Synthesis and in vitro evaluation of novel 2,6,9-trisubstituted purines acting as cyclin-dependent kinase inhibitors

Article abstract of DOI:10.1016/S0968-0896(99)00064-4

Novel C-2, C-6, N-9 trisubstituted purines derived from the olomoucine/roscovitine lead structure were synthesized and evaluated for their ability to inhibit starfish oocyte CDK1/cyclin B, neuronal CDK5/p35 and erk1 kinases in purified extracts. Structure

Full text of DOI:10.1016/S0968-0896(99)00064-4

Bioorganic & Medicinal Chemistry 7 (1999) 1281±1293
Synthesis and In Vitro Evaluation of Novel 2,6,9-Trisubstituted
Purines Acting as Cyclin-dependent Kinase Inhibitors
Michel Legraverend,a Odile Ludwig,a Emile Bisagni,a Sophie Leclerc,b Laurent Meijer,b
Nicole Giocanti,c Ramin Sadric,y and Vincent Favaudonc,*
aUMR 176 CNRS-IC, Institut Curie-Recherche, BaÃt. 110-112, Centre Universitaire, 91405 Orsay Cedex, France
bGroupe Cycle Cellulaire, UPR 9042 CNRS, Station Biologique, BP 74, 29682 Rosco€ Cedex, France
Â
Ã
Unite 350 INSERM, Institut Curie-Recherche, Bat. 110-112, Centre Universitaire, 91405 Orsay Cedex, France
c
Received 10 August 1998; accepted 23 November 1998
AbstractÐNovel C-2, C-6, N-9 trisubstituted purines derived from the olomoucine/roscovitine lead structure were synthesized and
evaluated for their ability to inhibit star®sh oocyte CDK1/cyclin B, neuronal CDK5/p35 and erk1 kinases in puri®ed extracts.
Structure±activity relationship studies showed that increased steric bulk at N-9 reduces the inhibitory potential whereas substitution
of the aminoethanol C-2 side chain by various groups of di€erent size (methyl, propyl, butyl, phenyl, benzyl) only slightly decreases
the activity when compared to (R)-roscovitine. Optimal inhibitory activity against CDK5, CDK1 and CDK2, with IC50 values of
0.16, 0.45 and 0.65 mM, respectively, was obtained with compound 21 containing a (2R)-pyrrolidin-2-yl-methanol substituent at the
C-2 and a 3-iodobenzylamino group at the C-6 of the purine. Compound 21 proved cytotoxic against human tumor HeLa cells
(LD50=6.7 mM versus 42.7 mM for olomoucine, 24-h contact). Furthermore, unlike olomoucine, compound 21 was e€ective upon
short exposure (LD50=25.3 mM, 2-h contact). The available data suggest that the anity for CDKs and the cytotoxic potential of
the drugs are inter-related. However, no straightforward cell cycle phase speci®city of the cytotoxic response to 21 was observed in
synchronized HeLa cells. With the noticeable exception of pronounced lengthening of the S-phase transit by 21 applied during
early-S in synchronized HeLa cells, and in striking contrast with earlier reports on studies using plant or echinoderm cells, olo-
moucine and compound 21 were unable to reversibly arrest cell cycle progression in asynchronous growing HeLa cells. Some irre-
versible block in G1 and G2 phase occurred at high olomoucine concentration, correlated with induced cell death. Moreover,
chronic exposure to lethal doses of compound 21 resulted in massive nuclear fragmentation, evocative of mitotic catastrophe with
minor amounts of apoptosis only. It was also found that olomoucine and compound 21 reversibly block the intracellular uptake of
nucleosides with high eciency. # 1999 Elsevier Science Ltd. All rights reserved.
Introduction
¯avopiridol,5,6 and 2,6,9-trisubstituted purines.7 The
®rst substituted purine derivative with high CDK inhi-
bitory activity, olomoucine (Scheme 1), was identi®ed
from screening against CDK1/cyclin B.8 Among 35
kinases tested, olomoucine was found to competitively
inhibit CDK1, CDK2, CDK5 and to a lesser extent
erk1.8 A more active 2,6,9-trisubstituted purine, ros-
covitine, has recently been described.9±11
Complexes of cyclins and cyclin-dependent kinases
(CDKs) play a central role in the control of the cell
division cycle.1 The frequent deregulation of cell cycle
progression in cancer2 has prompted an active search
for kinase inhibitors with high anity and speci®city
for CDKs. Three major classes of CDK-targeting
compounds have been identi®ed to date, including
butyrolactone I,3,4 polyhydroxylated ¯avones such as
The crystal structures of CDK2 complexes with olo-
moucine and (R)-roscovitine have been solved at 2.2±2.4
A resolution.9,12 Interestingly, the position of the purine
nucleus in the ATP binding pocket of CDK2 di€ers
markedly from that for the natural substrate, ATP. In
addition, previous structure±activity relationship (SAR)
studies8,9 suggested that further structural variations at
C-2, C-6 and N-9 might lead to more potent and more
selective inhibitors of CDK2 and other CDKs. We
have undertaken a new SAR program in order to test
these possibilities and bring about novel, CDK-target-
ing trisubstituted purines acting below the micromolar
Key words: Purines; olomoucine; cyclin-dependent kinases; cell cycle;
proliferation; human cells.
Abbreviations: BrdUrd, 5-bromo-20-deoxyuridine. EGTA, ethylene-
bis(oxonitrilo)tetraacetate. SAR, Structure±activity relationship.
EtOH, ethanol. CDK, cyclin-dependent kinase.
* Corresponding author. Tel.: 33-1-6986-3188; fax: 33-1-6986-3187;
e-mail: vincent.favaudon@curie.u-psud.fr
Present address: Department of Molecular and Chemical Endocri-
nology, St. Bartholomew's Hospital, 51±53 Bartholomew Close, Lon-
don EC1A 7BE, UK.
y
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(99)00064-4

