Conformationally restricted analogues of trimethoprim: 2,6-diamino-8- substituted purines as potential dihydrofolate reductase inhibitors from Pneumocystis carinii and Toxoplasma gondii

Article abstract of DOI:10.1021/jm970271t

Twenty-two 2,6-diamino-8-substituted purines (2-23) were synthesized, in which rotation around the two flexible bonds of trimethoprim (TMP), linking the pyrimidine ring to the side chain phenyl ring, was restricted by incorporation into a purine ring, in

Full text of DOI:10.1021/jm970271t

3032
J . Med. Chem. 1997, 40, 3032-3039
Con for m a tion a lly Restr icted An a logu es of Tr im eth op r im :
2,6-Dia m in o-8-su bstitu ted P u r in es a s P oten tia l Dih yd r ofola te Red u cta se
In h ibitor s fr om P n eu m ocystis ca r in ii a n d Toxopla sm a gon d ii1
Aleem Gangjee,*,† Anil Vasudevan,† and Sherry F. Queener‡
Division of Medicinal Chemistry, Graduate School of Pharmaceutical Sciences, Duquesne University,
Pittsburgh, Pennsylvania 15282, and Department of Pharmacology and Toxicology, Indiana University School of Medicine,
Indianapolis, Indiana 46202
Received April 23, 1997X
Twenty-two 2,6-diamino-8-substituted purines (2-23) were synthesized, in which rotation
around the two flexible bonds of trimethoprim (TMP), linking the pyrimidine ring to the side
chain phenyl ring, was restricted by incorporation into a purine ring, in an attempt to increase
the potency and selectivity of TMP against dihydrofolate reductase (DHFR) from the organisms
that often cause fatal opportunistic infections in patients with AIDS, i.e., Pneumocystis carinii
(pc) and Toxoplasma gondii (tg). The syntheses of analogues 2-20 were achieved via a one-
pot reaction of 2,4,5,6-tetraaminopyrimidine and the appropriately substituted benzaldehyde
or phenyl acetaldehyde, in acidic methoxyethanol. Analogues 21-23 were synthesized via
nucleophilic displacement of 2,6-diamino-8-(chloromethyl)purine with the appropriate anilines
or 2-naphthalenethiol. The compounds were evaluated as inhibitors of pcDHFR and tgDHFR
with rat liver (rl) DHFR as the mammalian reference enzyme. Compound 11, the 3′,4′-
dichlorophenyl analogue, was as potent as TMP and had a selectivity ratio of 13 for pcDHFR,
which ranked it as one of the three most selective inhibitors of pcDHFR (compared to rlDHFR)
known to date. It also displayed a selectivity ratio of 38 for tgDHFR. None of the other
analogues showed any improvement compared to TMP in potency or selectivity. In the
preclinical in vitro screening program of the National Cancer Institute, compound 11 showed
a GI50 of 10-6 M for the inhibition of the growth of 17 tumor cell lines.
Infections with Pneumocystis carinii (pc) and Toxo-
plasma gondii (tg) are a major cause of morbidity and
mortality in patients with acquired immunodeficiency
syndrome (AIDS), as well as in patients with other
immunosuppresive conditions.2,3 Regimens with docu-
mented efficacy for the treatment of these infections
include trimethoprim sulfamethoxazole,4 pentamidine,4
trimethoprim dapsone,5 and trimetrexate6 for P. carinii
infections and pyrimethamine plus a sulfonamide7 for
the treatment of T. gondii infections. The treatment of
P. carinii and T. gondii infections with antifolates such
as trimethoprim (TMP), trimetrexate (TMQ), and py-
central nervous system (CNS) where T. gondii infections
are a major cause of focal encephalitis.
TMP and pyrimethamine are weak inhibitors of
pcDHFR and tgDHFR and hence have to be used with
sulfonamides to augment their inhibition of folate
metabolism. TMQ and piritrexim (PTX), which are
100-10000 times more potent than TMP and py-
rimethamine against DHFR from P. carinii and T.
gondii, are also potent inhibitors of mammalian DHFR10
and hence lack selectivity, and their use is associated
with considerable toxicity. As a result, TMQ has to be
coadministered with leucovorin. Unfortunately, adverse
effects are often observed, and side effects associated
with the use of sulfa drugs often necessitate discontinu-
ation of therapy.11
As part of a program aimed at identifying potent and
selective inhibitors of P. carinii and/or T. gondii DHFR
as potential treatment of these infections, we have
reported the design and synthesis of a variety of 6-6
ring-fused analogues,12-16 as well as 6-5 ring-fused
analogues,17,18 with the objective of increasing the
selectivity of the significantly potent, nonselective bi-
cyclic DHFR inhibitors TMQ and PTX for pcDHFR and
tgDHFR. Some of these compounds15 were potent and
selective inhibitors of pcDHFR and tgDHFR, with one
analogue, 2,4-diamino-6-[N-(2′,5′-dimethoxybenzyl)-N-
methylamino]pyrido[2,3-d]pyrimidine (1, X ) H, Y )
NCH3, Z ) CH2, R ) 2,5-OCH3), displaying selectivity
ratios of 100 and 309 for pcDHFR and tgDHFR (com-
pared to human (h) DHFR), respectively. These com-
pounds were designed as analogues of TMQ and PTX,
with the specific objective of increasing the selectivity
of these drugs for pcDHFR and/or tgDHFR. An alter-
nate approach for the design of drugs useful for the
rimethamine takes advantage of the fact that these
organisms are permeable to lipophilic, nonclassical
antifolates and, unlike mammalian cells, lack the car-
rier-mediated active transport system(s) required for the
uptake of classical antifolates with polar glutamate side
chains.8 Thus host tissues can be selectively protected
by the coadministration of the reduced folate leucovorin,
which is selectively taken up by mammalian cells and
reverses toxicity associated with nonselective dihydro-
folate reductase (DHFR) inhibitors such as TMQ.9
Further, these lipophilic agents can penetrate into the
† Duquesne University.
‡ Indiana University School of Medicine.
X Abstract published in Advance ACS Abstracts, August 15, 1997.
S0022-2623(97)00271-9 CCC: $14.00 © 1997 American Chemical Society

Conformationally Restricted Analogues of Trimethoprim
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19 3033
F igu r e 1. (a) Observed mode of binding of pyrrolo[2,3-
d]pyrimidine antifolates to E. coli DHFR.24 (b) Possible mode
of binding of analogues 2-23 to DHFR.
treatment of these opportunistic infections is to attempt
to improve the potency of the selective, clinically useful
agent TMP. As mentioned previously, TMP is a selec-
tive but weak inhibitor of pcDHFR and hence has to be
used along with sulfa drugs for synergistic effects
against DHFR. Since the sulfa drug is the principal
culprit11 of adverse effects associated with TMP sul-
famethoxazole, a DHFR inhibitor which would combine
the selectivity of TMP and the high potency of TMQ or
PTX would be a viable single agent in the treatment of
these opportunistic infections.
The side chain of TMP can adopt a variety of different
conformations due to rotational flexibility around tor-
sion angles τ1 and τ2, several of which are unfavorable
for the inhibition of pcDHFR and/or tgDHFR. It is well
known that the conformation of the trimethoxyphenyl
side chain of TMP is responsible, in part, for its
selectivity for bacterial DHFR.19 One method of in-
creasing the potency and/or selectivity of TMP is to
introduce conformational restriction of the side chain
in an appropriate orientation to optimally interact with
pcDHFR and/or tgDHFR. The literature contains sev-
eral examples of TMP analogues with bridging atoms
other than a methylene moiety.20,21 Chan and Roth22
have reported elegant efforts to improve the antibacte-
rial activity of TMP by conformational restriction of the
side chain via incorporation into a dihydroindane ring.
However, no significant improvement in Escherichia coli
DHFR inhibition was obtained, presumably due to the
steric crowding caused by the extra atoms needed to
orient the side chain in the desired conformation.
Recently, Kuyper et al.23 reported attempts to restrict
rotation around τ1 of TMP by linking the 4-amino moiety
with the methylene bridge, to afford either a pyrrolo[2,3-
d]pyrimidine or a pyrido[2,3-d]pyrimidine. We elected
to restrict rotation around τ1 and τ2 by incorporation of
the C5-C7 and C7-C1′ bonds of TMP into a purine
should afford analogues with defined restricted geom-
etry. Such analogues could result in improved potency
while maintaining selectivity and were worth exploring.
