New purines and purine analogs as modulators of multidrug resistance

Article abstract of DOI:10.1021/jm960361i

A series of 36 purine and purine analog derivatives have been synthesized and tested for their ability to modulate multidrug resistance in vitro (P388/VCR-20 and KB-A1 cells) and in vive (P388/VCR leukemia). Compounds were compared to S9788, a triazine derivative which has already shown some activity during phase i clinical trials and also a limiting cardiovascular side effect possibly linked to its calcium channel affinity. The fact that active compounds increase adriamycin accumulation in the resistant KB-A1 cells, and not in the sensitive KB-3-1 cells, suggests they act predominantly by inhibiting the P-glycoprotein-catalyzed efflux of cytotoxic agents. No direct relation was found between the affinity for the phenylalkylamine binding site of the calcium channel and in vitro sensitization of resistant cells. In vivo, when administered po in association with vincristine (0.25 mg/kg), five compounds (3, 4, 9, 25, and 26), of very differing calcium channel affinities (K(i) from 5 to 560 nM), fully restored (T/V ≤ 1.4) the sensitivity of P388/VCR leukemia to vincristine.

Full text of DOI:10.1021/jm960361i

J . Med. Chem. 1996, 39, 4099-4108
4099
New P u r in es a n d P u r in e An a logs a s Mod u la tor s of Mu ltid r u g Resista n ce
Alain Dhainaut,* Gilbert Regnier, Andre Tizot, Alain Pierre, Stephane Leonce, Nicolas Guilbaud,
Laurence Kraus-Berthier, and Ghanem Atassi
Institut de Recherches Servier, 11 rue des Moulineaux, 92150 Suresnes, France
Received May 17, 1996X
A series of 36 purine and purine analog derivatives have been synthesized and tested for their
ability to modulate multidrug resistance in vitro (P388/VCR-20 and KB-A1 cells) and in vivo
(P388/VCR leukemia). Compounds were compared to S9788, a triazine derivative which has
already shown some activity during phase 1 clinical trials and also a limiting cardiovascular
side effect possibly linked to its calcium channel affinity. The fact that active compounds
increase adriamycin accumulation in the resistant KB-A1 cells, and not in the sensitive KB-
3-1 cells, suggests they act predominantly by inhibiting the P-glycoprotein-catalyzed efflux of
cytotoxic agents. No direct relation was found between the affinity for the phenylalkylamine
binding site of the calcium channel and in vitro sensitization of resistant cells. In vivo, when
administered po in association with vincristine (0.25 mg/kg), five compounds (3, 4, 9, 25, and
26), of very differing calcium channel affinities (Ki from 5 to 560 nM), fully restored (T/V g
1.4) the sensitivity of P388/VCR leukemia to vincristine.
Ch a r t 1
The resistance of tumor cells to cytotoxic drugs is a
major problem in cancer chemotherapy. The implicated
mechanisms are intrinsic or acquired. Among them,
multidrug resistance (MDR), associated with the mem-
brane P-glycoprotein overexpression, has been defined
as a cross-resistance to anticancer drugs especially
affecting vinca-alcaloides, anthracyclines, podophyllo-
toxins, actinomycin D, taxol, and colchicine.
By systematic screening, we found that almitrine (1),
a triazinylpiperazine currently used for respiratory
insufficiency, moderately sensitized the highly resistant
cell line DC-3F/AD to actinomycin D. This starting
point led the synthesis of a series of analogs in order to
select more potent modulator agents1 and provided
S9788 (2), which is currently in Phase 1 clinical trials.
Ch a r t 2
The preliminary clinical trials demonstrated the
interest of S9788 in modulating resistance. However,
the observed limiting toxicity is that of cardiovascular
side effects including prolonged QTc intervals.2,3 Con-
sequently, active plasma levels could not be achieved
in the patients. Moreover, S9788 was shown to have
calcium channel affinity comparable to that of verapamil
and markedly higher than that of almitrine. While the
clinical syndrome of long QT and Torsades de Pointes
is not generally associated with calcium channel block-
ers therapy,4,5 marked proarythmic effects were ob-
served with some calcium antagonists, such as bepridil,
that prolonged ventricular repolarization and QT in-
tervals.6,7
These results prompted us to search for second-
generation sensitizers, more efficient and more potent
than S9788 in vitro and in particular more active in vivo
by oral administration. These modulator agents also
would have to present a lower affinity for the phenyl-
alkylamine binding site of the calcium channel, to
attempt to reduce cardiovascular side effects observed
with S9788.
in vivo, and their calcium channel affinity. These new
compounds have the general formula shown in Chart
3. The formula partially respects structure-activity
relationships established from the first series of com-
pounds, including S9788.1 In the central part B, Z can
be CHNH (4-aminopiperidine) or N (piperazine), m ) 0
or 1. In part C, the two phenyl rings can eventually be
linked in a bridge including carbons and/or one or
several heteroatoms. T and V, alternatively or together,
represent H or halogen atoms or methoxy radicals. The
major modification concerns part A in which the S9788
triazine core is replaced by a purine core or an analog
in which the triaminotriazine arrangement remains
unchanged. The substituents on positions 2 and 6 can
be exchanged. R and R′ represent alternatively or
This paper describes the chemical properties of new
compounds, their ability to modulate MDR in vitro and
* To whom correspondence should be addressed.
X Abstract published in Advance ACS Abstracts, September 1, 1996.
S0022-2623(96)00361-5 CCC: $12.00 © 1996 American Chemical Society