1282
M. Legraverend et al. / Bioorg. Med. Chem. 7 (1999) 1281±1293
purine derivatives 3 using isoamylnitrite and a mixture
of CH2I2, I2 and CuI as the iodide source25,26 in either
THF or acetonitrile as solvent. When applied to cyclo-
pentenyl precursor 3f, this reaction led to a complex
mixture from which 4f could not be isolated. Nucleo-
philic substitution of the 6-chloro substituent was
then speci®cally achieved by reaction with the amine
below 45 ꢀC. Further substitution of the 2-iodo function
by another amine, required higher temperature (120±
140 ꢀC).
The structures of the compounds obtained in this way
are described in Table 1. As clearly shown by the var-
ious examples described in this work, the synthetic
strategy developed in this study allows unlimited varia-
tions at the 2, 6 and 9 positions of purine.
Scheme 1. Chemical structure of olomoucine and compound 21.
range, with the goal of providing new tools for cell cycle
studies as well as pharmacological applications. Indeed,
recent results point to both the antitumor potential of
this family of purines,13,14 and their ability to provide
selective protection of neurons against apoptotic cell
death.15,16 It should also be stressed that CDK inhibi-
tors might ®nd utilization in non malignant disorders
involving deregulated proliferation or metabolism
such as psoriasis, restenosis,17 arteriosclerosis, athero-
sclerosis, glomerulonephritis,18 proliferation of uni-
cellular parasites, cytomegalovirus infections19 or
Alzheimer's disease.20 One compound in the series dis-
played sub-micromolar IC50s against major CDKs. This
prompted us to compare the e€ects of this new deriva-
tive to those of the parent compound olomoucine in a
human cell line, including cytotoxicity studies, cell cycle
phase speci®city and redistribution, nucleoside import,
and interaction with ionizing radiation. HeLa cells were
chosen as a model system for this study, mostly because
they do not express multi-drug resistance, grow as well-
de®ned colonies and can be synchronized with great
ease and eciency. The results suggest that altered cell
cycle progression and induced cell death by trisubstituted
purines, may proceed through di€erent pathways.
Structure±activity relationships
Previous preliminary structure±activity studies8 of the
purine ring have established that the presence of C-2, N-6
and N-9 substituents on adenine are critical for CDK1
inhibitory properties. These studies allowed identi®ca-
tion of olomoucine8 and subsequently of (R)-roscov-
itine9,10 as new lead compounds that speci®cally inhibit
CDK1, CDK2, CDK5 and, to a lesser extent, erk1.
The crystal structure of the CDK2-olomoucine complex
at 2.2 A resolution12 and the CDK2-(R)-roscovitine
complex at 2.4 A resolution,9 revealed 16 and 20 van der
Waals interactions, respectively, between the drug N-6
benzyl substituent and the enzyme. We observed that
replacement of the N-6 benzyl group by either an N-
methyl allylamino group (7), or
a 1,2,3,6-tetra-
hydropyridinyl system (8) strongly decreased inhibitory
activity. However, increased anti-CDK1 and anti-
CDK5 activities of the C-6-(3-iodobenzylamino) deri-
vative 21 as compared to 10 should be noticed (Table 1).
Although an increased inhibitory e€ect against CDK1
was observed upon replacement of the 9-methyl sub-
stituent of olomoucine by an isopropyl group, as in (R)-
roscovitine,9 we found that, with the exception of the
cyclopentyl derivative 24 which exhibited the same
inhibitory activity as the isopropyl derivative 10, a
further increase in steric bulk and/or hydrophobicity at
N-9 strongly decreases the inhibitory activity (Table 3).
The C-2 side chain appears also to be essential in two
ways. First, a hydrogen bond was detected between
CDK2 Gln-131 and the hydroxyl group in olomoucine
and (R)-roscovitine CDK2 co-crystals.9,12 Second, sub-
stitution of the aminoethanol C-2 side chain by hydro-
phobic groups of varying size (methyl, propyl, butyl,
phenyl, benzyl) slightly decreased the activity as com-
pared to (R)-roscovitine (Table 1).
Results
General chemicals pathways
The preparation of 2,6-disubstituted purines has already
been described using solid-phase21 or solution-phase
methodology.22 The general synthetic pathway to tri-
substituted purines used in the present work is outlined
in Scheme 2 and employed 2,5-diamino-4,6-dichloro-
pyrimidine 123 as the starting material.* Selective C-4
monosubstitution of 1 was achieved by reaction with
one equivalent of the appropriate amine in ethanol
below 100 ꢀC. Acid-catalyzed cyclization of 2 to 3 was
carried out in N,N-dimethylacetamide at room tem-
perature using an excess of triethylorthoformate.
6-Chloro-2-iodopurine (4) was subsequently obtained
by diazotization-substitution of the 2 amino-6-chloro-
However, it is interesting to note the di€erence in activ-
ity between 9 and 10 in which the C-2 side chain is a 2-
(S)-pyrrolidine methanol and 2-(R)-pyrrolidine metha-
nol, respectively. Indeed, the 2-(R)-pyrrolidine methanol
derivative 10 exhibits more potent inhibitory activity
(IC50=0.65 mM) than the (S) derivative 9 (IC50=
2.1 mM). The R con®guration compound was always the
* The synthesis of some potent inhibitors of CDK2, using solid sup-
port technology, has been reported during the reviewing of this
paper.24

M. Legraverend et al. / Bioorg. Med. Chem. 7 (1999) 1281±1293
1283
Scheme 2. General pathways for the synthesis of the 2,6,9-trisubstituted purines in the series reported herein.
most active (Table 1), (i.e. 15>14; 17>16; 19>20;
10>9; and (R)-roscovitine (28)>(R,S)-roscovitine (27)).
also assayed on recombinant CDK5/p35. Thirdly,
because olomoucine and (R)-roscovitine inhibit erk1
and erk2 at micromolar concentrations (IC50s of 50 and
40 mM [olomoucine], 34 and 14 mM [(R)-roscovitine], for
erk1 and erk2, respectively), we tested all compounds on
puri®ed recombinant erk1. The results are summarized
in Table 1.
The most active compound in this series was the 6-(3-
iodobenzyl)-amino derivative 21 with an IC50 of 0.45,
0.65 and 0.16 mM against CDK1, CDK2 and CDK5,
respectively. In fact, 21 was much more active than the
lead compound olomoucine and as potent as (R)-ros-
covitine. As observed previously for olomoucine and
(R)-roscovitine, all the CDK1 and CDK5 inhibitors (9±
24) described in this paper were very weak or inactive
against erk1 MAP kinase (Table 1). The high activity of
21 suggests that adjunction of bulkier lipophilic groups
to the 6 benzylamino substituent may lead to other
more potent compounds. Further experiments will be
carried out to investigate this possibility.
Compound 21 emerged as the most potent inhibitor in
the series. It proved as ecient as (R)-roscovitine
against CDK5, and was still more active against CDK1.
However, in agreement with earlier results,8 the data
show that the inhibitory eciencies against CDK1 and
CDK5 are inter-related. With the exception of com-
pound 22 (17 times more potent inhibitor of CDK5)
and compound 15 (6 times more potent inhibitor of
CDK1), no CDK5- or CDK1-speci®c inhibitor could be
identi®ed, con®rming the close evolutionary relation-
ship between the two kinases. On the other hand, most
compounds at 10 mM concentration marginally inacti-
vated erk1 kinase only. We also determined recombi-
nant CDK2/cyclin E inhibition by olomoucine, (R)-
roscovitine and compound 21. Compound 21 again
proved over 10-fold more active than olomoucine, and
was as ecient as (R)-roscovitine in this assay (Table 2).
Kinase e€ects
All compounds in the series were routinely probed for
their ability to inhibit the CDK1/cyclin B complex from
star®sh oocytes. Due to the putative importance in Alz-
heimer disease of hyperphosphorylation of the micro-
tubule-associated protein tau27 and neuro®lament
subunits28 by neuronal CDK5/p35, all compounds were

