Polar, functionalized guanine-O6 derivatives resistant to repair by O6-alkylguanine-DNA alkyltransferase: implications for the design of DNA-modifying drugs

Article abstract of DOI:10.1016/j.ejmech.2005.11.007

The protein O6-alkylguanine-DNA alkyltransferase (Atase) is responsible for the repair of DNA lesions generated by several clinically important anti-cancer drugs; this is manifest as active resistance in those cancer cell lines proficient in Atase express

Full text of DOI:10.1016/j.ejmech.2005.11.007

European Journal of Medicinal Chemistry 41 (2006) 330–339
http://france.elsevier.com/direct/ejmech
Original article
Polar, functionalized guanine-O6 derivatives resistant to repair
by O6-alkylguanine–DNA alkyltransferase:
implications for the design of DNA-modifying drugs
Dimitrios Pletsas a,b, Richard T. Wheelhouse a,*, Vassiliki Pletsa c, Anna Nicolaou a,
Terence C. Jenkins d, Michael C. Bibby b, Soterios A. Kyrtopoulos c
a
School of Pharmacy, University of Bradford, Bradford, BD7 1DP, UK
b
Institute of Cancer Therapeutics, University of Bradford, Bradford, BD7 1DP, UK
Ethniko Idrima Erevnon 48, Vassileos Constantinou Av, 11635 Athens, Greece
c
d
School of Pharmacy and Pharmaceutical Sciences, Coupland III Building, The University of Manchester, Oxford Road, Manchester, M13 9PL, UK
Received 30 June 2005; received in revised form 3 November 2005; accepted 28 November 2005
Available online 03 February 2006
Abstract
The protein O6-alkylguanine–DNA alkyltransferase (Atase) is responsible for the repair of DNA lesions generated by several clinically im-
portant anti-cancer drugs; this is manifest as active resistance in those cancer cell lines proficient in Atase expression. Novel O6-substituted
guanine analogues have been synthesized, bearing acidic, basic and hydrogen bonding functional groups. In contrast to existing O6-modified
purine analogues, such as methyl or benzyl, the new compounds were found to resist repair by Atase even when tested at concentrations much
higher than O6-benzylguanine, a well-established Atase substrate active both in vitro and in vivo. The inactivity of the new purines as covalent
substrates for Atase indicates that agents to deliver these groups to DNA would represent a new class of DNA-modifying drug that circumvents
Atase-mediated resistance.
© 2006 Elsevier SAS. All rights reserved.
Keywords: Purine; O6-alkylguanine–DNA alkyltransferase; DNA repair; Drug design
1. Introduction
the modified guanine site of the parent strand, that generates
long-lived strand breaks [4]. In the absence of MMR, un-
checked guanine O6-alkylation causes G→A transition muta-
tions in the course of two cycles of DNA replication [5].
The extreme consequences of guanine O6-alkylation have
driven the evolution of another DNA-repair protein to specific-
ally cleave lesions from guanine-O6 positions. O6-alkylguan-
ine–DNA alkyltransferase (Atase; EC 2.1.1.63) plays a major
role in the cellular defense against alkylating agents: as a com-
ponent of DNA repair, it participates in a nucleophilic substitu-
tion reaction that transfers an alkyl group (e.g. methyl or 2-
chloroethyl) from the O6-position of guanine in DNA to a cyst-
eine residue in the protein [6]. The roles of the active site ami-
no acid residues determined to be essential for catalysis are
outlined in Scheme 1. Briefly, His146, by intervention of a
bridging water molecule, acts as the base to deprotonate
Cys145, generating a reactive thiolate nucleophile which is
able to attack at the O6-alkyl carbon of a modified guanine
The biological activity of many clinically-significant, DNA
major groove-alkylating drugs, including the methylating
agents (e.g. temozolomide and dacarbazine) and chloroethylat-
ing agents (e.g. bis(chloroethyl)nitrosourea [BCNU] and
chloroethyl(cyclohexyl)nitrosourea), is mediated via the forma-
tion of O6-alkylguanine adducts. Although these lesions
account for only a small percentage of the total DNA adducts
formed, they are potentially carcinogenic, mutagenic and cyto-
toxic [1,2]. Cytotoxicity arises from intervention of the mis-
match repair (MMR) enzymes that correct single base mispairs
following DNA replication [3]; in the case of O6-alkylguanine-
modified DNA, cell cycle arrest is triggered through a futile
cycle of deoxythymidine excision and replacement, opposite
* Corresponding author. Tel.: +44 1274 23 4710; fax: +44 1274 23 4660.
E-mail address: r.t.wheelhouse@brad.ac.uk (R.T. Wheelhouse).
0223-5234/$ - see front matter © 2006 Elsevier SAS. All rights reserved.
doi:10.1016/j.ejmech.2005.11.007

D. Pletsas et al. / European Journal of Medicinal Chemistry 41 (2006) 330–339
331
Scheme 1. Proposed reaction mechanism for hAtase. His146, by intervention of a bridging water molecule, deprotonates Cys145 to facilitate nucleophilic attack at
the O6-alkyl carbon with simultaneous protonation of N3 by Tyr114 (adapted from reference 7).
base; guanine-N3 is simultaneously protonated by Tyr114 [7].
In the course of the repair reaction, Atase is transformed into
an S-alkylcysteine form. As this process is irreversible, Atase is
a “suicide-protein” that cannot formally be classified as an en-
zyme. The single active site present means that Atase can re-
pair one O6-alkyl nucleobase before inactivation, so is con-
sumed stoichiometrically during the repair process. The
chemotherapeutic significance of Atase is being the determin-
ant of active resistance to drugs that alkylate in the major
groove of double-stranded DNA.
Overall, the response of any cell line (or tumor) to simple
alkylating agents depends on the relative activity of MMR and
Atase. Low levels of MMR expression result in a phenotype
that inherently lacks susceptibility to alkylating drugs (passive
resistance) and is likely to accumulate mutations; in contrast,
high Atase levels characterize the actively resistant phenotype.
A good response to alkylating chemotherapy therefore requires
the combination of efficient MMR with low levels of Atase.
Clinical strategies to overcome Atase-mediated resistance in-
clude divided dose scheduling of alkylating drugs [8–11], or
pre-administration of an Atase inactivator such as O6-benzyl-
guanine [12–15] or O6-(4-bromo)thienylguanine (Patrin 2)
[16]. For a recent comprehensive review of the inactivation of
Atase in cancer chemotherapy see reference [17].
boxymethylpurine and the active site may be postulated in
which the H-bonded network essential for generation of the
thiolate nucleophile is disrupted, thereby obfuscating the cata-
lytic activity of the protein.
A range of new O6-substituted guanines (1–5) bearing po-
lar, charged, or hydrogen bonding functional groups has been
prepared. Complementary molecular models of the target com-
pounds have been constructed and potential interactions with
the active site of the protein have been investigated (see
Appendix A). Enzyme assays determined the ability of Atase
to effect their repair and a new class of DNA lesion, refractory
to Atase-mediated repair, has been characterized. Details of the
macromolecular interactions that determine the properties of
O6-CMG-DNA have been highlighted, from which proposals
are made for lesions to be delivered by new DNA “alkylating”
agents. Such agents would exhibit a spectrum of antitumor act-
ivity independent of the mechanisms of active and passive res-
istance which have hitherto limited application of this class of
anticancer drug.
2. Results
2.1. Synthesis
An alternative to Atase inactivation would be drugs that de-
liver O6-guanine modifications to DNA which are not repair-
able by Atase. Investigation of carcinogens generated during
amino acid metabolism has identified O6-carboxymethyl-2′-
deoxyguanosine-modified DNA (O6-CMG-DNA) as an inter-
mediate in mutagenesis related to amino acid metabolism [18–
20]. O6-CMG-DNA appears to inhibit Atase but without trans-
fer of the O6-substituent to the protein active site, in marked
contrast to either O6-methyl- or O6-benzyl-guanine modifica-
tion [20,21]. Formation of an ionic complex between the car-
The synthetic strategy developed involved elaboration of the
N9-derivatized purine, aciclovir 6. Use of the 6-mesitylene-
sulfonyl leaving group (7) permitted ready introduction of
nucleophilic groups to the 6-position of the purine ring
(Scheme 2). Protection of the terminal hydroxy group of the
N9 sidechain was essential to direct sulfonylation to the O6
position. Thus, t-butyldimethylsilyl chloride (TBDMS-Cl)
was employed to protect the sidechain of aciclovir. The
TBDMS-protected aciclovir derivative 8 was obtained in good
yield (97%) after recrystallization from MeOH. Qian et al. [22]

332
D. Pletsas et al. / European Journal of Medicinal Chemistry 41 (2006) 330–339
prepared were the ethyl ester 13, amide 2, 2-(dimethylamino)-
ethyl 4 and 2-hydroxyethyl 5 derivatives.