The pKa of 2,6-diaminopurines (5.09)25 is similar to that
of 2,4-diaminopteridines (5.32) such as methotrexate
(MTX), which is a potent DHFR inhibitor, further
supporting the hypothesis that appropriately substi-
tuted 2,6-diaminopurine analogues could be potent
inhibitors of DHFR. Elion et al.26 reported several 2,6-
diamino-8-substituted purines in the 1950s and evalu-
ated them against the growth of L. casei cells in culture.
However, to our knowledge, no DHFR inhibition studies
against P. carinii or T. gondii were reported for these
analogues.
Molecular modeling using SYBYL 6.0,27 and its
SEARCH and MAXIMIN options, and superimposition
of the pyrimidine ring of the energy-minimized confor-
mations of analogues 2-19 onto the pyrimidine ring of
the E. coli DHFR-bound conformation of TMP19,28
indicated that the side chain aryl ring of these analogues
occupied similar but not identical orientations as the
side chain of TMP in its bound conformation. This
supports the premise that though there is an extra atom
between the pyrimidine ring and the side chain of
analogues 2-19 as compared to TMP, the distance
between the center of the pyrimidine ring and the
phenyl ring in TMP and analogues 2-19 is quite similar
but not identical. The crystal structure of TMP with
pcDHFR29 showed that TMP binds to pcDHFR in a
conformation similar to its binding mode in E. coli
DHFR; thus the use of E. coli DHFR-bound TMP as the
template was a good approximation for TMP bound to
pcDHFR. The coordinates of pcDHFR with TMP were
unavailable at the time the target compounds of this
study were designed.
Analogues 20-23 were synthesized with the intent
of extending the distance between the side chain phenyl
ring and the pyrimidine moiety, thus probing alternate
binding sites in pcDHFR and tgDHFR. Molecular
superimposition indicates that the 6-substituent of 6-5
ring-fused analogues lies between the 6- and 7-substit-
uents of a 6-6 ring-fused system. Most 6-substituted
6-6 ring-fused nonclassical compounds of the pyrido[2,3-
d]pyrimidine series synthesized in our laboratory12-16
have been found to potently inhibit DHFR. Hence, it
was reasoned that extension of the bridge length as in
analogues 20-23 would allow for an increased distance
between the heterocycle and the hydrophobic side chain
compared to that in analogues 2-19 and would ap-
proximate the distance in the pyrido[2,3-d]pyrimidine
analogues.
ring, as in analogues 2-23 (Table 1). The choice of the
ring structure (i.e., purine) for conformational restriction
was based on the knowledge that unsubstituted 2,6-
diaminopurines have been reported24 to be competitive
inhibitors of chicken liver (cl) DHFR, with Ki values
similar to that for 2,4-diamino-6-methylpteridines and
more than that of 2,4-diaminopyrimidines. Nonclassical
2,4-diamino-substituted pyrido[2,3-d]pyrimidine anti-
folates are potent inhibitors of pcDHFR and tgDHFR,
and since the potency against DHFR of unsubstituted
purines and pyrido[2,3-d]pyrimidines was similar,24 it
was reasoned that conformational restriction of the side
chain bridge of TMP via incorporation in a purine ring
Kuyper et al.24 have shown via X-ray crystallography
that 4-desamino pyrrolo[2,3-d]pyrimidine analogues
bind to E. coli DHFR in the ‘folate mode’ (Figure 1).
Analogues 2-23 may also bind to DHFR in the ‘folate
mode’ (Figure 1), with the N9-H functioning as the

3034 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19
Gangjee et al.
Sch em e 1a
a
Key: (a) methoxyethanol, HCl.
4-amino moiety and interacting with Ile 10 (pcDHFR
numbering). The N1-nitrogen (purine numbering) is
speculated30,31 to be the most basic nitrogen in the
purines. Using semiempirical methods32 for calculating
the electron density of each atom of a 2,6-diaminopurine,
the N1-nitrogen of the purine was calculated to have a
Mulliken charge density greater than the N3-nitrogen.
This suggested that binding of the purine analogues in
the ‘folate mode’ would enable the N1-nitrogen of the
purine, upon protonation, in conjunction with the
2-amino moiety, to participate in the salt bridge with
Glu 32 and Wat 403 (pcDHFR numbering),29 similar to
that observed in the binding of TMP with E. coli
DHFR28,33 and pcDHFR.29
The side chain substituents for the target analogues
include some substituents such as the methoxyphenyl
rings which we have found in prior studies12-16 to be
conducive to potency and/or selectivity. Other groups
including electron-withdrawing moieties and naphthyl
and fluorenyl moieties were included to provide a
diverse group of side chains, some of which we have also
reported to possess high potency and/or selectivity for
pcDHFR and/or tgDHFR in the pyrido-15 and pyr-
rolo[2,3-d]pyrimidine18 series. The use of different side
chains was important for exploring tgDHFR inhibition
in particular, since the crystal structure of tgDHFR is
not yet available.
raaminopyrimidine and a benzaldehyde without isola-
tion, in an acetate buffer at pH 4.5. However, they
isolated a number of other side products and reported
that this method did not have any preparative utility.
We decided to investigate the one-step cyclization of a
4,5-diaminopyrimidine and substituted benzaldehydes
in weakly acidic media with concomitant air oxidation
to afford the desired purines.
2,4,5,6-Tetraaminopyrimidine (24) was heated in
methoxyethanol, and at reflux 1 N hydrochloric acid was
added dropwise to solubilize the pyrimidine. This was
followed by the addition of the appropriately substituted
benzaldehyde, following which the reaction was refluxed
overnight. TLC analyses indicated the presence of an
intense, UV-absorbing product along with some unre-
acted aldehyde. In all the reactions performed (except
for the synthesis of analogues 12, 16, and 17), at the
end of the reaction period, a thick precipitate had
formed, which was filtered and collected to afford the
desired purines in 35-40% yield. No other products
were observed on TLC, and thus, analytically pure
target compounds were obtained in a single step, using
comparatively mild conditions. That purine ring forma-
tion had occurred was confirmed by the 1H NMR in
deuterated dimethyl sulfoxide, which indicated the
presence of an exchangeable proton at 11.80-13.00
ppm, corresponding to the purine N9-H. Further, the
presence of only the 2- and 4-amino proton, indicated
that the 5- and 6-amino moieties had participated in
the reaction. Had the reaction stopped at the Schiff
Ch em istr y
A search of the literature revealed that a variety of
methods exist for the synthesis of 8-substituted pu-
rines.34 Among the most widely used method is the
reaction of an appropriately substituted benzoic acid and
a 5,6-diaminopyrimidine followed by acid-catalyzed ring
closure.35 This methodology requires two steps for the
assembly of the purine ring, involving initial formation
of the 4-amino-5-(acylamino)pyrimidine, followed by
ring closure. The synthetic scheme utilized in this study
for the synthesis of 8-substituted purine analogues is
outlined in Scheme 1 and is a modification of the Traube
synthesis36 between a 4,5-diaminopyrimidine and a
substituted benzaldehyde.37 This method has been
widely used for the synthesis of benzimidazoles38 and
has been employed for purine synthesis only for xan-
thine derivatives.39 This methodology theoretically
requires two steps, with the initial formation of the
Schiff base between the 5-amino moiety and the alde-
hyde followed by oxidative cyclization in the presence
of ferric chloride or copper acetate.38,39 Kuznetsov and
Remizov40 reported an attempted one-step cyclization
of the intermediate Schiff base between 2,4,5,6-tet-
1
base, the H NMR would have indicated the presence
of three sets of amino groups along with the presence
of an imine proton of the Schiff base. The absence of
these features, as well as corroboration via mass
spectrometry, confirmed purine ring formation. For
analogues 12, 16, and 17, at the end of the reaction,
the solvent was evaporated and chromatographic puri-
fication was necessary to afford the desired analogues.
In all the compounds synthesized in this study, the
exhangeable proton at 11.80-13.00 ppm was assigned
to the N9-H rather than the N7-H based on the
comparison of the ground-state energies41 of an N9-H
purine as compared to an N7-H purine. Even though
the N7-H purine has been observed predominantly in
crystal structures of purines, the N9-H isomer is more
stable than the N7-H isomer by 2 kcal/mol.42
For the synthesis of analogue 20, a similar reaction
with phenylacetaldehyde which involved refluxing the
reaction mixture overnight in acidic methoxyethanol
resulted in a complex mixture of products on TLC.