1284
M. Legraverend et al. / Bioorg. Med. Chem. 7 (1999) 1281±1293
Table 1. Compared IC50s of purine analogues in the olomoucine series for inhibition of puri®ed erk1 kinase and p35/CDK5 or cyclinB/CDK1
complexes. Trivial names: compound 26, olomoucine; compound 27, (R,S)-roscovitine; compound 28, (R)-roscovitine
Compounda R1
R3
R2
IC50 CDK1 IC50 CDK5 erk1 activity
(%)b
(mM)
(mM)
3f
Cyclopent-3-ene
Isopropyl
Methyl
Isopropyl
Isopropyl
Cyclopentyl
Cyclohexyl
Benzyl
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
260
150
18
11
8.0
75
>100
>100
40
97
97
94
97
88
94
91
Ð
97
76
100
70
69
94
88
93
84
84
95
89
77
78
83
76
69
99
97
97
76
77
Ð
4b
5a
5b
5c
5d
5e
5f
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
N-Methyl, allylamino
N-Methyl, allylamino
Benzylamino
Benzylamino
Benzylamino
Benzylamino
N-Methyl, allylamino
N-Methyl, allylamino
1,2,3,6-Tetrahydropyridin-1-yl
Benzylamino
Benzylamino
Benzylamino
Benzylamino
Benzylamino
Benzylamino
Benzylamino
Benzylamino
Benzylamino
Benzylamino
Benzylamino
Benzylamino
3-Iodobenzylamino
Benzylamino
Benzylamino
Benzylamino
Benzylamino
Benzylamino
Benzylamino
Benzylamino
14
13
>100
>100
>100
>100
>100
>100
2.4
90
100
470
150
110
2.1
0.65
2.5
6.5
2.5
4.3
1.0
7.3
2.7
0.9
0.85
1.0
0.45
42
Methyl
2-Amino-ethanol
2-Amino-ethanol
2-Amino-ethanol
Isopropyl
Isopropyl
Isopropyl
Isopropyl
Isopropyl
Isopropyl
Isopropyl
Isopropyl
Isopropyl
Isopropyl
Isopropyl
Isopropyl
Isopropyl
Isopropyl
Isopropyl
Isopropyl
Cyclohexyl
Cyclopentyl
Benzyl
(2S)-Pyrrolidin-2-yl-methanol
(2R)-Pyrrolidin-2-yl-methanol
2-Benzyl-amino-ethanol
2-Amino-1-phenyl-ethanol
(2R,S)-2-Amino-hexanol
(2S)-2-Amino-2-phenyl-ethanol
(2R)-2-Amino-2-phenyl-ethanol
(2S)-2-Amino-2-benzyl-ethanol
(2R)-2-Amino-2-benzyl-ethanol
(2R,S)-2-Amino-pentanol
(2R)-2-Amino-propanol
(2S)-2-Amino-propanol
(2R)-Pyrrolidin-2-yl-methanol
(3R,S)-Pyrrolidin-3-ol
(2R)-Pyrrolidin-2-yl-methanol
(2R)-Pyrrolidin-2-yl-methanol
(2R)-Pyrrolidin-2-yl-methanol
2-Amino-ethanol
1.3
8.0
2.3
2.3
3.5
6.0
2.0
1.3
0.4
0.7
35
0.16
2.4
13
0.5
>100
40
0.7
200
7.0
0.65
0.45
Methyl
Isopropyl
Isopropyl
3.0
0.16
Ð
(2R,S)-2-Amino-butanol
(2R)-2-Amino-butanol
a See Scheme 2 for R1, R2, and R3.
b Percent residual activity of erk1 kinase in the presence of 10 mM of various compounds.
Table 2. Compared inhibitory concentrations of olomoucine, (R)-
roscovitine and compound 21 for the inactivation of puri®ed
recombinant CDK2/cyclin E complex (IC50), and for loss of the
clonogenic potential of asynchronously growing human tumor HeLa
Table 3. E€ect of the N-9 substituent on the IC50 value relative to
inhibition of CDK1/cyclin B activity
N-9 substituent
IC50 (mM)
cells (LD50
)
10
24
23
25
Isopropyl
Cyclopentyl
Cyclohexyl
Benzyl
0.65
0.70
40
Drug
IC50 (mM) Duration of contact
with cells
LD50 (mM)
200
Olomoucine
Olomoucine
(R)-Roscovitine
Compound 21
Compound 21
7.0
16 h
24 h
24 h
2 h
156
42.7
6.6
25.3
6.7
0.70
0.65
24 h
olomoucine, and was cytotoxic from the very beginning
of contact with drug (Table 2). Whatever the duration
of drug exposure, the reproductive ability of HeLa cells
decreased following a purely exponential dose±e€ect
relationship for both drugs (data not shown).
Cytotoxicity assays
The viability of proliferating cells following exposure to
olomoucine or compound 21, was assessed in clono-
genic assays. Even at the highest usable concentration,
close to the limit of solubility of the drug, olomoucine in
short (2 h) exposure did not produce any signi®cant
level of cytotoxicity. Olomoucine response increased
steadily with the duration of exposure for up to 24 h,
but further increase in the length of contact with drug
had a marginal e€ect only. The cytotoxic potential of 21
also reached a plateau after 24-h exposure. However,
compound 21 proved to be much more potent than
Cell cycle phase speci®city of compound 21 cytotoxicity
Cell cycle phase-speci®city of the cytotoxic response is
commonly observed with antitumor drugs, and may act
as a major determinant of drug outcome. In order to
investigate this issue, synchronized HeLa cells were
exposed for 1 h to 50 mM 21 at graded times following
release from a thymidine block. There was no signi®cant
change of HeLa cells killing by 21 with progression
through the cell cycle (Fig. 1).

M. Legraverend et al. / Bioorg. Med. Chem. 7 (1999) 1281±1293
1285
endoreduplication (polyploidy) or aberrant cytokinesis
(multinucleation) was found. Moreover, the amount of
hypodiploid nuclei remained to a basal level, less than
3% in terms of the DNA content as determined by ¯ow
cytometry, though less than 5% of the bulk cell popu-
lation was able to resume growth after drug removal
under these conditions.
The situation was substantially di€erent with compound
21. First of all, exposure to 25 mM compound 21
depressed the ability of cells to incorporate BrdUrd,
consistent with inhibition of DNA synthesis in S-phase
and/or of nucleoside uptake (see below). Second, ¯ow
cytometric analysis showed a high level of micronuclea-
tion yielding nuclei with random hypodiploid and
hypotetraploid DNA content (Fig. 3). Agarose gel elec-
trophoresis of DNA extracts did not get evidence of the
DNA ladders characteristic of apoptosis (data not
shown). The amount of cells undergoing nuclear frag-
mentation upon chronic exposure to 25 mM 21 was
considerably higher than with 100 mM olomoucine, and
reached completion upon 48-h incubation. This not-
withstanding, fragmentation did not occur at and below
12.5 mM compound 21 (48-h incubation), yet less than
10% of cells survived this treatment. Otherwise, S-phase
depletion was the main feature characterizing altered
cell cycle progression by compound 21. The potency of
21 to arrest cells in G2 seemed relatively minor com-
pared to that for olomoucine (Fig. 2).
Altered S-phase progression by compound 21 in
synchronized cells
Although cytotoxicity assays did not provide evidence
for cell cycle phase speci®city of HeLa cell killing, we
reasoned that if it occurred, the use of synchronized
HeLa cells could allow for a more precise determination
of altered cell cycle progression by olomoucine deriva-
tives. Synchronized HeLa cells were thus exposed brie¯y
(1 h) to 50 mM 21 at di€erent times following release
from a thymidine block, and collected 24 h after block
release for ¯ow cytometric analysis. Cells in untreated
samples were exactly halfway through the second S-
phase following thymidine block release, having succes-
sively traversed S-, G2-, M-, and G1-phases. Pulse con-
tact with 21 was found to slow down cell cycle traversal.
The e€ect was more pronounced as the drug was given
early in S-phase in the ®rst cycle, and disappeared pro-
gressively thereafter. Indeed, a 1-h exposure to drug in
early-S phase delayed entry into the next S-phase by as
much as 5 h (Fig. 4).
Figure 1. Cytotoxicity of compound 21 (50 mM, 1 h exposure) against
HeLa cells progressing synchronously through the cell cycle following
release from a double thymidine block. Flow cytometry (upper) and
survival assays (lower) were carried out in parallel. Drugs were intro-
duced at the times indicated. Survival data have been corrected for cell
multiplicity.54 Bars, SD.
Cell cycle phase redistribution by drugs
Asynchronous growing HeLa cells were exposed for
extended periods of time to olomoucine or compound
21 at concentrations ranging from 1 to 100 mM, then
collected after BrdUrd labeling, ®xed in 70% EtOH and
analyzed by ¯ow cytometry to determine drug-induced
cell cycle redistribution.
Incubation of HeLa cells with 100 mM olomoucine for
48 h resulted in pronounced S-phase depletion and
minor accumulation of cells in the G2 phase, preceded
by a transient block of cells in S-phase culminating at
24 h (Fig. 2). The G1 compartment evolved as the mir-
ror image of S-phase, suggesting that the cell fraction
exiting the G2 block through mitosis subsequently
accumulated in G1. Limited e€ects only on cell cycle
progression were observed at lower drug concentra-
tions, with S-phase cells being the most reactive. Even at
the highest olomoucine concentration (100 mM) and for
the longest incubation time (48 h), no evidence for
Inhibition of thymidine uptake by drugs
Olomoucine and compound 21 were tested for their
ability to alter uptake and incorporation of radio-
actively labeled thymidine by living cells, following a
procedure similar to that described by Buquet-Fagot et
al.29 Both drugs actually inhibited thymidine import
(Fig. 5). The process exactly ®tted an equilibrium equa-
tion, with Kd values in the same range as those for
CDKs in acellular extracts (Tables 1 and 2), and was
fully reversible upon drug removal.

