Resistance-modifying agents. 8. Inhibition of O6-alkylguanine-DNA alkyltransferase by O6-alkenyl-, O6-cycloalkenyl-, and O6-(2-oxoalkyl)guanines and potentiation of temozolomide cytotoxicity in vitro by O6-(1-cyclopentenylmethyl) guanine

Article abstract of DOI:10.1021/jm000961o

A series of O⁶-allyl- and O⁶-(2-oxoalkyl)guanines were synthesized and evaluated, in comparison with the corresponding O⁶-alkylguanines; as potential inhibitors of the DNA-repair protein O⁶-alkylguanine-DNA alkyltransferase (AGT). Simple O⁶-alkyl- and O⁶-cycloalkylguanines were weak AGT inactivators compared with O⁶-allylguanine (IC₅₀ = 8.5 ± 0.6 μM)with IC₅₀ values ranging from 100 to 1000 μM. The introduction of substituents at C-2 of the allyl group of O⁶-allylguanine reduced activity compared with the parent compound, while analogous compounds in the O⁶-(2-oxoalkyl)guanine series exhibited very poor activity (150-1000 μM). O⁶-Cycloalkenylguanines proved to be excellent AGT inactivators, with 1-cyclobutenylmethylguanine (IC₅₀ = 0.55 ± 0.02 μM) and 1-cyclopentenylmethylguanine(IC₅₀ = 0.39 ± 0.04 μM) exhibiting potency approaching that of the benchmark AGT inhibitor O⁶-benzylguanine (IC₅₀ = 0.18 ± 0.02 μM). 1-Cyclopentenylmethylguanine also inactivated AGT in intact HT29 human colorectal carcinoma cells (IC₅₀ = 0.20 ± 0.07 μM) and potentiated the cytotoxicity of the monomethylating antitumor agent Temozolomide by approximately 3- and 10-fold, respectively, in the HT29 and Colo205 tumor cell lines. The observation that four mutant AGT enzymes resistant to O⁶-benzylguanine also proved strongly cross-resistant to 1-cyclopentenylmethylguanine indicates that the O⁶-substituent of each compound makes similar binding interactions within the active site of AGT.

Full text of DOI:10.1021/jm000961o

J . Med. Chem. 2000, 43, 4071-4083
4071
Resista n ce-Mod ifyin g Agen ts. 8.1 In h ibition of O6-Alk ylgu a n in e-DNA
Alk yltr a n sfer a se by O6-Alk en yl-, O6-Cycloa lk en yl-, a n d O6-(2-Oxoa lk yl)gu a n in es
a n d P oten tia tion of Tem ozolom id e Cytotoxicity in Vitr o by
O6-(1-Cyclop en ten ylm eth yl)gu a n in e
Roger J . Griffin,*,† Christine E. Arris,‡ Christine Bleasdale,† F. Thomas Boyle,§ A. Hilary Calvert,‡
Nicola J . Curtin,‡ Christine Dalby,† Sreenivas Kanugula,| Nicola K. Lembicz,† David R. Newell,‡
Anthony E. Pegg,| and Bernard T. Golding†
Department of Chemistry, Bedson Building, The University, Newcastle upon Tyne NE1 7RU, U.K.; Cancer Research Unit,
Medical School, Framlington Place, Newcastle upon Tyne NE2 4HH, U.K.; AstraZeneca Pharmaceuticals, Alderley Park,
Macclesfield, Cheshire SK10 4TG, U.K.; and Departments of Cellular and Molecular Physiology and Pharmacology,
Pennsylvania State University College of Medicine, The Milton S. Hershey Medical Center, P.O. Box 850,
Hershey, Pennsylvania 17033
Received April 3, 2000
A series of O6-allyl- and O6-(2-oxoalkyl)guanines were synthesized and evaluated, in comparison
with the corresponding O6-alkylguanines, as potential inhibitors of the DNA-repair protein
O6-alkylguanine-DNA alkyltransferase (AGT). Simple O6-alkyl- and O6-cycloalkylguanines were
weak AGT inactivators compared with O6-allylguanine (IC50 ) 8.5 ( 0.6 µM) with IC50 values
ranging from 100 to 1000 µM. The introduction of substituents at C-2 of the allyl group of
O6-allylguanine reduced activity compared with the parent compound, while analogous
compounds in the O6-(2-oxoalkyl)guanine series exhibited very poor activity (150-1000 µM).
O6-Cycloalkenylguanines proved to be excellent AGT inactivators, with 1-cyclobutenylmeth-
ylguanine (IC50 ) 0.55 ( 0.02 µM) and 1-cyclopentenylmethylguanine (IC50 ) 0.39 ( 0.04 µM)
exhibiting potency approaching that of the benchmark AGT inhibitor O6-benzylguanine (IC50
) 0.18 ( 0.02 µM). 1-Cyclopentenylmethylguanine also inactivated AGT in intact HT29 human
colorectal carcinoma cells (IC50 ) 0.20 ( 0.07 µM) and potentiated the cytotoxicity of the
monomethylating antitumor agent Temozolomide by approximately 3- and 10-fold, respectively,
in the HT29 and Colo205 tumor cell lines. The observation that four mutant AGT enzymes
resistant to O6-benzylguanine also proved strongly cross-resistant to 1-cyclopentenylmeth-
ylguanine indicates that the O6-substituent of each compound makes similar binding interac-
tions within the active site of AGT.
In tr od u ction
or by direct administration of an AGT pseudosubstrate,
potentiates the cytotoxicity of DNA-damaging agents
which function by guanine O6-alkylation.5,8,9 Although
cellular AGT depletion has been achieved in vitro with
the purine free base O6-methylguanine (MG, 1)3,4 this
compound is devoid of activity in vivo.10 In contrast, O6-
benzylguanine (BG, 2) is a potent inactivator which
effects a rapid and pronounced AGT depletion, in vitro
and in vivo.6,11,12 Significant potentiation of methylating
and chloroethylating agent cytotoxicity has been ob-
served in both tumor cell lines and mice bearing human
tumor xenografts, following pretreatment with BG,5,6
and clinical trials with BG in combination with BCNU
(1,3-bis(2-chloroethyl)-1-nitrosourea) are currently in
progress.13,14
The DNA repair protein O6-alkylguanine-DNA alkyl-
transferase (AGT) acts on DNA in which guanine
residues have been converted into potentially muta-
genic, carcinogenic, and cytotoxic O6-alkylguanine ad-
ducts following exposure to alkylating agents.2-5 AGT
repairs O6-alkylguanine lesions by transferring the O6-
alkyl group to an active-site cysteine thiolate in a
stoichiometric reaction, which regenerates guanine and
produces a chemically stable S-alkylcysteine.5 Impor-
tantly, this essentially irreversible process results in the
inactivation of AGT, with the restoration of cellular AGT
levels requiring de novo resynthesis of the protein.
Tumor sensitivity to chemotherapeutic alkylating
agents has been found to correlate inversely with
cellular AGT activity, high levels of the protein confering
resistance to the cytotoxicity of these drugs.6,7 Conse-
quently, inactivation and depletion of AGT in tumor
cells, either by pretreatment with a methylating agent
* To whom correspondence should be addressed. Tel: +44 191 222
8591. Fax: +44 191 222 8591. E-mail: r.j.griffin@ncl.ac.uk.
† Department of Chemistry, Newcastle University.
‡ Cancer Research Unit, Newcastle University.
A number of point mutations in the human AGT
sequence that render the protein resistant to inactiva-
§ AstraZeneca Pharmaceuticals.
| Pennsylvania State University College of Medicine.
10.1021/jm000961o CCC: $19.00 © 2000 American Chemical Society
Published on Web 09/14/2000

4072 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Griffin et al.
Sch em e 1a
a
Reagents: (i) ROH, Na, reflux; (ii) ROH, NaH, THF, reflux; (iii) ROH, NaH, DMSO, 25 °C; (iv) AcOH or CF3CO2H, 25 °C.
tion by BG have been reported.15-18 The recently derived
crystal structure of human AGT19 provides a reasonable
explanation for the inability of these mutants to react
with BG, because it shows that the altered residues are
located in the BG-binding pocket and would tend to
exclude BG from this pocket. Although such alterations
have not yet been shown to occur in clinical trials of
patients treated with BG, these observations raise a
serious possibility that the development of resistance
to therapy with BG plus an alkylating agent may occur
relatively readily. The availability of additional inhibi-
tors that are able to inactive the resistant mutants is
therefore highly desirable.
A large number of analogues of BG have been
synthesized in order to investigate structure-activity
relationships (SARs) for AGT inhibition.5,20-22 In addi-
tion, thienylguanine 323 and guanosine derivatives (e.g.
4)24-26 have been investigated for enhanced potency and/
or water solubility. Interestingly, a series of 2,4-diamino-
4-benzyloxypyrimidines, exemplified by 5, were also
found to be significantly more active than BG and are
among the most potent AGT inactivators reported to
date.22,27 Direct administration of O6-benzyl-8-oxogua-
nine (6), a major metabolite of comparable potency to
BG, has also been considered as a means of minimizing
polymorphic variations in the metabolism of BG, al-
though 6 is very poorly soluble.28
As part of studies to develop novel AGT inactivators,
we have previously reported preliminary results with
a series of O6-(2-oxoalkyl)guanines, designed to probe
the stereoelectronic and hydrophobic requirements of
the AGT-mediated alkyl-transfer reaction.29 Although
these compounds exhibited only weak activity, several
of the corresponding O6-allyl derivatives, prepared for
comparative purposes, were relatively potent inactiva-
tors of AGT from human HT29 colorectal cells. Indeed,
O6-allylguanine 7, which has also been reported inde-
pendently by Moschel et al.,20 proved to be the most
potent AGT inactivator of the series (IC50 ) 8.5 ( 0.6
µM). While this may be attributed, in part, to stereo-
electronic acceleration of the SN2 substitution reaction
by the adjacent alkene, our results with O6-(2-oxoalkyl)-
guanines, presented more fully in this paper, and
observations by others5,21 suggest that steric factors are
also important. To explore this further, a series of
guanines bearing substituted O6-allyl groups, and also
derivatives which incorporate the allylic double bond
within a carbocyclic ring, have been synthesized. The
latter series of compounds was prepared with the
expectation that additional steric and hydrophobic
interactions, combined with enhanced SN2 reactivity
arising from conformational restriction of the allylic
group, would confer favorable AGT binding and inhibi-
tory activity, respectively. The analogous O6-alkyl- and
O6-cycloalkylguanines were also synthesized for com-
parison. In this paper we present the results of these
studies. A preliminary account of part of this work has
been published previously.29
Ch em istr y
The structures and properties of the compounds
evaluated as AGT inactivators are recorded in Table 1,
and their syntheses are outlined in Scheme 1. The
required O6-alkylguanines were synthesized from 2-
amino-6-chloropurine, either by direct reaction with the
appropriate sodium alkoxide, prepared by treating the
corresponding alcohol with sodium (7, 13-21) or sodium
hydride in THF (22-38),30,31 or following conversion into
the more reactive 2-aminotrimethylpurin-6-ylammo-
nium chloride32 (39-43). The target O6-(2-oxoalkyl)-
guanines (44-47) were prepared by acid-catalyzed
hydrolysis of the corresponding acetals (34-37).
Two approaches were investigated for the synthesis
of 1-cyclobutenemethanol (12), required for the prepara-
tion of O6-(1-cyclobutenylmethyl)guanine (38) (Scheme
2). Initial efforts to synthesize 12 by direct LiAlH4
reduction of the unstable 1-cyclobutenecarboxylic acid
(7a ), prepared by dehydrobromination of commercially
available ethyl 1-bromocyclobutanecarboxylate,33 were
unsuccessful, presumably as a consequence of the poor
solubility of the lithium salt of the acid in ether. Burger
et al.34 reported a low yield (15%) of 12, following
reduction of methyl 1-cyclobutenecarboxylate (8) with
LiAlH4. We have attempted to improve this reaction.
Esterification of 7a with N,N-diisopropyl-O-methyl-
isourea35 or 1 mol equiv of diazomethane afforded the
methyl ester 8 in excellent yield, whereas treatment
with excess diazomethane gave the pyrazoline ester 7b,
arising from 1,3-dipolar cycloaddition to the reactive
cycloalkene. Analysis (NMR and GC) of the reaction
mixture resulting from the reaction of 8 with LiAlH4
demonstrated complete consumption of the ester and
formation of the required alcohol 12 (57%), together with
cyclobutanemethanol (43%) arising from competing 1,4-
hydride addition to the cycloalkene. Systematic varia-
tion of the reaction conditions employed (nature and
quantity of hydride reagent, solvent, and duration of

