Inactivation of O6-alkylguanine-DNA alkyltransferase. 1. Novel O6-(hetarylmethyl)guanines having basic rings in the side chain.

Article abstract of DOI:10.1021/jm9708644

A number of novel guanine derivatives containing heterocyclic moieties at the O6-position have been synthesized using a purine quaternary salt which reacts with alkoxides under mild conditions. Initially O6-substituents were investigated in which the benzene ring of the known agent, O6-benzylguanine, was replaced by unsubstituted heterocyclic rings. The ability of these agents to inactivate the DNA repair protein O6-alkylguanine-DNA alkyltransferase (ATase), both as pure recombinant protein and in the human lymphoblastoid cell line Raji, has been compared with that of O6-benzylguanine. The present paper focuses on O6-substituents with basic rings, and under standard conditions several of them proved more effective than benzyl for inactivation of both recombinant and Raji ATase. Among the pyridine derivatives, the 2-picolyl compound 7 is not very active in contrast to the 3- and 4-picolyl compounds, and this influenced our choice of isomers of other basic ring systems for study. Since halogen substitution in the thiophene ring considerably increased the activity (17 versus 6), similar modifications in the pyridine series were examined. The more polar O6-substituents in this study are on the whole compatible with the stereochemical requirements of the ATase protein, and their pharmacological properties may be valuable in subsequent in vivo investigations, particularly the thenyl (6), 5-thiazolylmethyl (12), 5-bromothenyl (17), and 2-chloro-4-picolyl (21) derivatives.

Full text of DOI:10.1021/jm9708644

J . Med. Chem. 1998, 41, 5265-5271
5265
In a ctiva tion of O6-Alk ylgu a n in e-DNA Alk yltr a n sfer a se. 1. Novel
O6-(Heta r ylm eth yl)gu a n in es Ha vin g Ba sic Rin gs in th e Sid e Ch a in
R. Stanley McElhinney,*,† Dorothy J . Donnelly,† J oan E. McCormick,† J ane Kelly,‡ Amanda J . Watson,‡
J oseph A. Rafferty,‡ Rhoderick H. Elder,‡ Mark R. Middleton,‡ Mark A. Willington,‡ T. Brian H. McMurry,†
and Geoffrey P. Margison*,‡
University Chemical Laboratory, Trinity College, Dublin 2, Ireland, and CRC Section of Genome Damage and Repair,
Paterson Institute for Cancer Research, Christie Hospital (NHS) Trust, Manchester M20 4BX, United Kingdom
Received J uly 6, 1998
A number of novel guanine derivatives containing heterocyclic moieties at the O6-position have
been synthesized using a purine quaternary salt which reacts with alkoxides under mild
conditions. Initially O6-substituents were investigated in which the benzene ring of the known
agent, O6-benzylguanine, was replaced by unsubstituted heterocyclic rings. The ability of these
agents to inactivate the DNA repair protein O6-alkylguanine-DNA alkyltransferase (ATase),
both as pure recombinant protein and in the human lymphoblastoid cell line Raji, has been
compared with that of O6-benzylguanine. The present paper focuses on O6-substituents with
basic rings, and under standard conditions several of them proved more effective than benzyl
for inactivation of both recombinant and Raji ATase. Among the pyridine derivatives, the
2-picolyl compound 7 is not very active in contrast to the 3- and 4-picolyl compounds, and this
influenced our choice of isomers of other basic ring systems for study. Since halogen substitution
in the thiophene ring considerably increased the activity (17 versus 6), similar modifications
in the pyridine series were examined. The more polar O6-substituents in this study are on the
whole compatible with the stereochemical requirements of the ATase protein, and their
pharmacological properties may be valuable in subsequent in vivo investigations, particularly
the thenyl (6), 5-thiazolylmethyl (12), 5-bromothenyl (17), and 2-chloro-4-picolyl (21) derivatives.
The cytotoxic effects of the antitumor N-(2-chloro-
ethyl)-N-nitrosoureas (e.g., BCNU, CCNU, fotemustine)
and the related methylating agents (e.g., DTIC, procar-
bazine, temozolomide) are primarily a consequence of
their ability to alkylate DNA at the O6-position of
guanine.1-3 It is well-established that resistance to
these O6-alkylating chemotherapeutic agents can be
mediated by the DNA repair protein O6-alkylguanine-
DNA alkyltransferase (ATase), which removes the alkyl
group from the guanine in an autoinactivating, stoichio-
metric process.1-3 Tumor cells frequently express high
levels of ATase4,5 and are often resistant to the O6-
alkylating agents used in chemotherapy. There is
therefore much current interest in attenuating ATase
activity in order to sensitize such resistant cells to
killing by these agents.6
nucleus10-12 and in the 2-, 8-, and 9-positions of the
purine ring,13 as well as O6-allylic substituents and O6-
(2-oxoalkyl)guanines,14 and included monocyclic pyri-
midines with suitable 5- and 6-substituents.13
O6-Benzylguanine and its analogues inhibit ATase by
reacting with a specific cysteine residue (Cys 145 in the
human protein).15 The benzyl group is irreversibly
transferred to the cysteine thiol group to form the
S-benzyl derivative which can no longer react with an
O6-alkylated guanine residue in DNA, leaving such
lesions unrepaired. The O6-benzylguanine/thiol reaction
is a bimolecular displacement (SN2 type), with guanine
as leaving group, X; the thiol group, especially as the
anion, is a very good (soft) nucleophile (Nuc-) effective
in SN2 substitutions:
This was first achieved by prior O6-methylation of
guanine residues in DNA,7 but it soon transpired that
the simple compound O6-methylguanine8 effectively
inactivated ATase. Numerous other small molecules
have subsequently been designed, synthesized, and
assayed as inhibitors of the protein both in cell-free
extracts and in various cell lines. The most significant
agent to emerge was O6-benzylguanine (1), about 2000
times as effective as O6-methylguanine.9,10 The struc-
tural features of this compound responsible for the high
activity have been extensively explored. The variations
arose mainly from substitution in the O6-benzyl
RX + Nuc- f RNuc + X-
To design variations of the O6-benzylguanine molecule
which would provide still more effective ATase inacti-
vators, attempts to establish a reactivity sequence for
substituted benzyl reagents with thiolate anion are
relevant. For some chlorides (X ) Cl) and arenethio-
lates (Nuc ) ArS) characteristic U-shaped Hammett F
plots are observed, with both electron-donating and
electron-withdrawing benzylic ring substituents leading
to reactivity higher than that of benzyl chloride itself.16
For a wider range of O6-benzylguanines (X ) guanine),
Kohda et al.12 noted some anomalies and invoked a
possible SN1 mechanism in the case of O6-(1-naphthyl-
methyl)guanine. More disappointingly, their observed
* Corresponding authors.