1286
M. Legraverend et al. / Bioorg. Med. Chem. 7 (1999) 1281±1293
Figure 2. Time and drug concentration-dependence of cell cycle redistribution by olomoucine and compound 21 in asynchronously growing cultures
of HeLa cells. The length of contact with drug in the drug concentration dependence assay was 48 h. FRAG refers to total hypodiploid and hypo-
tetraploid nuclear fragments generated by the treatment and measured by cyto¯uorimetry in the same way as in Figure 3.
Altered radiation response by drugs
groups results in loss of the inhibitory potential. In
contrast, both C-2 and C-6 positions allow a large vari-
ety of variations, leading to substantial changes in the
inhibitory strength. Crystal structure analysis9,12 has
identi®ed the benzyl ring as the modi®cation responsible
for the selectivity of this family of compounds towards
CDKs; this motif interacts with amino acid residues
which are not conserved among kinases, and which do
not interact with ATP. This region of the molecule will
be the focus of future work. The C-2 position allows a
large variety of substitutions. Finally, optimal activity
was observed with a (R)-pyrrolidine methanol sub-
stituent at the C-2 of the purine whereas, as previously
observed,8,11 an isopropyl substituent at N-9 gave the
best activity. The high activity of 21 against CDK1,
CDK2 and CDK5 corroborated previous observa-
tions22,30 indicating that substitution of the 6-benzyl-
amino substituent by bulky lipophilic groups may lead
to more potent compounds. Our data show that human
HeLa cell response to substituted purines di€ers mark-
edly from what might be anticipated from studies on cell
extracts or puri®ed enzymes in the test tube, as well as
from e€ects found in algae, plants, echinoderms cells or
oocytes. In particular, olomoucine and/or (R)-roscov-
itine reportedly induce reversible arrest of plant cells in
late-G1 and late-G2 phases of the cell cycle,31,32 inhibit
Near-suppression of the initial slope (a) of the radiation
survival curve was observed as cells were exposed to
compound 21 (Fig. 6). Survival curves in the high dose
range with or without drug were parallel, and in fact the
presence of drug did not change signi®cantly the value
of the quadratic (b) parameter characterizing high dose
radiation survival. The same e€ect was observed with
100 mM olomoucine or 60 mM (R)-roscovitine (data not
shown). Therefore, short exposure to cytotoxic con-
centrations of olomoucine derivatives applied con-
comitantly with radiation results in very ecient
radiation protection selectively at low radiation doses,
but does not hinder radiation-induced cell kill at high
radiation doses.
Discussion
The series of purines described in this paper are close
structural analogues of the CDK inhibitors olomoucine
and (R)-roscovitine. Kinase assays for this new series
con®rm that C-2, C-6, and N-9 substituted purines are a
rich source of selective CDK-targeting inhibitors. Note,
however, that substitution of the N-9 position by bulky

M. Legraverend et al. / Bioorg. Med. Chem. 7 (1999) 1281±1293
1287
Figure 3. Dot plot cytograms showing cell cycle redistribution and micronucleation induced by compound 21 (25 mM) in HeLa cells as a function of
the length of contact with drug. Data were obtained from bivariate cyto¯uorimetric analysis, based on concomitant DNA ¯uorescence (propidium
iodide) and immuno¯uorescence (BrdUrd) detection (see Experimental). The propidium iodide signal (FL3) is proportional to the DNA content.
The BrdUrd signal (FL1) was converted through an 8-bits (256 channels) logarithmic ampli®er calibrated versus maximum ¯uorescence emission.
Each dot represents a particular nucleus or nuclear fragment. HDF, hypodiploid fragments; HTF, hypotetraploid fragments.
the maturation of star®sh oocytes,10,33 and arrest the
growth of Xenopus oocytes and sea urchin eggs.10 Both
purines thus have the ability to block cell cycle progres-
sion at de®nite checkpoints, suggesting that CDK2/
cyclin A, CDK2/cyclin E and CDK1/cyclin B complexes
are major targets for these drugs in these species. In
keeping with these results, we found that olomoucine
and compound 21 (the most active compound in the
series) did not produce reversible accumulation of HeLa
cells in G1 and G2 phase. When observed, such blocks
were irreversible and correlated with large amounts of
induced cell death. These results are in close agreement
with those reported by Schutte et al.34 and Buquet-
Fagot et al.29 Therefore, CDKs may not be the only
critical targets for substituted purine-induced cell death
in human cells. Brusselbach et al.35 working on the ¯avone
derivative ¯avopiridol, another potent CDK inhibitor,
recently reached similar conclusions. Furthermore,
HeLa cells did not show any particular cell cycle phase-
dependent cytotoxic response to the drug (Fig. 1). Dif-
ferences in the eciency of drug uptake might explain
such di€erential response of various species or cell lines
for induced killing or altered cell cycle regulation by
substituted purines. Indeed, as demonstrated by
Buquet-Fagot et al.29 exposure of growing cells to olo-
moucine or (R)-roscovitine results in rapid, ecient
inhibition of nucleoside intake. Figure 5 shows that this
inhibitory process obeys an equilibrium, with Kd values
in the same range as those determined for puri®ed
CDK1 (Table 1) or CDK2 (Table 2). This suggests that
purine derivatives in the olomoucine series act as com-
petitive inhibitors at the ATP binding site of a protein
involved in the active transport of nucleosides, in rather
the same way as for the ATP binding pocket of CDKs.
This may produce self-inhibition of purine import, and
contribute to drug-induced lethality upon prolonged

