.5% DMSO. We found that the
solubility was increased by the higher DMSO concen-
tration, as expected. Compounds 5a , 17a , and 34a all
showed an about 3-fold increase in solubility in aqueous
buffer solution containing 2.5% DMSO as compared to
1% DMSO (see Table 3). But this moderate increase in
solubility could still not fully explain our observations.
A further explanation could be the increase in solubility
at higher temperatures, since solubility was determined
at room temperature, while the A1 binding assays were
performed at 37 °C. Another explanation that has been
forwarded was that supersaturated solutions may be
formed.27 Our observations, however, appear to indicate
that the presence of membrane tissue in the assay
mixture leads to the observed increase in solubility of
the test compounds by acting as a kind of solubilizer.
Calculated partition coefficients for selected com-
pounds (Table 3) show that water solubility and log P
values cannot be correlated in the present series.
Chloro-substituted derivatives 4a and 5a show a better
water solubility than 1a despite their higher lipophi-
licity.
In addition to the high-affinity AR subtypes A1 and
A2a, at least two low-affinity AR subtypes exist, desig-
nated A2b and A3. The A2bAR shows a high degree of
homology to the high-affinity A2aAR.6 The relationship
between the A2AR subtypes is much closer than between
A1- and A2AR. Therefore, it is likely that AR antago-
nists selective for A1 versus A2a are also selective versus
the A2bAR subtype. One 7-deazahypoxanthine, 9-phen-
yl-9H-pyrimido[4,5-b]indol-4(3H)-one (HPPI), a com-
pound with reasonable A2aAR affinity,10 had been
investigated as antagonist at rat A2bAR.25 In fact, the
compound showed similar activity at A2aAR (KB ) 2.5
µM) and A2bAR (KB ) 3.4 µM) in functional assays
(antagonism of NECA stimulation of adenylate cyclase
in PC 12 (A2aAR) and NIH 3T3 (A2bAR) cell mem-
branes).25 To our knowledge, 7-deazaadenine and de-
rivatives have not been investigated at the novel A3AR
subtype so far. Two 7-deazahypoxanthines, HPPI and
its 5,6,7,8-tetrahydro derivative, showed less than 20%
inhibition of radioligand binding at a concentration of
10 µM at rat A3AR.36
Solu bility a n d Lip op h ilicity. The application of
in vitro assays for the development of AR ligands often
leads to compounds with high receptor affinity but, at
the same time, high lipophilicity and low water solubil-
ity. Due to their low bioavailability, such compounds
may be inactive in vivo and therefore of no interest for
drug development.26 Bruns and Fergus proposed that
the solubility-over-receptor affinity ratio (the so-called
Bruns-Fergus index or BF index) should be >100 for a
drug to exhibit in vivo activity.27 We used the A1AR
binding protocol developed by Bruns and Fergus27 to
determine solubilities of the potent A1-antagonists 5a ,
17a , and 34a . Solubilities of 1a and racemic 4 had been
determined earlier by the same method.11 It had been
shown that this biochemical method of solubility deter-
mination corresponds well with analytical methods,
such as UV-photometry, and it has the advantage of
being easily applicable to very insoluble compounds.11,27
Stock solutions of compounds in DMSO were prepared
and diluted into buffer to obtain a saturated solution
in buffer containing 1% DMSO. The earlier somewhat
surprising finding that a chloro substituent at the
2-phenyl ring could enhance solubility of 2-phenyl-7-
deazaadenines11 was confirmed in this study. Thus, the
chloro-substituted derivatives 4a and 5a of 1a showed
about 10-fold higher water solubility compared to 1a .
The very potent and highly selective meta-chloro de-
rivative 4a exhibited a solubility/A1 affinity ratio of 368
and was the best compound of the series with respect
to the BF index. Pyrimidoindole analog 34a of pyrrolo-
pyrimidine 1a was about 2-fold more soluble than 1a .
Since it was also more active at A1AR, the BF index was
improved from 55 in 1a to 185 in 34a . N4-Substituted
pyrrolopyrimidine 17a exhibited even higher water
In conclusion, new A1AR antagonists have been
developed which should be promising candidates for in
vivo studies.
Exp er im en ta l Section
Syn th etic P r oced u r es. NMR spectra were performed on
a Bruker WP-80 (1H, 80 MHz; 13C, 20 MHz), a Bruker AC-250
(1H, 250 MHz; 13C, 60 MHz), or a Varian Gemini 300 (1H, 300
MHz; 13C, 75 MHz) spectrometer, respectively. DMSO-d6 was
used as solvent, unless otherwise noted. The chemical shifts
of the remaining protons of the deuterated solvents served as
internal standards: δ (1H, DMSO-d6 ) 2.50, CDCl3 ) 7.24;
13C, DMSO-d6 ) 39.7, CDCl3 ) 77.0). All compounds were
checked for purity by TLC on 0.2 mm aluminum sheets with
silica gel 60 F254 (Merck); as eluent toluene, toluene:methanol
(9:1), toluene:acetone:formic acid (60:39:1), or ethyl acetate,
respectively, was used. Melting points were taken on a Bu¨chi
510 melting point apparatus and are uncorrected. Electron
ionization mass spectra were recorded on a Finnigan MAT
711A spectrometer. Optical rotation was measured with a
Perkin-Elmer 241 spectrometer in a cuvette of 10 cm length,
in a concentration of 1.00% in DMSO, unless otherwise noted.
Elemental analyses, NMR spectra, and mass spectra were
performed by the Institute of Chemistry, University of Tu¨b-
ingen, and the Institute of Organic Chemistry, University of
Leipzig, respectively. Additional NMR data are available as
Supporting Information.
Compounds 2a ,b, 19a ,b, and 22a ,b were obtained as
described.13-15 The synthesis of the final products 32-37 has

2488 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13
Mu¨ller et al.
