Structural studies on bioactive compounds. Part 29: Palladium catalysed arylations and alkynylations of sterically hindered immunomodulatory 2-amino- 5-halo-4,6-(disubstituted)pyrimidines

Article abstract of DOI:10.1016/S0968-0896(00)00017-1

The immunological agent bropirimine 5 is a tetra-substituted pyrimidine with anticancer and interferon-inducing properties. Synthetic routes to novel 5-aryl analogues of bropirimine have been developed and their potential molecular recognition properties analysed by molecular modelling methods. Sterically challenged 2-amino-5-halo-6-phenylpyrimidin-4-ones (halo=Br or I) are poor substrates for palladium catalysed Suzuki cross-coupling reactions with benzeneboronic acid because the basic conditions of the reaction converts the amphoteric pyrimidinones to their unreactive enolic forms. Palladium-mediated reductive dehalogenation of the pyrimidinone substrates effectively competes with cross-coupling. 2-Amino-5-halo-4-methoxy-6- phenylpyrimidines can be converted to a range of 5-aryl derivatives with the 5-iodopyrimidines being the most efficient substrates. Hydrolysis of the 2- amino-5-aryl-4-methoxy-6-phenylpyrimidines affords the required pyrimidin-4- ones in high yields. Semi-empirical quantum mechanical calculations show how the nature of the 5-substituent influences the equilibrium between the 1H- and 3H-tautomeric forms, and the rotational freedom about the bond connecting the 6-phenyl group and the pyrimidine ring. Both of these factors may influence the biological properties of these compounds. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(00)00017-1

Bioorganic & Medicinal Chemistry 8 (2000) 739±750
Structural Studies on Bioactive Compounds. Part 29:
Palladium Catalysed Arylations and Alkynylations of
Sterically Hindered Immunomodulatory
2-Amino-5-halo-4,6-(disubstituted)pyrimidinesy
Duncan R. Hannah, a Edward C. Sherer, a Roy V. Davies, b Roger B. Titman, b
Charles A. Laughton a and Malcolm F. G. Stevens a,*
aCancer Research Laboratories, School of Pharmaceutical Sciences, University of Nottingham, Nottingham NG7 2RD, UK
bKnoll Pharmaceuticals Research and Development Department, Pennyfoot Street, Nottingham NG1 1GF, UK
Received 30 August 1999; accepted 18 October 1999
AbstractÐThe immunological agent bropirimine 5 is a tetra-substituted pyrimidine with anticancer and interferon-inducing prop-
erties. Synthetic routes to novel 5-aryl analogues of bropirimine have been developed and their potential molecular recognition
properties analysed by molecular modelling methods. Sterically challenged 2-amino-5-halo-6-phenylpyrimidin-4-ones (halo=Br or I)
are poor substrates for palladium catalysed Suzuki cross-coupling reactions with benzeneboronic acid because the basic conditions
of the reaction converts the amphoteric pyrimidinones to their unreactive enolic forms. Palladium-mediated reductive dehalogena-
tion of the pyrimidinone substrates e€ectively competes with cross-coupling. 2-Amino-5-halo-4-methoxy-6-phenylpyrimidines
can be converted to a range of 5-aryl derivatives with the 5-iodopyrimidines being the most ecient substrates. Hydrolysis of the
2-amino-5-aryl-4-methoxy-6-phenylpyrimidines a€ords the required pyrimidin-4-ones in high yields. Semi-empirical quantum
mechanical calculations show how the nature of the 5-substituent in¯uences the equilibrium between the 1H- and 3H-tautomeric
forms, and the rotational freedom about the bond connecting the 6-phenyl group and the pyrimidine ring. Both of these factors
may in¯uence the biological properties of these compounds. # 2000 Elsevier Science Ltd. All rights reserved.
Introduction
N-methylformamide as a triply hydrogen-bonded duplex
(Fig. 1) in which one molecule of the 3H-tautomer
forms three hydrogens bonds to another molecule of the
1H-tautomer (Fig. 1) in the manner of a Watson±Crick
cytosine±guanine base pair. Indeed the lengths of the
H-bonds are almost precisely those in a cytosine±
guanine duplex in DNA.6 Thus bropirimine and poten-
tially other molecules which possess a 2-aminodiazin-4-
one fragment, and which display a divergent range of
biological properties, have the potential to recognise
guanine or cytosine residues in single-stranded RNA
or DNA by Watson±Crick bonding or base sequences
in double-stranded DNA by Hoogsteen bonds. Poss-
ibly, the immunomodulatory properties of bropirimine
are triggered by these encounters (discussed in ref 6).
Even if this is not the case, the crystal structure high-
lights two characteristics of bropirimine, either or both
of which may be implicated in its biological properties:
®rstly, its array of potential hydrogen bond donors
and acceptors; and secondly, its ability to adopt one of
two tautomeric forms in response to its molecular
environment.
Synthesis of the small molecule immunomodulatory
agent 2-amino-5-bromo-6-phenylpyrimidin-4(3H)-one
(bropirimine: 5) was ®rst reported in 1975.1 The drug,
which can induce endogenous liberation of inter-
ferons2,3 and provoke cytokine release leading to
enhanced natural killer cell activity,4 has been devel-
oped primarily as an alternative to the immunothera-
peutic agent BCG vaccine for the oral treatment of
bladder cancer.5 This type of agent might also have uses
to reduce the frequency of relapses in patients with
progressive multiple sclerosis.
An intriguing feature of bropirimine is exhibited in its
X-ray crystal structure. Bropirimine crystallises from
*Corresponding author. Tel.: +44-115-951-3414; fax: +44-115-951-
3412; e-mail: malcolm.stevens@nottingham.ac.uk
yPart 28. Stevens, M. F. G.; Phillip, K. S.; Rathbone, D. L.; O'Shea,
D. M.; Queener, S. F.; Schwalbe, C. J.; Lambert, P. A. J. Med. Chem.
1997, 40, 1986.
0968-0896/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00017-1

740
D. R. Hannah et al. / Bioorg. Med. Chem. 8 (2000) 739±750
Figure 1. Hydrogen bonding between 2-amino-5-bromo-6-phenylpyrimidin-4(3H)-one and 2-amino-5-bromo-6-phenylpyrimidin-4(1H)-one in the
crystal structure of the N-methylformamide solvate of bropirimine 5 (from ref 6). The 4(3H)-one tautomer is shown (left), the 4(1H)-one tautomer
(right) and the molecule of N-methylformamide (above).
The presence of a halogen atom at C(5) is crucial for
activity in this series of pyrimidinones.2±4 However, it is
dicult to provide an overall interpretation for the
structure±activity investigations that have been per-
formed on these compounds, since the results are
dependent on the biological activity sought.3
that the halide component should be electron de®cient.
However, under the basic conditions normally employed
in Suzuki couplings the halopyrimidinone substrate
would be in an electron rich anionic state. To o€set this
potential disadvantage it is known that 5-halo sub-
stituents in pyrimidin-4-ones are labile. For example,
debromination of bropirimine occurs readily in the
presence of strong nucleophiles such as aqueous alkali
and hydrazine hydrate12 where nucleophilic attack at
the electrophilic C(6) site presages displacement of the
halogen atom at C(5); also, debromination occurs
readily by thermal homolysis.13 However, the afore-
mentioned mechanisms are quite distinct from that of
a Suzuki reaction which would require the in situ for-
mation of an organopalladium (pyrimidinone-Pd-halide)
intermediate.
From the crystal structure of bropirimine it is apparent
that steric clashes between the bromine atoms and the
phenyl groups in the duplex are ameliorated by in-plane
bending so that both the Br-C(5)-C(6) and Ph-C(6)-C(5)
angles exceed 120ꢀ and by twisting between the pyr-
imidine and phenyl ring-planes of 52 and 47ꢀ in the 3H-
and 1H-tautomers, respectively (see Fig. 1).6
Replacement of the 5-bromo-6-phenyl fragment by a 5-
aryl-6-phenyl structure would be expected to induce
major perturbations in the disposition of substituents
around the pyrimidine ring and modulate both the
hydrogen-bonding potential of the molecules and the
relative stability of the 1H- and 3H-tautomers. 2-
Amino-5,6-diarylpyrimidin-4(3H)-ones have been little
studied: previous syntheses have employed the tradi-
tional approach to 2-aminopyrimidinone synthesis of an
initial condensation between guanidine carbonate and
a 3-carbon synthon such as diphenylacrylonitrile7 or
diphenylcyclopropenone.9 The Suzuki cross-coupling9
procedure o€ers the potential of a more adaptable
method to introduce a range of substituted phenyl
residues into the 5-position of bropirimine. Although
5-halopyrimidines have been coupled with arylboronic
acids under palladium-catalysed conditions, these
earlier e€orts have focused on neutral halopyrimidine
substrates.10 In contrast, 2-amino-5-halo-6-phenylpyr-
imidin-4-ones are amphoteric; the pKa values corre-
sponding to protonation at N(1) and deprotonation at
N(3) of bropirimine 5 are 3.18 and 8.53, respectively.11
Normally a requirement for ecient Suzuki coupling is
Then there is the steric question. Approach of an aryl-
boronic acid to the organopalladium intermediate
would be impeded by the ¯anking carbonyl group at
C(4) (or its enolate equivalent) and the bulky sub-
stituent at C(6). As the synthetic objectives of this work
were to develop high-yielding cross-couplings from
bropirimine and related pyrimidin-4-ones to yield their
5-aryl derivatives, this seemed to be a major impediment
to a successful outcome that would demand exploration
of a wide range of coupling conditions.