1288
M. Legraverend et al. / Bioorg. Med. Chem. 7 (1999) 1281±1293
Figure 6. g-Ray survival of asynchronous growing HeLa cells without
or with concomitant treatment with compound 21. Drug (40 mM) was
introduced 1 h prior to irradiation and was present for a total of 2 h.
Survival curves were ®tted to the linear-quadratic eq (2) (see Experi-
Figure 4. Characterization of the cell cycle delay produced by pulse
exposure to compound 21. HeLa cells were synchronized by a double
thymidine block and exposed to 21 (50 mM, 1 h contact) at di€erent
times following block release. At 24 h after block release, cells were
labeled with BrdUrd (10 mM, 15 min), trypsinized, ®xed in 70% cold
EtOH and subsequently examined by ¯ow cytometry. Untreated cells
at that time were exactly halfway through the second S-phase follow-
ing synchronization. Delay to the second S-phase was the larger as 21
was given closer to the beginning of the ®rst S-phase. Cell cycle pro-
gression in drug-free synchronized cells is shown in Figure 1.
2
mental). Found, a=0.186‹0.054 Gy 1 and b=0.0418‹ 0.0016 Gy
1
for radiation alone; a=0.0030‹0.0006 Gy and b=0.0475‹0.0074
Gy in combined treatment with 21; fractional survival to 21 alone
was 0.63. Bars, SD.
2
particular, compound 21 induced a prolonged, yet
reversible lag in progression through the S-phase of the
cell cycle in synchronized HeLa cells. Susceptibility to
this e€ect peaked in early-S, decreased with progression
through S-phase, and disappeared as cells reached the
S-G2 transition (Fig. 4). This gives circumstantial evi-
dence in favor of CDK2/cyclin E and CDK2/cyclin A
complexes being major targets of olomoucine analogues
in HeLa cells. As a matter of fact, both cyclin E and
cyclin A are considerably overexpressed in cells syn-
chronized at the G1-S boundary by thymidine,37 thus
putatively o€ering increased substrate concentration for
drug action. Moreover, early events in DNA replication
require CDK2/cyclin E38 and CDK2/cyclin A,39,40 while
CDK2/cyclin A is necessary to progression through S
phase.41 It is thus conceivable that inhibition of CDK2/
cyclin E or CDK2/cyclin A activity may result in slow-
ing down early-S-phase progression in some cell lines.
Furthermore, it has long been known that S-phase is the
most radioresistant compartment of the cell cycle.42±45
Reducing the rate of the S-phase transit is expected to
provide more time for radiation-induced DNA damage
repair, thus resulting in substantial enhancement of the
radiation resistance among drug survivors, consistently
with the e€ect we observed in the low radiation dose
range (Fig. 6).
drug exposure. Side e€ects from altered microtubule
dynamics by impaired kinase activity36 might also occur
and contribute to the lethal e€ect of the olomoucine
derivatives.
Nevertheless, there is a striking parallel between the
anity for CDKs and the cytotoxic potential of trisub-
stituted purines, and these appear to be able to target
speci®c cell cycle compartments in living HeLa cells. In
Experimental
Chemistry
Figure 5. Inhibition of thymidine uptake by olomoucine or 21. Asyn-
chronous growing HeLa cells (1.00‹0.05Â106 cells) were incubated
with drugs for 90 min prior to 30 min contact with [methyl-3H]thymi-
dine (see Experimental). The process was analyzed as a reversible
equilibrium,
Analytical methods. The melting points were taken on a
Ko¯er hot stage apparatus, and are uncorrected. Ele-
mental analyses were performed by the Service de
Microanalyse, CNRS, ICSN, Gif-sur-Yvette, France.
Proton nuclear magnetic resonance (1H NMR) spectra
were recorded at 200 MHz on a Bruker AC-200 spec-
trometer. Chemical shift values are reported in d (ppm),
relative to tetramethylsilane as an internal standard.
E ‡ D
ED
!
where E is a putative membrane receptor involved in the active trans-
port of thymidine, D the drug, and ED the enzyme±drug complex.
With the most active compound, thymidine intake was abolished at
high drug concentration, indicating that the ED complex is totally
unable to carry out thymidine import. Kd values were calculated at
24.1‹1.0 mM for olomoucine and 0.948‹0.014 mM for compound 21.

M. Legraverend et al. / Bioorg. Med. Chem. 7 (1999) 1281±1293
1289
Preparation of compounds 3a±f. In separate experi-
ments, methylamine, isopropylamine, cyclopentylamine,
cyclohexylamine, benzylamine (purchased from Aldrich)
or cyclopent-3-ene-1-yl-amine46 were treated with 1.1
equiv of 2,5-diamino-4,6-dichloropyrimidine (1) in n-
BuOH containing an excess of triethylamine at 100 ꢀC
for 48 h (argon atmosphere).47 Compounds 2 were then
cyclized to 3a±f at room temperature by treatment with
triethyl orthoformate in N,N-dimethylacetamide, con-
taining a catalytic amount of aqueous concentrated HCl.
8.75 (s, 1H), 4.44 (m, 1H), 2.00±1.44 (m, 10H). Anal.
(C11H12N4ClI) C,H,N.
6-Chloro-2-iodo-9-benzyl-9H-purine (4e). Yield 55%; mp
(heptane) 136±137 ꢀC; H NMR (DMSO-d6), d 8.73 (s,
1
1H), 7.30 (m, 5H), 5.48 (s, 2H). Anal. (C12H8N4ClI)
C,H,N.
Preparation of compounds 5a±h. Substitution of the 6-
chloro in 4a±e was routinely achieved by reaction with
1.1 equiv of the appropriate amine (N-methylallylamine,
benzylamine, 3-iodobenzylamine or 1,2,3,4-tetrahydro-
pyridine, purchased from Aldrich) in EtOH containing
N(Et)3 (4 equiv) under nitrogen at 45 ꢀC for 3 h. The
solvent was then removed in vacuo. The residue was
dissolved in CH2Cl2 and washed twice with water. The
organic phase was dried (MgSO4), evaporated to dry-
ness and crystallized. Flash chromatography puri®cation
on silica gel column eluting with 5% ethanol in CH2Cl2
was necessary in the case of 5b, 5c and 5f (Table 4).
2-Amino-6-chloro-9-methyl-ꢀ9H-purine (3a). Yield from 1
8.07 (s, 1H), 6.90 (s, 2H), 3.63 (s, 3H). Anal. (C6H6
N5Cl) C,H,N.
1
78%; mp sublimate at 240 C; H NMR (DMSO-d6), d
2-Amino-6-chloro-9-isopropyl-9H-purine (3b). Yield from
1 94%; mp (H2O) 139 ꢀC; H NMR (DMSO-d6), d 8.24
1
(s, 1H), 6.88 (s, 2H), 4.61 (m, 1H), 1.49 (d, 6H, 2ÂCH3,
J=6.7 Hz). Anal. (C8H10N5Cl) C,H,N.
2-Amino-6-chloro-9-cyclopentyl-9H-purine (3c). Yield
from 1 91%; mp (H2O) 142 ꢀC; 1H NMR (DMSO-d6), d
8.21 (s, 1H), 6.88 (s, 2H), 4.70 (m, 1H), 2.5±1.6 (m, 8H).
Anal. (C10H12N5Cl) C,H,N.
Preparation of compounds 6±25. Starting materials were
purchased from Aldrich: ethanolamine, (S)-(+)-2-pyr-
rolidine methanol, (R)-( )-2-pyrrolidine methanol, N-
benzylethanolamine, 2-amino-1-phenylethanol, dl-2-
amino-1-hexanol, (S)-(+)-2-phenylglycinol, (R)-( )-2-
phenylglycinol, (S)-( )-2-amino-3-phenyl-1-propanol,
(R)-(+)-2-amino-3-phenyl-1-propanol, dl-2-amino-1-
2-Amino-6-chloro-9-cyclohexyl-9H-purine (3d). Yield
from 1 80%; mp 165 ꢀC; H NMR (DMSO-d6), d 8.23
1
(s, 1H), 6.89 (s, 2H), 4.22 (m, 1H), 1.93±130 (m, 10 H).
Anal. (C11H14N5Cl) C,H,N.
pentanol,
(S)-(+)-2-amino-1-propanol,
(R)-( )-2-
amino-1-propanol, 3-pyrrolidinol.
2-Amino-6-chloro-9-benzyl-9H-purine (3e). Yield from 1
In a typical procedure, a mixture of 2-iodo purine deri-
vative 5c (2 mmol) in N,N-dimethylacetamide (40 mL),
tripropylamine (1 mL) and 1.5 equiv of (R)-( )-2-pyr-
rolidine methanol was stirred at 140 ꢀC for 24 h. In the
case of 12±19 and 20, the mixture was placed in a sealed
tube ®lled with glass beads, and was heated at 140 ꢀC.
After evaporation to dryness (oil pump), the residue was
dissolved in CH2Cl2, washed with water (twice) adsor-
bed on silica gel and chromatographed on a silica gel
column eluting successively with CH2Cl2 and 2±5%
EtOH in CH2Cl2. In general crystalline products were
obtained (see Table 4). NMR data for compounds 5a±h
and 6±25 are presented in Table 5.
ꢀ
1
82%; mp 210 C; H NMR (DMSO-d6), d 8.23 (s, 1H),
7.3 (m, 5H), 6.94 (s, 2H), 5.29 (s, 2H). Anal. (C12H10
N5Cl) C,H,N.
2-Amino-6-chloro-9-cyclopent-3-ene-1-yl-9H-purine (3f).
Yield from 1 82%; mp (ethanol) 149±150 ꢀC; H NMR
1
(DMSO-d6), d 8.04 (s, 1H), 6.89 (s, 2H), 5.87 (s, 2H)
5.09 (m, 1H), 2.95±2.62 (m, 4H). Anal. (C10H10N5Cl)
C,H,N.
Preparation of compounds 4a±e. Preparation of com-
pounds 4a±e from 3a±e was achieved as described by
Matsuda et al.26
Biological studies
6-Chloro-2-iodo-9-methyl-9H-purine (4a). Yield 42%;
mp (ethanol) 194±196 ꢀC; H NMR (DMSO-d6), d 8.04
1
Kinase assays. Unless otherwise speci®ed, kinase activ-
ities were assayed in bu€er C (see below), at 30 ꢀC, in
the presence of 15 mM ATP. Blank values were sub-
tracted from the data and activities were calculated as
pmoles of phosphate incorporated in protein acceptor
after 10 min incubation. Controls were performed with
appropriate dilutions of DMSO. In a few cases phos-
phorylation of the substrate was assessed by auto-
radiography after polyacrylamide gel electrophoresis in
the presence of lauryl sulfate.
(s, 1H), 6.89 (s, 2H), 5.87 (s, 2H) 5.09 (m, 1H), 2.95±2.62
(m, 4H). Anal. (C6H4N4ClI) C,H,N.
6-Chloro-2-iodo-9-isopropyl-9H-purine (4b). Yield 82%;
mp (heptane) 108±109 ꢀC; H NMR (DMSO-d6), d 8.76
1
(s, 1H), 4.83 (m, 1H), 1.55 (d, 6H, J=6.7 Hz). Anal.
(C8H8N4ClI) C,H,N.
6-Chloro-2-iodo-9-cyclopentyl-9H-purine (4c). Yield 56%;
mp (cyclohexane) 106±107 ꢀC; H NMR (DMSO-d6), d
1
8.73 (s, 1H), 4.92 (m, 1H), 2.52±1.71 (m, 8H). Anal.
(C10H10N4ClI) C,H,N.
CDK1/cyclin B was extracted in homogenization bu€er
from M phase star®sh (Marthasterias glacialis) oocytes
and puri®ed by anity chromatography on p9CKShs1
-
6-Chloro-2-iodo-9-cyclohexyl-9H-purine (4d). Yield 55%;
mp (cyclohexane) 133±134 ꢀC; H NMR (DMSO-d6), d
sepharose beads, from which it was eluted by free
p9CKShs1 as previously described.10,48 The kinase activity
1