However when the reaction was performed in acidic

Conformationally Restricted Analogues of Trimethoprim
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19 3035
Ta ble 1. Target Compounds of This Study
Sch em e 2a
compd
n
Ar
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
C6H5
3′,4′,5′-OCH3C6H2
2′,3′,4′-OCH3C6H2
2′,4′,6′-OCH3C6H2
2′,4′,5′-OCH3C6H2
2′,5′-OCH3C6H3
3′,5′-OCH3C6H3
3′,4′-OCH3C6H3
2′,4′-OCH3C6H3
3′,4′-Cl2C6H3
2′,6′-Cl2C6H3
a
Key: (a) 110 °C, 2 h; (b) 1 N NaOH; (c) SOCl2, DMF; (d) ArXH,
3′,5′-OCH3-4′-OHC6H2
2′-NO2-4′,5′-OCH3C6H2
4′-C6H5C6H4
4-C5H4N
2′-naphthyl
4′-OCH3-1′-naphthyl
9′-fluorenyl
C6H5
NH-3′,4′,5′-OCH3C6H2
NH-2′,5′-OCH3C6H3
S-2-naphthyl
K2CO3.
methoxyethanol for 6 h, followed by workup and puri-
fication, compound 20 was obtained in 12% yield.
The synthesis of analogues 21-23 required the ver-
satile and important intermediate, 2,6-diamino-8-(chlo-
romethyl)purine (27). Weinstock et al.43 have reported
an improved synthesis of the original method utilized
by Baker and Santi44 for the synthesis of this interme-
diate. Adoption of this method using 2,4,5,6-tetraami-
nopyrimidine (24) fused with glycolic acid for 4 h
(Scheme 2) followed by workup and basification afforded
2,4,6-triamino-5-(glycoamido)pyrimidine, 25. Base-
catalyzed ring closure followed by acidification afforded
2,6-diamino-8-(hydroxymethyl)purine (26) in 76% yield.
Chlorination of 26 with thionyl chloride in DMF af-
forded the desired 8-chloromethyl derivative 27, which
was sufficiently pure to be used in subsequent trans-
formations.
Ta ble 2. Inhibitory Concentrations (IC50, µM) and Selectivity
Ratios against P. carinii (pc), T. gondii (tg), and Rat Liver (rl)
DHFR45,46
compd
pcDHFR
153
rlDHFR
rl/pc
0.4
tgDHFR
rl/tg
2
3
4
5
6
7
8
9
59.9
13
3.7
1.88
7
3.5
32.4
52.3
0.9
252
11.0 (4%)
15.3
25 (14%)
30.1
18 (0%)
280
13 (15%)
29.8
1.6
34 (19%)
29 (15%)
107
133a
0.003a
0.0015a
28
2.2
1.8
0.84
1.5
1.7
3.0
14.2
2.4
6.7
11.0 (6%)
1
16.8
32.9
20 (14%)
27
13 (10%)
30.5
0.5
34 (5%)
49
2.1
5.9
2.1
2.2
4.7
2.0
10.8
3.7
0.38
38
59.4 (34%)
17.0 (15%)
31.5 (4%)
22 (10%)
37.4 (9%)
9.0 (17%)
12.0 (6%)
14.0 (11%)
19.5
3,4,5-Trimethoxyaniline and 2,5-dimethoxyaniline were
alkylated with 27 in N,N-dimethylacetamide using
potassium carbonate to afford, after purification, ana-
logues 21 and 22 in 56% and 52% yields, respectively.
Similarly, 2-naphthalenethiol was alkylated with 27 to
afford 23 in 65% yield.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
TMP
TMQ
PTX
13
11.0 (11%)
68 (6%)
25 (15%)
15.3
0.91
10.37
605
0.05
0.16
18 (5%)
113 (10%)
13 (0%)
35 (12%)
9.8
Biologica l Eva lu a tion a n d Discu ssion
Analogues 2-23 were evaluated against pcDHFR,
tgDHFR, and rlDHFR which served as the mammalian
source, and the results are reported in Table 2. These
assays were performed at 37 °C, under conditions of
substrate (90 µM dihydrofolic acid) and cofactor (119 µM
NADPH) saturation; tgDHFR and rlDHFR were as-
sayed in the presence of 150 mM KCl.45,46 Selectivity
ratios, determined as the ratio of the inhibition constant
(IC50) of rlDHFR vs pcDHFR or tgDHFR, are also
reported in Table 2. The corresponding values for TMP,
TMQ, and PTX are included in Table 2 for comparison.
Compound 2, which has an unsubstituted phenyl ring,
displayed micromolar activity against pcDHFR. Against
tgDHFR, it was 10-fold less potent than TMP and
showed a selectivity ratio of 2.1. This suggested that
substitution of suitable functional groups on the phenyl
ring may provide analogues with improved potency and/
or selectivity against pcDHFR and tgDHFR. Addition
of a 3,4,5-trimethoxy substituent on the phenyl ring to
afford compound 3 resulted in a 5-fold increase in
potency against rlDHFR compared to 2. Against tgDH-
FR, compound 3 was 12-fold more potent and twice as
selective as compound 2 and was as potent as TMP
1.0
3.2
34 (6%)
29 (7%)
105
1.02
11.1a
0.07a
0.048a
45
2.38
49a
12a
2.7a
0.042a
0.031a
0.010a
0.017a
0.29a
0.09a
a
Data from ref 47. Values in parentheses are the percent
inhibition at the highest concentration tested.
against tgDHFR. These encouraging results prompted
the synthesis of other trimethoxyphenyl-substituted
regioisomers (4-6). Compound 5, with a 2,4,6-tri-
methoxyphenyl substituent, was 33-fold more potent
than 2 against tgDHFR, which could be attributed to a
favorable interaction of the methoxy substituent at the
ortho positions with tgDHFR or the effect of conforma-
tional restriction induced by the ortho substituents, or
both. The activity of compounds 4-7 suggests that
substitution at the ortho position with a methoxy group
is beneficial for tgDHFR inhibition, since compounds
4-7 are more potent than 2 and 3, with compound 5
being the most potent tgDHFR inhibitor of this series.

3036 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19
Gangjee et al.
Comparison of the inhibition of compounds 7-10 indi-
cates the effect of dimethoxy-substituted compounds on
all three DHFRs. Once again, substitution at the ortho
position with a methoxy moiety, such as in compounds
7 and 10, showed better inhibitory potency against
rlDHFR and tgDHFR, compared to 8 and 9 which are
unsubstituted in the ortho positions. Compound 8, with
a selectivity ratio of 10.8 against tgDHFR, was 36- and
77-fold more selective than TMQ and PTX, respectively,
as an inhibitor of tgDHFR. Comparison of the activity
of compounds 3 and 8 suggests that substitution at the
para position with a methoxy group is somewhat
detrimental to the inhibition of rlDHFR, but not for
tgDHFR, resulting in a 2-fold increase in the selectivity
of 8 for tgDHFR.
Compounds 11 and 12 were synthesized to evaluate
the effect of electron-withdrawing chloro groups. Re-
placement of the methoxy groups of 9 with chloro groups
to afford compound 11 resulted in a 2-fold increase in
tgDHFR inhibition and a 5-fold decrease in rlDHFR
inhibition. As a result, compound 11 had selectivity
ratios of 13 and 38 for the inhibition of pcDHFR and
tgDHFR, respectively. This analogue is almost equi-
potent with TMP as an inhibitor of pcDHFR. Further,
it is one of the three most selective inhibitors of pcDHFR
(compared to rlDHFR) reported to date (the most
selective inhibitor of pcDHFR reported to date, 2,4-
diamino-6-[(thiophenyl)methyl]pteridine, has a selectiv-
ity ratio of 2547 under the conditions of the assay used
in these studies). The selectivity ratio of compound 11
for tgDHFR was also impressive, and it is the most
selective inhibitor of pcDHFR and tgDHFR of this entire
series. The solubility of 12 precluded its evaluation
beyond 11 µM. Replacement of the 4′-methoxy group
of compound 3 with a hydroxy group to afford 13 results
in a 2-fold increase in inhibition against tgDHFR.
Further, this analogue displays a 15-fold selectivity for
tgDHFR. Replacement of the o-methoxy group of
analogue 6 with an electron-withdrawing nitro group
(14) results in decreased inhibition against both rlDHFR
and tgDHFR. Substitution of a phenyl ring at the para
position of compound 2 also results in decreased inhibi-
tion of pcDHFR and tgDHFR. Further, replacement of
the phenyl ring of 2 with a 2-naphthyl ring as in
compound 17 results in a 5-fold increase in selectivity
against tgDHFR. On the other hand, replacing the
phenyl ring of 2 with a fluorenyl moiety in 19 results
in increased inhibition of rlDHFR, with no change
against tgDHFR.
F igu r e 2. Inhibition of P. carinii growth in culture. Test
compound or diluent was added to the medium at the time of
inoculation of monolayers of HEL cells with P. carinii tropho-
zoites, according to a standard protocol. Plates were harvested
at the times shown, and numbers of P. carinii were determined
by direct counting. Each point represents the mean of four
replicate wells ( standard error of the mean.
point for structural variation on account of its potency
against tgDHFR.