Chemistry
Arylation of 2-amino-5-halo-6-(substituted)pyrimidin-
4-ones
The halopyrimidinones required for this work were
synthesised in an ecient two-step process from guani-
dine carbonate and b-ketoesters (1: R=Ph, Me, Et).
The pyrimidinones 2±4 formed in near quantitative

D. R. Hannah et al. / Bioorg. Med. Chem. 8 (2000) 739±750
741
yields were either brominated (bromine in acetic acid)
or iodinated (iodine in chloroform±aqueous sodium
hydroxide) to a€ord a series of 5-bromo- and 5-iodo-
pyrimidinones 5±10 (Scheme 1).
When bropirimine (5) was used as substrate in DMF the
products were the required 5,6-diphenylpyrimidinone
(12) (17%) and the debrominated product (2) (83%)
(Table 1: Entry 1): use of triethylamine as base instead
of sodium carbonate in DMF gave no coupled product
(Table 1: Entry 2) whereas in toluene±ethanol a mixture
comprising the coupled product 12 (9%), the debromi-
nated pyrimidinone (2) (19%) and unreacted starting
material (5) (72%) was formed (Table 1: Entry 3).
Compound 12 was identical to an authentic sample
prepared from diphenylcyclopropenone and guanidine
carbonate.8 Although disappointing these initial results
showed, contrary to expectation, that the more steri-
cally-challenged bromophenylpyrimidinone 5 was a
marginally better substrate for coupling than the bro-
momethylpyrimidinone 7; also use of toluene±ethanol
was associated with less debromination. Based on this
information a series of couplings was performed in
toluene±ethanol in an e€ort to maximise the yield of
In order to minimise the steric impediment imposed by
the 6-substituent initial Suzuki couplings employed the
5-bromo-6-methylpyrimidinone (7). This was reacted
with benzeneboronic acid in degassed re¯uxing DMF in
the presence of tetrakis(triphenylphosphine)palladium
(0) [Pd(PPh3)4] (3.5 mol%) and aqueous sodium car-
bonate (3 mol equiv). The products were the desired
6-methyl-5-phenylpyrimidinone (11) (10%) and the
debrominated methylpyrimidinone (3) (85%). Applying
alternative `standard' conditions used by Wang and
Haseltine,14 which employ toluene±ethanol as solvent
and aqueous sodium carbonate (2 mol equiv) as base,
only a trace of coupling (TLC) was observed and
>90% starting material was recovered.
Scheme 1. Reagents: (a) EtOH re¯ux; (b) Br2 in AcOH (X=Br), or I2 in CHCl3-aq. NaOH (X=I); (c) Suzuki cross-coupling (see Table 1).
Table 1. Products of cross-coupling reactions on 2-amino-5-halo-4,6-(disubstituted)pyrimidines: variations in reaction conditionsa
Entry Starting pyrimidine
Boronic acid
Solventb,g
Catalyst
Base
Hours re¯ux
Products (% yield)c
Coupling Dehalogenation Unchanged
1
2
3
4
5
6
7
8
5
5
5
5
6
5
5
5
5
14
14
14
14
14
14
14
14
17
17
17
17
benzene
benzene
benzene
benzene
benzene
benzene
benzene
benzene
benzene
benzene
A
A
B
B
B
B
B
B
B
B
B
C
B
D
B
C
C
C
C
C
C
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)2
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd(dppf)(OAc)2 K3PO4
Pd(dppf)(OAc)2 K3PO4
Pd(OAc)2
Pd(PPh3)4
Pd(dppf)(OAc)2 K3PO4
Pd(dppf)(OAc)2 K3PO4
Pd(dppf)(OAc)2 K3PO4
Pd(dppf)(OAc)2 K3PO4
Pd(dppf)(OAc)2 K3PO4
Pd(dppf)(OAc)2 K3PO4
Na2CO3
NEt3
24
24
24
24
24
24
24
18
76
18
18
24
6
14
18
24
90
5
24
16
16
12 (17)
12 (0)
12 (9)
12 (4)
12 (1)
12 (15)
12 (16)
12 (10)
12 (21)
18 (75)d
19 (24)
19 (0)
2 (83)
2 (29)
2 (19)
2 (18)
2 (30)
2 (31)
2 (25)
2 (23)
5 (0)
5 (48)
5 (72)
5 (56)
6 (38)
5 (34)
5 (37)
5 (31)
Na2CO3
NaHCO3
Na2CO3
Na2CO3
Na2CO3
Na2CO3
Na2CO3
Na2CO3
Na2CO3
9
2 (46)
e
5 (20)
f
10
11
12
13
14
15
16
17
18
19
20
21
2,4-dichlorobenzene
2,4-dichlorobenzene
2,4-dichlorobenzene
2,4-dichlorobenzene
3-thienyl
16 (11)
16 (24)
16 (29)
16 (13)
16 (89)
16 (16)
16 (18)
16 (22)
16 (6)
14 (58)
14 (71)
14 (67)
14 (77)
14 (1)
14 (24)
14 (20)
17 (67)
17 (0)
19 (0)
19 (0)
20 (1)
Na2CO3
Na2CO3
3-thienyl
20 (50)d
21 (55)d
19 (0)
4-methoxybenzene
2,4-dichlorobenzene
4-methoxybenzene
4-chlorobenzene
3-nitrobenzene
21 (72)d
22 (78)d
23 (75)d
16 (4)
16 (3)
17 (0)
17 (0)
aSee Experimental for details of syntheses.
bIncluding water; see Experimental for details.
1
cRelative yields determined by H NMR analysis of partially puri®ed (column chromatography) reaction mixtures.
dIsolated by crystallisation from the reaction mixture.
eDe-halogenated product 2 detected (TLC) in the reaction mixture.
fStarting material 14 detected (TLC) in the reaction mixture.
gSolvents: A, dimethylformamide; B, toluene±ethanol; C, 1,2-dimethoxyethane; D, 95% ethanol.

742
D. R. Hannah et al. / Bioorg. Med. Chem. 8 (2000) 739±750
the required diphenylpyrimidinone 12. Product mixtures
were puri®ed by column chromatography and yields of
pyrimidine products were analysed by 1H NMR.
Results are summarised in Table 1. Replacement of the
sodium carbonate base with sodium bicarbonate led to
a reduced yield of the coupled product (Table 1: Entry
4). Surprisingly, use of the 5-iodo-6-phenylpyrimidinone
(6) as the halopyrimidinone component gave only a
trace (TLC) of the coupled product 12 (Table 1: Entry
5). In all cases (above) unreacted starting material was
recovered and considerable dehalogenation to 2 occurred
in a process that must be mediated by palladium since 5
is stable in the reaction medium (re¯uxing toluene±
ethanol containing sodium carbonate). Batch variation
in the tetrakis(triphenylphosphine)-palladium(0) catalyst
had a minimal bearing on the outcome of the coup-
lings (Table 1; Entries 3, 6, 7); in situ generation
of the presumed catalytic species, bis(triphenyl-
phosphine)palladium(0) [Pd(PPh3)2], formed by reaction
of palladium diacetate and triphenylphosphine15 did not
improve the outcome (Table 1: Entry 8). The highest
yield of 12 (21%) was formed in toluene±ethanol over
76 h with periodic additions of more benzeneboronic acid,
sodium carbonate and catalyst but the ratio of coupling:
debromination still remained only 1:2.2 (Table 1: Entry
9). Presumably the poor yield of coupled product, even
under the latter non-catalytic conditions, re¯ects the
small proportion of neutral halopyrimidine substrate
available for oxidative addition to the palladium(0)
species involved in the catalytic cycle.16
Arylation of 2-amino-5-halo-4-methoxy-6-phenylpyrimidines
Scheme 2. Reagents: (a) POCl3, re¯ux; (b) NaOMe, MeOH; (c) guan-
idinium sulphate, NaOMe, re¯ux; (d) N-iodosuccinimide, CHCl3;
(e) Suzuki reactions (see Table 1); (f) 10 M HCl, MeOH, re¯ux.
To explore the possibility that pre-derivatisation of the
pyrimidinone moiety might facilitate more e€ective
cross-coupling to 5-arylpyrimidines the 2-amino-5-halo-
4-methoxy-6-phenylpyrimidines 14 and 17 were syn-
thesised and subjected to Suzuki couplings (Scheme 2).
Bropirimine 5 was ®rst converted to the 4-chloropyr-
imidine 13 (67%) in re¯uxing phosphorus oxychloride.