1290
M. Legraverend et al. / Bioorg. Med. Chem. 7 (1999) 1281±1293
Table 4. Physico-chemical data for selected compounds
Compound
Mp (ꢀC)
% Yield
Formula
Analysis
Recrystallization solvent
5a
5b
5c
5d
5e
5f
5g
5h
6
122±123
Oil
182
213±214
210±212
184±185
176±178
94±98
Oil
Oil
Oil
116±117
Oil
124±126
172±174
149±151
132±134
130±131
148±150
150±151
172±174
112±114
112±113
152±154
167±168
151±152
147±149
129±131
98
86
70
90
88
77
93
98
57
91
69
C10H12N5I
12H16N5I
C15H16N5I
C17H18N5I
C18H20N5I
C19H16N5I
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N,I
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N,I
C,H,N
C,H,N
C,H,N
C,H,N
Cyclohexane
Ð
C
Cyclohexane
Dichloromethane
Dichloromethane
Dichloromethane
EtOH
EtOH/H2O
Ð
Ð
Ð
Diethyl ether
Ð
Cyclohexane
Dichloromethane
Dichloromethane
EtOH/H2O
EtOH/H2O
EtOH/H2O
Cyclohexane
Dichloromethane
Cyclohexane
Cyclohexane
EtOH
C
15H15N5I2
C13H16N5I
C
12H18N6O
C14H22N6O, 1/2H2O
C15H22N6O
7
8
9
C
20H26N6O
C20H26N6O
24H28N6O
C23H26N6O
21H30N6O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
62
52
85
58
47
90
73
60
55
43
47
84
37
50
87
89
C
C
C23H26N6O, 1/2H2O
C23H26N6O, 1/2H2O
C
24H28N6O
C24H28N6O
20H28N6O
C18H24N6O
18H24N6O
C20H25N6OI
19H24N6O
C
C
C
Cyclohexane
Diethyl ether
Diethyl ether
Diethyl ether
C23H30N6O
C22H28N6O
C
24H26N6O
a Yields refer to isolated and puri®ed materials.
b All compounds were analyzed within ‹0.4% of the theoretical values.
was assayed with 1 mg/mL histone H1 (Sigma type III-
S), in the presence of 15 mM [g-32P]ATP (3,000 Ci/mmol;
1 mCi/mL) in a ®nal volume of 30 mL. After 10 min
incubation at 30 ꢀC, 25 mL aliquots of supernatant were
spotted onto 2.5Â3 cm pieces of Whatman P81 phos-
phocellulose paper, and, after 20 s, the ®lters were
washed ®ve times (for at least 5 min each time) in a
solution of 10 mL phosphoric acid/liter of water. The
wet ®lters were transferred into 6 mL plastic scintillation
vials, 5 mL ACS (Amersham) scintillation ¯uid was
added and the radioactivity measured in a Packard
counter. The kinase activity was expressed in pmoles
phosphate incorporated in histone H1 per 10 min incu-
bation, or in % of maximal activity. CDK2/cyclin E
(provided by Dr. W. Harper) and His-tagged erk1
(provided by Dr. M. Cobb) were assayed as described.10
obtained from New England Nuclear Life Science Pro-
ducts. Rat monoclonal antibody directed against
BrdUrd was purchased from CeraLab. Fluorescein iso-
thiocyanate-conjugated goat anti-rat IgG (heavy chain
speci®c) was purchased from Southern Biotechnology
Associates.
Cell cultures. Human HeLa (cervix carcinoma) cells
were kindly provided by Dr. J. Coppey (Institut Curie).
Cells were checked to be free of mycoplasma con-
tamination, and routinely subcultured in 25 cm2 ¯asks
(Becton±Dickinson) every 4 days at a density of
1±2Â105 cells/¯ask in Dulbecco's modi®ed Eagle's
minimal essential medium with 4.5 g/L glucose, 0.1 g/L
pyruvate, 105 UI/L penicilline, 0.1 g/L streptomycine,
0.86 g/L Glutamax I and 10% v/v fetal calf serum
(37 ꢀC, 7% CO2). The number of passages from the
same primary subculture was kept below 7±8. The mean
doubling time (mid-log phase) was 21‹1 h. Control
plating eciencies were in the range of 70%.
Homogenization bu€er was made of 60 mM b-glycero-
phosphate, 15 mM p-nitrophenyl-phosphate, 25 mM 3-
(N-morpholino)propanesulfonic acid (pH 7.2), 15 mM
EGTA, 15 mM MgCl2, 1 mM dithiothreitol, 1 mM
sodium vanadate, 1 mM NaF, 1 mM phenylphosphate,
10 mg/mL leupeptin, 10 mg/mL aprotinin, 10 mg/mL soy-
bean trypsin inhibitor and 100 mM benzamidine. Bu€er
C was identical to the homogenization bu€er but did
not contain EGTA, NaF and protease inhibitors.
Synchronization of HeLa cell cultures at the G1-S
junction was performed using a double thymidine block
technique as described.49 The method does not a€ect the
viability of cells but requires correction of data for cel-
lular multiplicity in cytotoxicity assays. These correc-
tions were performed in the same way as reported.50,51
Reagents for biological investigations. Thymidine,
BrdUrd and propidium iodide came from Sigma.
Distilled DMSO (spectroscopy grade) was purchased
from Merck. All products for cell culture were from Life
Technologies/Gibco-BRL. [Methyl-3H]thymidine was
Cytotoxicity assays. HeLa cells from exponentially
growing subcultures were plated in triplicate in 25 cm2
¯asks (Becton±Dickinson) at a density of 600±1200 cells
per ¯ask, and incubated at 37 ꢀC for 4 h prior to drug