Ta ble 2. Adenosine Receptor Affinities of Pyrrolo[2,3-d]pyrimidines and Pyrimido[4,5-b]indoles
Ki ( SEM (µM)
A1-receptors
A2-receptors
vs [3H]NECA, A1 selectivity,
vs [3H]-(R)-PIA,
compd
R1
R2
R3
rat brain cortex
rat striatum
A2/A1
4-Aminopyrrolo[2,3-d]pyrimidines (I)
1a (ADPEP) (R)-1-phenylethyl
phenyl
H
0.00472
3.712
789
490
537
1b
3a
3b
4a
4b
5a
5b
6a
(S)-1-phenylethyl
(R)-1-phenylethyl
(S)-1-phenylethyl
(R)-1-phenylethyl
(S)-1-phenylethyl
(R)-1-phenylethyl
(S)-1-phenylethyl
(R)-1-phenylethyl
(S)-1-phenylethyl
(R)-1-phenylethyl
(R)-1-phenylethyl
(R)-1-phenylethyl
(S)-1-phenylethyl
H
phenyl
H
0.1652
812
2-chlorophenyl
2-chlorophenyl
3-chlorophenyl
3-chlorophenyl
4-chlorophenyl
4-chlorophenyl
3,4-dichlorophenyl
3,4-dichlorophenyl
3-methoxyphenyl
4-methoxyphenyl
H
H
H
H
H
H
H
H
H
0.0096 ( 0.0010
0.273 ( 0.028
0.0076 ( 0.0013
0.485 ( 0.015
0.050 ( 0.006
0.878 ( 0.153
0.33 ( 0.07
1.8 ( 0.1
5.2 ( 0.6
>30 (35%)a
16.3 ( 4.9
>30 (20%)a
>30 (23%)a
>30 (3%)a
>30 (8%)a
>30 (2%)a
>10 (41%)a
>10 (23%)a
>30 (13%)a
>30 (11%)a
>110
2145
>62
>600
>34
>91
>17
6b
7a
8a
10a
10b
17a (DPEAP)
H
H
0.083 ( 0.012
0.101 ( 0.004
2.65 ( 0.51
2.53 ( 0.61
>120
>99
(R)-1-phenylethyl
(S)-1-phenylethyl
(R)-1-phenylethyl
(S)-1-phenylethyl
>11
>12
H
phenyl
phenyl
0.0067 ( 0.0060 >30 (23%)a
>4478
>47
17b
H
0.64 ( 0.07
>30 (43%)a
4-Aminopyrimido[4,5-b]indoles (II)
25a
25b
26a
26b
32
(R)-1-(4-methylphenyl)ethyl phenyl
(S)-1-(4-methylphenyl)ethyl phenyl
0.014 ( 0.001
0.297 ( 0.005
0.57 ( 0.17
21.4 ( 4.4
.30 (0%)a
>30 (25%)a
>30 (15%)a
27.7 ( 5.9
8.3 ( 2.0
1529
.101
>53
>94
25
377
2385
>469
(R)-phenylisopropyl
(S)-phenylisopropyl
(rac)-1-phenylethyl
phenyl
phenyl
phenyl
H
phenyl
phenyl
phenyl
0.32 ( 0.29
1.13 ( 0.18
33
0.022 ( 0.002
0.0026 ( 0.0008
0.064 ( 0.017
34a (APEPI) (R)-1-phenylethyl
34b
6.2 ( 1.1
(S)-1-phenylethyl
>30 (15%)a
4-Oxopyrrolo[2,3-d]pyrimidines (III)
13a
13b
14a
14b
(R)-1-phenylethyl
(S)-1-phenylethyl
(R)-1-phenylethyl
(S)-1-phenylethyl
phenyl
2.98 ( 0.8
>30 (10%)a
.30 (0%)a
>30 (12%)a
>30 (5%)a
>10
phenyl
4-chlorophenyl
4-chlorophenyl
>30(12%)a
>30 (17%)a
>30 (24%)a
4-Oxopyrimido[4,5-b]indoles (IV)
phenyl
phenyl
30a
30b
(R)-1-phenylethyl
(S)-1-phenylethyl
>30 (19%)a
>30 (16%)a
>30 (12%)a
>30 (2%)a
Oxospiropyrrolo[2,3-d]pyrimidines (V)
phenyl
31a
35
(R)-1-phenylethyl
H
>30 (21%)a
127 ( 6
>30 (6%)a
346 ( 82
17.2 ( 13.4
88.4 ( 9.4
H
H
H
2.7
36
37
phenyl
(rac)-1-phenylethyl
1.3 ( 0.3
3.4 ( 0.1
13
26
Aminospiropyrrolo[2,3-d]pyrimidines (VI)
phenyl
27a
(R)-1-phenylethyl
0.18 ( 0.02
>30 (32%)a
>167
a
Percent inhibition at the indicated concentration; IC50 could not be determined due to limited solubility.
been described elsewhere.15 In cases of chiral compounds, the
R-configurated stereoisomer was designated a and the S-
configurated stereoisomer b after the compound number.
2-P h en yl-Su b st it u t ed Der iva t ives of 4-Am in o-5,6-d i-
m et h yl-7H -7-(1-p h en ylet h yl)p yr r olo[2,3-d ]p yr im id in es
3a ,b, 4a ,b, 5a ,b, 6a ,b , 7a , a n d 8a . General procedure: A
suspension of 2.4 g (10 mmol) of 2a or 2b,13 respectively, 1.1
g (20 mmol) of sodium methylate, and 10 mmol of the
appropriate benzonitrile derivative (see Table 1) in 20 mL of
2-propanol was refluxed for 8 h. After cooling to about 30 °C,
the mixture was diluted with EtOH to keep sodium methylate
in solution. After cooling, the precipitated crystals were
filtered and recrystallized from EtOH/H2O. Reactants, yields,
and some analytical data are given in Table 1. 3a : 1H NMR
δ 1.94 (d, J ) 7.3 Hz, 3H, CHCH3), 2.08 (s, 3H, CH3), 2.31 (s,
3H, CH3), 6.08 (q, J ) 7.3 Hz, 1H, CHCH3), 6.53 (s, 2H, NH2),
6.53-7.67 (m, 9H, aromatic).