The product was accompanied by an insoluble poly-
meric material which was reconverted to bropirimine
in boiling 1 M-hydrochloric acid. Possibly, this un-
identi®ed material was formed by condensation of the
reactive chloro group of 13 with the 2-amino group [or
N(1) or N(3) atoms] of another molecule. Methanolysis
of 13 with sodium methoxide±methanol a€orded the
5-bromo-4-methoxy-pyrimidine 14. The corresponding
5-iodopyrimidine (17) was not available by this route
since the 5-iodopyrimidinone (6) underwent quantita-
tive de-iodination in phosphorus oxychloride. Instead,
3,3-bis(methylthio)-1-phenyl-2-propene-1-one (15) was
cyclised to 2-amino-4-methoxy-6-phenylpyrimidine (16)
with guanidine sulphate in sodium methoxide solution
according to the method of Chauhan and Junjappa.17
Iodination of 16 with N-iodosuccinimide in chloroform
gave the required 5-iodopyrimidine (17).
a€orded the 5,6-diphenyl-pyrimidine 18 in 75% yield
(after crystallisation) (Table 1: Entry 10). But per-
versely, these conditions were not general and although
2,4-dichlorobenzeneboronic acid did yield the required
5-(2,4-dichlorophenyl)pyrimidine (19) in poor yield
(24%), a pure sample was obtained only after repeated
chromatographic fractionations to remove starting
material 14 and debrominated product 16 (Table 1:
Entry 11). The structure of 19, which was presumably a
mixture of enantiomeric (atropisomeric) pyrimidines non-
interconvertible by restricted rotation about the pivotal
pyrimidine±dichlorophenyl bond,18 was con®rmed by
1H and 13C NMR spectra and mass spectrometry.
Aqueous 1,2-dimethoxyethane (DME) has been recom-
mended by Gronowitz as a substitute for toluene in
troublesome couplings;19 alternative bases including
thallium (I) hydroxide and tripotassium phosphate have
been reported to produce dramatic rate and yield
enhancements in the synthesis of hindered biaryls.20
Also the nature and quality of the palladium catalyst is
often decisive: thus Thompson et al. have exploited the
catalyst 1,10-bis(diphenylphosphino)-ferrocenepalladium
(II) acetate [Pd(dppf)(OAc)2], prepared in situ, for
cross-couplings of hindered halopyrazines and halopyr-
Results of Suzuki couplings on halomethoxypyrimidines
14 and 17 are recorded in Table 1. Reaction between the
5-bromo-4-methoxypyrimidine (14) and benzeneboronic
acid in toluene±ethanol with sodium carbonate base
and tetrakis(triphenylphosphine)-palladium(0) catalyst

D. R. Hannah et al. / Bioorg. Med. Chem. 8 (2000) 739±750
743
idines where tetrakis(triphenylphosphine)palladium(0)
had failed;21 and Campi et al. have shown that catalytic
palladium acetate [Pd(OAc)2] in degassed 95% ethanol
containing sodium carbonate base can e€ect ecient
cross-couplings of ortho-substituted benzeneboronic
acids.22
Exploration of some of these variables to improve yields
of the hindered dichlorophenylpyrimidine (19) met with
no success (Table 1: Entries 12±14). However, coupling
of 14 and 3-thienylboronic acid, which had given only a
trace of product 20 with the toluene±ethanol±sodium
carbonate±Pd(PPh3)4 combination, gave a 50% yield of
20 using DME and catalytic Pd(dppf)(OAc)2 (Table 1:
Entries 15, 16). Similarly the 5-(4-methoxyphenyl)-pyr-
imidine 21 was isolated (55%), together with debromi-
nated by-product 16 and unreacted starting material
using the latter conditions (Table 1: Entry 17). Reac-
tions with the 5-iodo-4-methoxypyrimidine (17) and
four boronic acids were attempted in DME-tripo-
tassium phosphate-Pd(dppf)(OAc)2. Again, 2,4-dichloro-
benzeneboronic acid failed to give the desired product 19
(Table 1: Entry 18). However, reactions with 4-methoxy-,
4-chloro- and 3-nitrobenzeneboronic acids under these
conditions furnished excellent yields of the 5-aryl-6-
phenylpyrimidines 21±23, respectively (Table 1: Entries
19±21): not only was cross-coupling more rapid between
the 5-iodopyrimidine (17) and 4-methoxybenzene-boro-
nic acid than with the 5-bromo-pyrimidine (14), but also
selectivity for coupled (21) versus dehalogenated (16)
product was enhanced from 3:1, starting with 14, to 12:1
from 17.
Scheme 3. Reagents: (a) NEt3, DMF, 1-hexyne, Pd(PPh3)2Cl2-CuI,
50 ꢀC; (b) NEt3, 1-hexyne, Pd(PPh3)2Cl2-CuI, 50 ꢀC.
butyl chain and with the 20-proton on the 4-phenyl
group. Consistent with our experience with Suzuki
reactions the 4-alkoxy-5-iodopyrimidine (17) was a
better substrate for a palladium catalysed alkynylation.
Although the products from coupling of (17) and 1-
hexyne could not be separated completely, detection, by
1H NMR, of the novel alkynylpyrimidine 27 (27%)
together with recovered starting material 17 (55%),
con®rmed the potential of this route.
Quantum-Mechanical Calculations
Choice of modelling methods
The quantum-mechanical studies on bropirimine and
analogous structures considered the conformational
¯exibility of the 6-phenyl group, and the terms involved
in the formation of a hydrogen-bonded tautomeric
duplex, both in gas and solution phases. These terms
may be combined into a free energy cycle (Fig. 2). For
this work we chose to use semi-empirical PM3 calcula-
tions25 for gas-phase geometry optimizations and heat
of formation calculations, combined with the SM3 con-
tinuum solvation model for the prediction of free ener-
gies of hydration. The PM3 method has been shown to
model hydrogen bonding convincingly,26,27 and the
SMx approach has been found to accurately predict free
energies of solvation and partition coecients.28±31
Hydrolysis of the 4-methoxy-5-phenylpyrimidine (18) in
methanolic hydrochloric acid gave the pyrimidinone 12
(96%), identical to the sample originally prepared in
poor yield directly from a Suzuki reaction on bropir-
imine. Similarly the 5-(4-methoxyphenyl)pyrimidinone
(24) was formed (97%) by hydrolysis of the methoxy
precursor 21.
Alkynylation of 2-amino-5-iodo-6-phenylpyrimidines
Our preliminary results show that it is also possible to
exploit Heck reactions to introduce alkynyl groups into
sterically-hindered 2-amino-5-halo-4,6-(disubstituted)-
pyrimidines. Interaction of the iodopyrimidinone 6 with
1-hexyne was conducted under standard Heck condi-
tions 23 employing a mixture of triethylamine, DMF (to
dissolve the pyrimidinone), and a mixed Pd(PPh3)2Cl2-
CuI catalyst at 50 ꢀC. Identi®ed products included
starting iodopyrimidinone 6 (64%), deiodinated mate-
rial 2 (11%) and a mixture of non-acidic pyrimidines
which could not be separated completely. The major
non-acidic product had spectroscopic characteristics
consistent with a furo[2,3-d]pyrimidine structure 26.
Presumably the target 5-hexynylpyrimidinone (25)
cyclised to the furopyrimidine in the reaction conditions
(Scheme 3). Similar intramolecular Michael-type cycli-
sations have been reported in 5-alkynyluracils.24 The
proposed bicyclic structure was con®rmed by a NOE
experiment which showed the expected enhancements
between H(5) and the a- and b-methylene protons of the
Conformational analysis
One possible route by which variation in the 5-sub-
stituent in 2-aminopyrimidin-4-ones might a€ect bio-
logical activity is through its e€ect on the rotational
freedom of the 6-phenyl group. The PM3 calculations
indicated that up to the point where steric clashes
began, the phenyl group showed almost free rotation.
The e€ect of the variation in the bulk of the 5-sub-
stituent on the range of dihedral angles available for an
1
energy penalty <5 kcal mol is shown in Table 2. All
angles are accessible with 5-H and 5-F but the other
halogen groups restrict rotation to angles >20±40ꢀ
depending on their size. A 5-Me substituent has an
e€ect intermediate between 5-Br and 5-I. As predicted, a
5-Ph group restricts the 6-phenyl substituent to angles

744
D. R. Hannah et al. / Bioorg. Med. Chem. 8 (2000) 739±750
Figure 2. Free energy cycle used to calculate thermodynamic quantities. The quantities are ÁGtaut, free energy of tautomerism (3H to 1H); ÁGsolv
free energy of solvation; ÁGpair, free energy of hydrogen-bonded pair formation.
,
Table 2. Conformational analysis of 2-amino-6-phenyl-5-(substituted)-
pyrimidin-4-ones
tomer remains accessible providing the environmental
conditions are suitable. Theoretical predictions of
tautomer stabilities of a series of 5,6-diphenyl-pyr-
imidinones bearing substituents in the para position of
the 5-phenyl group (Table 3) show the relative stabilities
of the 1H-tautomers decrease in the order: OMe>
NO2>H>Cl, with no obvious relationship to the elec-
tronic characteristic of the substituent.