M. Legraverend et al. / Bioorg. Med. Chem. 7 (1999) 1281±1293
1291

1292
M. Legraverend et al. / Bioorg. Med. Chem. 7 (1999) 1281±1293
exposure. Following treatment, the ¯asks were rinsed
and cells returned to normal growth medium for 10±12
days. Colonies were ®xed with methanol, stained and
scored visually.
where D is the radiation dose, and a and b numerical
parameters characterizing the radiosensitivity of the cell
line under consideration.
Results have been expressed as mean‹standard devia-
tion. Least-squares regression analyses were performed
using a self-convergent Gauss±Levenberg±Marquardt
algorithm (Kaleidagraph 3.0.8, Abelbeck Software
Inc.).
All drugs were prepared as 50 mM sterile solutions in
pure DMSO and stored at 20 ꢀC for no more than 4
days. Drugs were adjusted extemporarily at the appro-
priate concentrations by successive dilutions in pure
DMSO and in growth medium through vortexing. The
®nal concentration of DMSO was low enough (ꢁ0.5%)
as not to alter either cell growth, or drug or radiation
response. Exposure to drugs was carried out in dim light
to prevent photodegradation of the compounds. Drug
was carefully removed through two washes with Hank's
balanced saline (37 ꢀC), after which cells were returned
to normal growth medium.
Acknowledgements
Thanks are due to Dr. Jacques Coppey (Institut Curie)
for providing HeLa cells. We are also indebted to Drs.
Wade Harper (Baylor College of Medicine, Houston)
and Melanie Cobb (University of Texas, Dallas) for the
generous gift of kinases, to Dany Rouillard (Institut
Curie) for ¯ow cytometric analyses, and to Dr. David
Grierson (Institut Curie) for rereading the manuscript.
We thank the ®shermen of the Station Biologique de
Rosco€ for collecting the star®sh. This research was
supported by ®nancial aid from the Institut National de
la Sante et de la Recherche Medicale (V.F.), by the
Ministere de l'Education Nationale de la Recherche et
de la Technologie (ACC-SV8 program), and by grants
from the Association pour la Recherche sur le Cancer
(ARC-9314) and the Conseil Regional de Bretagne
(L.M.). S.L. was supported by a fellowship from the
Conseil Regional de Bretagne; R.S. was the recipient of
a post-doctoral fellowship from the Institut National de
la Sante et de la Recherche Medicale.
g-Ray irradiation of cells with or without concomitant
drug exposure was performed at room temperature (21±
24 ꢀC) in medium saturated with 93% air±7% CO2,
using an IBL-637 (137Cs) irradiator (CIS-BioInterna-
tional). The dose rate was 0.9 Gy/min.
Thymidine uptake. For studies of altered thymidine
intake by drugs, 6Â105 HeLa cells were seeded in 10 cm2
Petri dishes and returned to the incubator for 16 h. All
dishes at that time contained 1.00‹0.05Â106 cells.
Dishes were washed with warm Hank's bu€er, then
fed with exactly 2 mL of complete medium and incu-
bated for 90 min with various amounts of the drugs of
interest. 2 mCi of [methyl-3H]thymidine were then intro-
duced and dishes allowed to stand in the incubator for a
further 30 min. Dishes were then carefully washed twice
with Hank's bu€er, and cells were lysed with 1 mL lysis
bu€er (20 mM Tris-HCl, 10 mM EDTA, 0.5% lauryl
sulfate, pH 7.8). The lysates were recovered, transferred
to scintillation vials and the radioactivity counted by
scintillation.
References
1. Chellappan, S. P.; Giordano, A.; Fisher, P. B. Curr. Top.
Microbiol. Immunol. 1998, 227, 57.
2. Bartkova, J.; Lukas, J.; Bartek, J. In Progress in Cell Cycle
Research; Meijer, L.; Guidet, S.; Philippe, M., Eds.; Plenum
Press: New York, 1997; pp 211±220.
3. Kitagawa, M.; Okabe, T.; Ogino, H.; Matsumoto, H.;
Suzuki-Takahashi, I.; Kokubo, T.; Higashi, H.; Saitoh, S.;
Taya, Y.; Yasuda, H.; Ohba, Y.; Nishimura, S.; Tanaka, N.;
Okuyama, A. Oncogene 1993, 8, 2425.
4. Kitagawa, M.; Higashi, H.; Suzuki-Takahashi, I.; Okabe,
T.; Ogino, H.; Taya, Y.; Nishimura, S.; Okuyama, A. Onco-
gene 1994, 9, 2549.
5. Carlson, B. A.; Dubay, M. M.; Sausville, E. A.; Brizuela,
L.; Worland, P. J. Cancer Res. 1996, 56, 2973.
6. De Azevedo, W. F.; Mueller-Dieckmann, H.-J.; Schulze-
Gahmen, U.; Worland, P. J.; Sausville, E.; Kim, S.-H. Proc.
Natl. Acad. Sci. USA 1996, 93, 2735.
Cyto¯uorimetric analysis. Cell cycle progression in sub-
cultures and in treated cells was monitored by dual
parameter ¯ow cytometry using a FACStarPLUS cyto-
¯uorometer (Becton±Dickinson). Cells were incubated
with BrdUrd (10 mM, 15 min) for pulse-labeling of S-
phase cells, then harvested by trypsinization, pelleted,
washed once with cold phosphate-bu€ered saline and
®xed in 70% ice-cold EtOH. Treatment of ®xed cells for
cyto¯uorimetric analysis and bivariate data acquisition
and processing were done as described.52
7. Meijer, L.; Kim, S.-H. Methods Enzymol. 1997, 283, 113.
8. Vesely, J.; Havlicek, L.; Strnad, M.; Blow, J. J.; Donella-
Deana, A.; Pinna, L.; Letham, D. S.; Kato, J.-Y.; Detivaud,
L.; Leclerc, S.; Meijer, L. Eur. J. Biochem. 1994, 224, 771.
9. De Azevedo, W. F.; Leclerc, S.; Meijer, L.; Havlicek, L.;
Strnad, M.; Kim, S.-H. Eur. J. Biochem. 1997, 243, 518.
10. Meijer, L.; Borgne, A.; Mulner, O.; Chong, J. P. J.; Blow,
J. J.; Inagaki, N.; Inagaki, M.; Delcros, J.-G.; Moulinoux, J.-P.
Eur. J. Biochem. 1997, 243, 527.
11. Havlicek, L.; Hanus, J.; Vesely, J.; Leclerc, S.; Meijer, L.;
Shaw, G.; Strnad, M. J. Med. Chem. 1997, 40, 408.
12. Shulze-Gahmen, U.; Brandsen, J.; Jones, H. D.; Morgan,
D. O.; Meijer, L.; Vesely, J.; Kim, S.-H. Proteins 1995, 22, 378.
13. Hajduch, M.; Kolar, Z.; Novotny, R.; Hanus, J.; Mihal,
Data handling. Drug survival ®tted a single-exponential
model,
S
a‰DrugŠ
ˆ exp
ꢀ1†
S0
where s=s0 is the surviving fraction relative to the plat-
ing eciency. On the other hand, radiation survival is
classically described using a linear-quadratic equation,53
S
ln
ˆ
aD bD2
ꢀ2†
S0