Synthesis of the O6-2-aminoethyl derivative 3, required pro-
tection of the amino group of the ethanolamine reagent since
reaction with the unprotected alcohol resulted solely in form-
ation of N6-hydroxyethyl derivative 14. Reaction with N-BOC-
aminoethanol gave O-coupled compound 15, which was trea-
ted first with TBAF to give BOC-intermediate 16 and then
with TFA to give the required amine salt 3. The protected 2-
chloroethyl derivative 17 was prepared in a similar manner
using 2-chloroethanol (Scheme 3). On warming the oil at
60 °C overnight, intramolecular cyclization resulted in form-
1
ation of the tricyclic ethano adduct 18. Indeed, H-NMR indi-
cated that conversion of the chloroethyl adduct, although slow,
started almost immediately at room temperature. The final
ethano salt 19 was obtained after TBAF deprotection.
2.2. Atase assays
The target purines were evaluated in two assays to estimate
their activity as suicide and competitive inhibitors of Atase.
The Atase assays used measure the transfer of radiolabeled
3
methyl groups from H-methylated DNA to protein, the level
of tritium detected being proportional to the number of methyl
groups transferred to the Cys145-thiol of the protein. The stan-
dard Atase assay was developed to measure the activity of sui-
cide inactivators of the protein [25,26]. In this assay, a 45 min
pre-incubation of Atase with the test compound is included
prior to addition of the [3H]-MeDNA; this provides sufficient
time for the protein to recognize the inhibitor and allow it to
occupy, and subsequently react with, the active site. With the
addition of [3H]-MeDNA and a further 2 h incubation, residual
Scheme 2. Reagents and conditions. (i) TBDMSCl, imidazole, DMF, 48 h, r.t.
(97 %); (ii) 2-mesitylenesulfonyl chloride, DMAP, Et3N, MeCN, 2 h, r.t.,
(80 %); (iii) DABCO, THF, 4 h, r.t.; (iv) ROH, DBU, THF, 3 h, r.t. (48–91 %);
(v) TBAF, THF, 30 min, r.t., (28–97 %); (vi) 0.1 M NaOH, 2 h, HCl, r.t.,
(88 %); (vii) NH3, r.t.; (viii) TFA, CH2Cl2, 10 min, r.t., (72 %).
3
have reported the synthesis of this compound using similar re-
action conditions. The mesitylene group was introduced at the
O6-position by adaptation of a published method [23] to give
sulfonate ester intermediate 7 in 80% isolated yield.
active Atase is allowed to repair H-methyl groups. The res-
idual activity of Atase is determined by isolating the protein
and comparing the radioactivity counts transferred to the pro-
tein against those of a control reaction in the absence of any
inhibitor. In this assay, an efficient inhibitor would not allow
Atase to react with the [3H]-MeDNA substrate. The established
suicide inhibitor O6-benzylguanine (0.5 μM) was used as the
positive control which showed 28 7% residual Atase activity
after 40 min reaction at 37 °C. This result is consistent with
data obtained by other investigators [26,27] so allows direct
comparisons to be made.
A variation on the standard suicide assay was developed to
estimate the potential of the test compounds as competitive in-
hibitors. The suicide assay was adapted to measure the activity
of Atase under virtual steady-state conditions, by eliminating
the pre-incubation time and thus give a “snapshot” estimate
The mesityl derivative 7 and the derived quaternary DAB-
CO salt 9 were key intermediates that gave access to the target
compounds; an approach also successfully applied by Laksh-
man et al. [24] using 2′-deoxyguanosine derivatives for the pre-
paration of O6-alkyl and O6-aryl ethers as well as N6-substi-
tuted diaminopurine nucleosides. The carboxymethyl group
was introduced as the methyl ester (10) by treatment with the
glycolate ester anion generated using DBU. Subsequent TBAF
deprotection of the TBDMS group gave the target ester 11.
Base hydrolysis yielded the carboxylate sodium salt, which
was converted to the free acid 1 by treatment with HCl; neut-
ralization with ammonia gave ammonium salt 12. Similarly
Scheme 3. Reagents and conditions. (i) ClCH2CH2OH, DBU, THF, 3 h, r.t.; (ii) heat 60 °C; (iii) TBAF, THF, 30 min, r.t., (53% over three steps).

D. Pletsas et al. / European Journal of Medicinal Chemistry 41 (2006) 330–339
333
of Atase activity in the presence of the inhibitor. For this assay,
the inhibitor concentration was fixed at 100 μM and O6-ben-
zylguanine (0.5 μM) was again used as a positive control, since
there are presently no established competitive inhibitors of
Atase.
cells is well established [21,29–32]. To assess interactions be-
tween the novel purines and Atase in cells a simple potentation
assay was performed, using the MCF 7 breast carcinoma cell
line, known to express high levels of Atase. BCNU generates
O6-alkylguanine-DNA adducts that are substrates for Atase [6,
33–35], therefore any purine–Atase interactions should trans-
late into an increase in the apparent chemosensitivity to
BCNU. On the basis of the cell-free Atase data, the carboxy-
methyl derivative 1 was selected for evaluation. Since the acid
would be ionized at the experimental pH, and anions do not
cross cell membranes by passive diffusion, the methyl ester
11 was also included as a more lipophilic prodrug form that
may hydrolyse to expose the carboxylate group once inside
cells. Cells were exposed to the purine derivatives for 1 h, foll-
owed by exposure to BCNU (1 h), data are presented in the
Table 1 as relative IC50 values [i.e. IC50 (BCNU)/IC50
(BCNU + purine)].
The results of both assays are presented in the Fig. 1. As
expected, O6-benzylguanine showed activity in both assays,
more so in the suicide assay with a pre-incubation period
(28% compared with 64% detectable residual Atase activity).
For most test compounds, there was little difference in the
Atase activity detected in either of the two distinct assays. In
neither assay was potent interaction or reaction with the novel
purines detected, despite using concentrations 200-times great-
er than the positive control compound, O6-benzylguanine. This
indicates that binding to Atase is weak compared with the in-
teraction between the protein and the second substrate, methy-
lated DNA. Although close to the limits of the experimental
error, O6-carboxymethyl- (1), aminoethyl- (3) and carboxa-
mide-substituted (2) purines showed some activity as reversible
inhibitors of Atase (Fig. 1).
3. Discussion
The failure to detect activity for ethano derivative 19 was
surprising as this compound was expected to be a suicide,
rather than a competitive, inhibitor of Atase; thus mimicking
the behavior of chloroethyl-modified guanine lesions generated
in DNA by chloroethylating drugs, such as BCNU. However,
in this case it seems likely that failure to detect activity was due
to rapid hydrolysis under aqueous conditions before the inves-
tigational compound was able to encounter its protein target
[28], rather than weakness of the purine–protein interaction.
In the course of this work, X-ray crystal structures of the
complex of Cys145S-mutant Atase with an O6-methylgua-
nine-bearing oligonucleotide and of active Atase with a sub-
strate analogue oligonucleotide were published [36]. Signif-
Table 1
Relative in vitro cytotoxicity of BCNU in MCF7 cells (Atase and MMR
proficient) in the absence and presence of selected purine derivatives
a
Test compound (100 μM)
BCNU relative IC50
Control (no additive)
O6-Benzylguanine
11
1
1.0b
1.8
1.6
1.4
2.3. Purine–Atase interactions in cells
a
The ability of Atase-binding purines to increase (potentate)
the toxicity of DNA alkylating drugs towards Atase-proficient
IC50(no additive)/IC50(in presence of test compound).
IC50(BCNU) = 236 μM.
b
Fig. 1. Alkyltransferase assay data. Plots of the suicide and competitive inhibition assay data for test compounds at 100 μM and the positive control, O6-
benzylguanine, at 0.5 μM. Data are the mean S.D. values of at least three separate determinations.

334
D. Pletsas et al. / European Journal of Medicinal Chemistry 41 (2006) 330–339
icantly, the guanine-binding active site structure of the DNA-
bound protein was little different from that reported for the free
protein (r.m.s. deviation 0.7–0.8 Å) and Cys-S-Me or Cys-S-
benzyl derivatives [7] used as starting structures for this study.
The catalytic amino acid residues, previously identified by
point mutation studies, and their disposition around the O6-
modified guanine substrate were confirmed. The proven struct-
ural and mechanistic details are entirely consistent with the
data and conclusions described herein.
11 (1.6-fold) and the acid 1 (1.4-fold), consistent with the small
Atase-inhibitory activity determined for carboxymethyl purine
1 in the cell-free assay. This apparent interaction with the
Atase protein must be interpreted with caution since O6-car-
boxymethylguanine has been shown to be a precursor of the
suicide substrate O6-methylguanine during amino acid metabo-
lism-derived carcinogenesis [19].