Angle (ꢀ)a
Thermodynamic analysis of base-pairing
1H-tautomer
3H-tautomer
On the basis of crystal structure data of bropirimine, the
1H-tautomers of the 2-amino-5-(substituted)-6-phenyl-
pyrimidin-4-ones can be regarded as being prototypical
targets for `base-pairing' recognition by 3H-tautomers,
and vice-versa. The calculation of free energies of pair-
ing gives an insight into the strength of this interaction,
and how it is a€ected by changes in the 5- and 6-sub-
stituents. In the gas phase, the free energy of pairing for
the 5-halo derivatives varies from 0.3 kcal mol 1 (for F)
H
F
Cl
0
0
27
0
0
21
Br (bropirimine)
I
Me
Ph
34 (47)b
39
39
32 (52)b
34
30
70
65
aMinimum dihedral angle between the 6-phenyl group and the pyr-
1
imidine rings accessible for an energy penalty <5 kcal mol
.
bBond angles in brackets are taken from the X-ray crystal structure of
the N-methylformamide solvate of bropirimine 5 (ref 6).
1
to 1.4 kcal mol (Cl). For the 5-aryl compounds,
both strongly electron-donating (OMe) and with-
drawing (NO2) attached groups disfavour pair forma-
tion. In solution, all free energies of pairing are positive
because the free energy of solvation of the individual
partners is greater than that of the H-bonded pair: the
weakest interactions are made by the 5-¯uoro and 5-(4-
methoxyphenyl) derivatives, and the strongest by the 5-
chloro compound.
>65±70ꢀ. In all cases the molecular modelling predicts
that the 1H-tautomers show more restricted rotation
(larger dihedral angle) than their 3H-counterparts pre-
sumably because of the steric clash between the pyr-
imidine 1H and the ortho hydrogen on the 6-phenyl
residue. This result contrasts with the dihedral angle in
the crystallographic structure of bropirimine 5 which is
greater in the 3H-tautomer (Fig. 1). Possibly perturba-
tions induced by the solvating N-methylformamide
within the crystal lattice account for these di€erences.
Conclusions
Attempts to rationalize the broad chemotherapeutic
properties of bropirimine and other bio-active molecules
bearing the 2-aminopyrimidin-4-one pharmacophore, in
terms of their abilities to participate in base-pairing
encounters, is thwarted by the fragmented nature of the
published bio-data.2±4 However, the following trends
can be discerned: for antitumour activity, bromine is
preferred to iodine at C(5), and phenyl to methyl at
C(6); for antiviral properties in the 6-phenyl derivatives,
activity of the 5-halo series follows the order
Cl>Br>I>F and groups smaller than chloro or larger
than propyl are disfavoured; when interferon induction
is considered as the biological end-point, most data ®t
the trends observed for antiviral activity.
Thermodynamic analysis of tautomerism
The free energy components of the thermodynamic
cycle (Fig. 2) are shown in Table 3. In the gas phase, the
3H-tautomers are always predicted to be more stable
than the 1H-tautomers. This is also observed in solu-
tion, but to a lesser extent, as the 1H-forms show more
favourable free energies of solvation. Amongst the 5-
halo derivatives, the relative stability of the 1H-tauto-
mers decreases in the order: F>Cl>I>Br. Despite this
ranking, even for a 5-Br substituent, it is clear from the
crystal structure data on bropirimine6 that the 1H-tau-

D. R. Hannah et al. / Bioorg. Med. Chem. 8 (2000) 739±750
745
Table 3. Free energy components in formation of hydrogen-bonded tautomeric pairs of 2-amino-6-phenyl-5-(substituted)pyrimidin-4-onesa,b
R
H
R1
H
Gf,(g)
ÁGpair,(g)
ÁGsolv
ÁGpair,(sol)
ÁGtaut,(g)
ÁGtaut(sol)
Ð
0.8
Ð
12.2
6.59
3.0
7.13(1H)
13.72(3H)
19.80(pair)
23.21(1H)
19.60(3H)
29.83(pair)
F
H
H
0.3
1.4
1.1
0.5
0.8
0.3
0.2
0.1
1.0
14.2
12.0
12.4
12.1
12.1
13.5
14.2
13.6
13.2
4.51
5.48
6.26
5.99
5.04
6.20
3.88
5.62
8.40
2.6
3.3
4.7
4.2
2.4
3.7
1.5
3.1
5.6
48.74(1H)
53.22(3H)
99.81(pair)
24.25(1H)
22.34(3H)
32.69(pair)
Cl
Br
I
12.83(1H)
18.31(3H)
30.67(pair)
23.03(1H)
20.90(3H)
30.51(pair)
H
2.61(1H)
3.65(3H)
0.28(pair)
23.34(1H)
21.83(3H)
31.71(pair)
H
23.14(1H)
17.15(3H)
41.62(pair)
21.85(1H)
20.02(3H)
29.31(pair)
Me
Ph
Ph
Ph
Ph
H
17.53(1H)
22.57(3H)
39.11(pair)
21.26(1H)
18.63(3H)
26.98(pair)
H
15.96(1H)
9.76(3H)
27.21(pair)
23.29(1H)
20.82(3H)
30.34(pair)
OMe
NO2
Cl
26.59(1H)
30.47(3H)
55.01(pair)
24.57(1H)
22.18(3H)
32.80(pair)
2.73(1H)
2.89(3H)
1.52(pair)
26.33(1H)
23.82(3H)
36.39(pair)
9.93(1H)
1.53(3H)
11.94(pair)
24.37(1H)
21.92(3H)
31.99(pair)
1
aValues in kcal mol
.
bÁGpair;ꢁg†=Gf;ꢁg†(pair) Gf;ꢁg†(1H) Gf;ꢁg†(3H); ÁGpair;ꢁsol†=ÁGpair;ꢁg†+ÁGsolv(pair) ÁGsolv(3H) ÁGsolv(1H); ÁGtaut;ꢁg†=Gf;ꢁg†(1H) Gfꢁg†(3H);
ÁGtautꢁsol†=ÁGtaut;ꢁg†+ÁGf;ꢁsol†(1H) ÁGf;ꢁsol†(3H).
In the present study we have considered a range of
compounds with di€ering conformational and electronic
characteristics which may contribute to their recognition
capabilities, possibly at speci®c DNA sequences. The most
bio-active agents identi®ed in the literature are bropir-
imine 5 and its 5-chloro congener which are also calcu-
lated to have a propensity for strong H-bonding
interactions. An ability to adopt the less favoured 1H-
tautomeric arrangement does not seem to be a critical
factor. However, the structural characteristics optimizing
H-bonding in 2-amino-5-halo-6-phenylpyrimidin-4-ones
in general also dictate that this type of molecule is almost
totally insoluble in water since the base-pairing phe-
nomenon in the solid state e€ectively presents only
hydrophobic halogen and phenyl groups to the aqueous
environment.11 Overall, despite its unfavourable phar-
maceutical characteristics, the modelling data supports the
selection of bropirimine 5 as the agent of clinical choice.
In our synthetic work we have shown that sterically-
hindered 2-amino-5-halo-6-phenylpyrimidin-4-ones are
only poor substrates for palladium catalysed couplings
but that 2-amino-5-iodo-4-methoxy-6-phenylpyrimidine
(17) can be converted eciently to 5-aryl derivatives
employing, optimally, a combination of an arylboronic
acid, Pd(dppf)(OAc)2 catalyst and tripotassium phos-
phate base in aqueous 1,2-dimethoxyethane as solvent.
The 4-methoxypyrimidines can then be processed
hydrolytically to the required 2-amino-5-aryl-6-phenyl-
pyrimidin-4-ones. This process o€ers a route to novel
structures which may facilitate a more detailed SAR
analysis of this intriguing series of compounds.

746
D. R. Hannah et al. / Bioorg. Med. Chem. 8 (2000) 739±750
Experimental
sodium carbonate (2 mol equiv) as a base, only a trace
of coupling (TLC) was observed and >90% starting
material 7 was recovered.
General methods
All melting points were recorded on a Gallenkamp
melting point apparatus, and are uncorrected. IR spectra
were taken on a Mattson 2020 GALAXY Series FT-IR
2-Amino-5,6-diphenylpyrimidin-4(3H)-one (12). Using
the conditions described (above) for the synthesis of
pyrimidinone 11, the bromopyrimidinone 5 was con-
verted to the debrominated pyrimidinone 2 (83%) and
the required diphenylpyrimidinone 12 (17%). A pure
sample of 12 had mp 312±314 ꢀC (lit.,7 mp 319 ꢀC); dH
([2H6]DMSO) 6.64 (2H, br s, NH2), 6.86±7.02 (2H, m,
ArH), 7.06±7.22 (8H, m, ArH) and 11.12 (1H, br s,
NH); dC([2H6]DMSO) 113.2 (C-5), 126.1 (CH, ArC),
127.5 (2ÂCH, ArC), 127.6 (2ÂCH, ArC), 128.2 (CH,
ArC), 129.4 (2ÂCH, ArC), 131.6 (2ÂCH, ArC), 135.7
(C, ArC), 139.7 (C, ArC), 154.3 (C), 162.1 (C) and 162.7
(C); HRMS calcd for C16H13N3O 263.1058, found
263.1038.