On the basis of its impressive selectivity for pcDHFR,
analogue 11 was further evaluated as an inhibitor of
the growth of P. carinii trophozoites in culture, accord-
ing to established methods.48 Compound 11 was inhibi-
tory to the growth of P. carinii at a concentration of 55
µM, and the inhibition was similar to that produced by
TMP/SMX at 170/990 µM. Compound 11 was not tested
in further culture studies due to its toxicity to host cells,
but analogues 8, 13, and 17 were evaluated in culture
studies, since they showed little evidence of cytotoxicity
toward host cells. IC50 values could not be determined
for 8 and 13 but were above 10 and 50 µM, respectively;
the IC50 value for 17 was 16 µM, a value similar to its
rlDHFR IC50, suggesting that this compound penetrates
well into infected cells.
Selected analogues were chosen to be evaluated in the
preclinical in vitro anticancer screening program of the
National Cancer Institute.49 While most of the ana-
logues displayed poor inhibition of the growth of tumor
cells in culture (GI50 > 10-5 M), reflecting their poor
inhibition of rlDHFR, analogue 11 displayed a GI50
value of 10-6 M for the inhibition of the growth of 17
tumor cell lines. This was surprising, since its IC50 for
the inhibition of rlDHFR was 252 µM, and suggests that
11 might have additional or alternate mechanism(s) of
action as an inhibitor of the growth of tumor cells.
In summary, conformational restriction of the side
chain of TMP analogues by incorporation of an imidazole
ring did not improve the activity relative to TMP. The
To study the effect of increasing the distance between
the purine nucleus and the hydrophobic side chain to
explore accessory binding sites on DHFR, analogues
20-23 were synthesized. Compound 20 displayed a 16-
fold increase in inhibition of pcDHFR and a 56-fold
increase in inhibition of tgDHFR compared to 2. Un-
fortunately, this improvement in potency did not trans-
late into selective inhibition of pcDHFR or tgDHFR,
since the inhibition of rlDHFR was also improved
compared to 2. However, since extension of the side
chain afforded an increase in potency, this suggested
that suitable extension of the side chain might provide
for analogues with not only improved potency but also
selectivity. Hence, the 2-atom-bridged analogues 21-
23 were synthesized. However, none of these analogues
were significantly potent against either pcDHFR or
tgDHFR. Compound 20 could be an interesting starting

Conformationally Restricted Analogues of Trimethoprim
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19 3037
2,3,4-trimethoxybenzaldehyde (0.43 g, 2.21 mmol) to afford
0.43 g (62%) as a yellow solid: TLC (CHCl3:CH3OH:NH4OH,
5:1:0.5) Rf ) 0.41; mp > 300 °C; H NMR (DMSO-d6) δ 3.78
(s, 3 H, 2′-OCH3), 3.90 (s, 3 H, 4′-OCH3), 4.08 (s, 3 H, 3′-OCH3),
6.86 (d, 1 H, 5′-H), 7.41 (s, 2 H, 2-NH2), 7.69 (d, 1 H, 6′-H),
8.31 (s, 2 H, 4-NH2), 12.31 (s, 1 H, NH). Anal. (C14H16N6O3‚
0.5H2O‚0.7HCl) C, H, N.
poor solubility of most of these analogues precluded
their evaluation as inhibitors of pcDHFR. This, in turn,
could be due to the increased hydrophilicity of these
analogues compared to TMP, which might, in turn,
contribute to expenditure of energy (and a subsequent
decrease in potency) required for desolvation of these
molecules to bind to the active site of DHFR. This effect
is more pronounced against pcDHFR as compared to
rlDHFR and tgDHFR, suggesting significant differences
in the active site of these DHFRs. Analogue 11 dis-
played excellent selectivity ratios for the inhibition of
both pcDHFR and tgDHFR and is superior to the
currently used combination of TMP/SMX in the inhibi-
tion of the growth of P. carinii cells in culture. Using
compound 11 as a lead analogue, studies are currently
underway to improve its potency and selectivity against
pcDHFR and tgDHFR by conformationally restricting
the side chain in 6-5 rings, other than the purine ring.
1
2,6-Dia m in o-8-(2′,4′,6′-t r im et h oxyp h en yl)p u r in e, 5.
Compound 5 was obtained from 24 (0.25 g, 1.79 mmol) and
2,4,6-trimethoxybenzaldehyde (0.35 g, 1.79 mmol) to afford
0.40 g (62%) as a cream solid: TLC (CHCl3:CH3OH:NH4OH,
1
6:1:0.5) Rf ) 0.38; mp > 300 °C; H NMR (DMSO-d6) δ 3.85
(s, 6 H, 2′,6′-OCH3), 3.90 (s, 3 H, 4′-OCH3), 5.75 (s, 2 H, 2-NH2),
6.05 (s, 2 H, 3′,5′-H), 6.75 (s, 2 H, 4-NH2), 12.55 (s, 1 H, NH).
Anal. (C14H16N6O3‚0.4H2O‚1.0HCl) C, H, N.
2,6-Dia m in o-8-(2′,4′,5′-t r im et h oxyp h en yl)p u r in e, 6.
Compound 6 was obtained from 24 (0.25 g, 1.79 mmol) and
2,4,5-trimethoxybenzaldehyde (0.35 g, 1.79 mmol) to afford
0.35 g (61%) as a cream solid: mp > 300 °C; 1H NMR (DMSO-
d6) δ 3.78 (s, 3 H, 2′-OCH3), 3.91 (s, 3 H, 4′-OCH3), 4.08 (s, 3
H, 5′-OCH3), 6.89 (s, 1 H, 3′-H), 7.34 (s, 2 H, 2-NH2), 7.68 (s,
1 H, 6′-H), 8.29 (s, 2 H, 4-NH2), 12.30 (s, 1 H, NH). Anal.
(C14H16N6O3‚0.3H2O) C, H, N.
Exp er im en ta l Section
2,6-Dia m in o-8-(2′,5′-d im eth oxyp h en yl)p u r in e, 7. Com-
pound 7 was obtained from 24 (0.25 g, 1.79 mmol) and 2,5-
dimethoxybenzaldehyde (0.29 g, 1.79 mmol) to afford 0.31 g
(61%) as a white solid: TLC (CHCl3:CH3OH:NH4OH, 5:1:0.5)
Melting points were determined on a Fisher-J ohns melting
point apparatus or a Mel Temp apparatus and are uncorrected.
Nuclear magnetic resonance spectra for proton (1H NMR) were
recorded on a Brucker WH-300 spectrometer (300 MHz). The
data were accumulated by 16K size with 0.5 s delay time with
internal standard TMS; s ) singlet, br s ) broad singlet, d )
doublet, t ) triplet, q ) quartet, m ) multiplet. Low-
resolution mass spectra were obtained on an LKB-9000
instrument. Thin layer chromatography was performed on
silica gel plates with fluorescent indicator and were visualized
with light at 254 and 366 nm, unless indicated otherwise.
Column chromatography was performed with 230-400 mesh
silica gel purchased from Aldrich Chemical Co., Milwaukee,
WI. Elution was performed using a gradient, and 10 mL
fractions were collected, unless mentioned otherwise. All
anhydrous solvents were purchased from Aldrich Chemical Co.
and were used without further purification. Samples for
microanalysis were dried in vacuo over phosphorous pentoxide
at 70 or 110 °C. Microanalyses were performed by Atlantic
Microlabs, Norcoss, GA.
Gen er a l P r oced u r e for th e Syn th esis of Com p ou n d s
2-20. To a suspension of 2,4,5,6-tetraaminopyrimidine (24)
in 25 mL of methoxyethanol at 80 °C was added concentrated
HCl (∼2-3 mL) dropwise to effect solution. To this solution
was added the appropriate aldehyde (substituted or unsub-
stituted benzaldehyde, substituted or unsubstituted naphthal-
dehyde, fluorene aldehyde, pyridinecarboxaldehyde or phen-
ylacetaldehyde), and the mixture was refluxed for 6 h to 2 days.
The precipitate formed was filtered and washed with hexanes
and ether to afford the desired compound. In the case of
compounds 12, 16, and 17, at the end of the reaction, 1.0 g of
silica gel was added to the reaction mixture and the solvent
evaporated to afford a dry plug. This plug was applied on the
surface of a silica gel column (1.05 in. × 23 in.) and eluted
with CHCl3:CH3OH (5:1). Fractions corresponding to the
desired product were pooled and evaporated to afford the
desired compound.