Resistance-Modifying Agents. 8
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4073
Sch em e 2a
a
Reagents: (i) KOH, PhMe, reflux; (ii) 1 mol equiv CH2N2, Et2O, 0 °C; (iii) LiAlH4, Et2O, 25 °C; (iv) excess CH2N2, Et2O, 0 °C; (v)
NaBH4, EtOH, 25 °C; (vi) cyclobutanone, ZnCl2, CCl4, 25 °C; (vii) t-BuLi, THF, -78 °C, then CH2O(g); (viii) NaIO4, aq MeOH, 25 °C; (ix)
Na2CO3, Et2O, reflux.
reaction) indicated that the reaction time is critical.
Thus, conducting the reaction with 0.6 mol equiv of
LiAlH4 in ether over 15 min afforded an optimal yield
(84%) of 12. All attempts to remove the residual 1-cy-
clobutanemethanol by chromatography or fractional
distillation resulted in extensive decomposition of 12,
and consequently, the mixture of alcohols was used
without further purification. Although the subsequent
reaction with 2-amino-6-chloropurine or its trimethyl-
ammonio derivative furnished an intractable mixture
of products, the use of ‘DABCO-purine’ (2-amino-6-
(azonia-4-azabicyclo[2.2.2-oct-1-yl]purine chloride, Scheme
1)36 enabled isolation of the target 1-cyclobutenylgua-
nine 38, albeit in low yield.
The alternative route to 12 utilized a selenoxide
fragmentation strategy. Treatment of cyclobutanone
with benzeneselenol, freshly prepared by reduction of
diphenyldiselenide with NaBH4,37 afforded 1,1-bis(phen-
ylseleno)cyclobutane (9) in excellent yield. Lithiation of
9 and reaction of the resulting R-selenocarbanion with
anhydrous methanal gave 1-hydroxymethyl-1-phenylse-
lenocyclobutane (10), which was smoothly converted
into the corresponding selenoxide 11 on treatment with
sodium periodate in aqueous MeOH.38 Final thermally
induced selenoxide fragmentation of 11 gave the req-
uisite 1-cyclobutenemethanol (12) in reasonable yield,
after fractional distillation.
Although the biological evaluation of compounds of
this type was hampered by their very poor solubility,
the effects of chain unsaturation and the insertion of a
ring system were also investigated. The O6-hexenylgua-
nine analogues (22-24) proved less active than the
parent O6-n-hexylguanine (18). The greater potency of
O6-((Z)-3-hexenyl)guanine (23) compared with the E-
stereoisomer (24) presumably arises as a result of a
more favorable interaction of 24 with the hydrophobic
pocket. The introduction of a cycloalkyl substituent at
the guanine O6-position was also detrimental to activity,
with the O6-cyclobutylmethyl (39), O6-cyclopentylmethyl
(40), and O6-cyclohexylmethyl (20) derivatives all ex-
hibiting IC50 values for AGT inactivation of >1 mM.
This poor activity presumably reflects the greater steric
bulk and reduced flexibility of an alicyclic substituent
compared with an open-chain alkyl group and suggests
that the hydrophobic pocket giving access to the AGT
active site is narrow.
The introduction of an allyl substituent at the guanine
O6-position (7) resulted in a marked increase in potency
(IC50 ) 8.5 µM) compared with the corresponding O6-
propyl analogue (14) (24% inactivation at 1.0 mM),
whereas O6-propargylguanine (31) (IC50 ) 20 µM) was
less potent than 7 as an AGT inactivator. These results
are consistent with the observations of Moschel et al.,20
who reported an IC50 value of 20 µM for 7. For SN2
substitution reactions a carbon-carbon double bond R
to the reaction center elicits a significant rate enhance-
ment, compared to a corresponding saturated group.39
However, and in keeping with previous reports,20,23 the
reactivities of O6-alkylguanines as AGT inactivators in
our studies did not parallel those reported for a simple
chemical system. Thus, the rank order of potency for
AGT inactivation is BG (2) > 7 . MG (1), with BG
proving approximately 50- and 1750-fold more active
than 7 and 1, respectively (Table 1). This is in contrast
to SN2 reactions, where benzyl, methyl, and allyl
chlorides have been reported to show comparable reac-
tivity toward iodide ion in acetone.40 These results are
consistent with the presence of an active-site hydropho-
bic pocket, which accommodates the benzyl group of 2
in an orientation so as to provide a stereoelectronic
acceleration of the SN2 reaction with the cysteine
thiolate of AGT.29
Resu lts a n d Discu ssion
SARs. The concentration of each compound required
to inactivate AGT extracted from HT29 human colorec-
tal tumor cells by 50% (IC50) is shown in Table 1. For
simple linear O6-alkylguanines (13-15, 17-19) homolo-
gation to ethyl (13) or n-propyl (14) resulted in a
substantial reduction in potency compared with 1, and
the poor activity of 13 is in agreement with the results
of Yarosh et al.2 Interestingly, although the n-pentyl
homologue 17 was only weakly active, O6-n-butylgua-
nine (15) and O6-n-hexylguanine (18) were of compa-
rable potency to 1. Homologation beyond an ethyl group
would not be expected to make any significant difference
to activity on electronic grounds, suggesting that hy-
drophobic interactions within the substrate binding site
are of importance. This suggestion is consistent with
the presence of a deep hydrophobic pocket or cleft
adjacent to the active site of AGT, as proposed previ-
ously.21,29
With a view to enhancing interactions with the
putative hydrophobic pocket of AGT, a series of substi-

4074 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Griffin et al.
Ta ble 1. Physical Data and Activity of Compounds as AGT Inactivators

Resistance-Modifying Agents. 8
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4075
Ta ble 1 (Continued)
a
b
See the Experimental Section. Recrystallization solvents: A, H2O; B, Me2CO; C, EtOAc-petrol; D, DMF-H2O; E, MeOH; F, EtOH.
Purified by column chromatography on silica: G, eluent ) DCM:EtOH (85:15); H, eluent ) DCM:MeOH (92:8). I, product crystallized
directly from reaction mixture on cooling and was triturated with ether. c Concentration of inhibitor required to reduce AGT activity to
d
50% of control rate. For details of assay see ref 29. Values for 2 independent IC50 determinations unless otherwise indicated. e O6-
g
h
i
j
Methylguanine. f Mean ( SD of 4 experiments. Mean ( SD of 3 experiments. O6-Thienylguanine. Lit.31 mp 205 °C. Lit.30 mp 293
°C. Lit.30 mp 208 °C. Lit.30 mp 175 °C. Determined at 100 µM. ND, IC50 value not determined. o Decomposes.
k
l
m
n
tuted O6-allylguanines (21, 27, 28, 30) was synthesized
and evaluated. Homologation from allyl (7) to methallyl
(21) resulted in an approximately 3-fold reduction in
potency, while a partial restoration of activity was
observed with O6-ethallylguanine (27), which was never-
theless some 2-fold less potent than 7. The introduction

4076 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Griffin et al.
Ta ble 2. Inactivation of Purified Wild-Type and Mutant Forms
of AGT by 2 and 41a
of a phenyl group (30) reduced potency approximately
10-fold compared with 7, whereas an isopropyl-
allyl group at the guanine O6-position (28) resulted in
a dramatic loss of potency (IC50 > 1 mM). Although none
of the substituted O6-allylguanines was as potent as the
parent compound 7, the relative activities of compounds
in this series followed a somewhat similar trend to that
observed for the simple O6-alkylguanines (1, 13-15,
17-19). This presumably reflects a balance between
favorable hydrophobic interactions and unfavorable
steric bulk and provides further evidence that the
putative hydrophobic pocket of AGT has strict con-
straints.
ED50 value (µM)
41
ED50 value (µM)
41
>1500 >1500
60 113
AGT protein
2
AGT protein
2
wild-type
G160R
Y158H
0.2
4.7
620
0.3
13
∼1500
P140K
G156A
a
See the Experimental Section.
have shown that phenacyl bromide is relatively more
reactive to charged and polarizable nucleophiles, includ-
ing thiolates.42 In light of these observations a series of
O6-(2-oxoalkyl)guanines (44-47) was synthesized and
evaluated for AGT inactivatory activity.
In contrast to the O6-cycloalkylmethylguanine series
(20, 39, 40), incorporation of the allylic double bond into
a ring system resulted in a large increase in potency
compared with the parent O6-allylguanine (7). Thus, the
potency of O6-(1-cyclohexenylmethyl)guanine (42) was
approximately 5-fold that of 7 and at least 1000-fold
greater than that of the corresponding cyclohexylmeth-
ylguanine 20. The effect of introducing an allylic double
bond was even more dramatic for the 1-cyclobutenyl-
methyl (38) and 1-cyclopentenylmethyl (41) derivatives,
which were potent AGT inactivators with IC50 values
of 0.55 ( 0.02 and 0.39 ( 0.04 µM, respectively. The
greater activity of the cycloalkenylguanines (38, 41, 42)
relative to 7 may be attributed to several factors. In
addition to an increased potential for hydrophobic
interactions within the active site of AGT, it is also
possible that the cycloalkenyl double bond adopts a more
favorable orientation, relative to that of O6-allylguanine
(7), for stabilization of the transition state arising from
attack on the R-CH2 by the active-site thiolate. The
chemical reactivity of O6-(1-cyclobutenylmethyl)guanine
(38) may also be further enhanced by ring strain arising
from the 4-carbon ring. However, the fact that O6-
cyclopentenylmethylguanine (41) proved the most po-
tent AGT inactivator in the cycloalkenyl series suggests
that a subtle combination of stereoelectronic, steric, and
lipophilic binding factors is the major determinant of
activity. Such factors would account for the reduction
in potency observed for the bulkier O6-cyclohexenyl-
methylguanine (42) and the detrimental effect of intro-
ducing an isopropenyl substituent onto the cyclohexenyl
ring in 43 (IC50 ) 2.6 ( 0.42 µM). Consistent with the
requirement for an allylic system, transposing the
cyclohexenyl double bond to the 3-position (32) dramati-
cally reduced potency, confirming that the influence of
the activating double bond does not extend to any
significant degree beyond the R-carbon atom.41
Comparison of the activities of the 2-oxoalkyl series
(44-47) with the isosteric allyl series (21, 27, 28, 30)
(Table 1) indicates that a minor structural modification,
notably replacing a methylene group by oxygen, has a
profound effect upon potency, with the O6-(2-oxoalkyl)-
guanines all proving considerably less active than their
O6-allylguanine counterparts. Interestingly, the overall
pattern of activity in the 2-oxoalkyl series paralleled
that observed in the allyl series, in that homologation
from methyl (compare 21 and 44) to ethyl (compare 27
and 45) resulted in a slight increase in potency. Simi-
larly, the presence of an isopropyl or phenyl substituent
in both the 2-oxoalkyl (46, 47) and allyl (28, 30) series
was not conducive to AGT inactivation.
It is evident that the presence of an R-carbonyl group
in compounds 44-47 confers no significant stereoelec-
tronic benefit, and this may be as a consequence of the
polar carbonyl group proving incompatible with the
hydrophobic pocket. Alternatively, the preferred binding
mode of compounds in this series may be such that the
orientation of the carbonyl function does not provide a
stereoelectronic acceleration of the cysteine alkylation
reaction. The failure of the 2-oxoalkyl substituent to
endow the reactivity observed in a simple chemical
system serves to illustrate the difficulty of rational
design of compounds intended to interact with a complex
biological system. With the exception of 36, which
exhibited weak activity, the O6-(2-oxoalkyl)guanine
precursors (34-37) were all essentially inactive.
In a ctiva tion of Wild -Typ e a n d Mu ta n t F or m s of
AGT by 2 a n d 41. The results in Table 1 show that 41
is only slightly inferior to 2 in the ability to inactivate
wild-type AGT and is therefore a viable potential
alternative to O6-benzylguanine (2). The small differ-
ence between these agents in this respect is probably
insignificant, and questions of pharmacokinetics would
determine their relative in vivo potential. However, the
results in Table 2 show clearly that 41 does not provide
a means to selectively inactivate mutants of AGT
resistant to 2. The four mutants tested represent several
sites at which point mutations interfere with the ability
of the AGT protein to react with 2, and in all cases, these
changes also impair reaction with 41. The mutants
tested include G160R, which is only moderately resis-
tant but is a naturally occurring, although rare, variant
of AGT,18 and three mutants (G156A, Y158H, and
P140K) which are more strongly resistant. It appears
from the crystal structure of the protein19 that the
extreme resistance of the P140K mutant to 2 is due to
the loss of the proline side chain, which forms a binding
Encouraged by the increased potency of O6-alkylgua-
nines bearing an R-alkenyl group, we considered alter-
native structural modifications predicted to enhance the
reactivity of the O6-alkyl group to nucleophilic attack.
Classical studies of SN2 reactions have demonstrated
that R-halo ketones are much more reactive to nucleo-
philes than alkyl or allyl halides. For example, in
reactions with iodide in acetone, phenacyl chloride and
chloroacetone exhibit reaction rates approximately
100 000 and 35 700 times, respectively, that of n-butyl
chloride, while benzyl chloride is only some 200-fold
more reactive than n-butyl.39 In addition, studies com-
paring the reactivity of phenacyl bromide and iodo-
methane with a variety of nucleophiles in acetonitrile