† Trinity College.
‡ Paterson Institute for Cancer Research.
10.1021/jm9708644 CCC: $15.00 © 1998 American Chemical Society
Published on Web 12/01/1998

5266 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 26
McElhinney et al.
Ch a r t 1. O6-Substituted Guanines and IC50 (µM) Values
sequence of reactivity with benzenethiolate was “not
obviously related” to the potentiation of ACNU cytotox-
icity through ATase depletion. Earlier work10,11 had
also revealed no apparent correlation between the
degree of ATase inactivation and the nature of the
benzyl substituent.
Evidently other factors are involved in the O6-
benzylguanine/ATase reaction. The available structural
data suggest that the reacting Cys residue is located in
a pocket deep within the protein,15 and the pattern of
activity of substituted O6-benzylguanines indicates that
this active site must have considerable hydrophobic
character. Further evidence for its nature was provided
by the inconsistencies in the relationship between
ATase-inactivating levels and simple chemical reactivity
recorded by Arris et al.14 In the SN2 substitution of
chlorides by iodide, methyl, allyl, and benzyl chlorides
react at roughly similar rates while phenacyl chloride
is vastly more reactive. However, O6-phenacylguanine
is devoid of ATase inhibitory activity and O6-allyl is
intermediate between O6-methyl and O6-benzyl.
A
hydrophobic pocket at the active site would accom-
modate the bulky substituent in the benzyl derivative
in the correct configuration to accelerate the reaction
with Cys thiolate, while the phenacyl derivative would
be rendered ineffective by its polarized carbonyl group.
Turning to heterocyclic examples of R other than
picolyl,11,12 it is known that the order of reactivity in
SN2 reactions (with aniline17 or halogen exchange18) is
3-thienylmethyl < benzyl < thenyl < furfuryl. The
value of furfuryloxycarbonyl as an N-protecting group
in some peptide synthetic sequences was advocated
since it is cleaved by acid very much more readily19 than
the familiar benzyloxycarbonyl (Z), and the analogous
thenyloxycarbonyl derivatives are even more reactive,20
although these comparisons involve SN1 reaction mech-
anisms and transient carbocations stabilized by the
heterocyclic ring.
described10-12,14 have usually been made, rather inef-
ficiently, by the former route (sometimes with tert-butyl
alcohol, dioxane, or tetrahydrofuran as cosolvents) since
preparation of a series of 2,4,5-triaminopyrimidines is
unattractive. We confirmed that the yields of O6-
benzylguanine from alkoxide and 2-amino-6-chloropu-
rine were poor; the product was difficult to purify, and
there was evidence for some N2-alkylation.22b,24 Neither
approach would have been suitable for synthesis of O6-
furfurylguanine with its sensitive furan ring. However,
the third method, outlined without experimental details
in a footnote,25 was much more promising since at room
temperature it apparently gave O6-benzylguanine in
92% yield.
Since none of the foregoing studies provided a clear
pointer, we felt that a comparison of O6-benzylguanine
with O6-furfurylguanine (2) and analogues with related
hetarylmethyl substituents would at least be instruc-
tive, and we therefore embarked on a program of
synthesizing and testing such molecules.21 An account
of the influence of ring substituents in the furan- and
thiophene-containing side chains is in preparation, and
in the present paper we concentrate on some O6-
substituted guanines with pyridine and other basic
heterocyclic rings in the side chain. If these types of
polar substituents were compatible with the require-
ments of the protein, they could confer some advantages
in terms of solubility and transport in vivo. It tran-
spired that several of these new drugs indeed possessed
activity as ATase inactivators, of the same order as
O6-benzylguanine.
The key reagent was the quaternary salt 5,26 pos-
sessing the much better leaving group +NMe3, using
dimethyl sulfoxide (DMSO) as solvent. Attempts to
prepare or utilize quaternary salts from 3 and triethyl-
amine or pyridine were unsuccessful, but trimethyl-
amine (bp 3 °C, readily available from Fluka as 250-
mL pressure tins) provided the salt 5 as described.
After some experimentation, we defined standard condi-
tions for displacement by alkoxide usually in good yield
with easy workup. Quaternary salts from 327 and its
9-benzyl derivative28 using 1,4-diazabicyclo[2.2.2]octane
(DABCO) were subsequently converted into O6-aralkyl-
guanines.
Ch em istr y
O6-Benzylguanine (1) (Chart 1) has been synthesized
in three ways. Displacement by alkoxide of halogen
from 2-amino-6-chloropurine (3) (in benzyl alcohol sol-
vent for 5-10 h at 130 °C)22 and annelation of an
imidazole ring to 6-(benzyloxy)-2,4,5-triaminopyrimidine
at 180 °C23 both employ very harsh experimental
conditions. The O6-substituted guanines 4 so far
Furfuryl and thenyl alcohols are of course available
commercially, as are all the picolyl alcohols and pyri-

Inactivation of ATase
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 26 5267
dine-3-methanol N-oxide. Others were synthesized by
reduction of the corresponding aldehydes or acid chlo-
rides (using sodium borohydride) or of esters (using
lithium aluminum hydride and working up by Fallab’s
method29 with carbon dioxide). 5-Bromo- and 6-chlo-
ronicotinic acids were commercial samples as was
5-bromothiophene-2-carboxaldehyde, the source of 5-bro-
mothenyl alcohol.30 Other starting materials, although
mentioned relatively infrequently in the literature, were
prepared as described.