(R)- a n d (S)-5,6-Dim eth yl-7H-7-(1-p h en yleth yl)p yr r olo-
[2,3-d ]p yr im id in -4(3H)-on e (9a ,b). A mixture of 9.57 g
(40mmol) of 2a or 2b, respectively, and 50 mL of formic acid
(85%) was refluxed for 5 h. After cooling, the precipitate was
collected by filtration, washed until the washings showed a
neutral pH value, and recrystallized from EtOH. Yield and
some analytical data are given in Table 1.
(R,R)- a n d (S,S)-N,7-Bis(1-p h en yleth yl)-5,6-d im eth yl-
7H-p yr r olo[2,3-d ]p yr im id in -4(3H)-on e (10a ,b). A mixture
of 5.3 g (20 mmol) of 9a or 9b, respectively, with 30 mL of
POCl3 was boiled for 1 h. Excess reagent was removed by
distillation. For removing residues of POCl3, the obtained oil
was heated with H2O and then extracted with CH2Cl2. The

Pyrrolopyrimidines and Pyrimidoindoles as AR Antagonists
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13 2489
Ta ble 3. Solubility and Solubility/Affinity Ratio of Selected Pyrrolo[2,3-d]pyrimidines and Pyrimido[4,5-b]indoles
ratio,c
solubility/
A1 affinity
calcd
log P
value
solubilitya (µM ( SEM)
compd
R1
R2
phenyl
3-chlorophenyl
4-chlorophenyl
phenyl
R3
1% DMSO
2.5% DMSO
Pyrrolo[2,3-d]pyrimidines (I)
1a
4a
5a
(R)-1-phenylethyl
(R)-1-phenylethyl
(R)-1-phenylethyl
H
H
H
H
0.2611
ndb
nd
7.2 ( 4.4
5.9 ( 3.8
55
368
44
3.8
4.4
4.4
3.5
2.811
2.2 ( 0.6
1.5 ( 0.5
17a
(R)-1-phenylethyl
224
Pyrimido[4,5-b]indoles (II)
0.48 ( 0.11
34a
(R)-1-phenylethyl
phenyl
1.5 ( 0.9
185
4.5
a
b
Determined in 50 mM TRIS-HCl buffer, pH 7.4, containing the indicated amount of DMSO. nd ) not determined. c Solubility in
buffer containing 1% (v/v) DMSO.
organic phase was dried with anhydrous Na2SO4 and evapo-
rated in vacuo. The remaining oil was refluxed for 96 h with
24.2 g (200 mmol) of (R)- or (S)-1-phenylethylamine in 20 mL
of EtOH. After removing excess solvent and reagent in vacuo,
the residue was triturated with a few drops of EtOH. The
crystalline product was filtered and recrystallized from EtOH/
H2O. Yield and some analytical data are given in Table 1.
10a : 1H NMR δ 1.56 (d, J ) 7.0 Hz, 3H, N4-CHCH3), 1.90 (d,
J ) 7.23 Hz, 3H, N7-CHCH3), 2.05 (s, 3H, CH3), 2.34 (s, 3H,
CH3), 5.50 (m, 1H, N4-CHCH3), 6.09 (q, J ) 7.3 Hz, 1H, N7-
CHCH3), 6.27 (d, J ) 8.0 Hz, NH), 7.10-7.47 (m, 10H,
aromatic), 8.00 (s, 1H, H-2).
(aromatic CH), 130.53 (C-6), 132.88, 141.80 (aromatic C),
147.00, 149.00, 159.65 (C-7a, C-2, C-4).
5,6-Dim e t h yl-2-p h e n yl-7H -p yr r olo[2,3-d ]p yr im id in -
4(3H)-on e (15). Compound 13 (6.87 g, 20 mmol) was sus-
pended in 100 g of polyphosphoric acid and heated for 2 h to
70-80 °C. The hot suspension was poured into 100 mL of H2O,
and the solution was brought to a pH value of 10 by the
addition of NH3 solution (25%). The precipitated product was
filtered off, washed with H2O until the washings showed a
neutral pH, and recrystallized from toluene. Yield and some
analytical data are listed in Table 1.
4-Ch lor o-5,6-d im eth yl-2-p h en yl-7H-p yr r olo[2,3-d ]p yr i-
m id in e (16). A mixture of 4.8 g (20 mmol) of 15 and 50
mL of POCl3 was refluxed for 1 h. Excess reagent was
removed in vacuo. The residue was poured on H2O, col-
lected by filtration, washed until neutral reaction, and recrys-
tallized from toluene. Yield and analytical data are given in
Table 1.
(R)- a n d (S)-N-(1-P h en yleth yl)-5,6-d im eth yl-2-p h en yl-
7H-p yr r olo[2,3-d ]p yr im id in -4-a m in e (17a ,b). A solution
of 1.55 g (6 mmol) of 16, 9.7 g (80 mmol) of (R)- or (S)-
phenylethylamine, respectively, and a catalytic amount of
concentrated HCl in 15 mL of EtOH was refluxed for 120 h.
Excess solvent and reagent were removed in vacuo by distil-
lation. The resulting oil was purified by column chromatog-
raphy on silica gel starting with toluene as eluent and
continuing with toluene:methanol (9:1). The product was
recrystallized from EtOH. 17a : 1H NMR δ 1.63 (d, J ) 7.0
Hz, 3H, CHCH3), 2.25 (s, 3H, CH3), 2.40 (s, 3H, CH3), 5.61 (m,
1H, CHCH3), 6.25 (d, J ) 7.3 Hz, 1H, NH), 7.18-8.29 (m, 10H,
aromatic), 11.26 (s, 1H, NH); 13C NMR δ CDCl3 11.23, 11.35
(CH3), 23.38 (CHCH3), 50.34 (CH), 102.97, 104.27 (C-4a, C-5),
126.68, 127.48, 128.50, 128.70 (aromatic CH), 128.83 (C-6),-
129.51 (aromatic CH), 140.10, 145.62 (aromatic C), 151.70,
156.24, 157.36 (C-7a, C-4, C-2).