1
spectrometer as KBr discs. H, 13C spectra were recor-
ded on a Bruker ARX250 spectrometer at 250.1 and
62.9 MHz, respectively, in solvents as speci®ed, with
tetramethylsilane or residual protic solvents as internal
standard, J values being in Hz. Low resolution mass
spectra were recorded on a Micromass Platform (AP+,
ES+). High resolution mass spectrometry (HRMS) were
performed by the EPSRC Mass Spectrometry Service
Centre, Swansea, UK. Silica gel TLC was performed on
60F-254 pre-coated sheets (E. Merck) and column
chromatography on silica gel C60 (60±120 mesh). Ele-
mental analyses were performed by Knoll Pharmaceu-
ticals Physical Chemistry Department, Nottingham,
UK. All solvents used in palladium catalysed cross-
coupling reactions were purged with nitrogen before use.
Synthesis of 2,4-disubstituted-5-halo-6-phenylpyrimidines.
The following pyrimidines were prepared.
Arylation of 2-amino-5-halo-6-(substituted)pyrimidin-4-
ones by cross-coupling reactions. Results of cross-coupl-
ing reactions between 2-amino-5-bromo-6-phenylpyr-
imidin-4(3H)-one (5) and 2-amino-5-iodo-6-phenyl-
pyrimidin-4(3H)-one 6 and benzeneboronic acid under a
range of conditions are summarised in Table 1 (Entries
1±9). Products were isolated by silica gel column
chromatography, with ethyl acetate:methanol:acetic
acid (16:4:1) as eluent, or the mixed pyrimidinone
2-Amino-5-bromo-4-methoxy-6-phenylpyrimidine (14).
Bropirimine 5 (3.725 g, 14.0 mmol) was heated to
re¯ux in phosphorous oxychloride (12 mL) for 45 min.
Following removal of the excess oxychloride in vacuo
the residue was triturated in aqueous ammonia-ice and
the solid collected. The product was warmed in di-
chloromethane (50 mL) and an insoluble by-product
was ®ltered o€. Evaporation of the dried (MgSO4) di-
chloromethane a€orded 2-amino-5-bromo-4-chloro-6-
phenylpꢀyrimidine 13 (2.68 g, ꢀ67%) as yellow crystals,
mp 135 C (lit.,12 mp 136±138 C).
1
products were analysed by H NMR spectroscopy in
[2H6]DMSO. Recorded below are details of syntheses
which gave fully characterised products.
A suspension of 13 (0.65 g) in anhydrous methanol
(25 mL) was re¯uxed (0.5 h) under nitrogen with a 25%
solution of sodium methoxide in methanol (2 mL).
Excess sodium methoxide was decomposed by solid
carbon dioxide and the mixture was diluted with water
(200 mL). The suspension was extracted with chloro-
form (3Â50 mL) and the combined organic extracts
were washed with water (25 mL), dried (MgSO4) and
reduced to give a white solid. The solid was crystallised
from hexane to furnish the methoxypyrimidine 14
(0.588 g, 93%), mp 174 ꢀC; nmax (KBr)/cm 1 3486, 3291,
3154, 1630, 1543, 1375, 1200 and 698; dH(CDCl3) 4.01
(3H, s, Me), 5.11 (2H, br s, NH2), 7.40±7.47 (3H, m,
ArH) and 7.61±7.65 (2H, m, ArH); dC(CDCl3) 54.8
(Me), 92.5 (C-5), 127.9 (2ÂCH, ArC), 128.8 (2ÂCH,
ArC), 129.2 (CH, ArC), 138.1 (C, ArC), 160.9 (C), 166.0
(C) and 166.8 (C); m/z M+ (ES+), 280, 282. Anal. calcd
for C11H10BrN3O: C, 47.2; H, 3.6; N, 15.0. Found: C,
47.1; H, 3.5; N, 14.8%.
2-Amino-6-methyl-5-phenylpyrimidin-4(3H)-one (11).
2-Amino-5-bromo-6-methylpyrimidin-4(3H)-one (7)
(0.414 g, 2.03 mmol) and tetrakis(triphenylphosphine)-
palladium(0) (0.080 g, 0.07 mmol) were stirred in DMF
(5 mL) for 5 min. Benzeneboronic acid (0.275 g, 2.25
mmol) and more DMF (10 mL) were added, followed
by sodium carbonate (0.550 g, 5.19 mmol) dissolved in
water (2 mL) and the mixture was heated to re¯ux for
24 h, under a gentle ¯ow of nitrogen. The cooled, ®l-
tered, mixture was evaporated in vacuo. The residue
was puri®ed by silica gel column chromatography, with
ethyl acetate:methanol:acetic acid (16:4:1) as eluent to
yield 2-amino-6-methylpyrimidin-4(3H)-one (3) (85%)
and the pyrimidinone 11 (0.040 g, 10%) as a white solid
with Rf 0.61. Compound 11 crystallised from ethanol
with mp >250 ꢀC (decomp.); nmax (KBr)/cm 3368,
1
3134, 2925, 1655, 1613, 1508, 1048 and 704; dH
([2H6]DMSO) 1.98 (3H, s, Me), 6.96 (2H, br s, NH2),
7.18±7.40 (5H, m, Ph) and 10.5±12.5 (1H, br s, NH); dC
([2H6]DMSO) 20.8 (Me), 114.1 (C-5), 126.8 (CH, ArC),
127.8 (2ÂCH, ArC), 130.6 (2ÂCH, ArC), 134.4 (C,
ArC), 153.1 (C), 157.0 (C) and 161.9 (C); HRMS calcd
for C11H11N3O 201.0902, found 201.0899. Anal. calcd
for C11H11N3O: C, 65.6; H, 5.7; N, 20.7. Found: C,
65.7; H, 5.5; N, 20.9%.
2-Amino-5-iodo-4-methoxy-6-phenylpyrimidine (17).
2-Amino-4-methoxy-6-phenylpyrimidine 16 (1.29 g,
6.41 mmol), prepared from 3,3-bis(methylthio)-1-phenyl-
2-propene-1-one and guanidine sulphate,17 was heated
to re¯ux with N-iodosuccinimide (1.88 g, 8.33 mmol) in
anhydrous chloroform (20 mL) for 2 h, under nitrogen.
After cooling, the red solution was diluted with more
chloroform (30 mL) and washed with water (2Â25 mL),
5% sodium metabisul®te solution (15 mL), water (10
Using the conditions of Wang and Haseltine,14 which
employ toluene±ethanol as a solvent and aqueous

D. R. Hannah et al. / Bioorg. Med. Chem. 8 (2000) 739±750
747
mL), then dried (MgSO4) and concentrated to dryness
in vacuo to provide the methoxypyrimidine 17 (2.07 g,
99%) as a pale yellow solid, mp 179±180 ꢀC (from hex-
7.09 (1H, dd, J=8 and 2, 50-H (5-Ar)), 7.17±7.31 (5 H,
m, Ph) and 7.40 (1H, d, J=2, 30-H (5-Ar)); dC(CDCl3)
54.0 (Me), 107.5 (C-5), 126.9 (CH, ArC), 127.9 (2ÂCH,
ArC), 128.6 (2ÂCH, ArC), 128.7 (CH, ArC), 129.1
(CH, ArC), 132.8 (C, ArC), 133.6 (CH, ArC), 133.7 (C,
ArC), 136.1(C, ArC), 138.1 (C, ArC), 162.1 (C), 165.7
(C) and 168.4 (C); MS (ES+) calcd for C17H13Cl2N3O
346, 348, 350, found 346, 348, 350.
1
ane); nmax (KBr)/cm 3476, 3150, 1628, 1545, 1362,
1198, 1005 and 696; dH(CDCl3) 3.96 (3H, s, Me), 5.46
(2H, br s, NH2) and 7.42±7.54 (5H, m, Ph); dC (CDCl3)
54.9 (Me), 64.9 (C-5), 127.9 (2ÂCH, ArC), 128.6
(2ÂCH, ArC), 129.0 (CH, ArC), 140.5 (C, ArC), 162.2
(C), 168.7 (C) and 170.2 (C); m/z M+(AP+), 328. Anal.
calcd for C11H10IN3O: C, 40.3; H, 3.0; N, 12.9. Found:
C, 40.7; H, 3.0; N, 12.9%.
2-Amino-4-methoxy-6-phenyl-5-(3-thienyl)pyrimidine (20).