The nicotinic acids were converted into the acid
chlorides using phosphorus oxychloride/pentachloride,31
and these reagents yielded 2-chloroisonicotinoyl chloride
directly from isonicotinic acid N-oxide.32 Reduction of
the acid chlorides with sodium borohydride31,33 afforded
respectively 5-bromopyridine-3-methanol,34 6-chloro-
pyridine-3-methanol,31 and 2-chloropyridine-4-metha-
nol.35
5,6-(Methylenedioxy)pyridine-3-carboxaldehyde (“5-
azapiperonal”) required a low-yielding synthetic route36
in which furfural was transformed by bromine/sulfamic
acid/concentrated HCl (fairly efficiently) into 5-bromo-
2,3-dihydroxypyridine, and subsequent ring closure with
bromochloromethane was followed by reaction with
butyllithium and N,N-dimethylformamide. The alcohol
has not been described previously, and indeed few
examples of this 1,3-dioxolo[4,5-b]pyridine ring system
are known.
Thiazole-5-carboxaldehyde, for thiazole-5-methanol,29
was obtained by deamination, using amyl nitrite, of the
2-amino derivative prepared from bromomalonalde-
hyde37 and thiourea.38 The synthesis of methyl isothia-
zole-4-carboxylate began from 2-chloro-5-nitrobenzal-
dehyde, readily converted into 5-nitrobenzisothiazole,39
followed by several steps including oxidative cleavage
of the fused benzene ring.40 This ester was reduced to
isothiazole-4-methanol in the manner applied for the
homologous 3-methylisothiazole-4-methanol.41 Methyl
oxazole-5-carboxylate42 was prepared from dimethyl
tartrate, via lead tetraacetate oxidation43 to methyl
glyoxylate, and tosylmethyl isocyanide;44 oxazole-5-
methanol has been described45 for use in the synthesis
of neooxazolomycin. Ethyl isoxazole-4-carboxylate re-
sulted from oximation of (ethoxycarbonyl)malonalde-
hyde;46 isoxazole-4-methanol was obtained by reduction,
similar to the homologous tertiary alcohol47 prepared
from this ester and methylmagnesium bromide. Methyl
1-methylimidazole-5-carboxylate,48 from sarcosine meth-
yl ester and Gold’s salt, yielded 1-methylimidazole-5-
methanol.49
F igu r e 1. Inactivation of pure recombinant human ATase
(closed symbols) or ATase activity in Raji cells (open symbols)
by increasing concentrations of the inactivators B.4203 (2;
squares) B.4205 (6; triangles), and B.4321 (21; circles). See
text for experimental details.
reflects differences in the assay procedures used by the
various groups. We found that in Raji cells the IC50
values for 1 and 2 are respectively 0.1 and 0.12 µM
(Table 3).
O6-Thenylguanine (6) has an IC50 of 0.018 and 0.03
µM, respectively, in cell-free extracts and in Raji cells.
As anticipated from the reactivity of these hetaryl-
methyl substituents, 2 and 6 are hydrolyzed to guanine
very quickly in acid solutions, much faster than 1. Even
in the assay medium, PBS at pH 8.3, the half-lives (t1/2
)
of 1, 2, and 6 are respectively >48h, 23 min, and 40
min. This reactivity will have significant consequences
for the behavior of the drugs in vivo.
It proved impossible to make the pyrrole (NH or NMe)
analogue of 2 and 6 since addition of sodium hydride to
the appropriate alcohols caused polymerization.50 Al-
though as confirmed later11,12 synthesis of the pyridine
analogue 7 presented no problems, its activity was
disappointingly poor (IC50 35 µM, lower even than the
22 µM value we found for O6-allylguanine). The chem-
istry of the 2-position in pyridine is somewhat special,
while the 3-position is more typical of a benzene ring.
As Kohda et al.12 found, the 3-isomer 8 proved fairly
active (IC50 0.41 µM), and perhaps surprisingly the
4-isomer 9 was more active still (IC50 0.13 µM). The
increased polarity and water solubility of the N-oxide
(10) of 8 was not advantageous (IC50 1.4 µM).
For investigation of other basic heterocyclic substit-
uents with five-membered rings, we concentrated on the
isomers where the alkoxy group was in a 1,3-relation-
ship to the nuclear N because of the poor activity of 7.
Again sulfur proved the most useful heteroatom, with
oxazole, thiazole, and N-methylimidazole yielding drugs
11, 12, and 13 having respective IC50 values of 0.35,
0.033, and 8.5 µM. The isoxazole and isothiazole
systems provide the option of a 1,4- as well as a 1,3-
relationship. While we were unable to prepare the
isoxazole 14 since sodium hydride caused decomposition
of the appropriate alcohol consistent with the known
effect of basic reagents on the isoxazole ring, the
isothiazole 15 proved almost as active as the thiazole
12. The synthesis of the 1,4-related 16 is forbidding
since the necessary alcohol requires several steps be-
yond the already arduous procedure to 15,40 and a
Resu lts a n d Discu ssion
Typical dose-response curves for the inactivation of
recombinant human ATase and ATase activity in Raji
cells by three of the inactivators considered are pre-
sented in Figure 1. IC50 values (Chart 1) were derived
using lines of best fit.
The level of activity of O6-furfurylguanine (2), ex-
pressed as the effective dose (IC50) required to produce
50% inactivation of ATase in cell-free extracts in
phosphate-buffered saline (PBS), was lower (IC50 0.08
µM) than for O6-benzylguanine (1), which, in our hands,
has an IC50 value of 0.043 ( 0.021 (n ) 21) µM. The
published10,14 value for O6-benzylguanine (0.2 µM)

5268 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 26
Ta ble 1. Spectral Data for Heterocyclic Alcohols
McElhinney et al.
a
b
UV spectra in MeOH; significant maxima > 240 nm are recorded. Unless otherwise noted, 1H NMR spectra in (CD3)2SO at 300
MHz; J values (Hz) in parentheses, otherwise singlets. c δ 3.61 (NCH3). NMR spectrum in CDCl3. e NMR spectrum at 80 MHz. f δ 6.06
(OCH2O); mp 82-84 °C, from EtOH. Anal. (C7H7NO3) C,H,N.
d
commercial source of isothiazole-5-carboxaldehyde is no
longer available.51 Of this group, the thiazole 12 is
fairly accessible and has properties which make further
biological investigation very desirable.
or substantially (7) more effective in Raji cells than with
the recombinant human ATase in vitro; however, the
majority of compounds were as effective or slightly less
effective in inactivating ATase in Raji cells than in vitro.