(R)- a n d (S)-2-Ben za m id o-4,5-d im eth yl-1-(1-p h en yleth -
yl)-1H-p yr r ole-3-ca r bon itr ile (11a ,b), (R)- a n d (S)-2-(4-
Ch lor ob en za m id o)-4,5-d im et h yl-1-(1-p h en ylet h yl)-1H -
p yr r ole-3-ca r bon itr ile (12a ,b), a n d (R)- a n d (S)-4,5,6,7-
Tetr a h yd r o-2-ben za m id o-1H-1-(1-p h en yleth yl)in d ole-3-
ca r bon itr ile (28a ,b). General procedure: To a solution of
30 mmol of 2a , 2b, 19a , or 19b, respectively, in 20 mL of CH2-
Cl2 was added 5 mL of pyridine followed by the addition of 32
mmol of the appropriate benzoyl chloride derivative upon
cooling. After 1 h of stirring the mixture in an ice bath, 10
mL of petroleum ether (bp 40-60 °C) was added to complete
the precipitation of the product. The precipitate was collected
by filtration and recrystallized from EtOH/H2O. Reagents,
yields, and some analytical data are given in Table 1.
(R)- a n d (S)-5,6-Dim eth yl-2-p h en yl-7H-7-(1-p h en yleth -
yl)p yr r olo[2,3-d ]p yr im id in -4(3H)-on e (13a ,b), (R)- a n d
(S)-2-(4-Ch lor op h en yl)-5,6-d im eth yl-7H-7-(1-p h en yleth -
yl)p yr r olo[2,3-d ]p yr im id in -4(3H)-on e (14a ,b), a n d (R)-
a n d (S)-5,6,7,8-Tetr a h yd r o-2-p h en yl-9H-9-(1-p h en yleth -
yl)p yr im id o[4,5-b]in d ol-4(3H )-on e
(29a ,b ).
General
procedure: H2O (4.5 g, 250 mmol) was added dropwise to a
mixture of 22.7 g (160 mmol) of phosphorus pentoxide and 20.2
g (160 mmol) of N,N-dimethylcyclohexylamine with stirring
and ice cooling.16 The viscous mixture was heated (to ca. 220
°C) until a homogenous solution was obtained. After cooling
to 190 °C, 8 mmol of the appropriate 2-benzamido-4,5-
dimethyl-1-(1-phenylethyl)-1H-pyrrole-3-carbonitrile (11a, 11b,
or 12, respectively) or compound 20a or 20b, respectively, was
added with stirring. The temperature was kept at 190-200
°C for 4 h. After allowing to cool to 90 °C, a pH value of 12
was adjusted by the addition of 2 N NaOH (ca. 150 mL), in
order to obtain a separated amine phase. The precipitate was
collected by filtration, washed with H2O and subsequently with
petroleum ether (bp 40-60 °C), and recrystallized from EtOH.
Reagents, yields, and some analytical data are given in Table
1. 13: 1H NMR δ 2.00 (d, J ) 7.2 Hz, 3H, CHCH3), 2.07 (s,
3H, CH3), 2.27 (s, 3H, CH3), 6.10 (q, J ) 7.1 Hz, 1H, CHCH3),
7.20-8.12 (m, 10H, aromatic), 11.95 (s, 1H, NH); 13C NMR δ
9.75, 10.43 (CH3), 19.44 (CHCH3), 51.83 (CH), 105.05, 110.02
(C-4a, C-5), 126.26, 127.04, 127.15, 128.24, 128.46, 128.58
1-Su bstitu ted 4,5,6,7-Tetr a h yd r o-2-a m in oin d ole-3-ca r -
bon itr iles 19a ,b, 20a ,b, a n d 21a ,b. General procedure: A
mixture of 5.65 g (50 mmol) of 2-hydroxycyclohexanone (18),
50 mmol of the appropriate amine13,28 (see Table 1), and a
catalytic amount of p-toluenesulfonic acid in 50 mL of cyclo-
hexane was refluxed using a Dean-Stark trap. After separa-
tion of the calculated amount of H2O, the solution was allowed
to cool to ca. 50 °C. After the addition of 2 mL of piperidine,
a solution of 3.3 g (50 mmol) of malononitrile in hot cyclohex-
ane was added dropwise, to keep the solution slightly boiling.
After cooling, the precipitate was collected by filtration and
recrystallized from EtOH. Reagents, yields, and some analyti-
cal data are given in Table 1. In contrast to the previously
described synthesis of racemic 19,15
which could be obtained
in crystalline form, the enantiomers of 19 as well as the
enantiomers of 21 were obtained as oils. These were used for
the subsequent step without prior purification.

2490 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13
Mu¨ller et al.
9-Su bstitu ted 5,6,7,8-Tetr ah ydr o-2-ph en yl-9H-pyr im ido-
[4,5-b]in d ol-4-a m in es 23a ,b a n d 24a ,b. General proce-
dure: A mixture of 20a , 20b, 21a , or 21b, respectively, sodium
methylate, and benzonitrile in 2-propanol was refluxed for
8-15 h. After cooling the solution to about 30 °C, a small
amount of EtOH and H2O was added to keep the sodium
methylate in solution. The precipitate was collected by filtra-
tion, washed until the washings showed a neutral pH value,
and recrystallized from EtOH. Amounts of reagents, yields,
and some analytical data are given in Table 1.