4. Conclusions
Nine target compounds were synthesized by modification of
the 6-oxopurine position via direct mesitylation and DABCO
salt formation (Scheme 2). This versatile, convenient and high-
yielding synthetic route did not necessitate extensive purifica-
tion of intermediates or target products. The modest activity of
the O6-carboxymethylguanine derivative 1 in both of the cell-
free Atase assays was surprising. Previous studies on O6-
CMG-DNA have indicated it to be a potent inactivator of the
reaction of Atase with [3H]-Me-DNA [20]; furthermore, if the
carboxymethyl group is 14C-radiolabeled, there is no evidence
of transfer of the O6-substituent group to the Atase protein
[20]. Thus it appears that the Atase–O6-CMG-DNA interaction
is sufficient to divert Atase from reaction with a [3H]-Me-DNA
co-substrate and disrupt the catalytic mechanism so that trans-
fer of the O6-carboxymethyl group to Cys145 does not take
place; i.e. O6-CMG-DNA is recognized but not repaired by
Atase. In contrast, for simple O6-substituted purines (e.g. O6-
benzylG, Patrin 2 and O6-MeG) protein–purine interactions
alone are sufficient for recognition and catalytic transfer of
the O6-substituent to Atase Cys145, albeit that the rate of reac-
tion is usually reduced relative to repair of the equivalent lesion
borne on a DNA duplex [37]. This provides the basis for the
development of drug-like, small-molecule inactivators of
Atase.
In the case of O6-CMG, it appears that the purine–protein
interactions alone are weak and the additional duplex DNA–
protein recognition and binding contacts are essential for the
formation of a stable complex. Notwithstanding the potential
for carboxylate–protein interactions (see Appendix A) the
structure of Atase bound to modified DNA shows the O6-sub-
stituent accommodated within a pocket composed of hydro-
phobic amino acid sidechains. Thus it appears that potential
polar or H-bonding interactions are unable to compensate for
the loss of hydrophobic contributions to the binding energy.
Subject to steric constraints in the active site, SAR for groups
cleaved by Atase have usually been interpreted in terms of the
propensity of the cleaved groups to leave during nucleophilic
substitution reactions; specifically, to stabilize the incipient
carbonium ion transition state [38]. Consideration must addi-
tionally be given to hydrophobic factors, which may in part
explain the divergence between O6-(2-pyridylmethyl)guanine
and benzylguanine (Atase IC50 = 55 and 0.2 μM, respectively)
in the same assay [31]. Similar arguments may also account for
the apparent weakness of interaction of the O6-methyl aciclovir
derivatives 2–5 with Atase.
This study has identified a family of novel and polar purine-
O6 derivatives, none of which are good substrates for the DNA
repair protein Atase, although carboxymethyl derivative 1 may
degrade in situ to an Atase-substrate form. The lack of sub-
strate activity may relate to poor hydrophobic interactions, in-
sufficient to drive molecular recognition within the active site
binding pocket of Atase. The novel purines are mimics of
DNA lesions and so the equivalent modifications on DNA
are also be expected to be resistant to repair.
Overall, two conclusions significant to the design of new
chemotherapeutic agents directed to guanine-O6 can be drawn.
Firstly, Atase-mediated drug resistance may be bypassed
through the development of agents to deliver the highly polar
groups investigated herein to guanine-O6 of DNA, rather than
simple alkyl groups (methyl, chloroethyl) as used currently. Of
particular interest are the O6-(2-hydroxyethyl) and O6-(2-di-
methylaminoethyl) modifications. Furthermore, the possibility
then exists for the formation of stable complexes between
Atase protein and modified DNA in cells, as apparently the
case with O6-CMG-DNA [20] — situation that would enhance
activity towards Atase-proficient cell lines by a mechanism
analogous to the ternary complex formation between HMG-
domain proteins and cisplatin-modified DNA, implicated in
the biological activity of cisplatin [39]. In this scenario, selec-
tive toxicity towards Atase-proficient cells would be manifest;
moreover, activity would not be dependent on expression of
the MMR proteins. On this basis, alkylating chemotherapy
against tumor lines hitherto inherently resistant (either actively
or passively) to such drugs could be achieved.
5. Experimental procedures
5.1. Synthesis
1H-NMR spectra were acquired at 270.17 MHz and 13C-
NMR spectra at 67.80 MHz with Me4Si as internal standard,
using a JEOL GX270 spectrometer. 13C assignments were
made using the DEPT135 experiment. Mass spectra were ob-
tained from the EPSRC Mass Spectrometry Service Centre at
the University of Wales, Swansea, UK. Elemental analysis data
were obtained from the Advanced Chemical and Materials
Analysis Unit at the University of Newcastle-upon-Tyne, UK.
Melting points (mp) were determined using an Electrothermal
IA9200 digital melting point apparatus. Infrared spectra were
obtained using a Perkin Elmer (Paragon 1000) FT-IR instru-
ment. TLC was performed on pre-coated plastic plates (silica
In the in vitro Atase assay, the presence of O6-benzylgua-
nine reduced the IC50(BCNU) 1.8-fold (from 236 → 131 μM).
A modest potentiation was detected in the presence of the ester

D. Pletsas et al. / European Journal of Medicinal Chemistry 41 (2006) 330–339
335
gel 60, F254, MERCK) and visualized under UV light or by
staining with I2.
5.1.3. [2-Amino-9-(2-hydroxyethoxymethyl)-9H-purin-6-yloxy]
acetic acid methyl ester (11)
To a solution of 7 (3.9 g, 7.4 mmol) in dry THF (60 ml),
DABCO (4.18 g, 37.3 mmol) was added and the mixture stir-
red at room temperature under N2. Progress of the reaction was
monitored by TLC (ethyl acetate). Disappearance of the start-
ing material and formation of a blue fluorescent spot at the
baseline indicated complete formation of the DABCO salt in-
termediate 9 after 4 h. Methyl glycolate (6.73 g, 74 mmol) and
DBU (3.4 g, 22.4 mmol) were added and the mixture stirred at
room temperature for another 3 h. Purification by column chro-
matography eluted with ethyl acetate yielded the TBDMS-prot-
ected methyl ester intermediate 10 (2.2 g, 72%) as a white so-
lid: m.p. 133.4 °C; 1H-NMR (CDCl3) δ 7.77 (s, 1H, 8-H), 5.52
(s, 2H, NCH2O), 5.04 (s, 2H, O6-CH2), 4.85 (s, 2H, NH2),
3.77 (s, 3H, OCH3), 3.74 and 3.72 (2 × t, J = 4.5 Hz, 4H, OC
H2CH2O), 0.88 (s, 9H, (CH3)3CSi), –0.05 (s, 6H, (CH3)2Si);
13C-NMR (DMSO-d6) δ 170.1 (CO), 160.4 (C-6), 156.3 (C-
2), 155.6 (C-4), 140.2 (C-8), 114.7 (C-5), 73.4 (NCH2O),
70.7 (OCH2CH2OTBDMS), 64.2 (OCH2CH2OTBDMS), 62.1
(O6-CH2), 51.7 (OCH3), 26.6 (SiC(CH3)3), 19.1 (SiC(CH3)3),
–4.6 (CH3Si). IR (KBr) 3340s (NH2), 3210s (NH2), 1720s
(CO), 1645s cm–1; MS (EI): m/z 250 (100%); 354 (80%);
412.2 (10%) [M + H]●+; MS (CI): m/z 120 (75%); 132
(95%); 159 (50%); 194 (100%); 412 (45%) [M + H]●+; HRMS
412.2019; C17H30SiN5O5 requires 412.2016.
To a solution of 10 (1.75 g, 4.3 mmol) in dry THF (30 ml),
TBAF (4.6 ml, 1 M in THF) was added and the solution stirred
at room temperature. The reaction was monitored by TLC
(ethylacetate) for the loss of 10 and formation of 11. The solu-
tion was evaporated in vacuo and the resulting oil was purified
by chromatography on silica gel, eluted with ethylacetate/
methanol 95:5. 11 was obtained as a pale orange crystalline
solid, recrystallized from methanol (0.91 g, 71%): m.p. 124–
126 °C.; 1H-NMR (DMSO-d6) δ 8.03 (s, 1H, 8-H), 6.5 (s, 2H,
NH2), 5.41 (s, 2H, NCH2O), 5.06 (s, 2H, O6-CH2), 4.65 (t, 1H,
J = 4.8 Hz, OCH2CH2OH), 3.68 (s, 3H, OCH3), 3.47 (s, 4H,
OCH2CH2OH); 13C-NMR (DMSO-d6) δ 169 (CO), 159.9
(C-6), 159.4 (C-2), 155 (C-4), 140.7 (C-8), 113.4 (C-5), 72.3
(NCH2O), 70.6 (OCH2CH2OH), 61.9 (O6-CH2), 60.1
(OCH2CH2OH), 52.1 (OCH3). IR (KBr) 3470 (OH), 3345s
(NH2), 3210s (NH2), 1720s (CO), 1645s cm–1; MS (FAB):
m/z 133 (37%); 142 (36%); 242 (100%); 298 (10%)
[M + H]●+; 320 (5%) [M + Na]●+. Anal. C, H, N.