The reduced effectiveness in Raji cells may be a conse-
quence of a number of factors including binding to
serum components, cellular uptake, nuclear transloca-
tion, or metabolic processes. Reasons for the increased
effectiveness in Raji cells are more obscure: chemical
instability does not seem to be an explanation since the
relatively short half-lives of some of the compounds did
not drastically affect their ability to inactivate ATase
in Raji cells.
The introduction of halogen to the thiophene nucleus
of 6 caused a very significant increase in the inhibitory
activity: the 5-bromothenyl derivative 17 has an IC50
value of 0.005 µM. In the pyridine series this sort of
modification was also helpful. For compounds where
N and alkoxy are 1,3-related, 5-bromo substitution (18)
gave IC50 0.25 µM and 6-chloro (19) 0.20 µM. The
electron-donating substituent in 20 (5,6-methylene-
dioxy, giving “azapiperonyl”) produced almost identical
inhibitory activity (IC50 0.23 µM). Like its parent 9 with
N and alkoxy 1,4-related, the 2-chloro derivative 21 was
more active (IC50 0.04 µM) and is another candidate for
further study as the best of O6-substituted guanines
with a pyridine ring in the side chain.
On the basis of IC50 values for the inactivation of pure
recombinant human ATase in vitro and ATase activity
in Raji cells, and the likelihood that these values would
reflect in vivo activity, four compounds (6, 12, 17, 21)
might, if they displayed no adverse toxicities, be con-
A few of the compounds were slightly (8, 11, 18-20)

Inactivation of ATase
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 26 5269
NMR δ (ppm from TMS)c
Ta ble 2. Preparation and Characterization of O6-Substituted Guanines
O6-substituted yield, mp, °C
UV λmax
guanine
%
(dec)
formulaa
(nm)b
H-9 H-8 NH2 CH2 H-2′
H-3′
H-4′
H-5′
H-6′
2, B.4203
85 from 185
C
10H9N5O2‚0.5H2O
243, 285
12.42 7.81 6.32 5.43
bs
6.66
d (3.1)
6.49
dd
7.71
d (1.5)
(3.1, 1.5)
7.04
dd
6, B.4205
7, B.4210
8, B.4211
72 from 160i C10H9N5OS‚0.33H2O 243, 284
12.47 7.84 6.30 5.67
bs
7.31
dd
7.56
dd
(3.5, 1.2) (5.1, 3.5) (5.1, 1.2)
71 220-221
53 226-228
C
C
11H10N6O‚0.5H2O
11H10N6O‚0.5H2O
247, 262
269sh, 285
12.47 7.84 6.27 5.58
bs
7.5
m
7.5
m
7.5
m
8.58
dd
(4.8, 1.2)
8.62
dd
245, 263sh
270sh, 285
12.48 7.87 6.36 5.58 8.80
bs
8.01
dd
7.49
dd
bs
(7.9, 1.8) (7.9, 4.8) (4.8, 1.8)
9, B.4277d
72 from 230 C11H10N6O
54 244-254 11H10N6O2‚H2O
244, 265sh
286
271, 286
[RCH2OH: 267]
243, 286
12.51 7.88 6.34 5.58 8.60
bs
12.52 7.87 6.41 5.48 8.42
7.47
d (5.7) d (5.7)
7.47
d (5.7)
7.47
m
8.60
d (5.7)
8.22
m
10, B.4282
11, B.4274d,e
12, B.4275d,e
C
7.47
m
32 180-215 C9H8N6O2‚0.25H2O
40 190-220 C9H8N6OS‚0.5H2O
12.48 7.85 6.38 5.56 8.45
bs
12.49 7.85 6.42 5.76 9.13
bs
12.44 7.82 6.28 5.48 7.66
12.47 7.84 6.41 5.58
12.47 7.85 6.36 5.62
bs
7.44
244, 286
8.14
7.11
13, B.4389f
15, B.4354
82 from 190i C10H11N5O
241, 284
244, 284
247, 284
28 from 200 C9H8N6OS‚0.75H2O
8.81
7.19
d (3.7)
9.22
17, B.4269d,e,g
14 170-180i C10H8BrN5OS
7.16
d (3.7)
8.26
18, B.4320
19, B.4319
56 from 220
58 from 215
C
C
11H9BrN6O‚0.5H2O
11H9ClN6O‚0.5H2O
242, 281
242, 276
12.50 7.86 6.41 5.53 8.73
8.78
d (2.2)
dd
d (1.8)
(2.2, 1.8)
8.05
12.48 7.87 6.38 5.53 8.64
d (2.4)
7.59
dd
d (8.2)
(8.2, 2.4)
7.48
d(1.8)
20, B.4271d,h
21, B.4321
63 230-240i C12H10N6O3.0.25EtOH 241, 290
12.46 7.83 6.39 5.40 7.83
bs
d (1.8)
10 from 234
C
11H9ClN6O
241, 272sh
285
12.56 7.91 6.36 5.58
7.61
bs
7.51
bs
8.44
bs
bs
bs
a
Anal. C,H,N in all cases except for 7 (N: calcd, 33.47; found, 32.79), 10 (C: calcd, 47.83; found, 47.22), 15 (N: calcd, 32.16; found,
31.53), 17 (N: calcd, 21.47; found, 20.81), and 21 (C, N: calcd, 47.75, 30.37; found, 47.15, 29.32, respectively). UV spectra in MeOH.c 1
H
b
NMR spectra in (CD3)2SO at 300 MHz; J values (Hz) in parentheses, otherwise singlets. ArCH2OH:5, 3:1. e Product extracted with
d
MeCN. f δ 3.68 (NCH3). 2.5 mL DMF (-10 °C): 1 mmol of 5: 3 mmol of ArCH2OH. δ 6.18 (OCH2O). Recrystallized from EtOH.
g
h
i
Ta ble 3. ATase-Inactivating Activity and Stability of
O6-Substituted Guanines
ASTM, Art. 7733). Microanalyses of all compounds were
within (0.4% of theory unless otherwise indicated.
Heterocyclic alcohols not commercially available were pre-
pared by literature methods, and UV and NMR spectral
characteristics are given in Table 1. The quaternary salt 5
was obtained as described,26 and O6-substituted guanines 4
were prepared as outlined25 for 1, using the following detailed
procedure.