9-Su b st it u t ed 2-P h en yl-9H -p yr im id o[4,5-b]in d ol-4-
a m in es 25a ,b a n d 26a ,b a n d (R)- a n d (S)-2-P h en yl-9H-9-
(1-p h en ylet h yl)p yr im id o[4,5-b]in d ol-4(3H )-on e (30a ,b ).
General procedure: A mixture of the appropriate tetrahydro-
pyrimido[4,5-b]indole 23a , 23b, 24a , 24b, 29a , or 29b, respec-
tively, and 0.1 g of Pd/C (10%) in 10 mL of 1-methylnaphtha-
lene was refluxed for 4 h (25, 26) or 5 h (30), respectively. The
hot solution was filtered. For the isolation of 25 and 26, the
solvent was removed by distillation in vacuo. Compounds 25
were purified by column chromatography as described for 17.
Compounds 26 were crystallized by trituration of the resulting
oil with a few drops of EtOH. Compounds 30 precipitated after
cooling of the filtrate and were collected by filtration and
recrystallized from EtOH. Amounts of reagents, yields, and
some analytical data are given in Table 1. 25: 1H NMR δ 2.00
(d, J ) 7.2 Hz, 3H, CHCH3), 6.45 (q, J ) 7.2 Hz, 1H, CHCH3),
7.18-7.31 (m, 8H, aromatic), 7.29 (s, 2H, NH2), 8.34 (m, 1H,
H-5), 8.36 (s, 1H, 2-H); 13C NMR δ 17.89 (CHCH3), 20.53
(phenyl-CH3), 50.54 (CH), 93.80 (C-4a), 110.96 (C-8), 120.22
(C-4b), 120.27 (C-6), 121.35 (C-5), 124.13 (C-7), 126.43, 127.77,
128.19, 128.96, 129.77 (aromatic CH), 136.23, 136.32, 138.02,
138.58 (C-8a, aromatic C), 156.17, 157.70, 159.61 (C-9a, C-4,
C-2).
described.29-31 2-Chloroadenosine (10 µM) was used in the A1
binding assay and theophylline (5 mM) in the A2 binding assay
to define nonspecific binding. The A1-selective adenosine
receptor agonist N6-cyclopentyladenosine (50 nM) was present
to block A1-adenosine receptors in the A2 binding assay.
Inhibition of the receptor-radioligand binding was determined
by a range of five to six concentrations of the compounds in
triplicate in at least three (A1) or two (A2) separate experi-
ments. The Cheng-Prusoff equation32 and KD values of 1 nM
for [3H]-(R)-PIA and 8.5 nM for [3H]NECA were used to
calculate the Ki values from the IC50 values, determined by
the nonlinear curve-fitting program InPlot, Version 4.03
(GraphPad, San Diego, CA). For selected compounds ad-
ditional A1AR binding studies with [3H]cyclohexyladenosine
([3H]CHA) as a radioligand were performed. A KD value of 1
nM was used for [3H]CHA for the calculation of Ki values.
Solu bility Deter m in a tion . Solubilities of selected com-
pounds were determined according to the method of Bruns and
Fergus.27 A 10 mM solution of compounds in DMSO was
prepared, diluted 1:100 in TRIS-HCl buffer, 50 mM, pH 7.4,
and allowed to reach equilibrium with shaking overnight at
room temperature. After centrifugation, the supernatant was
filtered through cotton. Several dilutions of these saturated
stock solutions were made in the buffer, on the basis of the
estimated solubility and the known affinity of each compound.
An A1 binding assay was performed with [3H]CHA as radio-
ligand. Data analysis was performed as described above using
a KD value of 1 nM for [3H]CHA. The solubility of each
compound was then calculated by dividing the IC50 value of a
compound, determined separately with [3H]CHA as described
above, by the fold dilution of the saturated solution required
to give 50% inhibition of [3H]CHA binding. Similar assays
were performed using a 1:40 dilution of the 10 mM stock
solution of the compounds to estimate their solubility in buffer,
containing 2.5% DMSO.
30: 1H NMR (CDCl3) δ 2.16 (d, J ) 7.2 Hz, 3H, CHCH3),
6.68 (q, J ) 7.2 Hz, 1H, CHCH3), 7.23-7.43 (m, 8H, aromatic),
7.64-7.66 (m, 3H, 2-phenyl-H), 8.28-8.41 (d, J ) 7.4 Hz, 1H,
H-5), 8.58-8.61 (m, 2H, 2-phenyl-H), 13.33 (s, 1H, NH); 13C
NMR δ 18.26 (CH3), 51.57 (CH), 98.36 (C-4a), 111.71 (C-8),
120.83, 121.44, 122.13, (C-5, C-4b, C-6), 124.02 (C-7), 126.54,
127.32, 127.97, 128.53, 128.75, 131.59 (aromatic CH), 132.45
(aromatic C), 135.49 (C-8a), 140.79 (aromatic C), 153.28,
154.21, 159.16 (C-9a, C-4, C-2).
(R)-4′-Am in o-2′-p h en yl-7′-(1-p h en ylet h yl)sp ir o[cyclo-
pen tan e-1,5′-[5H]pyr r olo[2,3-d]pyr im idin ]-6′(7′H)-on e (27a)
a n d (R)-2′-P h en yl-7′-(1-p h en yleth yl)sp ir o[cyclop en ta n e-
1,5′-[5H ]p yr r olo[2,3-d ]p yr im id in e]-4′,6′(3′H ,7′H )-d ion e
(31a ). General procedure: A suspension of 22a 15 (2.2 g, 6
mmol) in 100 mL of MeOH or 2.2 g (6 mmol) of 29a in 150 mL
of MeOH, respectively, was mixed with a solution of 6.0 g of
NaIO4 (28 mmol) in 50 mL of H2O, and the mixture was
refluxed for 6 h. Then the solvent was removed in vacuo, and
the aqueous phase was extracted with CH2Cl2 (3 × 20 mL).