Aciclovir was purchased from CHEFARMA (Athens,
Greece). Reagents and solvents were purchased from Aldrich
(Gillingham, UK), Lancaster Synthesis (Lancaster, UK) or
Avocado (Heysham, UK); other solvents were purchased from
Fisher Scientific (Loughborough, UK). O6-Benzylguanine was
prepared by adaptation of a published procedure [40].
5.1.1. 2-Amino-9-[2-(tert-butyldimethylsilyloxy)ethoxymethyl]-
1,9-dihydropurin-6-one (8)
To a solution of aciclovir 6 (3 g, 13.3 mmol) in DMF (75
ml), imidazole (2.72 g, 40 mmol) and TBDMS (4 g,
26.7 mmol) were added. The mixture was stirred for 48 h at
room temperature and then poured into distilled water (150
ml). The precipitate was collected by filtration, dried overnight
in a vacuum oven and then recrystallized from methanol to
yield 8 as colorless crystals (4.39 g, 97%); m.p. 208–212 °C.
1
NMR data consistent with reference [22]. H-NMR (DMSO-
d6) δ 10.58 (s, 1H, NH), 7.75 (s, 1H, 8-H), 6.41 (s, 2H,
NH2), 5.25 (s, 2H, NCH2O), 3.60 and 3.45 (2 × t, J = 4.6 Hz,
4H, OCH2CH2O), 0.75 (s, 9H, (CH3)3CSi), –0.11 (s, 6H, 2 × C
H3–Si); 13C-NMR (DMSO-d6) δ 156.5 (C-2), 154.7 (C-4),
151.2 (C-6), 137.5 (C-8), 116.8 (C-5), 72.5 (NCH2O), 70.1
(OCH2CH2OTBDMS), 62.3 (OCH2CH2OTBDMS), 26.1
(CH3CSi), 18.2 (Si–C(CH3)3), –5.0 (CH3Si). IR (KBr) 3480m
(NH), 3280s (NH2), 3190s (NH2), 1725s (CO), 1630 s cm–1;
MS (FAB): m/z 340 (10%) [M + H]●+; 362 (100%)
[M + Na]●+. Anal. C, H, N.
5.1.2. 2,4,6-Trimethylbenzenesulfonic acid 2-amino-9-[2-(tert-
butyldimethylsilyloxy)ethoxymethyl]-9H-purin-6-yl ester (7)
Protected purine 8 (3.83 g, 11.3 mmol), 2-mesitylenesulfo-
nyl chloride (4.92 g, 22.5 mmol), and DMAP (0.69 g, 5.65
mmol) were suspended in acetonitrile (40 ml). Triethylamine
(7.4 ml, 5.3 mmol) was added drop-wise and the suspension
stirred at room temperature. After 2 h, the volatile components
were removed under reduced pressure. The residue was applied
to a column of silica gel and eluted with CHCl3:EtOH (9:1, v/
v). Compound 7 was obtained as a white solid (4.67 g, 80%):
1
m.p. 160.3 °C; H-NMR (DMSO-d6) δ 7.85 (s, 1H, 8-H), 7.1
(s, 2H, Ar–H), 5.55 (s, 2H, NCH2O), 4.95 (s, 2H, NH2), 3.59
and 3.75 (2 × t, J = 4.5 Hz, 4H, OCH2CH2O), 2.78 (s, 6H,
2 × o-CH3-Ar), 2.32 (s, 3H, p-CH3–Ar), 0.85 (s, 9H, (C
H3)3CSi), –0.10 (s, 6H, 2 × CH3Si); 13C-NMR (DMSO-d6) δ
160.1 (C-6), 158.2 (C-2) 156.1(C-4), 144.8 (p-CCH3–Ar),
142.4(o-CCH3–Ar), 141.1 (C-8), 137.1 (CSO2–Ar), 133.5
(CH–Ar), 117 (C-5), 73.7 (NCH2O), 71.7(O
CH2CH2OTBDMS), 63.2 (OCH2CH2OTBDMS), 26.6 (ArCH3),
21.8 (SiC(CH3)3), 19.0 (SiC(CH3)3), –4.6 (CH3Si). IR (KBr)
3486s (NH), 3295s (NH2), 3180s (NH2), 1640s (NH2), 1360s
(SO), 1170s (SO); MS (FAB): m/z 522 (100%) [M + H]●+; 544
(15%) [M + Na]●+; 1043 (5%) (2 × M + H)●+; 1065 (5%) (2 × M
+ Na)●+. Anal. C, H, N.
5.1.4. [2-Amino-9-(2-hydroxyethoxymethyl)-9H-purin-6-yloxy]
acetic acid ethyl ester (13)
To a solution of 7 (0.38 g, 0.73 mmol) in dry THF (60 ml),
DABCO (0.4 g, 3.7 mmol) was added and the mixture stirred
at room temperature under N2. After 4 h, ethyl glycolate (0.76
g, 7.3 mmol) and DBU (0.34 g, 2.2 mmol) were added and the
reaction continued at room temperature for another 3 h. The
volatile components were removed in vacuo and the residue
dissolved in dry THF (5 ml). TBAF (0.6 ml, 1M in THF)
was added to the solution and the mixture stirred at room temp-
erature for 30 min. The solvent was removed in vacuo and the

336
D. Pletsas et al. / European Journal of Medicinal Chemistry 41 (2006) 330–339
residue purified by chromatography on silica gel, eluted with
of the DABCO salt 9 with glycolamide and DBU yielded
protected acetamide 21 as a white solid (1.1 g, 88%): m.p.
199.5 °C; H-NMR (CDCl3) δ 7.79 (s, 1H, 8-H), 6.62 (br s,
ethyl acetate-methanol 95:5 v/v. Ester 13 was obtained as a
1
1
white crystalline solid (0.07 g, 30%). m.p. 133–134 °C; H-
NMR (CDCl3) δ 7.74 (s, 1H, 8-H), 5.47 (s, 2H, NCH2O),
5.20 (s, 2H, NH2), 4.98 (s, 2H, O6-CH2), 4.23 (q, 2H,
J = 3.67 Hz, OCH2CH3), 3.71 and 3.62 (2 × t, 4H, J = 2.93
Hz, OCH2CH2OH), 1.23 (t, 3H, J = 7 Hz, OCH2CH3); 13C-
NMR (CDCl3) δ 168.4 (CO), 160.2 (C-6), 159.4 (C-2), 154.5
(C-4), 139.9 (C-8), 114.9 (C-5), 72.9 (NCH2O), 70.6
(OCH2CH2OH), 63.0 (OCH2CH3), 62.7 (O6-CH2), 61.2
(OCH2CH2OH), 14.0 (OCH2CH3). MS (FAB): m/z 312
(100%) (M + H)●+; 334 (55%) (M + Na)●+; 623 (10%) (2M
+ H)●+; 645 (10%) (2M + Na)●+. Anal. C, H, N.
2H, CONH2), 5.54 (s, 2H, OCH2N), 4.99 (s, 2H,O6-CH2),
4.94 (s, 2H, NH2), 3.75 and 3.61 (2 × t, J = 3.7 Hz, 4H, OC
H2CH2O), 0.88 (s, 9H, (CH3)3CSi), 0.05 (s, 6H, 2 × CH3Si);
13C-NMR (CDCl3) δ 170.9 (CO), 160.1 (C-6), 155.8 (C-2),
155.5 (C-4), 140.7 (C-8), 113.8(C-5), 73.6 (NCH2O), 71.6
(OCH2CH2OTBDMS), 65.3 (OCH2CH2OTBDMS), 63.2 (O6-
CH2), 26.6 (SiC(CH3)3), 19.0 (SiC(CH3)3), –4.6 (CH3Si). IR
(KBr) 3325s (NH2), 3220s (NH2), 1690s and 1620m
(CONH2), 1585s (C–O) cm–1; MS (FAB): m/z 397 (100%)
[M + H]●+; 419 (20%) [M + Na]●+. Anal. C, H, N.
Deprotection of 21 (1.25 g, 3 mmol) was carried as de-
scribed previously for the formation of the methyl ester (11)
using TBAF; acetamide 2 was obtained as a white solid (0.83
g, 93%). Alternatively, ester 11 (0.5 g, 1.68 mmol) was dis-
solved in aqueous ammonia solution (15 ml, 35%). The solu-
tion was stirred at room temperature and the reaction was mon-
itored with TLC (ethyl acetate/methanol 3:1 v/v). The product
precipitated from the reaction mixture; the precipitate was col-
lected by filtration, washed with distilled water and dried to
yield amide 2 (0.49 g, 98%) as a white solid: m.p. 245 °C;
1H-NMR (CDCl3) δ 8.02 (s, 1H, 8-H), 7.39 and 7.23 (2 × br
s, 2 × 1H, CONH2), 6.46 (br s, 2H, NH2), 5.42 (s, 2H, NC
H2O), 4.83 (s, 2H,O6-CH2), 4.65 (t, 1H, J = 4.0 Hz, OH),
3.46 (m, 4H, OCH2CH2OH); 13C-NMR (CDCl3) δ 169.8
(CO), 160.1 (C-6), 160.2 (C-2), 154.9 (C-4), 140.1 (C-8),
113.2 (C-5), 73.2 (NCH2O), 70.6 (OCH2CH2OH), 62.7 (O6-
CH2), 59.6 (OCH2CH2OH). IR (KBr) 3450s (OH), 3330s
(NH2), 3220s (NH2), 1690s and 1620m (CONH2), 1590s
cm–1; MS (FAB): m/z 283 (50%) [M + H]●+; 305 (20%) [M
+ Na]●+. Anal. C, H, N.