Palladium acetate (5.1 mg, 0.023 mmol) and 1,10-
(diphenylphosphino)ferrocene (dppf) (16.6 mg, 0.03
mmol) were heated in degassed 1,2-dimethoxyethane
(DME) (3 mL) for 15 min, under nitrogen. To the
cooled mixture 2-amino-5-bromo-4-methoxy-6-phenyl-
pyrimidine (14) (0.210 g, 0.75 mmol), thiophene-3-
boronic acid (0.106 g, 0.83 mmol) and tripotassium
phosphate (0.318 g, 1.5 mmol) were added, along with
DME (3 mL) and water (1 mL), and the mixture was
heated under gentle re¯ux for 22 h, under nitrogen. The
reaction mixture was cooled, diluted with water (10 mL)
and extracted with diethyl ether (3Â25 mL). The com-
bined ethereal extracts were washed with water (10 mL),
saturated brine (10 mL), dried (MgSO4) and reduced in
vacuo to leave a pale brown solid. Succesive crystal-
lisations of the brown solid from hexane±ethyl acetate
(5:1), then acetone, yielded the thienylpyrimidine 20
(0.055 g, 26%) as a crystalline white solid, mp 206±
Arylation of 2-amino-5-halo-4-methoxy-6-phenylpyrimi-
dines by cross-coupling reactions. A summary of the
results of cross-coupling reactions between 2-amino-
5-bromo-4-methoxy-6-phenylpyrimidin-4(3H)-one (14)
and 2-amino-5-iodo-4-methoxy-6-phenylpyrimidin-4(3H)-
one (17) and benzeneboronic acids under a range of
conditions are recorded in Table 1 (Entries 10±21). Lis-
ted below are details of syntheses which gave the opti-
mum yields of fully characterised products.
2-Amino-4-methoxy-5,6-diphenylpyrimidine (18). A solu-
tion of sodium carbonate (0.276 g, 2.61 mmol) in water
(2 mL) was added to a mixture of 2-amino-5-bromo-4-
methoxy-6-phenylpyrimidine 14 (0.287 g, 1.02 mmol),
benzeneboronic acid (0.138 g, 1.13 mmol) and tetra-
kis(triphenylphosphine)palladium(0) (0.040 g, 0.035
mmol) in toluene (2 mL) and ethanol (2 mL) and the
mixture was heated to re¯ux for 18 h, under nitrogen.
After cooling, the mixture was extracted with chloro-
form (3Â25 mL), and the combined extracts were
washed successively with water (10 mL), 0.5 N-sodium
hydroxide (2Â5 mL), water (10 mL) and saturated brine
(10 mL). The organic layer was then dried (MgSO4) and
evaporated to give a grey solid which was crystallised
from 10% ethyl acetate±hexane to yield the pyrimidine
18 (0.212 g, 75%), mp 185±185.5 ꢀC as a crystalline white
solid; nmax(KBr)/cm 1 3379, 3171, 1638, 1568, 1543, 1371,
1198 and 696; dH(CDCl3) 3.88 (3H, s, Me), 5.14 (2H, br
s, NH2) and 7.08±7.28 (10H, m, 2ÂPh); dC(CDCl3) 53.9
(Me), 110.9 (C-5), 126.7 (CH, ArC), 127.7 (2ÂCH,
ArC), 127.8 (2ÂCH, ArC), 128.3 (CH, ArC), 129.4
(2ÂCH, ArC), 131.2 (2ÂCH, ArC), 134.4 (C, ArC),
138.6(C, ArC), 161.3 (C), 165.2 (C) and 168.5 (C); m/z
M+(ES+), 278. Anal. calcd for C17H15N3O: C, 73.6; H,
5.4; N, 15.0. Found: C, 73.5; H, 5.4; N, 15.0%.
207 ꢀC; nmax(KBr)/cm 3461, 3144, 1630, 1549, 1479,
1
1452, 1344 and 1198; dH(CDCl3) 3.93 (3H, s, Me), 4.99
(2H, br s, NH2), 6.82 (1H, dd, J=1 and 5, 40-H (thie-
nyl)), 6.96 (1 H, dd, J=1 and 3, 20-H (thienyl)), 7.18
(1H, dd, J=3 and 5, 50-H (thienyl)) and 7.23±7.32 (5 H,
m, Ph); dC(CDCl3) 53.9 (Me), 115.4 (C-5), 124.7 (CH,
thienyl), 125.2 (CH, thienyl), 128.2 (2ÂCH, ArC), 129.0
(CH), 130.2 (2ÂCH, ArC), 131.2 (CH), 135.3 (C), 140.4
(C), 162.9 (C), 166.2 (C) and 169.3 (C). Anal. calcd for
C15H13N3OS: C, 63.6; H, 4.55; N, 14.8. Found: C, 63.6;
H, 4.55; N, 14.6%.
2-Amino-4-methoxy-5-(4-methoxyphenyl)-6-phenylpyri-
midine (21). Prepared according to the method for the
preparation of 20, with palladium acetate (5.1 mg, 0.023
mmol), dppf (16.6 mg, 0.03 mmol), 2-amino-5-iodo-4-
methoxy-6-phenylpyrimidine (17) (0.245 g, 0.75 mmol),
4-methoxybenzeneboronic acid (0.126 g, 0.83 mmol)
and tripotassium phosphate (0.318 g, 1.5 mmol), in
DME (6 mL) and water (1 mL), with a reaction time of
24 h. The crude residue from the organic extract was
puri®ed by silica gel chromatography, with chloroform:
diethyl ether (4:1) as eluent, to provide impure 21 (0.20
g). Crystallisation from acetone yielded white crystals
(0.165 g, 72%), mp 174 ꢀC; nmax(KBr)/cm 1 3484, 3140,
1545, 1371, 1246, 1042, 831 and 696; dH(CDCl3) 3.78
(3H, s, Me), 3.90 (3H, s, Me), 5.05 (2H, br s, NH2),
6.76±6.80 (2H, m, ArH), 7.00±7.04 (2H, m, ArH) and
7.19±7.31 (5H, m, Ph); dC(CDCl3) 54.0 (Me), 55.1 (Me),
110.5 (C-5), 113.4 (2ÂCH, ArC), 126.4 (C, ArC), 127.8
(2ÂCH, ArC), 128.2 (CH, ArC), 129.4 (2ÂCH, ArC),
132.2 (2ÂCH, ArC), 138.8 (C, ArC), 158.3 (C), 161.1
(C), 165.1 (C) and 168.7 (C). Anal. calcd for C18H17
N3O2: C, 70.3; H, 5.6; N, 13.7. Found: C, 70.4; H, 5.5;
N, 13.55%.
2-Amino-5-(2,4-dichlorophenyl)-4-methoxy-6-phenylpyri-
midine (19). Prepared as above, from 14 (0.210 g, 0.75
mmol) and 2,4-dichlorobenzeneboronic acid (0.158 g,
0.83 mmol), tetrakis(triphenylphosphine)palladium(0)
(0.030 g, 0.026 mmol) and sodium carbonate (0.203 g,
1.92 mmol) in re¯uxing toluene±aqueous ethanol (24 h),
the crude reaction mixture was subjected to three
sequential silica gel column puri®cations with chloro-
form±diethyl ether (85:15) as eluent and gave a white
solid (0.093 g) that by 1H NMR was a 1:1 mixture of 19
and starting pyrimidine 14. Crystallisation of this mix-
ture from acetone (1 mL) gave a sample 90% enriched
1
in 19 (0.021 g); nmax(KBr)/cm 3408, 3183, 1645, 1568,
1368, 1051, 814 and 698; dH(CDCl3) 3.89 (3H, s, Me),
5.23 (2H, br s, NH2), 6.92 (1H, d, J=8, 60-H (5-Ar)),

748
D. R. Hannah et al. / Bioorg. Med. Chem. 8 (2000) 739±750
2-Amino-5-(4-chlorophenyl)-4-methoxy-6-phenylpyrimidine
(22). Prepared according to the method for the pre-
paration of 20, with palladium acetate (5.1 mg, 0.023
mmol), dppf (16.6 mg, 0.03 mmol), 2-amino-5-iodo-4-
methoxy-6-phenylpyrimidine (17) (0.245 g, 0.75 mmol),
4-chlorobenzeneboronic acid (0.130 g, 0.83 mmol) and
tripotassium phosphate (0.318 g, 1.5 mmol), in DME
(6 mL) and water (1 mL), with a reaction time of 16 h.
The crude reaction product was crystallised from
hexane:acetone (approx. 2:1) to a€ord 22 (0.182 g, 78%)
water (10 mL), basi®ed with solid sodium hydroxide
(2.24 g) and re-acidi®ed with acetic acid to pH 6. The
precipitate was collected, washed with water followed
by a small portion of acetone, and dried to furnish 24
(0.16 g, 97%) which formed white crystals, mp 322±
326 ꢀC (from ethanol±ethyl acetate); nmax(KBr)/cm
1
3383, 3108, 1651, 1582, 1518, 1242, 993 and 704;
dH([2H6]DMSO) 3.68 (3H, s, Me), 6.61 (2H, br s, NH2),
6.68±6.75 (2H, m, ArH), 6.89±6.95 (2H, m, ArH),
7.16±7.23 (5H, m, Ph) and 11.10 (1H, br s, NH);
dC([2H6]DMSO) 55.1 (Me), 112.9 (C-5), 113.1 (2ÂCH,
ArC), 127.5 (2ÂCH, ArC), 127.7 (C, ArC), 128.1 (CH,
ArC), 129.4 (2ÂCH, ArC), 132.5 (2ÂCH, ArC), 139.9
(C, ArC), 154.1 (C), 157.6 (C), 161.7 (C) and 162.9 (C).