After drying with anhydrous Na2SO4, the solution was evapo-
rated to dryness. The remaining oil was crystallized with
EtOH and recrystallized from MeOH (27a ) or EtOH (31a ).
Yields and some analytical data are given in Table 1. 27a :
1H NMR δ 1.73-2.18 (m, 8H, -CH2-), 1.93 (d, J ) 7.6 Hz, 3H,
CHCH3), 5.67(q, J ) 7.3 Hz, 1H, CHCH3), 6.51 (s, 2H, NH2),
7.24-8.26 (m, 10H, aromatic); 13C NMR δ 17.52 (CH3), 25.72,
32.59, 32.71 (C-2, C-3, C-4, C-5), 49.39 (CH), 51.92 (C-5′), 99.27
(C-4a′), 126.88, 127.30, 127.73, 128.56, 130.40 (aromatic CH),
138.04, 141.45 (aromatic C), 156.66 (C-7a′), 161.71, 162.70 (C-
2, C-4′), 182.08 (C-6′).
Molecu la r Mod elin g. The molecular modeling studies
were carried out on an Evans & Sutherland ESV 3/32 graphics
workstation and a Silicon Graphics Iris Indigo workstation
using the software package Sybyl, program versions 6.0 and
6.1,33 and MOPAC 5.0.34 Structures were built using Sybyl.
Energy minimization was performed using the Sybyl force
field. Conformational analysis was done using the Random
Search option of Sybyl with 30° increments. Charges were
calculated with the AM1 method as implemented in MOPAC
5.0 without changing the previously found geometries. A low-
energy conformation of 1a was used as
a template for
compound 17a . Torsion angles of 17a were adjusted in
analogy to 1a to get a maximal overlap of the (R)-1-phenylethyl
residue. This conformation was again minimized by the Sybyl
force field. Fitting of the obtained structures was carried out
with the Multifit module in Sybyl, using the four nitrogens of
both molecules for the fit. Approximate log P values were
calculated as described by Broto et al.35 by summarizing the fi
values of each atom. Nondefined fi values for three fragmental
codes were extrapolated according to the method of Broto.35
Ack n ow led gm en t. We thank Dr. Tom Scior (Stras-
bourg) who provided guidance on molecular modeling
studies, Prof. Dr. W. Adam (Wu¨rzburg) for permission
to use the computer facility of his laboratory, and Prof.
Dr. G. Bringmann (Wu¨rzburg) for allowing us to work
in his isotope laboratory. K. Eger was supported by the
Deutsche Forschungsgemeinschaft. C. E. Mu¨ller is
grateful for support by the Fonds der Chemischen
Industrie.
31a : 1H NMR δ 2.00 (d, J ) 7.3 Hz, 3H, CHCH3), 2.06-
2.21 (m, 8H, -CH2-), 5.78 (q, J ) 7.3 Hz, 1H, CHCH3), 7.25-
8.31 (m, 10H, aromatic), 13.64 (s, 1H, NH); 13C NMR δ CDCl3
17.57 (CH3), 27.80, 35.84, 35.96 (C-2, C-3, C-4, C-5), 50.36 (CH),
52.88 (C-5′), 107.78 (C-4a′), 127.26, 127.43, 128.05, 128.39,
128.78 (aromatic CH), 131.91 (aromatic C), 132.22 (aromatic
CH), 140.83 (aromatic C), 157.92, 160.01, 162.57 (C-7a′, C-4′,
C-2′), 183.21 (C-6′).
Su p p or tin g In for m a tion Ava ila ble: 1H and 13C NMR
data of intermediate and final products synthesized (6 pages).
Ordering information is given on any current masthead page.
Ad en osin e Recep tor Assa ys. Inhibition of binding of
[3H]-(R)-N6-(phenylisopropyl)adenosine ((R)-PIA) to A1-adenos-
ine receptors of rat cerebral cortical membranes and inhibition
of [3H]-5′-(N-ethylcarbamoyl)adenosine (NECA) to A2-adenos-
ine receptors of rat striatal membranes were assayed as
Refer en ces
(1) Fredholm, B. B.; Abbracchio, M. P.; Burnstock, G.; Daly, J . W.;
Harden, T. K.; J acobson, K. A.; Leff, P.; Williams, M. VI.
Nomenclature and Classification of Purinoceptors. Pharmacol.
Rev. 1994, 46, 143-156.

Pyrrolopyrimidines and Pyrimidoindoles as AR Antagonists
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13 2491
(2) (a) Dubyak, G. R.; El-Moatassim, C. Signal transduction via P2-
purinergic receptors for extracellular ATP and other nucleotides.
Am. J . Physiol. 1993, 265, C577-C606. (b) Lazarowski, E. R.;
Harden, T. K. Identification of a Uridine Nucleotide-selective
G-protein-linked Receptor That Activates Phospholipase C. J .
Biol. Chem. 1994, 269, 11830-11836. (c) Seifert, R.; Schultz, G.
Involvement of pyrimidinoceptors in the regulation of cell
function by uridine and by uracil nucleotides. Trends Pharmacol.
Sci. 1989, 10, 365-369.
(17) Thompson, R. D.; Manning, M., J r.; Daly, J . W.; Olsson, R. A.
N6,9-Disubstituted adenines: effect on affinity for adenosine
receptors of a chiral center in the 9-substituent. Med. Chem. Res.
1992, 2, 474.
(18) Schwabe, U.; Trost, T. Characterization of adenosine receptors
in rat brain by (-)[3H]N6-phenylisopropyladenosine, a new
ligand for Ri adenosine receptors. Naunyn-Schmiedeberg’s Arch.
Pharmacol. 1980, 321, 84-87.
(19) Bruns, R. F. Adenosine receptor activation in human fibro-
blasts: nucleoside agonists and antagonists. Can. J . Physiol.
Pharmacol. 1980, 58, 673-691.