IC50 (µM)a
in vitro
compound
rhATase
in Raji cells
half-life (h)b
O6-benzylguanine
B.4269 (17)
B.4205 (6)
B.4275 (12)
B.4321 (21)
B.4354 (15)
B.4203 (2)
0.04
0.005
0.018
0.033
0.04
0.07
0.08
0.13
0.20
0.23
0.25
0.35
0.43
1.4
0.1
0.01
0.03
0.06
0.05
0.07
0.12
0.20
0.12
0.09
0.2
>48
1.95
0.67
>48
>48
>48
0.38
>48
>48
>48
>48
>48
>48
>48
0.19
>48
Sodium hydride (60% in oil; 0.8 g, 20 mmol) was added to a
solution of alcohol (56 mmol, ca. 5 mL) in DMSO (5 mL) and
the mixture stirred at room temperature for 1 h. For solid or
higher-molecular-weight alcohols, up to 10 mL of DMSO may
be used. 2-Amino-N,N,N-trimethyl-1H-purin-6-aminium chlo-
ride (5) (2.29 g, 10 mmol) was added, and stirring was
continued for a further 1 h. The change in UV spectrum was
then complete (λmax 312 f 284 nm), and the almost clear
solution was treated with acetic acid (1.7 mL). After cooling
and dilution with ether (300 mL), the mixture was set aside
(2 h) and the solid (A) collected. Trituration with water gave
the product. A second fraction can be obtained by evaporation
of the ether-DMSO filtrate and trituration of the residue
successively with ether and water. Alternatively the product
may be extracted from the solid (A) with warm acetonitrile.
The recrystallized compounds (usually from MeOH) showed
a single spot in TLC (C6H6-MeOH, 4:1) and were character-
ized by analysis and their NMR spectra, as indicated in Table
2. Frequently, they contained solvent of crystallization.
Biologica l Eva lu a tion . P u r ifica tion of Recom bin a n t
ATa ses. The cDNA cloning and overexpression of the human
ATase has been reported previously.52 Purification of the
recombinant proteins was achieved either by affinity chroma-
tography through a DNA-cellulose column as described53 or
by DEAE-cellulose ion-exchange chromatography. For the
latter, the ATase protein was partially purified by ammonium
sulfate precipitation (30-60%) and dialyzed against 10 mM
B.4277 (9)
B.4319 (19)
B.4271 (20)
B.4320 (18)
B.4274 (11)
B.4211 (8)
B.4282 (10)
B.4389 (13)
B.4210 (7)
0.22
0.22
5.4
8.5
35
21
6.0
a
IC50 is the concentration of inactivator required to produce
b
50% reduction in ATase activity. At 37 °C in phosphate-buffered
saline (PBS; pH 7-7.2).
sidered potential clinical alternatives to O6-benzylgua-
nine.
Exp er im en ta l Section
UV spectra were measured on a Unicam SP-800 spectro-
photometer and NMR spectra (80 MHz) on Bruker WP-80 or
(300 MHz) on MSL 300 instruments. Melting points, uncor-
rected, were determined in capillaries. Column chromatog-
raphy was performed on Merck silica gel 60 (35-70 mesh

5270 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 26
McElhinney et al.
(3) Pegg, A. E. Mammalian O6-alkylguanine-DNA alkyltrans-
ferase: regulation and importance in response to alkylating
carcinogenic and therapeutic agents. Cancer Res. 1990, 50,
6119-6129.
(4) Margison, G. P.; O’Connor, P. J .; Cooper, D. P.; Davis, J .; Hall,
N.; Redmond, S. M. S.; Buser, K.; Cerny, T.; Citti, L.; D’Incalci,
M. Triazenes: Chemical, Biological and Structural Aspects; In
Giraldi, T., Connors, T. A., Cartei, G., Eds.; Plenum Press: New
York, 1990; pp 195-206.
Tris-HCl (pH 7.5), 1 mM dithiothreitol (DTT), 2 mM EDTA,
10% glycerol, before loading onto a DEAE-cellulose column.
The ATase was then eluted with a 0-0.1 M NaCl gradient.
The purified human ATase protein retained activity for more
than 1 year when stored at high concentration at -20 °C in
buffer I [50 mM Tris-HCl (pH 8.3)/3 mM DTT/1 mM EDTA]
and could be thawed and refrozen several times without
substantial loss of activity.
(5) Citron, M.; Decker, R.; Chen, S.; Schneider, S.; Graver, M.;
Kleynerman, L.; Kahn, L. B.; White, A.; Schoenhaus, M.; Yarosh,
D. O6-Methylguanine-DNA methyltransferase in human normal
and tumor tissue from brain, lung and ovary. Cancer Res. 1991,
51, 4131-4134.
In cu ba tion w ith In a ctiva tor s a n d ATa se Assa y. Com-
pounds to be tested were dissolved in DMSO to a final
concentration of 10 mM and diluted just before use in buffer
I containing 1 mg/mL bovine serum albumin (IBSA). Recom-
binant ATase was diluted in IBSA and titrated in order that
the reaction be conducted under ATase, and not substrate,
limiting conditions. In each assay, fixed amounts of ATase
(50-60 fmol) were incubated with varying amounts of O6-
benzylguanine, or test compound, in a total volume of 200 µL
of IBSA containing 10 µg of calf thymus DNA at 37 °C for 1 h.
(6) Dolan, M. E.; Pegg, A. E. O6-Benzylguanine and its role in
chemotherapy. Clin. Cancer Res. 1997, 3, 8837-847.
(7) Zlotogorski, E.; Erickson, L. C. Pretreatment of human colon
tumor cells with DNA methylating agents inhibits their activity
to repair chloroethyl monoadducts. Carcinogenesis (London)
1984, 5, 83-87.
(8) (a) Dolan, M. E.; Morimoto, K.; Pegg, A. E. Reduction of
O6-alkylguanine-DNA alkyltransferase activity in HeLa cells
treated with O6-alkylguanines. Cancer Res. 1985, 45, 6413-
6417. (b) Yarosh, D. B.; Hurst-Calderone, S.; Babich, M. A.; Day,
R. S., III. Inactivation of O6-methylguanine-DNA methyltrans-
ferase and sensitisation of human tumor cells to killing by
chloroethylnitrosoureas by O6-methylguanine as a free base.