5.1.5. [2-Amino-9-(2-hydroxyethoxymethyl)-9H-purin-6-yloxy]
acetic acid (1)
Ester 11 (0.85 g, 2.7 mmol) was dissolved in aqueous
NaOH solution (15 ml, 0.1 M). The solution was stirred at
room temperature and the reaction was monitored by TLC.
The loss of 11 and formation of a blue fluorescent spot at the
baseline indicated the formation of the carboxylate sodium salt.
Dilute hydrochloric acid (0.1 M) was added drop-wise until the
free acid 1 precipitated from the reaction mixture. The precipi-
tate was collected by filtration, washed with distilled water and
dried in a vacuum oven to give 1 (0.71g, 88%) as a white solid:
m.p. 213–215 °C; 1H-NMR (DMSO-d6) δ 12.96 (s, 1H, CO2H
), 8.03 (s, 1H, 8-H), 6.5 (s, 2H, NH2), 5.43 (s, 2H, NCH2O),
4.98 (s, 2H, O6-CH2), 4.66 (br s, 1H, CH2OH), 3.47 (s, 4H,
OCH2CH2OH); 13C-NMR (DMSO-d6) δ 169.6 (CO), 159.8
(C-6), 159.5 (C-2), 154.7 (C-4), 140.4 (C-8), 113.4 (C-5),
72.1 (NCH2O), 70.5, (OCH2CH2OH), 59.9 (OCH2CH2OH),
61.7 (O6-CH2). IR (KBr) 3380s (OH), 3320s (NH2), 3210s
(NH2), 1650m (C=O), 1485m cm–1; MS (FAB): m/z 107
(100%); 152.1 (55%); 284 (35%) [M + H]●+; 306 (37%) [M
+ Na]●+. Anal. C, H, N.
5.1.8. 2-[2-Amino-6-(2-dimethylaminoethoxy)purin-9-
ylmethoxy]ethanol (4)
5.1.6. [2-Amino-9-(2-hydroxyethoxymethyl)-9H-purin-6-yloxy]
acetic acid ammonium salt (12)
This was prepared in an analogous series of reactions from
mesitylene derivative 7 (1.49 g, 2.84 mmol). The DABCO salt
9 was further reacted with 2-(dimethylamino)ethanol and DBU
to yield protected (dimethylamino)ethylaciclovir 22 as a white
Carboxylic acid 1 (0.1 g, 0.35 mmol) was dissolved in aque-
ous ammonia solution (15 ml, 35%) and the mixture evapor-
ated in vacuo. The residue was re-dissolved in ethanol and eva-
porated three times to give salt 12 as a white solid: m.p. 218–
1
solid (0.57 g, 48.2%): m.p. 130 °C; H-NMR (CDCl3) δ 7.72
(s, 1H, 8-H), 5.5 (s, 2H, NCH2O), 4.88 (br s, 2H, NH2), 4.60
(t, 2H, J = 6.2 Hz, O6-CH2CH2N(CH3)2), 3.72 and 3.58 (t, 2H,
J = 4.0 Hz, OCH2CH2OTBDMS), 2.80 (t, 2H, J = 6.2 Hz,
O6-CH2CH2N(CH3)2), 2.34 (s, 6H, N(CH3)2), 0.87 (s, 9H,
SiC(CH3)3), –0.05 (s, 6H, (CH3)2Si); 13C-NMR (CDCl3) δ
162.3 (C-6), 160.1 (C-2), 154.9 (C-4), 140.0 (C-8), 116.2
(C-5), 74.1 (NCH2O), 70.8 (OCH2CH2OTBDMS), 65.3 (O6-
CH2CH2N(CH3)2), 63.4 (OCH2CH2OTBDMS), 58.7 (O6-
CH2CH2N(CH3)2), 46.9 (N(CH3)2) 25.8, (SiC(CH3)3), 17.9 (Si
C(CH3)3), –6.4 (CH3Si). IR (KBr) 3300s (NH2), 3200s (NH2),
1660s cm–1; MS (FAB): m/z 411 (100%) [M + H]●+; 433
(10%) [M + Na]●+; 821 (5%) [2M + H]●+; 843 (2%) [2M +
Na]●+. Anal. C, H, N.
1
219 °C; H-NMR (DMSO-d6) δ 8.00 (s, 1H, 8-H), 6.4 (s, 2H,
NH2), 5.43 (s, 2H, NCH2O), 4.78 (s, 2H, O6-CH2), 3.47 (br s,
5H, OCH2CH2OH; becomes br s, 4H on adding D2O); 13C-
NMR (DMSO-d6) δ 170.1 (CO), 161.2 (C-6), 160.5 (C-2),
154.2 (C-4), 139.9 (C-8), 114.1 (C-5), 72.6 (NCH2O), 70.2, (O
CH2CH2OH), 60.0 (OCH2CH2OH), 63.8 (O6-CH2). IR (KBr)
3380 (OH), 3340s (NH2), 3210s (NH2), 1620m (C=O), 1485
(C–O) cm–1. MS (ES): m/z 282.1 (100%) (M)●+. Anal. C, H,
N.
5.1.7. 2-[2-Amino-9-(2-hydroxyethoxymethyl)-9H-purin-6-
yloxy]acetamide (2)
This was prepared in an analogous series of reactions from
mesitylene derivative 7 (1.49 g, 2.85 mmol). Further reaction
Deprotection of 22 (0.58 g, 1.35 mmol) was carried out in a
similar manner as described above using TBAF; dimethylami-

D. Pletsas et al. / European Journal of Medicinal Chemistry 41 (2006) 330–339
337
no derivative 4 was obtained as a white solid (0.12 g, 28%):
m.p. 132 °C; 1H-NMR (DMSO-d6) δ 7.97 (s, 1H, 8-H),
6.45 (br s, 2H, NH2), 5.4 (s, 2H, NCH2O), 4.68 (br s,
1H, OCH2CH2OH), 4.48 (t, 2H, J = 5.9 Hz, O6-
CH2CH2N(CH3)2), 3.45 (br s, 4H, OCH2CH2OH), 2.63 (t,
2H, J = 5.9 Hz, O6-CH2CH2N(CH3)2; 2.20 (s, 6H, N(CH3)2);
13C-NMR (DMSO-d6) δ 160.6 (C-6), 160.3 (C-2), 154.7 (C-4),
140.1 (C-8), 113.7 (C-5), 72.1 (NCH2O), 70.5 (OCH2CH2OH),
63.5 (O6-CH2CH2N(CH3)2), 59.9 (OCH2CH2OH), 57.6 (O6-
CH2CH2N(CH3)2), 45.5 (N(CH3)2). IR (KBr) 3300s (NH2),
3200s (NH2), 1660 s cm–1. MS (FAB): m/z 297 [M + H]●+.
Anal. C, H, N.
(NHCH2CH2OH), 25.8 (SiC(CH3)3), 17.9 (SiC(CH3)3), –5.3
(CH3Si). IR (KBr) 3440s (OH), 3340s (NH2, NH), 3225s
(NH2), 1600s, 1455s cm–1; MS (EI): m/z 255 (75%); 337
(75%); 353 (70%); 382 (100%) [M]●+; MS (CI): m/z 120
(90%); 159 (90%); 177 (100%); 383 (60%) [M + H]●+. Anal.
C, H, N.
TBAF deprotection of 14 (1.78 g. 4.46 mmol) gave deriva-
1
tive 20 as white plates (1.15 g, 91%): m.p. 167–170 °C; H-
NMR (DMSO-d6) δ 7.82 (s, 1H, 8-H), 7.03 (br s, 1H, N6-H),
5.89 (br s, 2H, NH2), 5.36 (s, 2H, NCH2O), 4.74 (br s, 1H,
OH), 4.65 (br s, 1H, OCH2CH2OH), 3.54 (br s, 4H, N6-
CH2CH2OH), 3.45 (br s, 4H, OCH2CH2OH); 13C-NMR
(DMSO-d6) δ 169.8 (CONH2), 160.1 (C-6), 160.2 (C-2),
154.9 (C-4), 140.1 (C-8), 113.2 (C-5), 73.2 (NCH2O), 70.6
(OCH2CH2OH), 62.7 (N6-CH2), 59.6 (OCH2CH2OH). IR
(KBr) 3450s (OH), 3330s (NH2), 3325s (NH), 3220s (NH2),
1605m, 1610m, 1465m (C–O) cm–1; MS (FAB): m/z 269
(100%) [M + H]●+. Anal. C, H, N.