(20) Peet, N. P.; Lentz, N. L.; Meng, E. C.; Dudley, M. W.; Ogden,
A.-M. L.; Demeter, D. A.; Weintraub, H. J . R.; Bey, P. A novel
synthesis of xanthines: support for a new binding mode for
xanthines with respect to adenosine at adenosine receptors. J .
Med. Chem. 1990, 33, 3127-3130.
(3) Mu¨ller, C. E.; Scior, T. Adenosine receptors and their modulators.
Pharm. Acta Helv. 1993, 68, 77-111.
(4) Londos, C.; Cooper, D. M.; Wolff, J . Subclasses of external
adenosine receptors. Proc. Natl. Acad. Sci. U.S.A. 1980, 77,
2551-2554.
(5) Linden, J . Cloned adenosine A3 receptors: pharmacological
properties, species differences and receptor function. Trends
Pharmacol. Sci. 1994, 15, 298-306.
(21) Dooley, M. J .; Quinn, R. J . The three binding domain model of
adenosine receptors: molecular modeling aspects. J . Med. Chem.
1992, 35, 211-216.
(6) Collis, M. G.; Hourani, S. M. O. Adenosine receptor subtypes.
Trends Pharmacol. Sci. 1993, 14, 360-366.
(7) (a) Ukena, D.; Padgett, W. L.; Hong, O.; Daly, J . W.; Daly, D.
T.; Olsson, R. A. N6-Substituted 9-methyladenines: a new class
of adenosine receptor antagonists. FEBS Lett. 1987, 215, 203-
208. (b) Thompson, R. D.; Secunda, S.; Daly, J . W.; Olsson, R.
A. N6,9-Disubstituted adenines: potent, selective antagonists at
the A1 adenosine receptor. J . Med. Chem. 1991, 34, 2877-2882.
(c) J acobson, K. A.; Siddiqi, S. M.; Olah, M. E.; J i, X.-d.; Melman,
N.; Bellamkonda, K.; Meshulam, Y.; Stiles, G. L.; Kim, H. O.
Structure-activity relationships of 9-alkyladenine and ribose-
modified adenosine derivatives at rat A3 adenosine receptors.
J . Med. Chem. 1995, 38, 1720-1735. (d) Lohse, M. J .; Klotz, K.-
N.; Diekmann, E.; Friedrich, K.; Schwabe, U. 2′,3′-Dideoxy-
N6-cyclohexyladenosine: an adenosine derivative with antago-
nist properties at adenosine receptors. Eur. J . Pharmacol. 1988,
156, 157-160.
(8) (a) Mu¨ller, C. E.; Shi, D.; Manning, M., J r.; Daly, J . W. Synthesis
of paraxanthine analogs (1,7-disubstituted xanthines) and other
xanthines unsubstituted at the 3-position: structure-activity
relationships at adenosine receptors. J . Med. Chem. 1993, 36,
3341-3349. (b) Grahner, B.; Winiwarter, S.; Lanzner, W.;
Mu¨ller, C. E. Synthesis and Structure-Activity Relationships of
Deazaxanthines: analogs of potent A1- and A2-adenosine recep-
tor antagonists. J . Med. Chem. 1994, 37, 1526-1534. (c) Mu¨ller,
C. E.; Grahner, B.; Heber, D. Amino-substituted 1,8-naphthy-
ridines and pyrido[2,3-d]-pyrimidines: new compounds with
affinity for A1- and A2-adenosine receptors. Pharmazie 1994, 49,
878-880. (d) Geis, U.; Grahner, B.; Pawlowski, M.; Drabczynska,
A.; Gorczyca, M.; Mu¨ller, C. E. Tricyclic theophylline derivatives
with high water solubility: structure-activity relationships at
adenosine receptors, phosphodiesterases, and benzodiazepine
binding sites. Pharmazie 1995, 50, 333-336.
(22) Cristalli, G.; Grifantini, M.; Vittori, S.; Balduini, W.; Cattabeni,
F. Adenosine and 2-chloroadenosine deaza-analogues as adenos-
ine receptor agonists. Nucleosides Nucleotides 1985, 4, 625-639.
(23) Mu¨ller, C. E.; Geis, U.; Grahner, B.; Lanzner, W.; Eger, K.
Pyrrolo[2,3-d]pyrimidines and pyrimido[4,5-b]indoles: Structure-
activity relationships of potent adenosine receptor antagonists.
Drug Dev. Res. 1994, 31, 301 (Abstract).
(24) (a) Biagi, G.; Giorgi, I.; Livi, O.; Scartoni, V.; Lucacchini, A.;
Martini, C.; Tacchi, P. N(6)-Substituted 2-phenyl-9-benzyl-8-
azaadenines. Affinity for adenosine A1 and A2 receptors. A
comparison with 2-N-butyl analogous derivatives. Farmaco
1994, 49, 187-191. (b) Biagi, G.; Giorgi, I.; Livi, O.; Scartoni,
V.; Lucacchini, A.; Martini, C.; Tacchi, P. Synthesis of new N(6)-
substituted 2-phenyl-8-azaadenosines. Their affinity for adenos-
ine A1 and A2 receptors. A comparison with the corresponding
2-phenyl-9-benzyl-8-azaadenines. Farmaco 1995, 50, 13-19.
(25) Brackett, L. E.; Daly, J . W. Functional characterization of the
A2b adenosine receptor in NIH 3T3 fibroblasts. Biochem. Phar-
macol. 1994, 47, 801-814.
(26) Ceccarelli, S.; Altobelli, M.; D’Alessandro, A.; Paesano, A. A novel
hydrophilic 8-cycloalkylxanthine derivative (IRFI 117) is a highly
selective antagonist at A1 adenosine receptors. Res. Commun.
Mol. Pathol. Pharmacol. 1995, 87, 101-102.
(27) Bruns, R. F.; Fergus, J . H. Solubilities of adenosine antagonists
determined by radioreceptor assay. J . Pharm. Pharmacol. 1989,
41, 590-594.