Anal. calcd for C17H15N3O2: C, 69.2; H, 5.2; N, 14.2.
Found: C, 60.1; H, 5.4; N, 14.1%.
as pale brown crystals, mp 196±198 ꢀC; nmax(KBr)/cm
1
3493, 3140, 1634, 1566, 1543, 1370, 833 and 694;
dH(CDCl3) 3.90 (3H, s, Me), 5.16 (2H, br s, NH2), 7.01±
7.06 (2H, m, 2ÂArH) and 7.17±7.30 (7H, m, ArH);
dC(CDCl3) 53.7 (Me), 107.5 (C-5), 127.7 (2ÂCH, ArC),
127.9 (2ÂCH, ArC), 128.4 (CH, ArC), 129.4 (2ÂCH,
ArC), 131.2 (C, ArC), 133.1 (2ÂCH, ArC), 134.0 (C,
ArC), 138.8 (C, ArC), 162.2 (C), 164.9 (C) and 167.8
(C). Anal. calcd for C17H13ClN3O: C, 65.5; H, 4.5; N,
13.5. Found: C, 65.6; H, 4.75; N, 13.4%.
Alkynylations of pyrimidines (6) and (17). 2-Amino-6-
butyl-4-phenylfuro[2,3-d]pyrimidine (26). Triethylamine
(8 mL) was added to a stirred suspension of 2-amino-5-
iodo-6-phenylpyrimidin-4(3H)-one (6) (0.626 g, 2.00
mmol), bis(triphenylphosphine)palladium dichloride
(0.028 g, 0.04 mmol) and cuprous iodide (0.100 g, 0.52
mmol) in DMF (8 mL) under nitrogen. A deep green
solution was formed. 1-Hexyne (0.329 g, 4.00 mmol)
was added and the mixture was stirred at room tem-
perature for 40 h. The reaction mixture was then heated
to 50 ꢀC for 4 h. Concentration of the reaction mixture
in vacuo gave a brown solid which was suspended in
diethyl ether (50 mL) and the organic layer was washed
successively with 1 N-sodium hydroxide (2Â10 mL),
water (10 mL), 1 N-hydrochloric acid (2Â10 mL), water
(10 mL) and saturated brine (10 mL). The ethereal layer
was dried (MgSO4) and reduced in vacuo to give a
brown oil, which was puri®ed by silica gel column
chromatography with diethyl ether:dichloromethane
(1:10), to give a sample of the ¯uorescent furopyridine
2-Amino-4-methoxy-5-(3-nitrophenyl)-6-phenylpyrimidine
(23). Prepared according to the method for the pre-
paration of 20, with palladium acetate (5.1 mg, 0.023
mmol), dppf (16.6 mg, 0.03 mmol), 2-amino-5-iodo-4-
methoxy-6-phenyl-pyrimidine (17) (0.245 g, 0.75 mmol),
3-nitrobenzeneboronic acid (0.139 g, 0.83 mmol) and
tripotassium phosphate (0.318 g, 1.5 mmol), in DME
(6 mL) and water (1 mL), with a reaction time of 16 h.
Crystallisation of the crude reaction product from
hexane:acetone (approx. 5:2) yielded 23 (0.197 g, 75%)
as its 2:1 solvate with acetone, mp 184±185 ꢀC; nmax
1
(KBr)/cm 3322, 3198, 1640, 1545, 1346, 1200, 1065
and 775; dH(CDCl3) 2.18 (3H, s, Me, acetone), 3.91 (3H,
s, Me), 5.24 (2H, br s, NH2), 7.16±7.31 (5H, m, Ph),
7.32±7.39 (2H, m, ArH) and 8.00±8.09 (2H, m, ArH);
dC(CDCl3) 30.9 (Me±acetone), 54.1 (Me), 108.6 (C-5),
121.6 (CH, ArC), 126.2 (CH, ArC), 128.2 (2ÂCH,
ArC), 128.6 (CH, ArC), 128.8 (CH, ArC), 129.4
(2ÂCH, ArC), 136.5 (C, ArC), 137.5 (CH, ArC), 137.8
(C, ArC), 147.9 (C, ArC), 161.7 (C), 165.9 (C), 168.2 (C)
and 207 (CO, acetone). Anal. calcd for C17H13N4O3: C,
63.2; H, 4.8; N, 15.9. Found: C, 63.0; H, 4.8; N, 16.0%.
1
(26) (0.021 g, 4%): nmax (CDCl3 ®lm)/cm 3530, 3422,
2961, 2934, 1616, 1460, 1371 and 698; dH(CDCl3) 0.95
(3H, t, J=7, CH2CH2CH2CH3), 1.43 (2H, quintet, J=7,
CH2CH2CH2CH3), 1.68±1.79 (2H, m, CH2CH2CH2
CH3), 2.74 (2H, td, J=7 and 1, CH2CH2CH2CH3), 5.20
(2H, br s, NH2), 6.52 (1H, t, J=1, H-5), 7.47±7.56 (3H,
m, ArH) and 7.95±8.01 (2H, m, ArH).
Hydrolysis of 2-amino-5,6-diphenyl-4-methoxypyrimidine
(18). 10 N-Hydrochloric acid (5 mL) was added to a
solution of 2-amino-5,6-diphenyl-4-methoxypyrimidine
(18) (0.044 g, 0.16 mmol) in methanol (5 mL), and the
mixture was heated to re¯ux for 1 h. After cooling,
the solution was basi®ed with solid sodium hydroxide
(2.27 g) and then re-acidi®ed with acetic acid to pH 6.
The precipitate was collected, washed with water and
dried in vacuo to furnish the pyrimidinone 12 (0.040 g,
96%) as a white solid, mp 314±316 ꢀC (from ethanol)
identical (1H NMR) to the sample prepared by a palla-
dium catalysed coupling between the bromopyr-
imidinone 5 and benzeneboronic acid (see earlier).
2-Amino-5-(1-hexyn-1-yl)-4-methoxy-6-phenylpyrimidine
(27). 1-Hexyne (0.14 mL, 1.2 mmol) was added to
a mixture of 2-amino-5-iodo-4-methoxy-6-phenylpyr-
imidine (17) (0.327 g, 1.0 mmol), bis(triphenylphos-
phine)palladium dichloride (0.014 g, 0.02 mmol) and
cuprous iodide (0.050 g, 0.26 mmol) in degassed (argon)
triethylamine (10 mL). The reaction mixture was then
heated to 50±60 ꢀC for 16 h, concentrated to dryness in
vacuo, suspended in water (50 mL) and the mixture
extracted with dichloromethane (3Â25 mL). The com-
bined organic extracts were washed with water (10 mL),
dried (MgSO4) and evaporated to leave an oily brown
solid. Silica gel column puri®cation of the crude product
with diethyl ether±chloroform (1:10) as eluent, gave a
band of Rf 0.36 (0.255 g, 82% recovery) comprising
starting pyrimidine 17 and the alkynylpyrimidine 27
Hydrolysis of 2-amino-4-methoxy-5-(4-methoxyphenyl)-
6-phenylpyrimidine (21). The 6-phenylpyrimidine (21)
(0.17 g, 0.55 mmol) was heated to re¯ux in a mixture of
methanol (5 mL) and 10 N-hydrochloric acid (5 mL) for
2 h. After cooling the reaction mixture was diluted with
1
(2:1). The H NMR spectrum of the weakly ¯uorescent
alkyne 27 (at 254 nm) showed dH(CDCl3) 1.22 (3H, t,

D. R. Hannah et al. / Bioorg. Med. Chem. 8 (2000) 739±750
749
J=7, CH2CH2CH2CH3), 1.26±1.56 (4H, m, CH2CH2
CH2CH3), 2.38 (2H, t, CH2CH2CH2CH3), 3.97 (3H, s,
OMe), 5.57 (2H, br s, NH2), 7.39±7.45 (3H, m, ArH)
and 7.88±7.92 (2H, m, ArH).
and Clinical Uses; Khan, A., Hill, N. O., Dorn, G. L., Eds.;
Wadley Press: Dallas, 1980; pp 325±326; Skulnick, H. I.;
Weed, D. S.; Eidson, E. E.; Renis, H. E.; Wierenga, W.;
Stringfellow, D. A. J. Med. Chem. 1985, 28, 1864; Milas, L.;
Hersh, E. M.; Stringfellow, D. A.; Hunter, N. J. Natl. Cancer
Inst. 1982, 86, 139; Stringfellow, D. A.; Vanderberg, H. C.;
Weed, S. D. J. Interferon Res. 1980, 1, 1; Hamilton, R. D.;
Wynalda, M. A.; Fitzpatrick, F. A.; Teagarden, D. L.;
Hamdy, A. H.; Snider, B. G.; Weed, S. D.; Stringfellow, D. A.