M. Legraverend et al. / Bioorg. Med. Chem. 7 (1999) 1281±1293
1293
V.; Hlobilkova, A.; Noskova, V.; Strnad, M. Anti-Cancer
Drugs 1997, 8, 1007.
14. Mgbonyebi, O. P.; Russo, J.; Russo, I. H. Anticancer Res.
1998, 18, 751.
15. Walker, D. H.; Parker, P. A.; Laethem, R. M. Proc. Amer.
Assoc. Cancer Res. 1998, 39, 1678.
16. Maas, J. W.; Horstmann, S.; Borasio, G. D.; Anneser,
J. M. H.; Shooter, E. M.; Kahle, P. J. J. Neurochem. 1998, 70,
1401.
A.; Belle, R.; Berger, F.; Bergounioux, C.; Brunn, G.; Buquet-
Fagot, C.; Fagot, D.; Glab, N.; Goudeau, H.; Goudeau, M.;
Guerrier, P.; Houghton, P. J.; Hendriks, H.; Kloareg, B.;
Lippai, M.; Marie, D.; Maro, B.; Meijer, L.; Mester, J.; Mul-
ner-Lorillon, O.; Poulet, S. A.; Schierenberg, E.; Schutte, B.;
Vaulot, D.; Verlhac, M. H. Biol. Cell 1995, 83, 105.
34. Schutte, B.; Nieland, L.; van Engeland, M.; Hen¯ing, M.
E. R.; Meijer, L.; Ramaekers, F. C. S. Exp. Cell Res. 1997,
236, 4.
17. Brooks, E. E.; Gray, N. S.; Joly, A.; Kerwar, S. S.; Lum,
R.; Mackman, R. L.; Norman, T. C.; Rosete, J.; Rowe, M.;
Schow, S. R.; Schultz, P. G.; Wang, X.; Wick, M. M.; Schi€-
man, D. J. Biol. Chem. 1997, 272, 29207.
35. Brusselbach, S.; Nettelbeck, D. M.; Sedlacek, H.-H.;
Muller, R. Int. J. Cancer 1998, 77, 146.
36. Howell, B.; Odde, D. J.; Cassimeris, L. Cell Motil. Cytos-
keleton 1997, 38, 201.
18. Pippin, J. W.; Qu, Q.; Meijer, L.; Shankland, S. J. J. Clin.
Invest. 1997, 100, 2512.
37. Gong, J.; Traganos, F.; Darzynkiewicz, Z. Cell Growth
Di€er. 1995, 6, 1485.
19. Bresnahan, W. A.; Boldogh, I.; Chi, P.; Thompson, E. A.;
Albrecht, T. Virology 1997, 231, 239.
38. Jackson, P. K.; Chevalier, S.; Philippe, M.; Kirschner,
M. W. J. Cell Biol. 1995, 130, 755.
20. Yakisich, J. S.; Siden, A.; Vargas, V. I.; Eneroth, P.; Cruz,
M. Biochem. Biophys. Res. Commun. 1998, 243, 674.
21. Nugiel, D. A.; Cornelius, L. A. M.; Corbett, J. W. J. Org.
Chem. 1997, 62, 201.
22. Schow, S. R.; Mackman, R. L.; Blum, C. L.; Brooks, E.;
Horsma, A. G.; Joly, A.; Kerwar, S. S.; Lee, G.; Shi€man, D.;
Nelson, M. G.; Wang, X.; Wick, M. M.; Zhang, X.; Lum, R. T.
Bioorg. Med. Chem. Lett. 1997, 7, 2697.
23. Legraverend, M.; Boumchita, H.; Bisagni, E. Synthesis
1990, 587.
24. Gray, N. S.; Wodicka, L.; Thunissen, A. M. W. H.; Nor-
man, T. C.; Kwon, S. J.; Espinoza, F. H.; Morgan, D. O.;
Barnes, G.; LeClerc, S.; Meijer, L.; Kim, S. H.; Lockhart, D.
J.; Schultz, P. G. Science 1998, 281, 533.
25. Nair, V.; Richardson, S. G. Synthesis 1982, 670.
26. Matsuda, A.; Shinozaki, M.; Yamaguchi, T.; Homma, H.;
Nomoto, R.; Miyasaka, T.; Watanabe, Y.; Abiru, T. J. Med.
Chem. 1992, 35, 241.
27. Imahori, K.; Uchida, T. J. Biochem. 1997, 121, 179.
28. Qi, Z.; Tang, D.; Zhu, X.; Fujita, D. J.; Wang, J. H. J.
Biol. Chem. 1998, 273, 2329.
39. Cardoso, M. C.; Leonhardt, H.; Nadal-Ginard, B. Cell
1993, 74, 979.
40. Sobczak-Thepot, J.; Harper, F.; Florentin, Y.; Zindy, F.;
Brechot, C.; Puvian, E. Exp. Cell Res. 1993, 206, 43.
41. Fotedar, A.; Cannella, D.; Fitzgerald, P.; Rousselle, T.;
Gupta, S.; Doree, M.; Fotedar, R. J. Biol. Chem. 1996, 271,
31627.
42. Terasima, T.; Tolmach, L. J. Biophys. J. 1963, 3, 11.
43. Sinclair, W. K.; Morton, R. A. Radiat. Res. 1966, 29, 450.
44. Wilson, W. R.; Whitmore, G. F. Radiat. Res. 1981, 87,
121.
45. Skwarchuk, M. W.; Wouters, B. G.; Skarsgard, L. D. Int.
J. Radiat. Biol. 1993, 64, 601.
46. Murdock, K. C.; Angier, R. B. J. Org. Chem. 1962, 27,
2395.
47. Boumchita, H.; Legraverend, M.; Bisagni, E. Heterocycles
1991, 32, 1785.
48. Azzi, L.; Meijer, L.; Ostvold, A. C.; Lew, J.; Wang, J. H.
J. Biol. Chem. 1994, 269, 13279.
49. Tsao, Y.-P.; D'Arpa, P.; Liu, L. F. Cancer Res. 1992, 52,
1823.
29. Buquet-Fagot, C.; Lallemand, F.; Montagne, M.-N.;
Mester, J. Anti-Cancer Drugs 1997, 8, 623.
50. Hennequin, C.; Giocanti, N.; Favaudon, V. Cancer Res.
1996, 56, 1842.
30. Norman, T. C.; Gray, N. S.; Koh, J. T.; Schultz, P. G. J.
Am. Chem. Soc. 1996, 118, 7430.
31. Glab, N.; Labidi, B.; Qin, L.-X.; Trehin, C.; Bergounioux,
C.; Meijer, L. FEBS Lett. 1994, 353, 207.
32. Planchais, S.; Glab, N.; Trehin, C.; Perennes, C.; Bureau,
J.-M.; Meijer, L.; Bergounioux, C. Plant J. 1997, 12, 191.
33. Abraham, R. T.; Acquarone, M.; Andersen, A.; Asensi,
51. Albright, N. Radiat. Res. 1987, 112, 331.
52. Demarcq, C.; Bastian, G.; Remvikos, Y. Cytometry 1992,
13, 416.
53. Hall, E. J. Radiobiology for the Radiologist, 4th ed.; J. B.
Lippincott Co.: Philadelphia, 1994; Chapter 3.
54. Szechter, A.; Kowalsky, W. P.; Schwartz, G. Radiat. Res.
1978, 75, 432.