Resistance-Modifying Agents. 8
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4077
Ta ble 3. Comparison of the IC50 Values for the Cytotoxicity of
Temozolomide Alone and Temozolomide Plus 10 µM
O6-Benzylguanine (2) or O6-Cyclopentenylmethylguanine (41)a
cell line Temozolomide Temozolomide + 2 Temozolomide + 41
HT29
Colo205
951 ( 69b
343 ( 38
86 ( 8.6
14 ( 5
299 ( 38
36 ( 1
a
b
See the Experimental Section. The data represent the IC50
for Temozolomide (mean ( SD) from 3 independent experiments.
(0.065 µM in intact HT29 cells (Table 1). Potentiation
of the monomethylating antitumor agent Temozolo-
mide43 by 41 was investigated, in comparison with 2,
in the HT29 and Colo205 cell lines. These cell lines were
selected owing to their similarly high levels of AGT
activity (908 ( 76 and 1160 ( 78 fmol/mg cell protein,
respectively). Data for the potentiation of Temozolomide
cytotoxicity by 10 µM 41 in HT29 cells and Colo205 cells
are summarized in Table 3, which shows that there was
an approximate 3- and 10-fold increase in Temozolomide
activity against HT29 and Colo205 cells, respectively,
in the presence of the novel AGT inactivator. Compara-
tive experiments were performed with 10 µM 2, and
these data are also shown in Table 3. Control incubation
with 10 µM 2 or 41 alone did not cause any detectable
cytotoxicity (data not shown). Examination of the Te-
mozolomide IC50 values shown in Table 3 reveals that
41 was less effective at modulating the cytotoxicity of
Temozolomide than 2, in keeping with the greater
potency of 2 as an AGT inactivator in cell extracts and
whole cells (Table 1 and Figure 1).
F igu r e 1. Comparison of AGT inactivation by selected
compounds in intact HT29 cells and HT29 extracts. Data are
from Table 1. The IC50 for MGMT inactivation in intact HT29
cells is the concentration of inactivator required to reduce
whole cell MGMT activity by 50%, following incubation of
intact cells with the inactivator for 4 h.
surface for the benzyl group and its replacement by the
charged lysine side chain which discourages binding.
Similarly, the phenyl ring of Y158 appears to take part
in an aromatic stacking reaction with the phenyl ring
of 2, which is prevented when Y158 is replaced by
histidine. Replacement of G156 by any other residues
including the G156A change produces a distortion of the
active site that would prevent access of 2. The G160R
change would also be expected to provide some steric
interference to the binding of 2. The cross-resistance of
all four mutant AGT proteins to 41 suggests that this
compound binds in an analogous fashion to 2 in the
active site of AGT and that interactions with the
cyclopentenyl ring of 41 are broadly similar to those of
the phenyl ring of 2.
Com p a r a tive In a ctiva tion of AGT by Alk ylgu a -
n in es in HT29 Cells a n d HT29 Cell Extr a cts. The
ability of selected guanine derivatives to inactivate AGT
in intact HT29 cells was determined and is shown in
Table 1 and Figure 1. There was a strong linear
relationship between the potency of individual com-
pounds as AGT inactivators in whole cells and their
potency against AGT in cell extracts (r ) 0.998, p )
<0.00001). The fact that the relative AGT inactivating
potencies of the substrate analogues were the same in
whole cells and in cell free extracts suggests that there
is no barrier to the entry of the compounds into cells.
The finding that the IC50 value for AGT inactivation in
intact cells is approximately one-half that observed in
cell extracts is consistent with similar studies of other
AGT-inactivating substrate analogues.20-22 However,
exceptions have been reported, e.g. 8-aza-O6-benzylgua-
nine and O6-benzyl-8-bromoguanine22 and 2-amino-6-
benzyloxy-9-(carboxymethyl)purine, sodium salt.20 These
latter compounds are relatively inactive as AGT inac-
tivators in intact cells, suggesting that they have poor
membrane permeability.
Con clu sion s
The clinical value of the AGT inactivator BG (2) as a
potentiator of methylating and chloroethylating agent
cytotoxicity is currently being assessed. However, the
likelihood that resistance will arise to this agent neces-
sitates the development of new AGT depletors active
against resistant mutants. With a view to identifying
such compounds, we have synthesized novel O6-allyl-
and O6-(2-oxoalkyl)guanines and elucidated SARs for
AGT inactivation, in comparison with their O6-alkyl-
guanine counterparts. As expected, simple O6-alkyl- and
O6-cycloalkylguanines were weak AGT inactivators.
Perhaps more surprisingly, the introduction of substit-
uents onto O6-allylguanine (7) reduced activity com-
pared with the parent compound, while analogous
compounds in the O6-(2-oxoalkyl)guanine series exhib-
ited very poor activity. These results are consistent with
the existence of a narrow hydrophobic pocket adjacent
to the active site of AGT, which limits the size and
polarity of the O6-substituent and which may accom-
modate this substituent in an orientation which may
impede or assist the cysteine alkylation reaction.
By contrast, O6-cycloalkenylguanines proved to be
excellent AGT inactivators, with 1-cyclobutenylmeth-
ylguanine (38) and 1-cyclopentenylmethylguanine (41)
exhibiting potency approaching that of 2 against protein
from HT29 colorectal tumor cell extracts. More detailed
studies conducted with 41 demonstrated that this
compound inactivated AGT in intact HT29 cells and
potentiated the cytotoxicity of the monomethylating
agent Temozolomide in the HT29 and Colo205 tumor
cell lines, to a degree comparable with that of 2.
Unfortunately, the four mutant AGT enzymes resistant
P oten tia tion of th e Cytotoxicity of Tem ozolo-
m id e by O6-Cyclop en ten ylm eth ylgu a n in e (41) in
HT29 a n d Colo205 Hu m a n Color ecta l Tu m or Cell
Lin es. The most potent inactivator of AGT identified
in these studies was O6-cyclopentenylmethylguanine
(41), a compound with an IC50 of 0.39 ( 0.34 µM for
the depletion of AGT in HT29 cell extracts and 0.204