J. Interferon Res. 1982, 2, 317; Li, H.; Wallace, T. L.; Wier-
enga, W.; Skulnick, H. I.; DeKoning, T. F. J. Biol. Response
Modi®ers 1987, 6, 44.
3. Wierenga, W. Pharmacol. Therapeutics 1985, 30, 67.
4. Lotsova, E.; Savary, C. A.; Pollack, S.; Hanna, N. Cancer
Res. 1986, 46, 5004; Li, L. H.; DeKoning, T. F.; Wallace, T. L.
Cancer Res. 1987, 47, 5894; Scheringa, M.; Ijzermans, J. N.
M.; Jeekel, J.; Marquet. Cancer Immunol. Immunother. 1990,
32, 251; Li, L. H.; Wallace, T. L.; Richard, K. A.; Tracey, D.
E. Cancer Res. 1985, 45, 532; Lotzova, E.; Savary, C. A.;
Stringfellow, D. A. J. Immunol. 1983, 130, 965.
Molecular modelling studies
The structure of bropirimine 5 was taken from the
reported X-ray crystal structure.6 The (solvating) N-
methylformamide in the unit cell was removed and
initial geometry optimizations were started on the pyr-
imidinone base-pair. Individual tautomer structures
were also extracted from this structure. Analogues of
bropirimine were constructed by substituting appro-
priate atoms or functional groups at the 5- and 6-posi-
tions. All structures were geometry optimized in the gas
phase at the PM3 level. and calculations run on Silicon
Graphics workstations using SPARTAN 4.0.32
5. Wierenga, W. In Immunomodulating Drugs, Georgiev, V. S.,
Yamaguchi, H., Eds.; New York Academy of Sciences: New
York, 1993; Vol. 695, pp 296±300; Sarosdy, M. F. In Immu-
nomodulating Drugs, Georgiev, V. S., Yamaguchi, H., Eds.;
New York Academy of Sciences: New York, 1993; Vol. 695,
pp 301±306; Sarosdy, M. F.; Lamm, D. L.; Williams, R. D.;
Moon, T. D.; Flanigan, R. C.; Drawford, E. D.; Wilks, N. E.;
Earhart, R. H.; Merritt, J. A. J. Urol. 1992, 147, 31; Sarosdy,
M. F.; Lowe, B. A.; Schellhammer, P. F.; Lamm, D. L.;
Graham, S. D.; Grossman, H. B.; See, W. A.; Peabody, J. O.;
Moon, T. D.; Flanigan, R. C.; Crawford, E. D.; Morganroth,
J. Urology 1996, 48, 21.
Gas heats of formation (HOF) were corrected for
vibrational entropy and enthalpy, and the zero point
energy (ZPE) was subtracted to arrive at a value for
ÁG(g) (eq. (1)). The vibrational values were calculated at
298.15 K from a frequency analysis within SPARTAN.
After a solvation calculation was run using SM3, the
reported energy (ESM3) was a combination of the gas
heat of formation plus the solvation free energy. Sub-
traction of the heat of formation from the SM3 energy
provides ÁGsolv (eq. (2)). Solution phase ÁG values
(ÁGsol) were obtained by adding ÁG(g) and ÁGsolv and
correcting for the di€erences in standard states (eqs (3)
and (4)).33
6. Stevens, M. F. G.; Schwalbe, C. H.; Patel, N.; Gate, E. N.;
Bryant, P. K. Anti-Cancer Drug Des. 1995, 10, 203.
7. Hitchings, G. H.; Russell, P. B.; Whittaker, N. J. Chem.
Soc. 1956, 1019
8. Eicher, T.; Franke, G.; Abdesaken, F. Tetrahedron Lett.
1977, 4067
9. Miyaura, N.; Yanagi, T.; Suzuki, A. Synth. Commun. 1981,
11, 513; Kalinin, V. N. Synthesis 1992, 413.
10. Crisp, G. T.; Macolino, V. Synth. Commun. 1990, 20, 413;
Gronowitz, S.; Hornfeldt, A.-B.; Kristjansson, V.; Musil, T.
Chem. Scr. 1986, 26, 305; Stavenuiter, J.; Hamzink, M.; van
der Hulst, R.; Zomer, G.; Westra, G.; Kriek, E. Heterocycles
1987, 26, 2711; Peters, D.; Hornfeldt, A.-B.; Gronowitz, S. J.
Heterocycl. Chem. 1990, 27, 2165.
11. Alpar, H. O.; Whitmarsh, S. J.; Ismail, H.; Slack, J. A.;
Belaid, K. A.; Stevens, M. F. G. Drug Development and
Industrial Pharmacy 1986, 12, 1795.
ÁG ˆ HOF ‡ Hvib ‡ Hrot ‡ Htran ZPE
ꢀg†
ꢀ1†
‰TꢀSvib ‡ Srot ‡ Stran†Š
ÁGsolv ˆ ESM3 ÁG
ꢀ2†
ꢀg†
1
ꢀ ˆ RTlnꢀ1=22:4L† ˆ 1:83 kcal mol
ꢀ3†
ꢀ4†
ÁG
ˆ ÁG ‡ ÁGsolv ‡ ꢀ
ꢀsol†
ꢀg†
12. Stevens, M. F. G.; Baig, G. U.; Gate, E. N.; Wheelhouse,
R. T. Anti-Cancer Drug Des. 1995, 10, 215.
13. Irwin, W. J.; Iqbal, M. Int. J. Pharmaceutics 1988, 41, 41.
14. Wang, D.; Haseltine, J. J. Heterocycl. Chem. 1994, 31,
1637.
Acknowledgements
This work was supported by Knoll Pharmaceuticals
Research and Development Department, Nottingham,
UK. We also thank the Fulbright Commission for
support (to E. C. S.) and Dr. M. A. Searle for use of
SPARTAN 4.0.
15. Smith, G. B.; Dezeny, G. C.; Hughes, D. L.; King, A. O.;
Verhoeven, T. R. J. Org. Chem. 1994, 59, 8151.
16. Martin, A. R.; Yang, Y. Acta Chem. Scand. 1993, 47, 221.
17. Chauhan, S. M. S.; Junjappa, H. Synthesis 1974, 880.
18. Grin, R. J.; Stevens, M. F. G. J. Chem. Soc., Perkin
Trans. 1 1994, 3311.
19. Gronowitz, S.; Lawitz, K. Chem. Scr. 1983, 22, 265;
Gronowitz, S.; Bobosik, B.; Lawitz, K. Chem. Scr. 1984, 23,
120.
References
20. Anderson, J. C.; Namli, H. Synlett. 1995, 765.
21. Thompson, W. J.; Jones, J. H.; Lyle, P. A.; Thies, J. E. J.
Org. Chem. 1988, 53, 2052.
1. Brown, T. B.; Stevens, M. F. G. J. Chem. Soc., Perkin
Trans. 1 1975, 1023.
2. Wierenga, W.; Skulnick, H. I.; Stringfellow, D. A.; Weed,
D. S.; Renis, H. E.; Eidson, E. E. J. Med. Chem. 1980, 23, 237;
Stringfellow, D. A.; Weed, S. D. In Interferon: Properties
22. Campi, E. M.; Jackson, W. R.; Marcuccio, S. M.; Naes-
lund, C. G. M. J. Chem. Soc., Chem. Commun. 1994, 2395.
23. Tao, W.; Nesbitt, N.; Heck, R. F. J. Org. Chem. 1990, 55, 63.

750
D. R. Hannah et al. / Bioorg. Med. Chem. 8 (2000) 739±750
24. Robins, M. J.; Barr, P. J. Tetrahedron Lett. 1981, 22,
421; Robins, M. J.; Barr, P. J. J. Org. Chem. 1983, 48,
1854.
25. Stewart, J. P. P. J. Comp. Chem. 1989, 10, 209.
26. Lively, T.; Jurema, M. W.; Shields, G. C. Int. J. Quan.
Chem. 1994, S21, 95.
29. Cramer, C. J.; Truhlar, D. G. J. Comp. Chem. 1992, 13,
1089.
30. Giesen, D. J.; Chambers, C. C.; Cramer, C. J.; Truhlar, D.
G. J. Phys. Chem. B 1997, 101, 5087.
31. Eksterowicz, J. E.; Miller, J. L.; Kollman, P. A. J. Phys.
Chem. B 1997, 101, 10971.
27. Jurema, M. W.; Shields, G. C. J. Comp. Chem. 1993, 14,
89.
32. SPARTAN SGI Version 3.1.3 GL, # by Wavefunction
Inc., 1991±1994.
28. Cramer, C. J.; Truhlar, D. G. J. Comput.-Aid. Mol. Des.
1992, 6, 629.
33. Cieplak, P.; Kollman, P. A. J. Am. Chem. Soc. 1988, 110,
3734.