2,6-Dia m in o-8-p h en ylp u r in e, 2. Compound 2 was ob-
tained from 24 (0.10 g, 0.71 mmol) and benzaldehyde (0.07 g,
0.71 mmol) to afford 0.11 g (68%) as a white solid: TLC
(CHCl3:CH3OH:NH4OH, 5:1:0.5) Rf ) 0.51; lit.40 mp > 300 °C;
1H NMR (DMSO-d6) δ 5.72 (s, 2 H, 2-NH2), 6.67 (br s, 2 H,
4-NH2), 7.10-7.60 (m, 5 H), 12.56 (s, 1 H, NH). Anal.
(C11H10N6‚0.2H2O) C, H, N.
2,6-Dia m in o-8-(3′,4′,5′-t r im et h oxyp h en yl)p u r in e, 3.
Compound 3 was obtained from 24 (0.10 g, 0.71 mmol) and
3,4,5-trimethoxybenzaldehyde (0.14 g, 0.71 mmol) to afford
0.15 g (58%) as a light yellow solid: TLC (CH2Cl2:CH3OH:
NH4OH, 5:2:0.5) Rf ) 0.43; mp > 300 °C; 1H NMR (DMSO-d6)
δ 3.70 (s, 3 H, 4′-OCH3), 3.84 (s, 6 H, 3′,5′-OCH3), 5.71 (s, 2 H,
2-NH2), 6.66 (s, 2 H, 4-NH2), 7.37 (s, 2 H, 2′,6′-H), 12.50 (s, 1
H, NH). Anal. (C14H16N6O3‚0.2H2O‚0.2 HCl) C, H, N.
2,6-Dia m in o-8-(2′,3′,4′-t r im et h oxyp h en yl)p u r in e, 4.
Compound 4 was obtained from 24 (0.31 g, 2.21 mmol) and
1
Rf ) 0.58; mp > 300 °C; H NMR (DMSO-d6) δ 3.79 (s, 3 H,
2′-OCH3), 4.02 (s, 3 H, 5′-OCH3), 7.13 (d, 1 H, 4′-H), 7.23 (d, 1
H, 3′-H), 7.39 (s, 2 H, 2-NH2), 7.71 (s, 1 H, 6′-H), 8.39 (s, 2 H,
4-NH2), 12.35 (s, 1 H, NH). Anal. (C13H14N6O2‚0.1H2O‚0.1HCl)
C, H, N.
2,6-Dia m in o-8-(3′,5′-d im eth oxyp h en yl)p u r in e, 8. Com-
pound 8 was obtained from 24 (0.10 g, 0.71 mmol) and 3,5-
dimethoxybenzaldehyde (0.12 g, 0.71 mmol) to afford 0.14 g
(68%) as a white solid: TLC (CHCl3:CH3OH:NH4OH, 5:1:0.5)
1
Rf ) 0.53; mp > 300 °C; H NMR (DMSO-d6) δ 3.83 (s, 6 H,
3′,5′-OCH3), 6.65 (s, 1 H, 4′-H), 7.28 (s, 2 H, 2′,6′-H), 7.50 (s, 2
H, 2-NH2), 8.50 (s, 2 H, 4-NH2), 12.90 (s, 1 H, NH). Anal.
(C13H14N6O2‚0.8H2O‚0.7HCl) C, H, N.
2,6-Dia m in o-8-(3′,4′-d im eth oxyp h en yl)p u r in e, 9. Com-
pound 9 was obtained from 24 (0.31 g, 2.21 mmol) and 3,4-
dimethoxybenzaldehyde (0.37 g, 2.21 mmol) to afford 0.38 g
(60%) as a yellowish solid: TLC (CHCl3:CH3OH:NH4OH, 5:1:
0.5) Rf ) 0.51); mp > 300 °C; 1H NMR (DMSO-d6) δ 3.91 (s, 3
H, 3′-OCH3), 3.96 (s, 3 H, 4′-OCH3), 5.72 (s, 2 H, 2-NH2), 6.73
(s, 2 H, 4-NH2), 6.80 (d, 1 H, 5′-H), 7.50-7.60 (m, 2 H, 2′-H,
6′-H), 12.60 (s, 1 H, NH). Anal. (C13H14N6O2‚0.3HCl) C, H,
N.
2,6-Diam in o-8-(2′,4′-dim eth oxyph en yl)pu r in e, 10. Com-
pound 10 was obtained from 24 (0.30 g, 2.14 mmol) and 2,4-
dimethoxybenzaldehyde (0.36 g, 2.14 mmol) to afford 0.36 g
(59%) as a cream solid: TLC (CHCl3:CH3OH:NH4OH, 5:1:0.5)
1
Rf ) 0.52; mp > 300 °C; H NMR (DMSO-d6) δ 3.87 (s, 2 H,
2′-OCH3), 4.06 (s, 3 H, 4′-OCH3), 6.75 (s, 1 H, 3′-H), 6.82 (d, 1
H, 5′-H), 7.29 (s, 2 H, 2-NH2), 8.13 (d, 1 H, 6′-H), 8.29 (s, 2 H,
4-NH2), 12.00 (s, 1 H, NH). Anal. (C13H14N6O2‚0.5HCl) C, H,
N.
2,6-Dia m in o-8-(3′,4′-d ich lor op h en yl)p u r in e, 11. Com-
pound 11 was obtained from 24 (0.50 g, 3.57 mmol) and 3,4-
dichlorobenzaldehyde (0.62 g, 3.57 mmol) to afford 0.36 g (59%)
as a fluffy, white solid: TLC (CHCl3:CH3OH:NH4OH, 5:1:0.5)
1
Rf ) 0.65; mp > 300 °C; H NMR (DMSO-d6) δ 7.37 (s, 2 H,
2-NH2), 7.62 (s, 1 H, 2′-H), 7.89 (d, 1 H, 6′-H), 8.03 (d, 1 H,
5′-H), 8.72 (s, 2-NH2), 12.96 (s, 1 H, NH). Anal. (C11H8N6Cl2‚
0.5H2O) C, H, N, Cl.
2,6-Dia m in o-8-(2′,6′-d ich lor op h en yl)p u r in e, 12. Com-
pound 12 was obtained from 24 (0.50 g, 3.57 mmol) and 2,6-
dichlorobenzaldehyde (0.62 g, 3.57 mmol) to afford 0.48 g (46%)
as a light yellow solid: TLC (CHCl3:CH3OH:NH4OH, 5:1:0.5)
Rf ) 0.7; TLC (CHCl3:CH3OH:NH4OH, 4:1:0.5) Rf ) 0.69; mp
> 300 °C; 1H NMR (DMSO-d6) δ 6.43 (br m, 4 H, 2-NH2,
4-NH2), 6.93 (s, 1 H, 4′-H), 7.11 (s, 2 H, 3′,5′-H), 12.61 (s, 1 H,
NH). Anal. (C11H8N6Cl2‚0.4H2O) C, H, N, Cl.
2,6-Dia m in o-8-(3′,5′-d im eth oxy-4′-h yd r oxyp h en yl)p u -
r in e, 13. Compound 13 was obtained from 24 (1.0 g, 7.14
mmol) and 3,5-dimethoxy-4-hydroxybenzaldehyde (vanillin)

3038 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19
Gangjee et al.
(1.3 g, 7.14 mmol) to afford 0.60 g (28%) as a white solid: TLC
(CHCl3:CH3OH:NH4OH, 5:1:0.5) Rf ) 0.37; mp > 300 °C; H
NMR (DMSO-d6) δ 3.93 (s, 6 H, 3′,5′-OCH3), 5.75 (s, 2 H,
2-NH2), 6.35 (s, 1 H, OH), 6.75 (s, 2 H, 4-NH2), 7.00 (s, 2 H,
2′,6′-H), 12.60 (s, 1 H, NH). Anal. (C13H14N6O3‚1.0H2O) C,
H, N.
amide was added 3.3 mL (5.6 g, 0.05 mol) of thionyl chloride,
dropwise over a 10 min period. The resulting solution was
heated at 60-70 °C for 3 h and allowed to stand at room
temperature overnight. The solvents were evaporated, and
25 mL of anhydrous ether was added to the residue which
precipitated a yellow solid: 1H NMR (DMSO-d6) δ 4.87 (s, 2
H, CH2), 6.74 (br s, 2 H, 2-NH2), 7.67 (br s, 2 H, 4-NH2), 12.4
(s, 1 H, NH). Due to the hygroscopic nature of this intermedi-
ate, it was not purified further but carried directly to the next
step.