5.1.9. 2-[2-Amino-6-(2-hydroxyethoxy)purin-9-ylmethoxy]
ethanol (5)
This was prepared in an analogous series of reactions from
mesitylene derivative 7 (0.24 g, 0.46 mmol). The DABCO salt
9 was further reacted with ethanediol and DBU to yield the
protected hydroxyethyl derivative 23 as a white solid (0.3 g,
1
5.1.11. 2-[2-Amino-6-(2-aminoethoxy)purin-9-ylmethoxy]
ethanol (3)
68%): m.p. 134 °C; H-NMR (CDCl3) δ 7.77 (s, 1H, 8-H),
5.52 (s, 2H, NCH2O), 4.89 (br s, 2H, NH2), 4.65 (t, 2H,
J = 4.4 Hz, O6-CH2CH2OH), 4.00 (t, 2H, J = 4.4 Hz, O6-
CH2CH2OH), 3.75 and 3.59 (2 × t, 4H, J = 4.4 Hz, OC
H2CH2OTBDMS), 0.88 (s, 9H, SiC(CH3)3), 0.05 (s, 6H, (C
H3Si); 13C-NMR (CDCl3) δ 161.9 (C-6), 160.2 (C-2), 155.0
(C-4), 140.4 (C-8), 115.8 (C-5), 73.7 (NCH2O), 71.5 (O6-
CH2CH2OH), 69.7 (OCH2CH2OTBDMS), 63.2 (O6-
CH2CH2OH), 62.2 (OCH2CH2OTBDMS), 26.5, (SiC(CH3)3),
19.0 (SiC(CH3)3), –4.5 (CH3Si); MS (low FAB): m/z 384.3
(100%) [M + H]●+; 406.2 (15%) [M + Na]●+. Anal. C, H, N.
TBAF deprotection of 23 (0.12 g, 0.31 mmol) gave 5 as a
This was prepared in an analogous series of reactions from
mesitylene derivative 7 (2.1 g, 4 mmol). The DABCO salt 9
was further reacted with N-BOC-aminoethanol and DBU to
yield fully-protected intermediate 15 as a colorless oil (91%):
1H-NMR (DMSO-d6) δ: 7.97 (s, 1H, 8-H), 7.02 (br t, 1H, NH-
BOC), 6.44 (br s, 2H, NH2), 5.41 (s, 2H, NCH2O), 4.35 (t, 2H,
J = 5.5 Hz, O6-CH2), 3.62 and 3.51 (2 × t, J = 4.8 Hz, 4H, OC
H2CH2OTBDMS), 3.33 (br t, 2H, O6-CH2CH2NH-BOC), 1.36
(s, 9H, OC(CH3)3), 0.80 (s, 9H, (CH3)3CSi), –0.02 (s, 6H,
CH3Si); 13C-NMR (DMSO-d6) δ 160.3 (CO), 160.0 (C-6),
155.7 (C-2), 154.5 (C-4), 139.9 (C-8), 113.5 (C-5), 77.8
(OC(CH3)3), 72.1 (NCH2O), 70.2 (OCH2CH2OTBDMS),
64.4 (O6-CH2CH2N), 61.9 (OCH2CH2OTBDMS), 40.4 (O6-
CH2CH2N), 28.2 (OC(CH3)3), 25.8, (SiC(CH3)3), 17.9 (Si
C(CH3)3), –6.4 (CH3Si). IR (KBr) 3515s (OH), 3445s (NH),
3300s (NH2), 3190s (NH2), 1710s (CO), 1630s, 1580s cm–1:
MS (ES): m/z 107 (75%); 153 (100%); 191 (80%); 505 (30%)
[M + Na]●+; 987 (5%) [2M + Na]●+. Anal. C, H, N.
1
white solid (0.086 g, 97%): m.p. 134 °C; H-NMR (CDCl3) δ
7.98 (s, 1H, 8-H), 6.45 (br s, 2H, NH2), 5.41 (s, 2H, NCH2O),
4.88 (t, 1H, J = 5.5 Hz, O6-CH2CH2OH), 4.65 (br t, 1H,
OCH2CH2OH), 4.41 (t, 2H, J = 5.5 Hz, O6-CH2CH2OH),
3.74 (m, 2H, O6-CH2CH2OH), 3.45 (br s, 4H, OC
H2CH2OH); 13C-NMR (CDCl3) δ 160.5 (C-6), 160.0 (C-2),
154.4 (C-4), 139.9 (C-8), 113.5 (C-5), 72.0 (NCH2O), 70.4 (O
CH2CH2OH), 67.5 (O6-CH2CH2OH), 59.9 (OCH2CH2OH),
59.3 (O6-CH2CH2OH). IR (KBr) 3320s (NH2, OH), 1612m,
TBAF deprotection of 15 (1.4 g, 3.6 mmol) gave BOC-
protected amine 16 as a white solid (0.95 g, 98%): m.p. 167–
1470m (C–O) cm–1; MS (FAB): m/z 269 (100%) [M + H]●+
;
1
291 (30%) [M + Na]●+. Anal. C, H, N.
170 °C; H-NMR (DMSO-d6) δ 8.0 (s, 1H, 8-H), 7.06 (br t,
1H, NH–BOC), 6.49 (br s, 2H, NH2), 5.42 (s, 2H, NCH2O),
4.70 (br t, 1H, OCH2CH2OH), 4.41 (t, 2H, J = 5.9 Hz, O6-C
H2CH2NH), 3.47 (br s, 4H, OCH2CH2OH), 3.35 (br t, 2H, O6-
CH2CH2NH), 1.38 (C(CH3)3); 13C-NMR (DMSO-d6) δ 160.3
(CO), 160.0 (C-6), 155.7 (C-2), 154.5 (C-4), 139.9 (C-8),
113.5 (C-5), 77.8 (C(CH3)3), 72.0 (NCH2O), 70.4 (O
CH2CH2OH), 64.4 (O6-CH2CH2NH), 59.9 (OCH2CH2OH),
40.4 (O6-CH2CH2NH–BOC), 28.2 (C(CH3)3 (BOC)). IR
(KBr) 3510s (OH), 3440s (NH), 3300s (NH2), 3200s (NH2),
1715s (CO), 1630s, 1590s, 1470s (C–O) cm–1; MS (FAB):
m/z 165 (30%); 226 (25%); 369 (10%) [M + H]●+; 391 (5%)
[M + Na]●+. Anal. C, H, N.
5.1.10. 2-[2-Amino-9-(2-hydroxyethoxyethyl)-9H-purin-6-
ylamino]ethanol (20)
Prepared analogously from mesitylene derivative 7 (1.35 g,
2.6 mmol). The DABCO salt intermediate 9 was further re-
acted with 2-aminoethanol and DBU to yield protected inter-
1
mediate 14 as a white solid (1.1 g, 88%): m.p. 129 °C; H-
NMR (CDCl3) δ 7.63 (s, 1H, 8-H), 6.2 (br s, 1H, N6-H),
5.47 (s, 2H, NCH2O), 4.73 (br s, 2H, NH2), 3.86 (t, 2H,
J = 4.8 Hz, NHCH2), 3.73 (m, 5H, OCH2CH2OTBDMS,
OH), (t, 2H, J = 4.8 Hz, NHCH2CH2OH), 0.88 (s, 9H,
SiC(CH3)3), 0.05 (s, 6H, (CH3Si)); 13C-NMR (DMSO-d6) δ
160.5 (C-6), 155.0 (C-2), 151.5 (C-4), 137.4 (C-8),
113.0 (C-5), 71.7 (NCH2O), 70.0 (OCH2CH2OTBDMS),
61.9 (NHCH2CH2OH), 60.1 (OCH2CH2OTBDMS), 42.3
For BOC deprotection, 16 (0.225 g, 0.58 mmol) was dis-
solved in TFA (10 ml, 10% in CH2Cl2) and the solution stirred
at room temperature for 10 min. The reaction was monitored

338
D. Pletsas et al. / European Journal of Medicinal Chemistry 41 (2006) 330–339
by TLC (ethyl acetate/methanol, 95:5) for the loss of 16 and
formation of 3. The reaction volume was reduced by /3 and
UK). E. coli K12 strain JM101 for the overproduction of the
human Atase protein (hAtase) was provided by Dr. P.C.E.
Moody (Department of Biochemistry, University of Leicester,
UK) with the agreement of Dr. P. Karran (Clare Hall Labora-
tories, Cancer Research UK). For centrifugation, a Sorval GS3
centrifuge was used. A Packard (1900CA TRI-CARB) liquid
scintillation analyzer and Optiphase Hisafe 3 scintillation cock-
tail (Perkin Elmer, Beaconsfield, UK) were used for H count-
ing. Absorbance measurements were carried out using a Beck-
man DU-640 UV spectrophotometer.