Cancer Res. 1986, 46, 1663-1668.
(9) Dolan, M. E.; Moschel, R. C.; Pegg, A. E. Depletion of mammalian
O6- alkylguanine-DNA alkyltransferase activity by O6-benz-
ylguanine provides a means to evaluate the role of this protein
in protection against carcinogenic and therapeutic alkylating
agents. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 5368-5372.
(10) Moschel, R. C.; McDougall, M. G.; Dolan, M. E.; Stine, L.; Pegg,
A. E. Structural features of substituted purine derivatives
compatible with depletion of human O6-alkylguanine-DNA
alkyltransferase. J . Med. Chem. 1992, 35, 4486-4491.
(11) Chae, M.-Y.; McDougall, M. G.; Dolan, M. E.; Swenn, K.; Pegg,
A. E.; Moschel, R. C. Substituted O6-benzylguanine derivatives
and their inactivation of human O6-alkylguanine-DNA alkyl-
transferase. J . Med. Chem. 1994, 37, 342-347.
(12) Kohda, K.; Terashima, I.; Koyama, K.-I.; Watanabe, K.; Mineura,
K. Potentiation of the cytotoxicity of chloroethylnitrosourea by
O6-arylmethylguanines. Biol. Pharm. Bull. 1995, 18, 424-430.
(13) Chae, M.-Y.; Swenn, K.; Kanugula, S.; Dolan, M. E.; Pegg, A.
E.; Moschel, R. C. 8-substituted O6-benzylguanine, substituted
6(4)-(benzyloxy)pyrimidine, and related derivatives as inactiva-
tors of human O6-alkylguanine-DNA alkyltransferase. J . Med.
Chem. 1995, 38, 359-365.
(14) Arris, C. E.; Bleasdale, C.; Calvert, A. H.; Curtin, N. J .; Dalby,
C.; Golding, B. T.; Griffin, R. J .; Lunn, J . M.; Major, G. N.;
Newell, D. R. Probing the active site and mechanism of action
of O6-methylguanine-DNA methyltransferase with substrate
analogues (O6-substituted guanines). Anti-Cancer Drug Des.
1994, 9, 401-408.
3
The H-methylated DNA substrate (100 µL containing 4 µg of
DNA and 100 fmol of O6-methylguanine) was added and
incubation continued at 37 °C for 1 h, until the reaction was
completed. Following acid hydrolysis of the DNA as previously
described,54 the 3H-methylated protein was recovered and
quantitated by liquid scintillation counting. Samples were
typically assayed in duplicate and experiments repeated
several times. IC50 is the concentration of inactivator required
to produce a 50% reduction in ATase activity.
Cell Tr ea tm en t a n d P r ep a r a tion of Extr a cts. Raji cells
(a human lymphoblastoid cell line derived from a Burkitt’s
lymphoma) were grown in suspension culture in RPMI me-
dium supplemented with 10% horse serum. Cells were diluted
to 106/mL in culture medium containing either the appropriate
concentration of inactivator or an equivalent volume of vehicle
(DMSO). Following incubation at 37 °C for 2 h, the cells were
harvested by centrifugation and washed twice with PBS and
the resulting cell pellets (between 1 and 2 × 107 per pellet)
stored at -20 °C. Cell pellets were resuspended in cold (4 °C)
buffer I containing 5 µg/mL leupeptin and sonicated for 10 s
at 216-µm peak-to-peak amplitude. After the mixture cooled
in ice, 0.01 volume of 8.7 mg/mL phenylmethanesulfonyl
fluoride in ethanol was added, and the sonicates were centri-
fuged at 15000g for 10 min at 4 °C to pellet cell debris. ATase
activity in the supernatant was determined as described above
and expressed as the percentage activity remaining based on
that present in the untreated controls (usually 350-450 fm/
mg).
Sta bility of In a ctiva tor s a t 37 °C. Inactivators (10 mM
in DMSO) were diluted to 0.01 mM in prewarmed degassed
PBS. Samples were immediately transferred to a CARY13
spectrophotometer (cuvette block held at 37 °C) and scanned
at an appropriate wavelength (according to the spectral
properties of the compound) at 3-10-min intervals for up to
20 h. The results were expressed as percentage absorbance
change versus time and t1/2 values (half-life) extrapolated from
this.
(15) (a) Moore, M. H.; Gulbis, J . M.; Dodson, E. J .; Demple, B.; Moody,
P. C. E. Crystal structure of a suicidal DNA repair protein: the
Ada O6-methylguanine-DNA methyltransferase from E. coli.
EMBO J . 1994, 13, 1495-1501. (b) Wibley, J . E. A.; McKie, J .
H.; Embrey, K.; Marks, D. S.; Douglas, K. T.; Moore, M. H.;
Moody, P. C. E. A homology model of the three-dimensional
structure of human O6-alkylguanine-DNA alkyltransferase based
on the crystal structure of the C-terminal domain of the Ada
protein from Escherichia coli. Anti-Cancer Drug Des. 1995, 10,
75-95.
(16) Westaway, K. C.; Waszczylo, Z. Isotope effects in nucleophilic
substitution reactions. IV. The effect of changing a substituent
at the R carbon on the structure of SN2 transition states. Can.
J . Chem. 1982, 60, 2500-2520.
(17) Arcoria, A.; Maccarone, E.; Musumarra, G.; Tomaselli, G. A.
Nucleophilic substitution in the side-chain of five-membered
heterocycles - I. Reactions of furfuryl, 2-thenyl and benzyl
chlorides with aniline in acetonitrile and in benzene. Tetrahe-
dron 1975, 31, 2523-2527.
Ack n ow led gm en t. The authors are grateful to the
Cancer Research Campaign (U.K.) and the Health
Research Board (Ireland) for financial support. They
greatly appreciate a generous gift for the project, in
memory of William G. Ashmore, from Mrs. J oyce Ash-
more. Christophe Carola, Dr. Kevin Crowley, and Paul
Murray are thanked for the preparation of some inter-
mediates, and Dr. J ohn O’Brien is thanked for providing
NMR data.
(18) O¨ stman, B. Secondary R-deuterium isotope effects and relative
rates in the halogen exchange reactions of benzyl and thenyl
chlorides. J . Am. Chem. Soc. 1965, 87, 3163-3166.