4078 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Griffin et al.
3-H), 2.30-2.50 (4H, m, 2-H and 4-H), 7.31-7.39 (6H, m, Ph),
7.61-7.72 (4H, m, Ph); MS (EI) 368 (M+). Anal. (C16H16Se2)
C, H.
to 2 also proved strongly cross-resistant to 41, suggest-
ing that the phenyl and cyclopentenyl groups of these
two compounds make similar binding interactions within
the active site of AGT. In light of these observations, it
appears that the development of AGT inactivators active
against benzylguanine-resistant mutants will require
more dramatic structural modifications to be made to
the O6-alkylguanine pharmacophore.
1-Hyd r oxym eth yl-1-p h en ylselen ocyclobu ta n e (10). A
solution of 9 (1.5 g, 3.96 mmol) in anhydrous THF (redistilled
from LiAlH4) (14 mL) was stirred at -78 °C under argon
(previously rendered oxygen and water free by sequential
passage through chromous solution, concentrated sulfuric acid,
and solid potassium hydroxide). tert-Butyllithium (2.6 mL, 4.0
mmol, 1.54 M in THF) was added over 30 min and the mixture
was stirred at -78 °C for a further 6 h. Methanal, generated
by heating polyacetal (2.0 g),44 was passed directly into the
reaction mixture and the reaction was allowed to warm to room
temperature over 12 h. After removal of solvents under
reduced pressure, saturated ammonium chloride solution (50
mL) was added, and the solution was extracted with ether (3
× 30 mL). The combined ether extracts were dried (MgSO4)
and the ether removed to yield the product as a brown solid.
Purification by chromatography on silica, employing petrol:
ether (7:3) as eluent, afforded the title compound as a pale
Exp er im en ta l Section
Melting points were obtained on a Kofler hot stage ap-
paratus and are uncorrected. Infrared spectra (IR) were
recorded as KBr disks on a Nicolet 20 PC Fourier transform
spectrometer. Mass spectra were determined on a Kratos
MS80 spectrometer in electron impact (EI) mode or fast atom
bombardment (FAB) mode using a m-nitrobenzyl alcohol
matrix. Proton (1H) and carbon (13C) nuclear magnetic reso-
nance (NMR) spectra were recorded at 200 and 50 MHz,
respectively, on a Bruker WP 200 spectrometer employing
TMS or the solvent as internal standard. Unless indicated
otherwise, spectra were recorded in [2H6]DMSO as solvent, and
J values are given in hertz (Hz). NH signals appeared as broad
singlets (br s) exchangeable with D2O.
1
brown oil (0.37 g, 41%): IR 3400, 3100, 2900, 1500 cm-1; H
NMR (CDCl3) δ 2.00 (2H, m, 3-H), 2.20 (4H, m, 2-H and 4-H),
3.43 (2H, s, CH2O), 7.10-7.30 (3H, m, Ph), 7.40-7.50 (2H, m,
Ph); MS (EI) 242 (MH+).
1-(Hydr oxym eth yl)cyclobu tyl P h en yl Selen oxide38 (11).
Sodium periodate (2.31 g, 10.82 mmol) was added to a solution
of 10 (0.62 g, 2.58 mmol) in 80% aqueous methanol (60 mL).
The mixture was stirred for 5 min, the methanol was removed
in vacuo, and saturated sodium chloride solution (40 mL) was
added. The mixture was extracted with dichloromethane (8 ×
40 mL), and the organic layers were combined, washed with
water (10 mL), and dried (MgSO4). The solvent was evaporated
under reduced pressure to yield the selenoxide as a white solid
The TLC systems employed Merck 1.05554 aluminum
sheets precoated with Kieselgel 60F254 (0.2 mm) as the
adsorbent and were visualized with UV light at 254 and 365
nm. Column chromatography was conducted under medium
pressure on silica (Kieselgel 60, 240-400 mesh). Gas chroma-
tography was performed using a Pye Unicam GCD chromato-
graph with a Carbowax 1000 (poly(ethylene glycol)) column
and Shimadzu Chromatopac CE1B integrator. Elemental
analyses were performed in house on a Carlo-Erba Instru-
mentazione 1106 analyzer or by Butterworth Laboratories,
Middlesex, U.K., and are within (0.4% of theory unless
otherwise specified. Reagents were purchased from Aldrich
Chemical Co., Gillingham, U.K., and used as received unless
otherwise stated. Ethanol and methanol were dried using Mg/
I2 and stored over 4 Å molecular sieves. Diethyl ether and
tetrahydrofuran were predried over CaCl2 and distilled from
sodium/benzophenone. Petrol (petroleum ether) refers to that
fraction in the boiling range 40-60 °C.
Meth yl 1-Cyclobu ten eca r boxyla te (8). An ethereal solu-
tion of diazomethane (3.32 mmol) was added to 1-cyclobutene-
carboxylic acid34 (0.30 g, 3.06 mmol) in ether (15 mL). After
30 min any residual diazomethane was removed by passing
N2 through the reaction mixture for 1 h, and the solvent was
evaporated under reduced pressure to afford 8 in quantitative
yield: IR 1731 cm-1; 1H NMR δ 2.60 (2H, m, 3-H),), 2.82 (2H,
t, 4-H), 3.81 (3H, s, OCH3), 7.00 (1H, t, J ) 1.3, 2-H); HRMS
(EI) m/z found 112.0523, (M+) calcd for C6H8O2 112.0524.
(0.54 g, 81%): IR 3190, 3046, 2990, 1560 cm-1 1H NMR
;
(CDCl3) δ 1.78-2.63 (6H, m, cyclobutyl), 3.78 (1H, d, J ) 12.5,
CH2O), 4.12 (1H, d, J ) 12.5, CH2O), 7.46-7.49 (3H, m, Ph),
7.65-7.70 (2H, m, Ph); MS (EI) 314 (M+).
1-Cyclobu ten em eth a n ol (12): By Ester Red u ction .
Lithium aluminum hydride (1.0 M solution in ether) (34.0 mL,
34.0 mmol) was added dropwise to a solution of 8 (6.34 g, 56.6
mmol) in ether (30 mL) at room temperature. After 15 min
water (10 mL) was cautiously added to destroy the excess of
lithium aluminum hydride, followed by hydrochloric acid (1.5
M, 10 mL). The product was extracted into ether (3 × 20 mL)
and dried (MgSO4) and the bulk of the ether removed using a
long fractionating column to afford a yellow residue (7.6 g, 83%
ether from NMR integrals): IR 3409 cm-1 1H NMR δ (1-
;
cyclobutenemethanol) 1.40 (1H, t, OH), 2.25 (2H, m, 3-H), 2.35
(2H, t, 4-H), 3.85 (2H, m, CH2O), 5.75 (1H, m, 2-H); (cyclobu-
tanemethanol) 1.60-2.20 (7H, m, cyclobutyl), 3.59 (2H, m,
CH2O); GLC (100 °C) tR (min) 20.94 (cyclobutanemethanol,
16%), 30.78 (1-cyclobutenemethanol 84%).
1-Cyclobu ten em eth a n ol (12): By Selen oxid e F r a gm en -
ta tion . A suspension of 11 (0.60 g, 2.33 mmol) and sodium
carbonate (1.24 g, 11.7 mmol) in ether (50 mL) was heated
under reflux for 12 h. The bulk of the ether was removed by
atmospheric distillation using a long fractionating column, and
the residual colorless oil was further purified by vacuum
distillation (25 °C; 0.01 mmHg) to afford the title compound
Compound 8 was also prepared by treatment of 1-cy-
clobutenecarboxylic acid (8.0 g, 82 mmol) with N,N-diisopropyl-
O-methylisourea35 (14.3 g, 90 mmol) under N2 at room
temperature for 1 h. After addition of anhydrous ether (25 mL),
the reaction mixture was stirred for 48 h, cooled to -15 °C,
and filtered. Removal of solvent gave a yellow oil which was
purified by chromatography on silica, eluting with petrol:
ether (99:1), to afford the ester 8 (7.47 g, 81%) identical to that
prepared above.
1
(0.152 g 55%): IR 3409 cm-1; H NMR (CDCl3) 1.40 (1H, t, J
) 7, OH), 2.41 (2H, m, 3-CH2), 2.51 (2H, t, J ) 3, 4-CH2), 4.01
(2H, m, OCH2), 5.87 (1H, m, 2-H): MS (EI) 83 (M+-H).
O6-Alk ylgu a n in es 7 a n d 13-21: Met h od I. Gen er a l
P r oced u r e. To a stirred solution of sodium (5.0 mol equiv) in
the appropriate alcohol (50 mL) was added 2-amino-6-chloro-
purine (1.0 mol equiv). The mixture was heated to reflux under
N2 until TLC analysis confirmed the absence of starting
materials. After cooling, the reaction mixture was neutralized
with glacial acetic acid, and the solvent was removed in vacuo.
The product was purified by chromatography on silica and/or
recrystallization from an appropriate solvent.
1,1-Bis(p h en ylselen o)cyclobu ta n e (9).37 Cyclobutanone
(2.12 g, 30.3 mmol) was added to a solution of benzeneselenol
(10.0 g, 63.70 mmol) and zinc chloride (2.19 g, 16.10 mmol) in
anhydrous carbon tetrachloride (50 mL), and the reaction
mixture was stirred at room temperature for 12 h. The reaction
mixture was partitioned between ether (90 mL) and aqueous
HCl (0.5 M, 30 mL), and the organic layer was washed
sequentially with saturated NaHCO3 solution (30 mL) and
water (30 mL) and dried (MgSO4). Removal of the solvent gave
a yellow oil which was purified by chromatography on silica,
employing petrol as eluent, to afford the title compound as a
yellow solid (10.44 g, 94%): 1H NMR (CDCl3) δ 1.94 (2H, m,
O6-Allylgu a n in e (7). Compound 7 was prepared from
2-amino-6-chloropurine (0.96 g, 5.66 mmol), sodium (0.50 g,
22.1 mmol) and allyl alcohol (13.5 mL) according to method I:

Resistance-Modifying Agents. 8
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4079
IR 3441, 3310, 3185, 2986, 2784 cm-1; 1H NMR δ 4.94 (2H, d,
OCH2), 5.28 (1H, m, CHdCH2), 5.42 (1H, m, CHdCH2), 6.19-
6.03 (1H, m, CHdCH2), 6.27 (2H, s, NH2), 7.85 (1H, s, 8-H);
MS (EI) 191 (M+).
O6-Alk ylgu a n in es 22-38: Meth od II. Gen er a l P r oce-
d u r e. The appropriate alcohol (2.5 mol equiv), dissolved in dry
THF (5-10 mL), was added to a solution of sodium hydride
(2.0 mol equiv) in dry THF (20 mL) and the mixture was
stirred for 30 min. 2-Amino-6-chloropurine (1.0 mol equiv) was
added and the mixture was heated to reflux under N2 until
TLC analysis confirmed the absence of starting material. The
reaction mixture was neutralized with glacial acetic acid, and
the solvent was removed in vacuo. The product was purified
by chromatography on silica and/or recrystallization from an
appropriate solvent.
O6-Eth ylgu a n in e (13). Compound 13 was prepared ac-
cording to method I from 2-amino-6-chloropurine (0.75 g, 4.42
mmol), sodium (0.5 g, 22 mmol) and ethanol (50 mL): IR 3505,
3484, 3432, 3324, 3191, 3110, 2984, 2901, 2705, 2544 cm-1
;
1H NMR δ 1.35 (3H, t, J ) 7.1, CH2CH3), 4.44 (2H, q, J ) 7.1,
OCH2), 6.22 (2H, br s, NH2), 7.81 (1H, s, 8-H); MS (EI) m/z
179 (M+).
O6-(5-Hexen yl)gu a n in e (22). Compound 22 was prepared
according to method II from 5-hexen-1-ol (1.47 g, 14.7 mmol),
sodium hydride (0.14 g, 5.9 mmol) and 2-amino-6-chloropurine
(0.50 g, 2.95 mmol): IR 3483, 3302, 3181 cm-1; 1H NMR δ 1.60
(2H, m, O(CH2)2CH2CH2CHdCH2), 1.87 (2H, m, OCH2CH2),
2.22 (2H, q, CH2CH), 4.49 (2H, t, J ) 6.55, OCH2), 5.1 (2H, m,
CH2dCH), 5.94 (1H, m, CH2dCH), 6.33 (2H, s, NH2), 7.93 (1H,
s, 8-H), 12.50 (1H, br s, NH); 13C NMR δ (50.3 MHz, DMSO-
d6) 160.1, 138.9, 115.3, 65.6, 33.2, 28.3, 25.1; MS (EI) 233 (M+).
Anal. (C11H15N5O) C, H, N.
O6-n -P r op ylgu a n in e (14). Compound 14 was prepared
from 2-amino-6-chloropurine (0.50 g, 2.95 mmol), sodium (0.35
g, 15.2 mmol) and propan-1-ol (30 mL) according to method I:
IR 3490, 3301, 3173, 2975, 2886, 2780, 2539 cm-1; 1H NMR δ
0.97 (3H, t, J ) 7, CH3), 1.76 (2H, sextet, J ) 7, CH2CH2CH3),
4.33 (2H, t, J ) 7, OCH2), 6.22 (2H, br s, NH2), 7.80 (1H, s,
8-H); HRMS (EI) m/z found 193.0938, (M+) calcd for C8H11N5O
193.0912.
O6-n -Bu tylgu a n in e (15). Compound 15 was prepared by
method I from 2-amino-6-chloropurine (0.50 g, 2.95 mmol),
sodium (0.34 g, 15.0 mmol) and butan-1-ol (20 mL): IR 3501,
3374, 3106, 2955, 2874, 2803 cm-1; 1H NMR δ 0.93 (3H, t, J )
7.2, CH3), 1.42 (2H, m, CH2CH3), 1.73 (2H, m, CH2CH2CH2),
4.38 (2H, t, J ) 6.7, OCH2), 6.22 (2H, br s, NH2), 7.80 (1H, s,
8-H); MS (EI) m/z 207 (M+).
O6-[(Z)-3-Hexen yl]gu a n in e (23). Compound 23 was pre-
pared from (Z)-3-hexen-1-ol (1.47 g, 14.7 mmol) sodium hydride
(0.14 g, 5.9 mmol) and 2-amino-6-chloropurine (0.50 g, 2.95
1
mmol) according to method II: IR 3501, 3399, 3206 cm-1; H
NMR δ 1.02 (3H, t, CH2CH3), 2.15 (2H, p, CH2CH3), 4.48 (2H,
t, J ) 6.9, OCH2), 5.57 (2H, m, CHdCH), 6.32 (2H, s, NH2),
7.92 (1H, s, 8-H), 12.58 (1H, br s, NH); 13C NMR δ (50.3 MHz,
DMSO-d6) 160.0, 138.2, 134.2, 124.8, 65.2, 26.9, 20.5, 14.4;
HRMS (EI) m/z 233.1271, (M+) calcd for C11H15N5O 233.1277.
O6-[(E)-3-Hexen yl]gu a n in e (24). Compound 24 was pre-
pared by method II from (E)-3-hexen-1-ol (1.47 g, 14.7 mmol)
sodium hydride (0.14 g, 5.9 mmol) and 2-amino-6-chloropurine
(0.50 g, 2.95 mmol): IR 3500, 3190, 3005 cm-1; 1H NMR δ 1.03
(3H, t, J ) 6.4, CH3), 2.09 (2H, p, J ) 6.9, CH2CH3), 2.50-
2.60 (m, CH2CHCH), 5.40-5.70 (2H, m, CHdCH), 6.35 (2H,
s, NH2), 7.91 (1H, s, 8-H), 12.50 (1H, s, NH); MS (EI) 233 (M+).
O6-(3-Hexyn yl)gu a n in e (25). Compound 25 was prepared
according to method II from hex-3-yn-1-ol (1.44 g, 14.7 mmol),
sodium hydride (0.14 g, 5.9 mmol) and 2-amino-6-chloropurine
(0.50 g, 2.95 mmol): IR 3503, 3314, 3284 cm-1; 1H NMR δ 1.13
(3H, t, J ) 7.5, CH2CH3), 2.24 (2H, q, J ) 7.5, CH2CH3), 2.75
(2H, t, J ) 7.0, OCH2CH2), 4.48 (2H, t, J ) 7.0, OCH2), 6.41
(2H, s, NH2), 7.98 (1H, s, 8-H); HRMS (EI) m/z 231.1123, (M+)
calcd for C11H13N5O 231.1120.
O6-(2-P h en yleth yl)gu a n in e (26). Compound 26 was pre-
pared according to method II from 2-phenylethanol (3.0 g, 25.0
mmol), sodium hydride (0.27 g, 11.0 mmol) and 2-amino-6-
chloropurine (0.75 g, 4.42 mmol): IR 3495, 3366, 3127, 2982,
2801 cm-1; 1H NMR δ 3.08 (2H, t, J ) 7.2, CH2Ph), 4.58 (2H,
t, J ) 7.2, OCH2), 6.27 (2H, s, NH2), 7.22-7.37 (5H, m, Ph),
7.80 (1H, s, 8-H); MS (EI) 255 (M+). Anal. (C13H13N5O) C, H,
N.
O6-(3-Meth ylbu tyl)gu a n in e (16). Compound 16 was syn-
thesized according to method I from 2-amino-6-chloropurine
(1.5 g, 8.84 mmol), sodium (1.0 g, 44.2 mmol) and 3-methyl-
1-butanol (60 mL): IR 3505, 3310, 3183, 2961, 2919, 2708,
1
2552 cm-1; H NMR δ 0.93 (6H, d, J ) 6.3, CH(CH3)2), 1.65
(2H, dt, J ) 6.7, J ) 6.3, CHCH2CH(CH3)2), 1.74 (1H, septet,
J ) 6.3, CH(CH3)2), 4.41 (2H, t, J ) 6.7, OCH2), 6.23 (2H, br
s, NH2), 7.79 (1H, s, 8-H); HRMS (EI) m/z found 221.1266, (M+)
calcd for C10H15N5O 221.1255. Anal. (C10H15N5O) C, H, N.
O6-n -P en tylgu a n in e (17). Compound 17 was prepared
according to method I from 2-amino-6-chloropurine (0.75 g,
4.42 mmol), sodium (0.50 g, 22.1 mmol) and pentan-1-ol (20
mL): IR 3497, 3364, 3107 cm-1 1H NMR δ 0.99 (3H, t,
;
CH2CH3), 1.47 (4H, m, CH2CH2CH3), 1.82 (2H, m, OCH2CH2),
4.47 (2H, t, J ) 6.7, OCH2), 6.32 (2H, br s, NH2), 7.92 (1H, s,
8-H), 12.58 (1H, br s, NH); MS (EI) m/z 221 (M+). Anal.
(C10H15N5O) C, H, N.
O6-n -Hexylgu a n in e (18). Compound 18 was synthesized
following method I, from 2-amino-6-chloropurine (0.75 g, 4.42
mmol), sodium (0.50 g, 22.1 mmol) and hexan-1-ol (20 mL):
IR 3499, 3360, 3181 cm-1 1H NMR δ 0.99 (3H, t, J ) 6.7,
;
CH2CH3), 1.42 (6H, m, (CH2)3CH3), 1.80 (2H, m, OCH2CH2),
4.05 (2H, t, J ) 6.6, OCH2), 6.32 (2H, s, NH2), 7.92 (1H, s,
8-H), 12.50 (1H, br s, NH); MS (EI) 235 (M+). Anal. (C11H17N5O)
C, H, N.
O6-n -Hep tylgu a n in e (19). Compound 19 was prepared
according to method I from 2-amino-6-chloropurine (0.75 g,
4.42 mmol), sodium (0.50 g, 22.1 mmol) and heptan-1-ol (20
O6-(2-Eth yla llyl)gu a n in e (27). Compound 27 was pre-
pared by method II from 2-ethylallyl alcohol45 (1.0 g, 11.6
mmol) sodium hydride (0.60 g, 25 mmol) and 2-amino-6-
chloropurine (1.0 g, 5.90 mmol): IR 3465, 3306, 3200, 3137,
2965, 2940, 2915, 2882, 2803, 1630, 1584 cm-1; 1H NMR δ 1.06
(3H, t, J ) 7.4, CH2CH3), 2.13 (2H, q, J ) 7.4, CH2CH3), 4.92
(2H, s, OCH2), 4.96 (1H, s, dCH2), 5.12 (1H, s, dCH2), 6.26
(2H, br s, NH2), 7.83 (1H, s, 8-H); 13C NMR δ 160, 155.5, 146.5,
138.1, 113.8, 110.9, 67.8, 25.8, 12.1; HRMS (EI) m/z 219.1121,
(M+) calcd for C10H13N5O 219.112.
1
mL): IR 3499, 3300, 3179 cm-1; H NMR δ 0.96 (3H, t, J )
6.4, CH3), 1,38 (8H, m, (CH2)4CH3), 1.84 (2H, m, OCH2CH2),
4.47 (2H, t, J ) 6.6, OCH2), 6.33 (2H, s,NH2), 7.89 (1H, s, 8-H),
12.50 (1H, br s, NH); MS (EI) 249 (M+). Anal. (C12H19N5O) C,
H, N.
O6-(Cycloh exylm eth yl)gu a n in e (20). Compound 20 was
prepared according to method I from 2-amino-6-chloropurine
(0.50 g, 2.95 mmol), sodium (0.5 g, 22 mmol) and cyclohexyl-
methanol (6 mL): IR 3507, 3301, 3189, 3137, 2924, 2851, 2799
cm-1; H NMR δ 1.06-1.83 (11H, m, C6H11), 4.21 (2H, d, J )
O6-(2-Isop r op yla llyl)gu a n in e (28). Compound 28 was
prepared according to method II from 2-isopropylallyl alcohol45
(1.77 g, 17.7 mmol), sodium hydride (0.60 g, 25 mmol) and
2-amino-6-chloropurine (1.0 g, 5.90 mmol): IR 3322, 3189,
1
6.2, OCH2), 6.24 (2H, s, NH2), 7.81 (1H, s, 8-H); MS (EI) 247
(M+). Anal. (C12H17N5O) C, H, N.
O6-(2-Meth yla llyl)gu a n in e (21). Compound 21 was pre-
pared by method I from 2-amino-6-chloropurine (0.50 g, 2.95
mmol), sodium (0.4 g, 17.4 mmol) and 2-methyl-2-propen-1-ol
(10 mL), with the addition of THF (10 mL): IR 3494, 3314,
3185, 2978, 2782 cm-1; 1H NMR δ 1.78 (3H, s, CH3), 4.86 (2H,
s, OCH2), 4.93 (1H, s, dCH2), 5.10 (1H, s, dCH2), 6.28 (2H, s,
NH2), 7.84 (1H, s, 8-H); HRMS (EI) m/z 205.0967, (M+) calcd
for C10H12N2O3 205.0972.
1
2963, 2872, 2789 cm-1; H NMR δ 1.07 (6H, d, J ) 6.8, CH-
(CH3)2), 2.39 (1H, septet, J ) 6.8, CH(CH3)2), 4.95 (2H, s,
OCH2), 4.97 (1H, d, J ) 1, dCH2), 5.10 (1H, d, J ) 1, dCH2),
6.26 (2H, br s, NH2), 7.81 (1H, s, 8-H); HRMS (EI) m/z
233.1268, (M+) calcd for C11H15N5O 233.1260.
O6-(2-Meth yl-2-bu ten ylm eth yl)gu a n in e (29). Compound
29 was prepared following method II, from 2-methyl-2-buten-