(28) (a) Ingersoll, A. W.; Burns, F. B. Solubility relationships among
optically isomeric salts II. The camphorates of R para tolylethy-
lamine. J . Am. Chem. Soc. 1932, 54, 4712-4715. (b) Leithe, W.
Die Konfiguration der Ephedrin-Basen. (The configuration of the
ephedrine bases.) Chem. Ber. 1932, 65, 660-666. (c) Pratesi, P.;
La Manna, A. Azione di Alcune Basi Otticamente Attive sugli
Acidi Canfocarbonici Enantiomorfi: L’Inversione di Effetto
cinetico. (Actions of Some Optically Active Bases on Enantio-
meric Camphoric Acids: The Inversion of the Kinetic Effect.)
Farmaco Ed. Sci. 1956, 11, 33-45.
(29) Bruns, R. F.; Daly, J . W.; Snyder, S. H. Adenosine receptors in
brain membranes: binding of N6-cyclohexyl[3H]adenosine and
1,3-diethyl-8-[3H]phenylxanthine. Proc. Natl. Acad. Sci. U.S.A.
1980, 77, 5547-5551.
(9) (a) J acobson, K. A.; Trivedi, B. K.; Churchill, P. C.; Williams,
M. Novel therapeutics acting via purine receptors. Biochem.
Pharmacol. 1991, 41, 1399-1410. (b) Suzuki, F. Adenosine A1
antagonists; A new therapeutic approach to cognitive deficits
and acute renal failure. Drug News Perspect. 1992, 5, 587-591.
(c) Kanda, T.; Shiozaki, S.; Shimada, J .; Suzuki, F.; Nakamura,
J . KF17837: a novel selective adenosine A2a receptor antagonist
with anticataleptic activity. Eur. J . Pharmacol. 1994, 256, 263-
268.
(10) Daly, J . W.; Padgett, W. L.; Eger, K. 7-Deaza-9-phenyladenines
a
New Class of Adenosine Receptor Antagonists. Biochem.
(30) Bruns, R. F.; Lu, G. H.; Pugsley, T. A. Characterization of the
A2 adenosine receptor labeled by [3H]NECA in rat striatal
membranes. Mol. Pharmacol. 1986, 29, 331-346.
Pharmacol. 1988, 37, 3749-3753.
(11) Mu¨ller, C. E.; Hide, I.; Daly, J . W.; Rothenha¨usler, K.; Eger, K.
7-Deaza-2-phenyladenines: structure-activity relationships of
potent A1 selective adenosine receptor antagonists. J . Med.
Chem. 1990, 33, 2822-2828.
(31) J acobson, K. A.; Ukena, D.; Kirk, K. L.; Daly, J . W. [3H]Xanthine
amine congener of 1,3-dipropyl-8-phenylxanthine, an antagonist
radioligand for adenosine receptors. Proc. Natl. Acad. Sci. U.S.A.
1986, 83, 4089-4093.
(12) (a) Roth, H. J .; Eger, K. Synthese von 2-Amino-3-cyano-pyrrolen.
(Synthesis of 2-amino-3-cyanopyrroles.) Arch. Pharm. (Wein-
heim, Ger.) 1975, 308, 179-185. (b) Roth, H. J .; Eger, K.
Synthese von Pyrrolo[2,3-d]pyrimidinen. (Synthesis of pyrrolo-
[2,3-d]pyrimidines.) Arch. Pharm. (Weinheim, Ger.) 1975, 308,
252-258.
(13) Pichler, H.; Folkers, G.; Roth, H. J .; Eger, K. Synthese von
7-unsubstituierten 7H-Pyrrolo[2,3-d]pyrimidinen. (Synthesis of
7-unsubstituted 7H-pyrrolo[2,3-d]pyrimidines.) Liebigs Ann.
Chem. 1986, 1485-1505.
(14) Eger, K.; Pfahl, J . G.; Folkers, G.; Roth, H. J . Selected reactions
on the o-aminonitrile system of substituted pyrroles. J . Hetero-
cycl. Chem. 1987, 24, 425-430.
(15) Eger, K.; Lanzner, W.; Rothenha¨usler, K. Synthesis of Substi-
tuted Indoles and Pyrimido[4,5-b]indoles by Dehydrogenation
of Tetrahydroindoles and Tetrahydro-pyrimidoindoles. Liebigs
Ann. Chem. 1993, 465-470.
(16) Girgis, N. S.; J o¨rgensen, A.; Pedersen, E. B. Phosphorus Pen-
toxide in Organic Synthesis; XI. A New Synthetic Approach to
7-Deazahypoxanthines. Synthesis 1985, 101-104.
(32) Cheng, Y. C.; Prusoff, W. H. Relationship between the inhibition
constant (Ki) and the concentration of inhibition which causes
50 percent inhibition (I50) of enzyme reaction. Biochem. Phar-
macol. 1973, 22, 3099-3108.
(33) Tripos Associates, Inc. 1993, 1994 Molecular Modelling Software
Sybyl, Version 6.0 and 6.1; St. Louis, MO 63144.
(34) Stewart, J . J . P. MOPAC Version 5.0 (QCPE No. 455).
(35) Broto, P.; Moreau, G.; Vandycke, C. Molecular structures:
perception, autocorrelation descriptor and sar studies. System
of atomic contributions for the calculation of the n-octanol/water
partition coefficients. Eur. J . Med. Chem. 1984, 19, 71-78.
(36) Siddiqi, S. M.; J i, X.-d.; Melman, N.; Olah, M. E.; J ain, R.; Evans,
P.; Glashofer, M.; Padgett, W. L.; Cohen, L. A.; Daly, J . W.; Stiles,
G. L.; J acobson, K. A. A survey of nonxanthine derivatives as
adenosine receptor ligands. Nucleosides Nucleotides 1996, 15,
693-717.
J M960011W