2
ether added. A white precipitate resulted; the reaction solvents
were decanted and ethyl acetate (10 ml) was added. The solid
was stirred with ethyl acetate for 30 min and then filtered to
collect TFA salt 3 as a white solid (0.12 g, 72%): m.p. 245 °C;
1H-NMR (DMSO-d6) δ 8.09 (NH3 ), 8.04 (s, 1H, 8-H), 6.55
+
3
(br s, 2H, NH2), 5.43 (s, 2H, NCH2O), 4.58 (t, 2H, J = 5.5 Hz,
+
O6-CH2CH2NH3 ), 3.45 (br s, 4H, OCH2CH2OH), 3.29 (br s,
2H, O6-CH2CH2NH3 ); 13C-NMR (DMSO-d6) δ 159.9 (C-6),
+
159.8 (C-2), 154.7 (C-4), 140.3 (C-8), 113.4 (C-5), 72.1
+
5.2.1. Preparation of hAtase protein
(NCH2O), 70.6 (OCH2CH2OH), 62.8 (O6-CH2CH2NH3 ),
+
E. coli K12 strain JM101 were grown in L-broth medium
containing ampicillin and crude cell extracts containing hAtase
were prepared following published procedures [41,42]. The fi-
nal hAtase protein stock solution was brought to 15% v/v gly-
cerol, aliquoted and stored at –70 °C; protein content was de-
termined (2.53 mg ml–1) using a Protein Assay Kit (BioRad)
with a BSA standard based upon the method of Bradford [43].
59.9 (OCH2CH2OH), 38.4 (O6-CH2CH2NH3 ). IR (KBr)
3390s (NH2), 3340s (NH2), 3220s (NH2), 2990s (NH2),
1650s, 1590s, 1470s cm–1; MS (EI): m/z 269.2 (100%) [M +
H]●+; MS (CI): 113 (100%) (CF3CO2). Anal. C, H, N.
5.1.12. 5-Amino-3-(2-hydroxyethoxymethyl)-7,8-dihydro-3H-
oxazolo[2,3-i]purin-6-ylium chloride (19)
This was prepared in an analogous series of reactions from
mesitylene derivative 7 (0.54 g, 1.03 mmol). The DABCO salt
9 was further reacted with 2-chloroethanol and DBU to yield a
mixture of 6-(2-chloroethoxy) intermediate 17 and the cyclized
product 18 as an oil after flash column chromatography on
silica eluted with ethyl acetate. The oil was warmed overnight
at 60 °C to complete the conversion. The crude product was
analyzed and used without further purification:
5.2.2. Preparation of the [3H]-methyl labeled DNA substrate
[3H]-MeDNA was prepared by adaptation of a published
method [44]. Calf thymus DNA (200 mg) was treated with
N-[3H]-methylnitrosourea (0.2 mCi, 22 mCi mmol–1) in 40
ml buffer (85 mM Tris–Cl, pH 8.0) at r.t. for 2 h. DNA was
precipitated with NaCl (final conc. 0.1 M) and 2.5 volumes of
ice-cold ethanol, washed three times with 70% ethanol and
once with absolute ethanol. The [3H]-MeDNA was dissolved
in assay buffer (30 ml, 10 mM Tris–HCl, pH 8.0; 1 mM EDTA
pH 8.0) and dialysed against Buffer A (0.1 M NaCl, 0.01 M
sodium citrate, 0.01 M potassium phosphate, pH 7.4, 4 °C,
24 h). The DNA solution was incubated at 80 °C for 16 hours
and then dialysed sequentially against Buffer B (1 M NaCl,
0.1 M Tris–HCl, 1 mM EDTA, pH 8.0, 4 °C, 24 h), and Buffer
C (10 mM Tris–HCl, 1 mM EDTA, pH 8.0, 4 h, 4 °C). The
final [3H]-MeDNA stock solution (173 μg ml–1, determined
spectrophotometrically at 260 nm) with a specific activity of
41.86 cpm μg–1, was aliquoted and stored at –20 °C.
1
17: H-NMR (CDCl3) δ 7.75 (s, 1H, 8-H), 5.51 (s, 2H, NC
H2O), 4.99 (br s, 2H, NH2), 4.73 (t, 2H, J = 6.2 Hz, O6-
CH2CH2Cl), 3.87 (t, 2H, J = 6.2 Hz, O6-CH2CH2Cl), 3.73
and 3.57 (2 × t, 4H, 5.1 Hz, OCH2CH2O), 0.87 (s, 9H, (C
H3)3CSi), 0.04 (s, 6H, (CH3)2Si);
18: 1H-NMR (CDCl3) δ: 7.68 (s, 1H, 8-H), 5.46 (s, 2H, NC
H2O), 5.12 (br s, 2H, NH2), 4.41 (t, 2H, J = 5.5 Hz, O6-CH2),
4.00 (t, 2H, J = 5.5 Hz, CH2N+), 3.75 and 3.61 (2 × t, 4H,
J = 4.4 Hz, OCH2CH2OTBDMS), 0.89 (s, 9H, (CH3)3CSi),
0.06 (s, 6H, (CH3)2Si); MS (FAB): m/z 366 (25%) [M–Cl]●+,
402 (65%) [M + H]●+, 424 (60%) [M + Na]●+. Anal. C, H, N.
TBAF deprotection of 18 (0.273 g, 1.22 mmol) was carried
out under anhydrous conditions, as described above, to yield
19 as a white crystalline solid (0.1 g, 53%); 1H-NMR (DMSO-
d6) δ 7.88 (br s, 2H, NH2), 7.81 (s, 1H, 8-H), 5.34 (s, 2H, NC
H2O), 4.65 (br t, 1H, OCH2CH2OH), 4.04 (t, 2H, J = 8.6 Hz,
O6-CH2CH2N+), 3.64 (t, 2H, J = 8.6 Hz, O6-CH2CH2N+),
3.45 (br s, 4H, OCH2CH2OH); 13C-NMR (DMSO-d6) δ
156.8 (C-6), 155.5 (C-2), 151.1 (C-4), 137.8 (C-8), 116.8 (C-
5), 72.1 (NCH2O), 70.4 (OCH2CH2OH), 59.9 (OCH2CH2OH),
42.3 (O6-CH2CH2N+), 40.5 (O6-CH2CH2N+). IR (KBr) 3230s
br (NH2, OH), 1680s cm–1; MS (CI): m/z 252 (80%) [M – Cl]
●+. Anal. C, H, N.
5.2.3. Suicide inhibition assay
The activity of the newly synthesized compounds was ass-
ayed using a modified version of a published method [25]. As-
say mixtures (total volume 650 μL) contained 225 μl assay
buffer (40 mM DTT, 4 mM EDTA, 200 μM spermidine,
20% glycerol, 200 mM Hepes, 10 μg ml–1 BSA), 266 μl
Tris–HCl buffer (50 μM Tris–HCl pH 7.5, 0.1 mM EDTA,
1 mM DTT), BSA (50 μl, 10 μg ml–1), 100 μl hAtase stock
solution and 9 μl test compound (7.22 mM in DMSO, to give a
final concentration of 100 μM in the assay mixture). The assay
mixture was pre-incubated at 37 °C for 45 min; reaction was
then initiated by addition of [3H]-MeDNA (250 μl) and incuba-
tion continued for a further 2 h. The enzyme reaction was ter-
minated by addition of ice-cold HClO4 (300 μl, 1 M), the sam-
ples were vortexed rapidly, stored on ice for 30 min and
centrifuged (9000 × g) for 15 min. The supernatants were dis-
carded and the pellets washed twice with HClO4 (300 μl, 0.25
M) followed by centrifugation (9000 × g, 10 min, 4 °C) after
5.2. Alkyltransferase assays
N-[3H]-methylnitrosourea was purchased from Amersham
Biosciences, UK. Calf thymus DNA, ampicillin, isopropyl β-
D-thiogalactopyranoside (IPTG), PMSF, polyethyleneimine
and dithiothreitol (DTT) were purchased from Sigma (Poole,

D. Pletsas et al. / European Journal of Medicinal Chemistry 41 (2006) 330–339
339
[11] M.A. Rudek, R.C. Donehower, P. Statkevich, V.K. Batra, D.L. Cutler, S.
D. Baker, Pharmacotherapy 24 (2004) 16–25.
[12] J.A. Quinn, A.H. Friedman, J. Pluda, D.D. Bigner, A.E. Pegg, Neurology
each wash. After removing all traces of the supernatant, the
washed pellets were treated with HCl (300 μl, 0.1 M) and the
resulting suspension heated at 70 °C for 30 min, with frequent
mixing, to hydrolyze residual [3H]-MeDNA. The resulting
samples were centrifuged (9000 × g, 10 min) and the hydroly-
sis step repeated on the pellets. Finally, the washed pellets con-
taining the [3H]-Me-Atase were dissolved in warm NaOH (400
μl, 0.1 M) and incubated at 60 °C for 5 min. The resulting
mixtures were transferred to scintillation vials, diluted with
scintillation (4 ml) cocktail and measured in a scintillation
counter. Optimization experiments were undertaken to deter-
mine appropriate protein concentration and incubation time
for the assay; positive control experiments contained 0.5 μM
O6-benzylguanine in place of the test compound.