4080 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Griffin et al.
1-ol (2.0 g, 23 mmol),45 sodium hydride (0.37 g, 15.4 mmol)
and 2-amino-6-chloropurine (1.31 g, 7.7 mmol): IR 3500 3350
O6(1-Cyclobu ten ylm eth yl)gu a n in e (38). Compound 38
was synthesized according to method II from 1-cyclobutene-
methanol 12 (0.09 g, 1.07 mmol), sodium hydride (0.10 g, 2.5
mmol) and 2-amino-6-chloropurine (0.12 g, 0.71 mmol): 1H
NMR (CD3OD) δ 2.48 (2H, m, 3′-H), 2.68 (2H, t, 4′-H), 4.98
(2H, s, CH2O), 6.03 (1H, s, 2′-H), 7.98 (1H, s, 8-H); HRMS (EI)
m/z 217.0975, (M+) calcd for C10H11N5O 217.0987.
Compound 38 was also prepared as follows: A mixture of
12 and cyclobutylmethanol (84:16) (0.84 g, 10 mmol) was added
to sodium hydride (0.082 g, 3.41 mmol) in DMSO (0.3 mL) and
the mixture was stirred for 1 h. 2-Amino-6-(azonia-4-azabicy-
clo[2.2.2-oct-1-yl]purine chloride (DABCO-purine; 0.50 g, 1.72
mmol)36 was added and the reaction mixture was stirred for a
further 12 h at room temperature. After addition of acetic acid
(0.06 mL), the mixture was adsorbed onto neutral alumina and
purified by chromatography on neutral alumina, employing
7% MeOH in CH2Cl2 as eluent, affording 38 as a white solid
(0.05 g, 13%).
O6-Alk ylgu a n in es 39-43: Meth od III. Gen er a l P r oce-
d u r e. A mixture of the appropriate alcohol (5.5 mol equiv) and
sodium hydride (2.0 mol equiv) in anhydrous DMSO was
stirred for 1 h, and 2-aminotrimethylpurin-6-ylammonium
chloride32 (1.0 mol equiv) was added. The reaction mixture was
stirred under N2 at room temperature until TLC analysis
confirmed the absence of starting material. The reaction
mixture was neutralized with glacial acetic acid, and the
solvent was evaporated under reduced pressure. The product
was purified by chromatography on silica and/or recrystalli-
zation from an appropriate solvent.
1
3200 1610 cm-1; H NMR δ 1.80 (6H, m, 2 × CH3), 4.95 (2H,
s, OCH2), 5.75 (1H, q, J ) 5.5, 3′-H), 6.35 (2H, s, NH2), 7.90
(1H, s, 8-H), 12.58 (1H, br s, NH); MS (EI) 220 (MH+). Anal.
(C10H13N5O) C, H, N.
O6-(2-P h en yla llyl)gu a n in e (30). Compound 30 was pre-
pared from 3-hydroxy-2-phenyl-1-propene (820 mg, 6.12 mmol),46
sodium hydride (0.45 g, 7.90 mmol) and 2-amino-6-chloro-
purine (0.70 g, 4.13 mmol) according to method II: IR 3484,
1
3326, 3189, 2787, 1622, 1586 cm-1; H NMR (methanol-d4) δ
5.65 (2H, s, OCH2), 5.73 (1H, s, dCH2), 5.82 (1H, s, dCH2),
7.46-7.54 (3H, m, Ph), 7.57-7.75 (2H, m, Ph), 8.02 (1H, s,
8-H); MS (EI) 267 (M+).
O6-P r op a r gylgu a n in e (31). Compound 31 was prepared
by method II from propargyl alcohol (3 mL), sodium hydride
(0.50 g, 22.0 mmol) and 2-amino-6-chloropurine (1.0 g, 5.90
mmol): IR 3501, 3397, 3173, 2982, 2787, 2124 cm-1; 1H NMR
δ 3.59 (1H, t, J ) 2.4, dCH), 5.11 (2H, d, J ) 2.4, CH2CdC),
6.34 (2H, s, NH2), 6.56 (1H, s, 8-H); HRMS (EI) m/z 189.0644,
(M+) calcd for C8H7N5O 189.0638.
O6-(Cycloh ex-3-en -1-ylm eth yl)gu a n in e (32). Compound
32 was prepared according to method II from cyclohex-3-en-
1-ylmethanol (6.16 g, 55 mmol), sodium hydride (0.53 g, 22
mmol) and 2-amino-6-chloropurine (1.87 g, 11 mmol): IR 3340,
1
3200, 2900, 1620, 1580 cm-1; H NMR δ 1.40 (1H, m, 1′-H),
1.85-2.10 (6H, 3 × m, 2′-H, 5′-H, 6′-H), 4.30 (2H, d, J ) 7,
OCH2), 5.70 (2H, s, 3′-H, 4′-H), 6.15 (2H, s, NH2), 7.76 (1H, s,
8-H), 12.35 (1H, br s, NH); MS (FAB) 246 (MH+). Anal.
(C12H15N5O‚0.25MeOH) C, H, N.
O6-(Cyclobu tylm eth yl)gu a n in e (39). Compound 39 was
prepared following method III, from cyclobutanemethanol (5.00
g, 58 mmol), sodium hydride (0.51 g, 21 mmol) and 2-amino-
trimethylpurin-6-ylammonium chloride (2.41 g, 11 mmol) in
O6-(2,2-Dim eth yl-1,3-dioxolan -4-ylm eth oxy)gu an in e (33).
Compound 33 was prepared by method II from 2,2-dimethyl-
1,3-dioxolan-4-methanol (12 g, 90.80 mmol), sodium hydride
(1.27 g, 10.61 mmol) and 2-amino-6-chloropurine (3.0 g, 17.69
mmol): IR 3413, 3179, 2990 cm-1; 1H NMR δ 1.31 (3H, s, Me),
1.37 (3H, s, Me), 3.77 (1H, dd, J ) 5.4, J ) 8.1, OCH2CHCH2),
4.11 (1H, dd, J ) 5.4, J ) 8.1, OCH2CHCH2),), 4.39-4.47 (1H,
m, OCH2CH), 4.43 (2H, d, J ) 2.4, OCH2), 6.27 (2H, s, NH2),
7.85 (1H, s, 8-H); HRMS (EI) m/z 265.1174, (M+) calcd for
C11H15N5O3 265.1173.
DMSO (6 mL): IR 3469, 3333, 3164, 2977, 1671, 1629 cm-1
;
1H NMR δ 1.8-2.3 (6H, m, 2′-CH2, 3′-CH2, 4′-CH2), 2.85 (1H,
m, (1′-H), 4.47 (2H, d, J ) 7.0, OCH2), 6.31 (2H, s, NH2), 7.89
(1H, s, 8-H), 12.50 (1H, br s, NH); MS (EI) 219 (M+). Anal.
(C10H13N5O) C, H, N.
Compound 39 was also prepared according to method II
from cyclobutanemethanol (1.2 g, 14.7 mmol), sodium hydride
(1.4 g, 5.9 mmol), and 2-amino-6-chloropurine (0.50 g, 2.95
mmol).
O6-(Cyclop en tylm eth yl)gu a n in e (40). Compound 40 was
prepared according to method III from cyclopentanemethanol
(0.20 g, 1.99 mmol), sodium hydride (17 mg, 0.71 mmol) and
2-aminotrimethylpurin-6-ylammonium chloride (2.41 g, 11
mmol) in DMSO (0.4 mL): IR 3481, 3352, 3204, 2957, 1626,
1581 cm-1; 1H NMR δ 1.2-2.0 (8H, m, 2′-CH2, 3′-CH2, 4′-CH2,
5′-CH2), 2.47 (1H, m, 1′H), 4.38 (2H, d, J ) 7, OCH2), 6.35
(2H, s, NH2), 7.93 (1H, s, 8-H), 12.50 (1H, s, NH); MS (EI) 233
(M+). Anal. (C11H15N5O‚0.5H2O) C, H, N.
O6-(2,2-Dieth oxyp r op -1-yl)gu a n in e (34). Compound 34
was prepared according to method II from 2,2-diethoxy-1-
propanol (3 mL), sodium hydride (0.35 g, 14.74 mmol) and
2-amino-6-chloropurine (1.0 g, 2.95 mmol): IR 3465, 3314,
1
2923 cm-1; H NMR δ 1.09 (3H, t, J ) 7.1, OCH2CH3), 1.37
(3H, s, CH3), 3.48 (1H, q, J ) 7.1, OCH2CH3), 3.48 (1H, q, J )
7.1, OCH2CH3), 4.39 (2H, s, OCH2), 6.29 (2H, s, NH2), 7.83
(1H, s, 8-H); MS (EI) 281 (M+). Anal. (C12H19N5O3) C, H, N.
O6-(2,2-Dim eth oxybu t-1-yl)gu a n in e (35). Compound 35
was prepared according to method II from 2,2-dimethoxy-1-
butanol47 (0.94 g, 7 mmol), sodium hydride (0.48 g, 20 mmol)
and 2-amino-6-chloropurine (1.0 g, 5.9 mmol): IR 3480, 3434,
3318, 3206, 3104, 2973, 2840, 2795 (cm-1); 1H NMR δ 0.82 (3H,
t, J ) 7.6, CH2CH3), 1.71 (2H, q, J ) 7.6, CH2CH3), 3.14 (6H,
s, OCH3), 4.38 (2H, s, OCH2), 6.29 (2H, s, NH2), 7.82 (1H, s,
8-H); MS (EI) 267 (M+) Anal. (C11H17N5O3) C, H, N.
Compound 40 was also prepared according to method II
from cyclopentanemethanol (1.45 g, 14.5 mmol), sodium hy-
dride (139 mg, 5.8 mmol) and 2-amino-6-chloropurine (0.50 g,
2.9 mmol).
O6-(1-Cyclop en ten ylm eth yl)gu a n in e (41). Compound 41
was prepared by method III from 1-cyclopentenemethanol49
(6.57 g, 67 mmol), sodium hydride (0.57 g, 24 mmol) and
2-aminotrimethylpurin-6-ylammonium chloride (2.74 g, 12
O6-(3-Met h yl-2-oxob u t -1-yl)gu a n in e E t h ylen e Acet a l
(36). Compound 36 was prepared according to method II from
1-hydroxy-3-methylbutan-2-one ethylene acetal48 (0.95 g, 6.48
mmol), sodium hydride (0.26 g, 11 mmol) and 2-amino-6-
chloropurine (0.73 g, 4.32 mmol): IR 3459, 3343, 3223, 3133,
1
mmol) in DMSO (6 mL): IR 3460, 3300, 1280, 1150 cm-1; H
NMR δ 1.90 (2H, q, J ) 7.4, 4′-CH2), 2.35 (4H, m, 3′-CH2, 5′-
CH2), 5.00 (2H, s, OCH2), 5.75 (1H, s, 2′-H), 6.15 (2H, s, NH2),
7.75 (1H, s, 8-H), 12.35 (1H, br s, NH); MS (FAB) 232 (MH+).
Anal. (C11H13N5O‚0.4EtOH) C, H, N.
O6-(1-Cycloh exen ylm eth yl)gu a n in e (42). Compound 42
was prepared according to method III from 1-cyclohexene-
methanol50 (2.04 g, 18.2 mmol), sodium hydride (0.16 g, 6.6
mmol) and 2-aminotrimethylpurin-6-ylammonium chloride
(0.75 g, 3.3 mmol) in DMSO (6 mL): IR 3457, 3295, 3186, 2931,
1698, 1631 cm-1; 1H NMR δ 1.68-2.14 (8H, 2 × m, 3′-CH2 and
6′-CH2, 4′-CH2, 5′-CH2), 4.88 (2H, s, OCH2), 5.93 (1H, s, 2′-H),
6.33 (2H, s, NH2), 7.92 (1H, s, 8-H), 12.50 (1H, br s, NH); MS
(EI) 245 (M+). Anal. (C12H15N5O‚0.85H2O) C, H, N.
1
2980, 2878, 2799 cm-1; H NMR δ 0.94 (6H, d, J ) 6.9, CH-
(CH3)2), 2.11 (1H, septet, J ) 6.9, CH(CH3)2), 3.86-4.09 (4H,
m, OCH2CH2O), 4.42 (2H, s, OCH2), 6.27 (2H, s, NH2), 7.83
(1H, s, 8-H); MS (EI) 279 (M+) Anal. (C12H17N5O3) C, H, N.
O6-(2,2-Dim eth oxy-2-p h en yleth yl)gu a n in e (37). Com-
pound 37 was prepared by method II from 2,2-dimethoxy-2-
phenylethanol48 (0.68 g, 3.74 mmol), sodium hydride (0.23 g,
9.6 mmol) and 2-amino-6-chloropurine (0.32 g, 1.87 mmol): IR
3497, 3438, 3310, 3206, 2946, 2832, 1626 cm-1; 1H NMR δ 3.18
(6H, s, OCH3), 4.68 (2H, s, CH2), 6.29 (2H, br s, NH2), 7.34-
7.38 (3H, m, Ph), 7.53-7.57 (2H, m, Ph), 7.77 (1H, s, 8-H);
MS (EI) 315 (M+). Anal. (C15H17N5O3) C, H, N.