(19) Losse, G.; J eschkeit, H.; Willenberg, E. Der Furfuryloxycarbonyl-
Rest, eine acidolytisch leicht abspaltbare N-Schutzgruppe fu¨r
Peptidsynthesen. (The furfuryloxycarbonyl residue, an N-protec-
tive group for peptide synthesis readily cleaved by acidolysis.)
Angew. Chem. 1964, 76, 271. J eschkeit, H.; Losse, G.; Neubert,
K. Peptidsynthesen mit den Furfuryloxycarbonylrest als Ami-
noschutzgruppe. (Peptide synthesis using the furfuryloxycarbo-
nyl residue for protection of the amino group.) Chem. Ber. 1966,
99, 2803-2812.
Refer en ces
(1) D’Incalci, M.; Citti, L.; Taverna, P.; Catapano, C. V. Importance
of the DNA repair enzyme O6-alkylguanine alkyltransferase (AT)
in cancer chemotherapy. Cancer Treat. Rev. 1988, 15, 279-292.
(2) Margison, G. P.; O’Connor, P. J . Biological consequences of
reactions with DNA: Role of specific lesions. Handbook Exp.
Pharmacol. 94/ 1.

Inactivation of ATase
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 26 5271
(20) Bla´ha, K.; Rudinger, J . Amino acids and peptides. XLVII. Rates
of fission of some substituted benzyloxycarbonylglycines and two
heterocyclic analogues with hydrogen bromide in acetic acid.
Collect. Czech. Chem. Commun. 1965, 30, 585-598.
(21) McMurry, T. B. H.; McElhinney, R. S.; McCormick, J . E.; Elder,
R. H.; Kelly. J .; Margison, G. P.; Rafferty, J . A.; Watson, A. J .;
Willington, M. A. Preparation of O6-substituted guanine deriva-
tives as antitumor agents. Int. Pat. Appl. WO 94/29312, 1994;
Chem. Abstr. 1995, 122, 239458e.
(22) (a) Bowles, W. A.; Schneider, F. H.; Lewis, L. R.; Robins, R. K.
Synthesis and antitumor activity of 9-(tetrahydro-2-furyl)purine
analogues of biologically important deoxynucleosides. J . Med.
Chem. 1963, 6, 471-480. (b) Frihart, C. R.; Leonard, N. J . Allylic
rearrangement from O6 to C-8 in the guanine series. J . Am.
Chem. Soc. 1973, 95, 7174-7175.
(23) Robins, M. J .; Robins, R. K. Purine nucleosides. XXIV. A new
method for the synthesis of guanine nucleosides. The preparation
of 2′-deoxy-R- and â-guanosines and the corresponding N2-methyl
derivatives. J . Org. Chem. 1969, 34, 2160-2163.
(24) (a) Spassova, M.; Dvora´kova´, H.; Holy, A.; Budes´ınky, M.;
Masoj´ıdkova´, M. Synthesis of N-(3-azido-2-hydroxypropyl), N-(3-
phthalimido-2-hydroxypropyl) and N-(3-amino-2-hydroxypropyl)
derivatives of heterocyclic bases. Collect. Czech. Chem. Commun.
1994, 59, 1153-1174. (b) Koyama, K.; Hitomi, K.; Terashima,
I.; Kohda, K. Mechanism of formation of N2-benzylguanine in
the reaction of 2-amino-6-chloropurine with sodium benzyloxide,
and benzylation of nucleic acid bases. Chem. Pharm. Bull. 1996,
44, 1395-1399.
(25) McCoss, M.; Chen, A.; Tolman, R. L. Synthesis of the chiral
acyclonucleoside antiherpetic agent (S)-9-(2,3-dihydroxy-1-pro-
poxymethyl)guanine. Tetrahedron Lett. 1985, 26, 1815-1818.
(26) Kiburis, J .; Lister, J . H. Nucleophilic displacement of the
trimethylammonio-group as a new route to fluoropurines. J .
Chem. Soc. C 1971, 3942-3947.
(27) Lembicz, N. K.; Grant, S.; Clegg, W.; Griffin, R. J .; Heath, S. L.;
Golding, B. T. Facilitation of displacements at the 6-position of
purines by the use of 1,4-diazabicyclo[2.2.2]octane as leaving
group. J . Chem. Soc., Perkin Trans. 1 1997, 185-186.
(28) Linn, J . A.; McLean, E. W.; Kelley, J . L. 1,4-Diazabicyclo[2.2.2]-
octane (DABCO) - catalysed hydrolysis and alcoholysis reactions
of 2-amino-9-benzyl-6-chloro-9H-purine. J . Chem. Soc., Chem.
Commun. 1994, 913-914.
(38) Sawhney, I.; Wilson, J . R. H. Preparation of 5-(benzoylmethyl)-
thiazoles and related compounds as agrochemical fungicides.
Eur. Pat. Appl. EP 395,174, 1990; Chem. Abstr. 1991, 114,
143410s.
(39) Hagen, H.; Markert, J .; Ziegler, H. 1,2-Benzisothiazoles. Ger.
Pat. DE 3,018,108, 1981; Chem. Abstr. 1982, 96, 68980g. The
reaction in dimethylformamide at 70 °C proceeds very satisfac-
torily without an autoclave; cf. Markert, J .; Hagen, H. Prepara-
tion of 1,2-benzisothiazoles and some secondary reactions.
Liebigs Ann. Chem. 1980, 768-778.
(40) Adams, A.; Slack, R. Isothiazole: a new mononuclear heterocyclic
system. J . Chem. Soc. 1959, 3061-3072.
(41) Hatanaka, M.; Ishimaru, T. Synthetic penicillins. Heterocyclic
analogs of ampicillin. Structure-activity relationships. J . Med.
Chem. 1973, 16, 978-984.
(42) Maquestiau, A.; Flammang, R.; Ben Abdelouahab, F.-B. Flash-
vacuum pyrolysis of 1,2,4-triazolides: a new synthesis of func-
tionalized azoles. Heterocycles 1989, 29, 103-114.
(43) Wolf, F. J .; Weijlard, J . n-Butyl glyoxylate. Organic Syntheses;
J ohn Wiley: New York, 1963; Coll. Vol. 4, pp 124-125.