1
2,6-Dia m in o-8-(2′-n it r o-4′,5′-d im e t h oxyp h e n yl)p u -
r in e, 14. Compound 14 was obtained from 24 (0.30 g, 2.14
mmol) and 2-nitro-4,5-dimethoxybenzaldehyde (0.45 g, 2.14
mmol) to afford 0.32 g (45%): mp > 300 °C; 1H NMR (DMSO-
d6) δ 3.63-3.65 (overlapping s, 6 H, 4′,5′-OCH3), 5.00 (s, 2 H,
2-NH2), 6.10 (s, 2-H, 4-NH2), 6.87 (s, 1 H, 6′-H), 8.04 (s, 1 H,
3′-H), 11.9 (s, 1 H, NH). Anal. (C13H13N7O4‚0.5H2O) C, H, N.
2,6-Dia m in o-8-(4′-bip h en ylyl)p u r in e, 15. Compound 15
was obtained from 24 (0.41 g, 2.92 mmol) and 4-phenylbenz-
aldehyde (0.53 g, 2.92 mmol) to afford 0.45 g (51%): TLC
2,6-Dia m in o-8-[(3′,4′,5′-tr im eth oxya n ilin o)m eth yl]p u -
r in e, 21. To a solution of 27 (0.50 g, 2.71 mmol) in 20 mL of
anhydrous N,N-dimethylacetamide was added 3,4,5-trimethoxy-
aniline (0.49 g, 2.71 mmol) followed by potassium carbonate
(0.30 g). The reaction mixture was stirred at room tempera-
ture under nitrogen for 24 h. TLC analyses (CHCl3:CH3OH,
5:1) indicated the presence of a new product at a Rf ) 0.6 along
with trace amounts of both starting materials and a baseline
spot. The reaction mixture was filtered, 1.0 g of silica gel
added to the filtrate, and the solvent evaporated to afford a
dry plug. This plug was eluted with CHCl3:CH3OH on a 1.05
in. × 23 in. silica gel column, using a gradient elution (95:5 to
80:20), and fractions containing the product were pooled and
evaporated to afford pure 21 as a white solid (0.52 g, 56%):
mp ) 298 °C dec; 1H NMR (DMSO-d6) δ 3.50 (s, 3 H, 4′-OCH3),
3.66 (s, 6 H, 3′,5′-OCH3), 4.24 (d, 2 H, CH2), 5.55 (s, 2 H,
2-NH2), 5.82 (t, 1 H, NH), 5.96 (s, 2 H, 2′,6′-H), 6.48 (s, 2 H,
4-NH2), 11.95 (s, 1 H, NH). Anal. (C15H19N7O3‚1.1H2O) C, H,
N.
1
(CHCl3:CH3OH:NH4OH, 5:1:0.5) Rf ) 0.59; mp > 300 °C; H
NMR (DMSO-d6) δ 5.81 (s, 2 H, 2-NH2), 6.70 (s, 2 H, 4-NH2),
7.4-7.5 (m, 3 H, 3′′,4′′,5′′-H), 7.70 (m, 2 H, 6′′,2′′-H), 7.78 (d,
2 H, 3′, 5′-H), 8.00 (d, 2 H, 2′, 6′-H), 12.56 (s, 1 H, NH). Anal.
(C17H15N6‚1.0H2O) C, H, N.
2,6-Dia m in o-8-(4′-p yr id in yl)p u r in e, 16. Compound 16
was obtained from 24 (1.0 g, 7.14 mmol) and pyridine-4-
carboxaldehyde (0.85 g, 7.14 mmol) to afford 0.38 g (22%) as
a yellow solid: TLC (CHCl3:CH3OH:NH4OH, 5:1:0.5) Rf ) 0.3;
mp > 300 °C; 1H NMR (DMSO-d6) δ 5.65 (s, 2 H, 2-NH2), 6.75
(s, 2 H, 4-NH2), 7.72 (d, 2 H, 2′,6′-H), 8.90 (d, 2 H, 3′,5′-H),
12.55 (s, 1 H, NH). Anal. (C10H9N7‚0.1HCl) C, H, N.
2,6-Dia m in o-8-(2′-n a p h t h yl)p u r in e, 17. Compound 17
was obtained from 24 (0.21 g, 1.5 mmol) and 2-naphthaldehyde
(0.17 g, 1.5 mmol) to afford 0.25 g (60%) as a cream solid: TLC
2,6-Dia m in o-8-[(2′,5′-d im e t h oxya n ilin o)m e t h yl]p u -
r in e, 22. To a solution of 27 (0.40 g, 1.65 mmol) in 20 mL of
anhydrous N,N-dimethylacetamide was added 2,5-dimethoxy-
aniline (0.25 g, 1.65 mmol) followed by potassium carbonate
(0.30 g). The reaction mixture was stirred at room tempera-
ture under nitrogen for 24 h. TLC analyses (CHCl3:CH3OH,
5:1) indicated the presence of a new product at a Rf ) 0.7 along
with trace amounts of 27 and a baseline spot. The reaction
mixture was filtered, 1.0 g of silica gel added to the filtrate,
and the solvent evaporated to afford a dry plug. This plug
was placed on a silica gel column and eluted with CHCl3:
CH3OH using gradient elution (95:5 to 80:20), and fractions
containing the product were pooled and evaporated to afford
pure 22 (0.27 g, 52%) as a light yellow solid: mp ) 290 °C
dec; 1H NMR (DMSO-d6) δ 3.59 (s, 3 H, 2′-OCH3), 3.74 (s, 3 H,
5′-OCH3), 4.30 (d, 2 H, CH2), 5.42 -5.55 (overlapping peaks,
3 H, 2-NH2, NH), 6.10 (m, 2 H, 4′,6′-H), 6.40 (s, 2 H, 4-NH2),
6.69 (d, 1 H, 3′-H), 11.89 (s, 1 H, NH). Anal. (C14H17N7O2‚
0.3H2O) C, H, N.
1
(CHCl3:CH3OH:NH4OH, 5:1:0.5) Rf ) 0.72; mp > 300 °C; H
NMR (DMSO-d6) δ 5.71 (s, 2 H, 2-NH2), 6.70 (s, 2 H, 4-NH2),
7.31-7.40 (m, 3 H), 7.66-7.74 (m, 4 H), 12.50 (s, 1 H, NH).
Anal. (C15H12N6‚0.5H2O) C, H, N.
2,6-Diam in o-8-(4′-m eth oxy-1′-n aph th yl)pu r in e, 18. Com-
pound 18 was obtained from 24 (0.28 g, 2.0 mmol) and
4-methoxy-1-naphthaldehyde (0.31 g, 2.0 mmol) to afford 0.31
g (51%) as a white solid: mp > 300 °C; 1H NMR (DMSO-d6) δ
3.71 (s, 3 H, 4′-OCH3), 5.91 (s, 2 H, 2-NH2), 6.11 (d, 1 H, 3′-H),
6.21 (d, 1 H, 2′-H), 6.61 (d, 1 H, 5′-H), 6.66 (d, 1 H, 8′-H), 6.71
(m, 2 H, 6′, 7′-H), 7.14 (s, 2 H, 4-NH2), 12.10 (s, 1 H, NH).
Anal. (C10H14N6O‚0.5H2O) C, H, N.
2,6-Dia m in o-8-(9′-flu or en yl)p u r in e, 19. Compound 19
was obtained from 24 (0.48 g, 3.42 mmol) and 9-fluorenecar-
boxaldehyde (0.66 g, 3.42 mmol) to afford 0.52 g (48%) as a
yellow solid: TLC (CHCl3:CH3OH:NH4OH, 4:1:0.5) Rf ) 0.61;
1
mp > 300 °C; H NMR (DMSO-d6) δ 4.20 (s, 2 H, CH2), 7.46
(m, 4 H, 2-NH2, 4′,7′-H), 7.58 (d, 1 H, 8′-H), 8.01 (d, 1 H, 9′-
H), 8.28 (m, 2 H, 5′,6′-H), 8.40 (s, 1 H, 2′-H), 8.60 (s, 2 H,
4-NH2), 12.66 (s, 1 H, NH). Anal. (C18H13N6‚0.5H2O) C, H,
N.