Resistance-Modifying Agents. 8
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4081
O6-[(S )-(4-Isop r op e n ylcycloh e x-1-e n yl)m e t h yl]gu a -
n in e (43). Compound 43 was synthesized according to method
III from (S)-perillyl alcohol (3.66 g, 24 mmol), sodium hydride
(0.21 g, 8.8 mmol) and 2-aminotrimethylpurin-6-ylammonium
chloride (1.00 g, 4.4 mmol) in DMSO (6 mL): IR 3460, 3404,
DMSO. HT29 cells were seeded at 5 × 106 cells in 7-cm radius
Petri dishes in EMEM (Gibco-GBL Life Technologies Ltd.,
Paisley, Scotland) medium supplemented with 10% (v/v) fetal
calf serum (FCS; Globepharm, Esher, U.K.). After 3 days the
growth medium was replaced with serum free EMEM contain-
ing 1% DMSO ( AGT inactivators for 4 h. Cells were harvested
by trypsination, resuspended in serum-containing EMEM, and
centrifuged for 5 min at 1750g at 4 °C. The pellet was washed
twice with EMEM and resuspended in EMEM, and the cells
were disaggregated by passing through a 25 G 5/8 hypodermic
needle. Viable cell number was determined by trypan blue
excluding haemocytometer counts. 107 Cells were pelleted by
centrifugation, snap-frozen in liquid nitrogen and stored at
-80 °C. Extracts were prepared by resuspending cell pellets
in 3 mL of extraction buffer (50 mM Tris-HCl, pH 8.3, 0.5 mM
EDTA disodium salt, 1 mM dithiothreitol, 200 mM NaCl) and
homogenizing with a T25 Ultra-Turrax homogenizer (SH
Scientific, Northumberland, U.K.) fitted with an 8-mm head
for 15 s at 8000 rpm, 3 × 5 s, each separated by a 30-s period
during which the sample was allowed to cool on ice. The
homogenizer head was then rinsed using 2 mL of extraction
buffer, which was added to the cell homogenate, and the total
volume was centrifugated for 5 min at 1750g, 4 °C. The
supernatant liquid was snap-frozen in liquid nitrogen and
stored at -80 °C. The AGT activity of the extracts thus
obtained was subsequently determined by the AGT assay
described previously,29 and from these data the concentration
of AGT inactivator required to deplete whole HT29 cell AGT
activity by 50% was calculated.
1
3315, 3205, 2964, 1626, 1584 cm-1; H NMR δ 1.60 (1H, m,
4′-H), 1.82 (3H, s, CH3), 1.9-2.2 (6H, m, 3′-CH2, 5′-CH2, 6′-
CH2), 4.82 (2H, s, OCH2), 4.91 (2H, s, CH2)), 5.96 (1H, s, 2′-
H), 6.35 (2H, s, NH2), 7.93 (1H, s, 8-H), 12.50 (1H, br s, NH);
MS (EI) 285 (M+). Anal. (C15H19N5O‚0.2MeOH) C, H, N.
O6-(2-Oxo-2-a lk yl)gu a n in es 44-47: Meth od IV. Gen -
er a l P r oced u r e. A solution of the appropriate O6-(2-oxo-2-
alkyl)guanine acetal in aqueous acetic or trifluoroacetic acid
(5-20 mL) was stirred at ambient temperature until TLC
analysis confirmed the absence of starting material. Solvents
were removed under reduced pressure and the product was
purified by chromatography on silica and/or recrystallization
from the appropriate solvent.
O6-(2-Oxop r op yl)gu a n in e (44). Compound 44 was pre-
pared according to method IV from O6-(2,2-diethoxypropyl)-
guanine (34) (0.50 g, 1.78 mmol) by treatment with aqueous
acetic acid (1 M, 12 mL): IR 3355, 3119, 2780, 1734 cm-1; 1H
NMR δ 2.18 (3H, s, COCH3), 5.07 (2H, s, OCH2), 6.27 (2H, s,
NH2), 7.90 (1H, s, 8-H); HRMS (EI) m/z 207.0765, (M+) calcd
for C8H9N5O2 207.0756). Anal. (C8H9N5O2‚0.25AcOH‚0.8H2O)
C, H, N.
O6-(2-Oxobu tyl)gu a n in e (45). Compound 45 was prepared
from O6-(2,2-dimethoxybutyl)guanine (35) (0.08 g, 0.3 mmol)
and 50% aqueous acetic acid (5 mL) according to method IV:
1
IR 3488, 3380, 3322, 3110, 2982, 2942, 2782, 1732 cm-1; H
Effect of AGT In a ctiva tor s on Tem ozolom id e Cyto-
toxicity in HT29 a n d Colo205 Hu m a n Tu m or Cells.
Potentiation of Temozolomide activity by 2 and 41 was
determined by means of a colony-forming assay. Cells were
seeded in 6-well dishes (Nunc) in EMEM + 10% FCS, and after
72 h exponentially growing cultures were exposed to medium
supplemented with 10 µM 2 or 41, in a final concentration of
0.1% DMSO (v/v). After a 1-h preincubation with the AGT
inactivator, varying concentrations of Temozolomide dissolved
in DMSO were added to give a final concentration range of
0-1000 µM in 1% DMSO (v/v), control cells being treated with
1% v/v DMSO only. After incubation at 37 °C for 4 h the cells
were harvested by trypsinisation, resuspended in 2 mL of
EMEM, disaggregated and counted as described above. A
known number of HT29 cells was transferred into 10-cm
plastic Petri dishes containing EMEM. Colo205 cells have poor
cloning efficiency on plastic and thus were dispensed into
tissue culture tubes containing 0.15% (w/v) agarose (Seakem
ME; FMC Bioproducts, Rockland, ME) in EMEM. The AGT
inactivators were also present throughout the cloning period
at a concentration of 10 µM (0.1% v/v DMSO final concentra-
tion). After approximately 2 weeks adherent HT29 colonies
were fixed with acetic acid:methanol (3:1 v/v), stained with
0.4% crystal violet (N-hexamethyl-p-rosaniline; Sigma) (w/v)
for 1 min, rinsed with water, air-dried and counted by eye.
Colonies of Colo205 cells were visualized by adding 1 mL 0.5
mg/mL of MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltet-
razolium bromide; Sigma), incubating at 37 °C overnight and
counting by eye. The percentage cloning efficiencies for
untreated HT29 and Colo205 cells were 62 ( 11% (n ) 6) and
50 ( 15% (n ) 6), respectively. From the cloning efficiencies
relative to control the % cell survival at each concentration of
Temozolomide ((10 µM AGT inactivator) was calculated. The
Temozolomide concentration associated with 50% inhibition
of colony formation was determined from a point-to-point curve
plot using GraphPad PRISM software.
NMR δ 0.98 (3H, t, J ) 7.3, CH2CH3), 2.54 (2H, q, J ) 7.3,
CH2CH3), 5.06 (2H, s, OCH2), 6.22 (2H, s, NH2), 7.86 (1H, s,
8-H); HRMS (EI) m/z 221.0937, (M+) calcd for C9H11N5O2
221.0962.
O6-(3-Meth yl-2-oxobu tyl)gu a n in e, TF A Sa lt (46). Com-
pound 46 was prepared by method IV from O6-(3-methyl-2-
oxobutyl)guanine ethylene acetal (36) (0.2 g, 0.72 mmol) and
80% aqueous trifluoroacetic acid (10 mL): IR 3854, 3501, 3320,
1
3183, 2977, 2940, 2791, 1732 cm-1; H NMR δ 1.09 (6H, d, J
) 6.9, CH(CH3)2), 2.80 (1H, septet, J ) 6.9, CH(CH3)2), 5.26
(2H, s, OCH2), 6.62 (2H, s, NH2), 8.23 (1H, s, 8-H); HRMS (EI)
m/z 235.1063, (M+) calcd for C10H13N5O2 235.1057.
O6-(2-Oxo-2-p h en yleth yl)gu a n in e (47). Compound 47
was prepared by method IV from O6-(2,2-dimethoxy-2-phenyl-
ethyl)guanine (37) (0.09 g, 0.29 mmol) and aqueous acetic acid
(3 M, 20 mL): IR 4390, 3391, 3061, 2973, 2924, 1690 cm-1; 1H
NMR δ 5.90 (2H, s, CH2), 6.18 (2H, br s, NH2), 7.55-7.71 (3H,
m, Ph), 7.87 (1H, s, 8-H), 8.00-8.04 (2H, m, Ph); MS (EI) 270
(MH+). Anal. (C13H11N5O2) C, H, N.
In a ctiva tion of AGT in HT29 Cell Extr a cts. The inac-
tivation of AGT in HT29 human colorectal tumor cell extracts
by guanine derivatives, and the calculation of IC50 concentra-
tions, was performed as previously described.29
In a ctiva tion of P u r ified Wild -Typ e a n d Mu ta n t AGT.
Recombinant wild-type and mutant AGT proteins P140K,
Y158H, G160R, and G156A were produced in Escherichia coli
and purified as previously described by immobilized metal
affinity chromatography.15,51 (The amino acid sequence at the
amino terminus of the protein was changed from M- to M-GS-
(H)6GS- to allow this purification.) The purified AGT protein
was incubated for 30 min at 37 °C with the concentrations of
2 or 41 indicated in 0.5 mL of 50 mM Tris-HCl, pH 7.5, 0.1
mM EDTA, 5 mM dithiothreitol and 50 µg of hemocyanin in a
total volume of 0.5 mL. The residual AGT activity was then
determined by a further 30-min incubation with a methylated
DNA substrate.16 The results were expressed as the percentage
of the AGT activity lost in response to the inhibitor and an
ED50 value obtained from a plot of this value against the
amount of inhibitor added. Due to solubility considerations,
the maximum inhibitor concentration used was 1.2 mM.
Dep letion of AGT in In ta ct HT29 Cells by AGT In a c-
tiva tor s. All inhibitors were dissolved in 100% dry DMSO at
100 or 1000 times the final desired concentration to allow
dilution in medium to give a final concentration of 1% or 0.1%
Ack n ow led gm en t. We thank the Cancer Research
Campaign, Cancer Research Campaign Technology, and
AstraZeneca Pharmaceuticals for generous support and
also the Engineering and Physical Sciences Research
Council, U.K., for the award of a studentship to C.D.
This work was supported in part by Award CA-71976
(A. E. Pegg, PI).