54 (2000) A32–A33.
[13] R.L. Schilsky, M.E. Dolan, D. Bertucci, R.B. Ewesuedo, N.J. Vogelzang,
S. Mani, L.R. Wilson, M.J. Ratain, Clin. Cancer Res. 6 (2000) 3025–
3031.
[14] H.S. Friedman, J. Pluda, J.A. Quinn, R.B. Ewesuedo, L. Long, A.
H. Friedman, I. Cokgor, O.M. Colvin, M.M. Haglund, D.M. Ashley, J.
N. Rich, J. Sampson, A.E. Pegg, R.C. Moschel, R.E. McLendon, J.
M. Provenzale, E.S. Stewart, S. Tourt-Uhlig, A.M. Garcia-Turner, J.
E. Herndon, D.D. Bigner, M.E. Dolan, J. Clin. Oncol. 18 (2000) 3522–
3528.
[15] K. Neville, S. Blaney, M. Bernstein, P. Thompson, D. Adams, A. Alek-
sic, S. Berg, Clin. Cancer Res. 10 (2004) 5072–5075.
[16] M.R. Middleton, M. Ranson, S.M. Lee, J. Bridgewater, R.
S. McElhinney, T.B.H. McMurry, N. Thatcher, G. Halbert, S. Arm-
strong, G.P. Margison, Clin. Cancer Res. 6 (2000) 538.
[17] R.S. McElhinney, T.B.H. McMurry, G.P. Margison, Mini Rev. Med.
Chem. 3 (2003) 471–486.
5.2.4. Competitive inhibition assay
All assay mixtures were prepared as described for the sui-
cide inhibition assay but the reaction mixtures were not pre-
incubated. [3H]-MeDNA was added at the same time as the
other reagents and incubations were performed at 37 °C for
45 min. The reactions were terminated and worked up as de-
scribed above.
[18] B.C. Cupid, Z.T. Zeng, R. Singh, D.E.G. Shuker, Chem. Res. Toxicol.
17 (2004) 294–300.
[19] K.L. Harrison, R. Jukes, D.P. Cooper, D.E.G. Shuker, Chem. Res. Tox-
icol. 12 (1999) 106–111.
[20] D.E.G. Shuker, G.P. Margison, Cancer Res. 57 (1997) 366–369.
[21] R.C. Moschel, M.G. McDougall, M.E. Dolan, L. Stine, A.E. Pegg, J.
Med. Chem. 35 (1992) 4486–4491.
[22] M. Qian, R. Glaser, J. Am. Chem. Soc. 127 (2005) 880–887.
[23] K.L. Harrison, N. Fairhurst, B.C. Challis, D.E.G. Shuker, Chem. Res.
Toxicol. 10 (1997) 652–659.
Acknowledgments
[24] M.K. Lakshman, F.N. Ngassa, J.C. Keeler, Y.Q.V. Dinh, J.H. Hilmer, L.
D.P. was supported by a University of Bradford postgradu-
ate studentship; M.C.B. is supported by Cancer Research UK,
grant number A2580; T.C.J. was supported by a Yorkshire
Cancer Research (YCR) programme grant. The authors thank
Dr. P.C.E. Moody (Department of Biochemistry, University of
Leicester, UK) and Dr. P. Karran (Clare Hall Laboratories,
Cancer Research UK) for a gift of E. coli K12 strain JM101
over-expressing hAtase, and Dr. L.J. Adams (YCR Laboratory
of Drug Design, University of Bradford, UK) for helpful assist-
ance with molecular modeling.
M. Russon, Org. Lett. 2 (2000) 927–930.
[25] A.E. Pegg, M. Roberfroid, C. Vonbahr, R.S. Foote, S. Mitra, H. Bresil,
A. Likhachev, R. Montesano, Proc. Natl. Acad. Sci. USA 79 (1982)
5162–5165.
[26] M.E. Dolan, R.C. Moschel, A.E. Pegg, Proc. Natl. Acad. Sci. USA 87
(1990) 5368–5372.
[27] M.E. Dolan, M.Y. Chae, A.E. Pegg, J.H. Mullen, H.S. Friedman, R.
C. Moschel, Cancer Res. 54 (1994) 5123–5130.
[28] D.M. Noll, N.D. Clarke, Nucleic Acids Res. 29 (2001) 4025–4034.
[29] R.S. McElhinney, D.J. Donnelly, J.E. McCormick, J. Kelly, A.J. Watson,
J.A. Rafferty, R.H. Elder, M.R. Middleton, M.A. Willington, T.B.
H. McMurry, G.P. Margison, J. Med. Chem. 41 (1998) 5265–5271.
[30] M.E. Dolan, R.B. Mitchell, C. Mummert, R.C. Moschel, A.E. Pegg, Can-
cer Res. 51 (1991) 3367–3372.
Appendix A. Supplementary material
[31] M.Y. Chae, M.G. McDougall, M.E. Dolan, K. Swenn, A.E. Pegg, R.
C. Moschel, J. Med. Chem. 37 (1994) 342–347.
Supplementary Material Available: doi:10.1016/j.ej-
mech.2005.11.007. Molecular models of O6-functionalized tar-
get purine molecules 1–5 with hAtase, tabulated data from the
Atase assays.
[32] M.R. Middleton, J. Kelly, N. Thatcher, D.J. Donnelly, R.S. McElhinney,
T.B.H. McMurry, J.E. McCormick, G.P. Margison, Int. J. Cancer 85
(2000) 248–252.
[33] J. Hall, H. Bresil, M. Serres, G. Martelplanche, C.P. Wild, R. Montesano,
Cancer Res. 50 (1990) 5426–5430.
[34] T. Lindahl, B. Sedgwick, M. Sekiguchi, Y. Nakabeppu, Annu. Rev. Bio-
chem. 57 (1988) 133–157.
References
[35] K.W. Kohn, Cancer Res. 37 (1977) 1450–1454.
[1] S.A. Kyrtopoulos, L.M. Anderson, S.K. Chhabra, V.L. Souliotis, V. Plet-
sa, C. Valavanis, P. Georgiadis, Cancer Detect. Prev. 21 (1997) 391–405.
[2] G.P. Margison, M.F. Santibanez-Koref, Bioessays 24 (2002) 255–266.
[3] R.D. Kolodner, Trends Biochem. Sci. 20 (1995) 397–401.
[4] S. Ceccotti, E. Dogliotti, J. Gannon, P. Karran, Biochemistry 32 (1993)
13664–13672.
[5] C. Rewinski, M.G. Marinus, Nucleic Acids Res. 15 (1987) 8205–8215.
[6] P.F. Schendel, Crit. Rev. Toxicol. 8 (1981) 311–362.
[7] D.S. Daniels, C.D. Mol, A.S. Arvai, S. Kanugula, A.E. Pegg, J.
A. Tainer, EMBO J. 19 (2000) 1719–1730.
[36] D.S. Daniels, T.T. Woo, K.X. Luu, D.M. Noll, N.D. Clarke, A.E. Pegg,
J.A. Tainer, Nat. Struct. Mol. Biol. 11 (2004) 714–720.
[37] A.E. Pegg, S. Kanugula, S. Edara, G.T. Pauly, R.C. Moschel, K. Good-
tzova, J. Biol. Chem. 273 (1998) 10863–10867.
[38] C.E. Arris, C. Bleasdale, A.H. Calvert, N.J. Curtin, C. Dalby, B.
T. Golding, R.J. Griffin, J.M. Lunn, G.N. Major, D.R. Newell, Antican-
cer Drug Des. 9 (1994) 401–408.
[39] U.M. Ohndorf, M.A. Rould, Q. He, C.O. Pabo, S.J. Lippard, Nature 399
(1999) 708–712.
[8] J.K.V. Willson, J.R. Haaga, J.E. Trey, T.A. Stellato, J. Clin. Oncol. 13
[40] N.K. Lembicz, S. Grant, W. Clegg, R.J. Griffin, S.L. Heath, B.
T. Golding, J. Chem. Soc, Perkin Trans. I (1997) 185–186.
[41] J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning: A Laboratory
Manual, 2nd ed, Cold Spring Harbor Laboratory Press, New York, 1989.
[42] M.C. Psaroudi, S.A. Kyrtopoulos, Mutat. Res. 447 (2000) 257–265.
[43] M. Bradford, Anal. Biochem. 71 (1976) 248–254.
[44] A.E. Pegg, W. Perry, R.A. Bennett, Biochem. J. 197 (1981) 195–201.
(1995) 2301–2308.
[9] E.S. Newlands, G.R.P. Blackledge, J.A. Slack, G.J.S. Rustin, D.
B. Smith, N.S.A. Stuart, C.P. Quarterman, R. Hoffman, M.F.G. Stevens,
M.H. Brampton, A.C. Gibson, Br. J. Cancer 65 (1992) 287–291.
[10] J.J. Raizer, M.G. Malkin, M. Kleber, L.E. Abrey, Neuro-oncol. 6 (2004)
247–252.