2,6-Dia m in o-8-[(2′-n a p h th ylth io)m eth yl]p u r in e, 23. To
a solution of 27 (0.50 g, 2.71 mmol) in 20 mL of anhydrous
N,N-dimethylacetamide was added 2-naphthalenethiol (0.35
g, 2.71 mmol) followed by potassium carbonate (0.30 g). The
reaction mixture was stirred at room temperature under
nitrogen for 12 h. TLC analysis (CHCl3:CH3OH, 5:1) indicated
the presence of a new product at a Rf ) 0.8 along with trace
amounts of 27 and a baseline spot. The reaction mixture was
filtered, 1.0 g of silica gel added to the filtrate, and the solvent
evaporated to afford a dry plug. This plug was loaded on a
silica gel column and eluted with CHCl3:CH3OH using gradient
elution (95:5 to 80:20), and fractions containing the product
were pooled and evaporated to afford pure 23 as a light yellow
2,6-Dia m in o-8-ben zylp u r in e, 20. Compound 20 was ob-
tained from 24 (1.0 g, 7.14 mmol) and phenylacetaldehyde
(0.86 g, 7.14 mmol) to afford 0.21 g (12%): TLC (CHCl3:
CH3OH:NH4OH, 5:1:0.5) Rf ) 0.54; mp > 300 °C; 1H NMR
(DMSO-d6) δ 3.69 (s, 2 H, CH2), 5.72 (s, 2 H, 2-NH2), 6.67 (s,
2 H, 4-NH2), 7.10-7.50 (m, 5 H), 12.56 (s, 1 H, NH). Anal.
(C12H12N6‚0.5HCl) C, H, N, Cl.
2,6-Dia m in o-8-(h yd r oxym eth yl)p u r in e, 26. A mixture
of 2,4,5,6-tetraaminopyrimidine (24) (4.8 g, 20 mmol) and
glycolic acid (7.6 g, 100 mmol) was heated at 110 °C for 2 h.
At the end of this period, the syrupy reaction mixture was
washed with 2 × 50 mL portions of ethyl ether, and the residue
was dissolved in 30 mL of water. The pH of the solution was
adjusted to 9 with NH4OH and the precipitated solid collected
and dried. The product was recrystallized from water to afford
2,4,6-triamino-5-(glycolamido)pyrimidine, 25.
1
solid (0.50 g, 68%): mp ) 282 °C dec; H NMR (DMSO-d6) δ
4.36 (s, 2 H, CH2), 5.61 (s, 2 H, 2-NH2), 6.53 (s, 2 H, 4-NH2),
7.49 (m, 3 H), 7.89 (m, 3 H), 7.96 (s, 1 H, 1′-H), 12.12 (s, 1 H,
NH). Anal. (C16H14N6S‚0.3H2O) C, H, N, S.
Ack n ow led gm en t. This work was supported in part
by the National Institutes of Health Grant GM 40998
(A.G.) and NIH contracts N01-AI-87240 (S.F.Q.) and
N01-AI-35171 (S.F.Q.). Michael Mohutsky and Karen
Balzer (1994-1995 undergraduate research partici-
pants) are acknowledged for assistance with some of the
syntheses. The culture tests for P. carinii were carried
out in the laboratories of Professors J . W. Smith and
M. S. Bartlett at Indiana University School of Medicine.
A solution of 25 (3.5 g, 0.018 mol) in 25 mL of 2 N NaOH
was refluxed for 8 h. The reaction mixture was cooled and
acidified to pH 5 with acetic acid. The precipitated solid was
collected to afford 2.3 g (70%) of 26 as a white solid: mp >
1
360 °C; H NMR (DMSO-d6) δ 3.57 (br s, 2 H, CH2), 4.5 (s, 1
H, OH), 5.57 (s, 2 H, 2-NH2), 6.49 (s, 2 H, 4-NH2), 11.89 (s, 1
H, NH). Anal. (C6H8N6O‚1.2H2O) C, H, N.
2,6-Dia m in o-8-(ch lor om eth yl)p u r in e, 27. To a suspen-
sion of 26 (3.00 g, 0.016 mol) in 20 mL of N,N-dimethylform-

Conformationally Restricted Analogues of Trimethoprim
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 19 3039
(21) Kompis, J .; Then, R.; Wick, A.; Montavon, M. 2,4-Diamino-5-
benzylpyrimidines as Inhibitors of Dihydrofolate Reductase. In
Enzyme Inhibitors; Brodbeck, U., Ed.; Verlag Chemie: Wein-
heim, Germany, 1980; pp 177-189.
(22) Chan, J . H.; Roth, B. 2,4-Diamino-5-benzylpyrimidines as
Antibacterial Agents. 14. 2,3-Dihydro-1-(2,4-diamino-5-pyrim-
idyl)-1H-indenes as Conformationally Restricted Analogs of
Trimethoprim. J . Med. Chem. 1991, 34, 550-555.
(23) Kuyper, F. L.; Garvey, J . M.; Baccanari, D. P.; Champness, J .
N.; Stammers, D. K.; Beddell, C. R. Pyrrolo[2,3-d]pyrimidines
and Pyrido[2,3-d]pyrimidines as Conformationally Restricted
Analogues of the Antibacterial Agent Trimethoprim. Bioorg.
Med. Chem. 1996, 4, 593-602.
(24) Zakrzewski, S. F. The Mechanism of Binding of Folate Analogues
by Folate Reductase. J . Biol. Chem. 1963, 238, 1485-1490.
(25) Roth, B.; Strelitz, J . Z. The Protonation of 2,4-Diaminopyrim-
idines. I. Dissociation Constants and Substituent Effects. J . Org.
Chem. 1963, 34, 821-836.
(26) Elion, G. B.; Burgi, E.; Hitchings, G. H. Studies on Condensed
Pyrimidine Systems. VII. Some 8-Aryl Purines. J . Am. Chem.
Soc. 1951, 73, 5235-5239.
(27) Tripos Associates, Inc., 1699 S. Hanley Rd., Suite 303, St. Louis,
MO 63144.
Refer en ces
(1) (a) Taken in part from the thesis submitted by A.V. to the
Graduate School of Pharmaceutical Sciences, Duquesne Uni-
versity, in partial fulfillment of the requirements for the degree
of Doctor of Philosophy, J une 1996. (b) Presented in part at the
209th American Chemical Society National Meeting, Anaheim,
CA, April 2-6, 1995; MEDI 125.
(2) Haverkos, H. W. Assessment of Therapy for Toxoplasma En-
cephalitis. Am. J . Med. 1987, 82, 907-914.
(3) Kovacs, J . A.; Hieminz, J . W.; Macher, A. M.; Stover, D.; Murray,
H. W.; Shelhamer, J .; Lane, H. C.; Urmacher, U.; Honig, C.;
Longo, D. L.; Parker, M. M.; Natanson, C.; Parillo, J . E.; Fauci,
A. S.; Pizzo, P. A.; Masur, H. Pneumocystis carinii Pneumonia:
A Comparison between Patients with the Acquired Immunode-
ficiency Syndrome and Patients with Other Immunodeficiencies.
Ann. Intern. Med. 1984, 100, 663-671.
(4) Wharton, J . M.; Coleman, D. L.; Wofsy, C. B.; Luce, J . M.;
Blumenfield, W.; Hadley, W. K.; Ingram-Drake, L.; Volberding,
P. A.; Hopwell, P. C. Trimethoprim-sulfamethoxazole or Pen-
tamidine for Pneumocystis carinii Pneumonia in the Acquired
Immunodeficiency Syndrome. Ann. Intern. Med. 1986, 105, 37-
44.
(28) Baker, D. J .; Beddell, C. R.; Champness, J . N.; Goodford, P. J .;
Morrington, F. E. A.; Smith, D. R.; Stammers, D. K. The Binding
of Trimethoprim to Bacterial Dihydrofolate Reductase. FEBS
Lett. 1981, 126, 49-52.
(29) Champness, J . N.; Achari, A.; Ballantine, S. P.; Bryant, P. K.;
Delves, C. J .; Stammers, D. K. The Structure of Pneumocystis
carinii Dihydrofolate Reductase to 1.9 Å Resolution. Structure
1994, 2, 915-924.
(30) Pugmire, R.; Grant, D. M. Carbon-13 Magnetic Resonance. XIX.
I Benzimidazole, Purine, and Their Anionic and Cationic Species.
J . Am. Chem. Soc. 1971, 93, 1880.
(31) Wagner, R.; Philipsborn, W. Protonation of Purine, Adenine and
Guanine-NMR Spectrum and Structure of the Mono-, Di- and
Tri-Cations. Helv. Chim. Acta 1971, 54, 1543-1558.
(32) The Mulliken and Electrofit charges were calculated using the
MacSpartan semiempirical program (release 1.0).
(5) Leoung, G. S.; Mills, J .; Hopewell, P. C.; Hughes, W.; Wofsy, C.
Dapsone-trimethoprim for the Treatment of Pneumocystis carinii
Pneumonia in the Acquired Immunodeficiency Syndrome. Ann.
Intern. Med. 1986, 105, 45-48.