4082 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Griffin et al.
(21) 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.
(22) 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.
(23) McElhinney, S.; Donnelly, J . D.; McCormick, J . E.; Kelly, J .;
Watson, A. J .; Rafferty, J . A.; Elder, R. H.; Middleton, M. R.;
Willington, M. A.; McMurry, T. B. H.; Margison, G. P. Inactiva-
tion of O6-alkylguanine-DNA alkyltransferase. 1. Novel O6-
(hetarylmethyl)guanines having basic rings in the side chain.
J . Med. Chem. 1998, 41, 5265-5271.
(24) Cussac, C.; Rapp, M.; Mounetou, E.; Madelmont, J . C.; Maurizis,
J . C.; Godeneche, D.; Dupuy, J . M.; Sauzieres, J .; Baudry. J . P.;
Veyre, A. Enhancement by O6-benzyl-N-acetylguanosine deriva-
tives of chloroethylnitrosourea antitumor action in chloroeth-
ylnitrosourea-resistant human malignant melanocytes. J . Phar-
macol. Exp. Ther. 1994, 271, 1353-1358.
(25) Cussac, C.; Mounetou, E.; Rapp, M.; Madelmont, J . C.; Maurizis,
J . C.; Labarre, P.; Chollet, P.; Chabard, J . L.; Godeneche, D.;
Baudry. J . P.; Veyre, A. Disposition and metabolism of O6-
alkylguanine-DNA alkyltransferase inhibitor in nude mice bear-
ing human melanoma. Drug Metab. Dispos. 1994, 22, 637-642.
(26) Schold, S. C.; Kokkinakis, D. M.; Rudy, D. M.; Moschel, R. C.;
Pegg, A. E. Treatment of human brain tumor xenografts with
O6-benzyl-2′-deoxyguanosine and BCNU. Cancer Res. 1996, 56,
2076-2081.
Refer en ces
(1) For part 7, see: Barlow, H. C.; Bowman, K. J .; Curtin, N. J .;
Calvert, A. H.; Golding, B. T.; Huang, B.; Loughlin, P. J .; Newell,
D. R.; Smith, P. G.; Griffin, R. J . Resistance modifying agents.
7. 2,6-Disubstituted-4,8-dibenzylaminopyrimido[5,4-d]pyrim-
idines that inhibit nucleoside transport in the presence of R1-
acid glycoprotein (AGP). Bioorg. Med. Chem. Lett. 2000, 585-
589.
(2) Yarosh, D. B. The role of O6-methylguanine methyltransferase
in cell survival, mutagenesis and carcinogenesis. Mutat. Res.
1985, 145, 1-16.
(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) Pegg, A. E.; Byers, T. L. Repair of DNA containing O6-alkylgua-
nine. FASEB J . 1992, 6, 2302-2310.
(5) Pegg, A. E.; Dolan, M. E.; Moschel, R. C. Structure, function and
inhibition of O6-alkylguanine-DNA alkyltransferase. Prog. Nu-
cleic Acid Res. Mol. Biol. 1995, 51, 167-223.
(6) Dolan, M. E.; Pegg, A. E. O6-Benzylguanine and its role in
chemotherapy. Clin. Cancer. Res. 1997, 3, 837-847.
(7) Belanich, M.; Pastor, M.; Randall, T.; Guerra, D.; Kibitel, J .; Alas,
L.; Li, B.; Citron, M.; Wasserman, P.; White, A.; Eyre, H.;
J aeckle, K.; Shulman, S.; Rector, D.; Prados, M.; Coons, S.;
Shapiro, W.; Yarosh, D. B. Retrospective study of the correlation
between the DNA repair protein alkyltransferase and survival
of brain tumor patients treated with carmustine. Cancer Res.
1996, 56, 783-788.
(8) Baer, J . C.; Freeman, A. A.; Newlands, E. S.; Watson, A. J .;
Rafferty, J . A.; Margison, G. P. Depletion of O6-alkylguanine-
DNA alkyltransferase correlates with potentiation of Temozo-
lomide and CCNU toxicity in human tumor cells. Br. J . Cancer
1993, 67, 1299-1302.
(9) Dolan, M. E. Inhibition of DNA repair as a means of increasing
the antitumor activity of DNA reactive agents. Adv. Drug
Delivery Rev. 1997, 26, 105-118.
(10) Dolan, M. E.; Larkin, G. L.; English, H. F.; Pegg, A. E. Depletion
of O6-alkylguanine-DNA alkyltransferase activity in mammalian
tissues and human tumor xenografts in nude mice by treatment
with O6-methylguanine. Cancer Chemother. Pharmacol. 1989,
25, 103-108.
(11) Dolan, M. E.; Moschel, R. C.; Pegg, A. E.; Depletion of mam-
malian O6-alkylguanine-DNA alkyltransferase activity by O6-
benzylguanine 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.
(12) Pegg, A. E.; Boosalis, M.; Samson, L.; Moschel, R. C.; Byers, T.
L.; Swenn, K.; Dolan, M. E. Mechanism of inactivation of human
O6-alkylguanine-DNA alkyltransferases by O6-benzylguanine.
Biochemistry 1993, 32, 11998-12006.
(13) Spiro, T. P.; Gerson, S. L.; Liu, L.; Majka, S.; Haaga, J .; Hoppel,
C. L.; Ingalls, S. T.; Pluda, J . M.; Willson, J . K. V. O6-
Benzylguanine: A clinical trial establishing the biochemical
modulatory dose in tumor tissue for alkyltransferase-directed
DNA repair. Cancer Res. 1999, 59, 2402-2410.
(14) Friedman, H. S.; Kokkinakis, D. M.; Pluda, J .; Friedman, A. H.;
Cokgor, I.; Haglund, M. M.; Ashley, D. M.; Rich, J .; Dolan, M.
E.; Pegg, A. E.; Moschel, R. C.; McLendon, R. E.; Kerby, T.;
Herndon, J . E.; Bigner, D. D.; Schold, S. C., J r. Phase 1 trial of
O6-benzylguanine for patients undergoing surgery for malignant
glioma. J . Clin. Oncol. 1999, 16, 3570-3575.
(15) Crone, T. M.; Goodtzova, K.; Edara, S.; Pegg, A. E. Mutations
in O6-alkylguanine-DNA alkyltransferase imparting resistance
to O6-benzylguanine. Cancer Res. 1994, 54, 6221-6227.
(16) Xu-Welliver, M.; Kanugula, S.; Pegg, A. E. Isolation of human
O6-alkylguanine-DNA alkyltransferase mutants highly resistant
to inactivation by O6-benzylguanine. Cancer Res. 1998, 59,
1936-1945.
(27) Terashima, I.; Kohda, K. Inhibition of human O6-alkylguanine-
DNA alkyltransferase and potentiation of the cytotoxicity of
chloroethylnitrosourea by 4(6)-benzyloxy)-2,6(4)-diamino-5-(nitro
or nitroso)pyrimidine derivatives and analogues. J . Med. Chem.
1998, 41, 503-508.
(28) Dolan, M. E.; Chae, M.; Pegg, A. E.; Mullen, J . H.; Friedman,
H. S.; Moschel, R. C. Metabolism of O6-benzylguanine, an
inactivator of O6-alkylguanine-DNA alkyltransferase. Cancer
Res. 1994, 54, 5123-5130.
(29) 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-alkylguanines). Anti-Cancer Drug Des. 1994, 9,
401-408.
(30) Balsinger, R. W.; Montgomery, J . A. Synthesis of potential
anticancer agents XXV. Preparation of 6-alkoxy-2-aminopurines.
J . Org. Chem. 1960, 25, 1573-1575.
(31) Frihart, C. R.; Leonard, N. J .; Intramolecular mechanism of the
allylic rearrangement from O6 to C-8 in the guanine series.
Double labeling experiments. J . Am. Chem. Soc. 1974, 96, 5894-
5903.
(32) Kiburis, J .; Lister, J . H. Nucleophilic displacement of the
trimethylammonio-group as a new route to fluoropurines. J .
Chem. Soc. C 1971, 3942-3947.
(33) Campbell, A.; Rydon, H. N. The synthesis of caryophyllenic acid.
J . Chem. Soc. 1953, 3002-3008.
(34) Burger, A.; Standridge, R. T.; Stjernstrøm, N. E.; Marchini, P.
Alicyclic alkylamines and alkanolamines. J . Med. Pharm. Chem.
1961, 4, 517-534.
(35) Mathias, L. J . Esterification and alkylation reactions employing
isoureas. Synthesis 1979, 561-576.
(36) Linn, J . A.; McClean, E. W.; Kelley, J . L. 1,4-Diazabicyclo[2.2.2]-
octane (DABCO)-catalyzed hydrolysis and alcoholysis reactions
of 2-amino-9-benzyl-6-chloro-9H-purine. J . Chem. Soc., Chem.
Commun. 1994, 913-914. 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.
(37) Reich, H. J .; Cohen, M. L. Organoselenium chemistry. Dealky-
lation of amines with benzeneselenol. J . Org. Chem. 1979, 44,
3149-3155.
(38) Halazy, S.; Krief, A. R-Selenocyclobutyllithium as a 2-lithio-1,3-
diene equivalent. Regioselective (100%) synthesis of ipsenol.
Tetrahedron Lett. 1980, 1997-2000.
(39) Conant, J . B.; Kirner, W. R. The relation between the structure
of organic halides and the speed of their reaction with inorganic
iodides. I. The problem of alternating polarity in chain com-
pounds. J . Am. Chem. Soc. 1924, 46, 232-253. Conant, J . B.;
Hussey, R. E. The relation between the structure of organic
halides and the speed of their reaction with inorganic iodides.
II. A study of the alkyl chlorides. J . Am. Chem. Soc. 1925, 47,
476-488. Conant, J . B.; Kirner, W. R.; Hussey, R. E. The relation
between the structure of organic halides and the speed of their
(17) Xu-Welliver, M.; Leitao, J .; Kanugula, S.; Pegg, A. E. Alteration
of the conserved residue tyrosine-158 to histidine renders human
O6-alkylguanine-DNA alkyltransferase insensitive to the inhibi-
tor O6-benzylguanine. Cancer Res. 1999, 59, 1514-1519.
(18) Xu-Welliver, M.; Leitao, J .; Kanugula, S.; Meehan, W. J .; Pegg,
A. E. The role of codon 160 in sensitivity of human O6-
alkylguanine-DNA alkyltransferase to O6-benzylguanine. Bio-
chem. Pharmacol. 1999, 58, 1279-1285.
(19) Wibley, J . E. A.; Pegg, A. E.; Moody, P. C. E. Crystal structure
of the human O6-alkylguanine-DNA alkyltransferase. Nucleic
Acid Res. 2000, 28, 393-401.
(20) 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 alkyl-
transferase. J . Med. Chem. 1992, 35, 4486-4491.

Resistance-Modifying Agents. 8
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4083
reaction with inorganic iodides. III. The influence of unsaturated
groups. J . Am. Chem. Soc. 1925, 47, 488-501.
(40) de la Mare, P. B. D.; Hughes, E. D. Mechanism of substitution
at a saturated carbon atom. Part L. Kinetic effects of phenyl
and halogen substituents in alkyl halides on their reactions with
halide ions in acetone. J . Chem. Soc. 1956, 845.
(47) Mori, K.; Ishikura, M.; Seu, Y.-B.; Preparation of (2S,3S)-1-
phenylthio-2,3-pentanediol and (2S,3S)-1-phenylsulfonyl-2,3-
pentanediol by yeast reduction of the corresponding diketones,
and conversion of the former to (+)-endo-brevicomin. Synthesis
1991, 487-490.
(48) Moriarty, R. M.; Hou, K.-C.; Prakash, I.; Arora, S. K. Hydroxy-
lation of a ketone using o-iodosylbenzoic acid: R-hydroxyaceto-
phenone via the R-hydroxy dimethylacetal. In Organic Syntheses;
Freeman, J . P., Ed.; Wiley: New York, 1990; Collect. Vol. VII,
pp 263-266.
(49) Pal, P. R.; Skinner, C. G.; Dennis, R. L.; Shive, W. Cyclopentane-
alanine and 1-cyclopentene-1-alanine, inhibitory analogues of
leucine and phenylalanine. J . Am. Chem. Soc. 1956, 78, 5116-
5118.
(50) Lythgoe, B.; Trippett, S.; Watkins, J . C. Calciferol and its
relatives. Part II. An alternative synthesis of trans-1,2′-cyclo-
hexylidene-ethylidine-2-methylenecyclohexane. J . Chem. Soc.
1956, 4060-4065.
(41) Bordwell, F. G.; Brannen, W. T. The effect of the carbonyl and
related groups on the reactivity of halides in SN2 reactions. J .
Am. Chem. Soc. 1964, 86 4645-4650.
(42) Halvorsen, A.; Songstad, J . The reactivity of 2-bromo-1-phenyl-
ethanone (phenacyl bromide) towards nucleophilic species. J .
Chem. Soc., Chem. Commun. 1978, 327-331.
(43) Newlands, E. S.; Stevens, M. F. G.; Wedge, S. R.; Wheelhouse,
R. T.; Brock, C. Temozolomide:
a review of its discovery,
chemical properties, pre-clinical development and clinical trials.
Cancer Treat. Rev. 1997, 23, 35-61.
(44) Rodriguez, L.; Lu, N.; Yang, N.-L. Anhydrous molecular form-
aldehyde solution as a synthetic reagent. Synlett 1990, 227-
228.
(51) Edara, S.; Kanugula, S.; Goodtzova, K.; and Pegg, A. E.
Resistance of the human O6-alkylguanine-DNA alkyltransferase
containing arginine at codon 160 to inactivation by O6-ben-
zylguanine. Cancer Res. 1996, 56, 5571-5575.
(45) Green, M. B.; Hickinbottom, W. J . The rearrangement of Râ-
unsaturated alcohols to saturated aldehydes and ketones. Part
I. The preparation of Râ-unsaturated alcohols and 1:2-diols and
their prototropic change. J . Chem. Soc. 1957, 3262-3270.
(46) Hatch, L. F.; Patton, T. L. Allylic chlorides. XXI. 3-Chloro-2-
phenyl-1-propene. J . Am. Chem. Soc. 1954, 78, 2705-2707.
J M000961O