(44) Van Leusen, A. M.; Hoogenboom, B. E.; Siderius, H. A novel and
efficient synthesis of oxazoles from tosylmethyl isocyanide and
carbonyl compounds. Tetrahedron Lett. 1972, 2369-2372.
(45) Kende, A. S.; Kawamura, K.; DeVita, R. J . Enantioselective total
synthesis of neooxazolomycin. J . Am. Chem. Soc. 1990, 112,
4070-4072.
(46) Panizzi, L. Investigations of â-dialdehydes. I. Carbethoxy-
malonaldehyde. Gazz. Chim. Ital. 1946, 76, 56-65. Panizzi, L.
Heterocyclic syntheses. VIII. 4-Isoxazolecarboxylic acid and its
isomers. Gazz. Chim. Ital. 1947, 77, 206-214; Chem. Abstr.
1946, 40, 7163-7164; 1948, 42, 903-904.
(47) Kusumi, T.; Chang, C. C.; Wheeler, M.; Kubo, I.; Nakanishi, K.
Isolation, structure and synthesis of 4-hydroxyisoxazole (trium-
ferol), a seed germination inhibitor from an African plant.
Tetrahedron Lett. 1981, 22, 3451-3454.
(48) Kirchlechner, R.; Casutt, M.; Heywang, U.; Schwarz, M. W. New
synthesis of 1,5-disubstituted imidazoles. Synthesis 1994, 247-
248.
(49) Link, H.; Bernauer, K. U¨ ber die Synthese der Pilocarpus-
Alkaloide Isopilosin und Pilocarpin, sowie die absolute Konfigu-
ration des (+)-Isopilosins. (Synthesis of the Pilocarpus alkaloids
isopilosine and pilocarpine, and the absolute configuration of (+)-
isopilosine.) Helv. Chim. Acta 1972, 55, 1053-1062.
(50) Briggs, A. G.; Czyzewski, J .; Reid, D. H. Studies of heterocyclic
compounds. Part 26. Synthesis of 1,6,6aλ4-triheterapentalenes
from isothiazole-5-carbaldehyde. J . Chem. Soc., Perkin Trans.
1 1979, 2340-2345.
(51) Doyle, F. P.; Mehta, M. D.; Sach, G. S.; Pearson, J . L. Pharma-
codynamic compounds. Part I. Some antispasmodics derived
from substituted 2-pyrrolidinylalkanols. J . Chem. Soc. 1958,
4458-4466.
(29) Fallab, S. U¨ ber Thiazolyl-(5)-carbinol. (On Thiazolyl-5-carbinol.)
Helv. Chim. Acta 1952, 35, 215-216.
(30) Cama, L. D.; Wildonger, K. J .; Guthikonda, R.; Ratcliffe, R. W.;
Christensen, B. G. Total synthesis of thienamycin analogs. III.
Tetrahedron 1983, 39, 2531-2549.
(31) Ziegler, F. E.; Sweeny, J . G. Synthetic studies related to the
yohimbine alkaloids. J . Org. Chem. 1969, 34, 3545-3548.
(32) De Wet, C. R.; De Villiers, P. A. New synthesis of 2-hydroxy-
isonicotinic acid and its ethyl ester. Tydskr. Natuurwet. 1974,
14, 70-72; Chem. Abstr. 1976, 84, 30822w.
(33) Kelly, T. R.; Howard, H. R.; Koe, B. K.; Sarges, R. Synthesis
and dopamine autoreceptor activity of a 5-(methylmercapto)-
methyl-substituted derivative of (()-3-PPP (3-(3-hydroxyphenyl)-
1-n-propylpiperidine). J . Med. Chem. 1985, 28, 1368-1371.
(34) Kauffmann, T.; Fischer, H. Hetarine,XVI. Seitenkettencyc-
lisierung bei 3,4-Dehydropyridin-Derivaten. (Side chain cycliza-
tion in 3,4-dehydropyridine derivatives.) Chem. Ber. 1973, 106,
220-227.
(35) Hamana, M.; Yamazaki, M. Tertiary amine oxides. IX. Reaction
of 2-substituted pyridine-1-oxides with acetic anhydride. J .
Pharm. Soc. J pn. 1961, 81, 574-578; Chem. Abstr. 1961, 55,
24743.
(36) Dallacker, F.; Fechter, P.; Mues, V. 1,3-Dioxolohetarenes, 3.
Preparation and reactions of pyrido [2,3-d]- and pyrido[3,4-d]-
[1,3]dioxoles. Z. Naturforsch. 1979, 34b, 1729-1736. This paper
is not recorded in Chem. Abstr. but is quoted in Dallacker, F.;
J ouck, W. Derivatives of 1,3-benzdioxoles, 53. Z. Naturforsch.
1984, 39b, 1598-1606; Chem. Abstr. 1985, 102, 166561h.
(37) Trofimenko, S. Dihalomalonaldehydes. J . Org. Chem. 1963, 28,
3243- 3245.
(52) Fan, C. Y.; Potter, P. M.; Rafferty, J . A.; Watson, A. J .; Cawkwell,
L.; Searle, P. F.; O′Connor, P. J .; Margison, G. P. Expression of
a human O6-alkylguanine-DNA-alkyltransferase cDNA in hu-
man cells and transgenic mice. Nucleic Acids Res. 1990, 18,
5723-5727.
(53) Wilkinson, M. C.; Potter, P. M.; Cawkwell, L.; Georgiadis, P.;
Patel, D.; Swann, P. F.; Margison, G. P. Purification of the E.
coli ogt gene product to homogeneity and its rate of action on
O6-methylguanine, O6-ethylguanine and O4-methylthymine in
dodecadeoxyribonucleotides. Nucleic Acids Res. 1989, 17, 8475-
8484. Wilkinson, M. C.; Cooper, D. P.; Southan, C.; Potter, P.
M.; Margison, G. P. Purification to apparent homogeneity and
partial amino-acid sequence of rat liver O6-alkylguanine-DNA-
alkyltransferase. Nucleic Acids Res. 1990, 18, 13-16.
(54) Morten, J . E. N.; Margison, G. P. Increased O6-alkylguanine
alkyltransferase activity in Chinese hamster V79 cells following
selection with chloroethylating agents. Carcinogenesis (London)
1988, 9, 45-49.
J M9708644