(6) News. FDA approves Trimetrexate as a second line therapy for
Pneumocystis carinii. Am. J . Hosp. Pharm. 1994, 51, 591-592.
(7) Georgiev, V. S. Opportunistic/Nosocomial Infections. Treatment
and Developmental Therapeutics. Toxoplasmosis. Med. Res. Rev.
1993, 13 (5), 529-568.
(8) Allegra, C. J .; Kovacs, J . A.; Drake, J . C.; Swan, J . C.; Chabner,
B. A.; Masur, H. Activity of Antifolates against Pneumocystis
carinii Dihydrofolate Reductase and Identification of a Potent
New Agent. J . Exp. Med. 1987, 165, 926-931.
(9) Masur, H.; Polis, M. A.; Tuazon, C. U.; Ogata-Arakaki, D.;
Kovacs, J . A.; Katz, D.; Hilt, D.; Simmons, T.; Feurstein, I.;
Lundgren, B.; Lane, H. C.; Chabner, B. A.; Allegra, C. J . Salvage
Trial of Trimetrexate-Leucovorin for the Treatment of Cerebral
Toxoplasmosis. J . Infect. Dis. 1992 167, 1422-1426.
(10) Kovacs, J . A.; Allegra, C. A.; Swan, J . C.; Drake, J . C.; Parrillo,
J . E.; Chabner, B. A.; Masur, H. Potent Antipneumocystsis and
Antitoxoplasma Activities of Piritrexim, A Lipid-soluble Anti-
folate. Antimicrob. Agents Chemother. 1988, 32, 430-433.
(11) Gordin, F. M.; Simon, G. L.; Wofsy, C. B.; Mills, J . Adverse
Reactions to Trimethoprim-sulfamethoxazole in Patients with
the Acquired Immunodeficiency Syndrome. Ann. Intern. Med.
1984, 100, 495-499.
(12) Gangjee, A.; Shi, J .; Queener, S. F.; Barrows, L. R.; Kisliuk, R.
L. Synthesis of 5-Methyl-5-Deaza Nonclassical Antifolates as
Inhibitors of Dihydrofolate Reductases and as Potential Anti-
pneumocystis, Antitoxoplasma, and Antitumor Agents. J . Med.
Chem. 1993, 36, 3437-3443.
(13) Gangjee, A.; Vasudevan, A.; Queener, S. F.; Kisliuk, R. L.
6-Substituted 2,4-Diamino-5-methylpyrido[2,3-d]pyrimidines as
Inhibitors of Dihydrofolate Reductases from Pneumocystis carinii
and Toxoplasma gondii and as Antitumor Agents. J . Med. Chem.
1995, 38, 1778-1785.
(14) Gangjee, A.; Zaveri, N.; Kothare, M.; Queener, S. F. 6-Substi-
tuted 2,4-diamino-5,6,7,8-Tetrahydroquinazolines as Potential
Inhibitors of Dihydrofolate Reductases. J . Med. Chem. 1996, 39,
1448-1456.
(15) Gangjee, A.; Vasudevan, A.; Queener, S. F.; Kisliuk, R. L. 2,4-
Diamino-6-substituted Pyrido[2,3-d]pyrimidine Antifolates as
Potent and Selective Nonclassical Inhibitors of Dihydrofolate
Reductases. J . Med. Chem. 1996, 39, 1448-1456.
(16) Gangjee, A.; Zhu, Y.; Queener, S. F.; Francome, P.; Broom, A.
D. Nonclassical 2,4-Diamino-8-deazafolate Analogues as Inhibi-
tors of Dihydrofolate Reductases from Rat Liver, Pneumocystis
carinii, and Toxoplasma gondii. J . Med. Chem. 1996, 39, 1836-
1845.
(17) Gangjee, A.; Devraj, R.; McGuire, J . J .; Kisliuk, R. L.; Queener,
S. F.; Barrows, L. R. Classical and Nonclassical Furo[2,3-
d]pyrimidines as Novel Antifolates: Synthesis and Biological
Activities. J . Med. Chem. 1994, 37, 1169-1176.
(18) Gangjee, A.; Mavandadi, F.; Queener, S. F.; McGuire, J . J . Novel
2,4-Diamino-5-substituted-pyrrolo[2,3-d]pyrimidines as Classical
and Nonclassical Inhibitors of Dihydrofolate Reductases. J . Med.
Chem. 1995, 38, 2158-2165.
(19) Mathews, D. A.; Bolin, J . T.; Burdge, J . M.; Filman, K. W.;
Kaufman, B. T.; Bedell, C. R.; Champness, J . N.; Stammers, D.
N.; Kraut, J . Refined Crystal Structure of E. coli and Chicken
Liver DHFR Containing Bound Trimethoprim. J . Biol. Chem.
1985, 260, 381-391.
(20) Stogryn, E. L. Synthesis of Trimethoprim Variations. Replace-
ment of CH2 by Polar Groupings. J . Med. Chem. 1972, 15, 200-
201.
(33) Bolin, J . T.; Filman, D. J .; Mathews, D. A.; Hamlin, R. C.; Kraut,
J . Crystal Structures of E. coli and L. casei DHFR Refined at
1.7 Å Resolution. J . Biol. Chem. 1982, 257, 13650-13662.
(34) Lister, J . H, Ed. The Purines; J ohn Wiley and Sons, Inc.: New
York, 1996; Vol. 54 (Suppl. I), pp 21-60.
(35) Fu, S. C. J .; Chinopros, J . 8-Arylpurines and Their UV Absorp-
tion Spectra (I). J . Heterocycl. Chem. 1966, 3, 476.
(36) Traube, W. The Synthesis of Ureas, Xanthines, Theobromines,
Theophyllines and Caffeines from Cyanoacetic acid. Chem. Ber.
1900, 3, 3035-3056.
(37) Traube, W.; Nithak, W. On the Reactions of Aldehydes with
ortho-diamines. Chem. Ber. 1906, 39, 227-235.
(38) Weidenhagen, R. A New Synthesis of Benzimidazole Derivatives.
Chem. Ber. 1936, 69, 2263-2272.
(39) J erchel, D.; Kracht, M.; Krucker, K. Examination of Benzimid-
azole. Ann. 1954, 590, 232-241.
(40) Kuznetsov, V. S.; Remizov, A. L. Synthesis of 2,6-Diamino-8-
arylpurines Using Aromatic Aldehydes. Khim. Geterotsikl. Soed.
1972, 10, 1430-1434.
(41) Hjalmars, F. I.; Chaudhari, J . N. Electronic Structure and
Spectra of Imidazole, Purine and Some Aminopurines. Acta
Chem. Scand. 1969, 23, 2963-2977.
(42) Pullman, B.; Berthod, H.; Caillet, J . On the Crystal Structure
of Purine. Theor. Chim. Acta 1968, 10, 43-46.
(43) Weinstock, L. T.; Grabowski, B. F.; Cheng, C. C. Folic Acid
Analogs. II. p-{[(2,6-Diamino-8-purinyl)methyl]amino}-benzoyl-
L-glutamic Acid and Related Compounds. J . Med. Chem. 1970,
13, 995-997.
(44) Baker, B. R.; Santi, D. V. Irreversible Enzyme Inhibitors. XCIV.
Inhibition of Dihydrofolate Reductase with Derivatives of 2,6-
Diaminopurines. J . Heterocycl. Chem. 1967, 4, 216-223.
(45) Broughton, L.; Queener, S. F. Pneumocystis carinii Dihydrofolate
Reductase Used to Screen Potential Antipneumocystis Drugs.
Antimicrob. Agents Chemother. 1991, 35, 1348-1355.
(46) Chio, L.; Queener, S. F. Identification of Highly Potent and
Selective Inhibitors of Toxoplasma gondii Dihydrofolate Reduc-
tase. Antimicrob. Agents Chemother. 1993, 37, 1914-1923.
(47) Piper, J . R.; J ohnson, C. A.; Hosmer, C. A.; Carter, R. L.;
Pfefferkorn, E. R.; Borotz, S. E.; Queener, S. F. Lipophilic
Antifolates a Candidates Against Opportunistic Infections. Adv.
Exp. Med. Biol. 1993, 338, 429-433.
(48) Bartlett, M. S.; Shaw, M.; Navaran, P.; Smith, J . Q.; Queener,
S. F. Evaluation of Potent Inhibitors of Dihydrofolate Reductase
in a Culture Model for Growth of Pneumocystis carinii. Antimi-
crob. Agents Chemother. 1995, 39, 2436-2441.
(49) We thank the Developmental Therapeutics Program of the
National Cancer Institute for perfoming the in vitro anticancer
evaluation.
J M970271T