4100 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20
Dhainaut et al.
Ta ble 1. Target Compounds: (Alkylamino)purines and
Ch a r t 3
Analogs
together an alkyl or alkenyl group and X and Y a carbon
or nitrogen atom.
Ch em istr y
The target compounds listed in Tables 1-3 have been
prepared according to the Schemes 1-3. Table 4 and
Schemes 4 and 5 are related to the starting materials.
Alkylation of 2,6-dichloropurine by means of Mitsuno-
bu’s method8 afforded a mixture of 7- and 9-isomers
which were separated and purified by flash chromatog-
raphy (Scheme 4). To avoid reproduction of this isomer
mixture in the case of triazolopyrimidines, the alkyla-
tion was performed on the precursor pyrimidine before
cyclization (Scheme 5). Dichloropyrrolopyrimidine9 and
dichloropyrazolopyrimidine,10 which have been already
described, were alkylated by the same Mitsunobu
method (Scheme 4).
Resu lts a n d Discu ssion
Str u ctu r e-Activity Rela tion sh ip s (SAR). For the
discussion of SAR, the general formula has been divided
into three parts (see Chart 3). To obtain a potent and
optimal MDR modulation on P388/VCR-20 and/or KB-
A1 cells lines at nontoxic concentrations in this series
(see Table 5), the following criteria have emerged.
(1) Part A should preferably be a pyrrolopyrimidine
core (XY ) CC, 18) rather than a purine (XY ) NC, 3),
pyrazolopyrimidine (XY ) CN, 13), or triazolopyrimi-
dine core (XY ) NN, 27), other components being
constant. R and R′ groups (see Chart 3) can be equally
allyl and/or propyl ones, the reversing activity on both
cells lines being of the same order (compare 3, 24, 25,
and 26). Permutation of substituents on positions 2 and
6 (see Chart 3) do not greatly modify compound activi-
ties (compare 4 vs 29, 6 vs 30, 7 vs 31, 8 vs 32, 9 vs 33,
and 10 vs 34, respectively).
(2) Part B should be a 4-(methylamino)piperidine
group (m ) 1) rather than a 4-aminopiperidine (m ) 0)
(compare 8 vs 16 and 3 vs 19), and when Part B is a
piperazine, the compounds are almost inactive (see 36-
38).
(3) Concerning part C, steric factors do not seem to
be determinant as activities of the same order are
obtained with a tricyclic (10), a bicyclic (7, 21), a
a
All compounds except 9, 10, 33, and 34 were purified by flash
chromatography. C, H, N, analyses were within 0.4% of theoretical
values for the formulae given, except for the following products.
8: N calcd, 15.37; found, 14.95. 11: Cl calcd, 15.69; found 15.28.
14: H calcd, 6.58; found, 7.05. 21: H calcd, 7.03; found, 6.58. All
compounds exhibited NMR consistent with assigned structures.
Synthesis described in the Experimental Section. c For starting
b
materials, see the Experimental Section. Fumaric acid. e Maleic
d
acid. f [R]D ) +34.6° (c ) 0.5%, EtOH, t ) 21 °C). [R]D ) -34.8°
g
(c ) 0.5%, EtOH, t ) 21 °C).
(1) Part A: Compound affinities depend on the nature
of the heteroaromatic core, other components being
constant. The order of decreasing affinities is 3 (XY )
NC) > 18 (XY ) CC) > 13 (XY ) CN) > 27 (XY ) NN).
When R and R′ (see Chart 3) are the same group, the
highest affinity is observed for the allyl group, and for
the propyl group, the affinity is actually divided by
about 12 (compare 26 vs 3). Except for 33 vs 9,
permutation of the residues at positions 2 and 6 (see
benzhydryl (5), or a triphenylmethyl group (20).
A
tricyclic group with a heteroatomic bridge affords slightly
better compounds than a carbon bridge, at least in the
case of KB-A1 cells (compare 8 and 17 vs 3 and 10).
For the discussion concerning calcium channel affin-
ity, the general formula has again been divided into the
same three parts. Compounds 16, 19, 29, a n d 35-38,
which are less active than S9788, will not be discussed.

Purines as Modulators of MDR
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20 4101
Ta ble 2. Target Compounds: 6-Piperidinopurines
Since the cytotoxicity of these products can be different,
we chose, for the purpose of comparison, the two
relatively low concentrations of 1 µM for P388/VCR-20
cells and 2.5 µM for KB-A1 cells. Table 5 gives the fold
reversal values obtained with these concentrations.
Some compounds were nearly inactive (36-38), and
others were significantly more active than S9788 (9, 28
on P388/VCR-20 cells; 8, 17, 18, 20, 32 on KB-A1 cells),
the other ones being approximately as active as S9788.
The effect of some selected compounds on ADR
accumulation by KB-A1 cells was studied by flow
cytometry. Under these experimental conditions (5 h
of incubation), all the compounds were devoid of cyto-
toxicity at 5 µM. Figure 1b shows that 2 and 9
increased ADR accumulation more efficiently than VRP
and CsA. This result suggests that these compounds
act mainly by inhibiting the Pgp-mediated efflux of
cytotoxic drugs, as already discussed for S9788 and
derivatives.13 This is in agreement with the inability
of the compounds to increase the sensitivity of the P388
and KB-3-1 cell lines (not shown).
The compounds which were at least as active as S9788
in one cell line in vitro were tested in vivo in association
with VCR on the P388/VCR murine leukemia. This
leukemia is resistant to VCR, the T/C obtained with 0.25
mg/kg VCR ip (QDX4.1) being on average 141% versus
208% on the sensitive P388. Three doses of tested
compounds (25, 50, and 100 mg/kg po) were routinely
used in association with 0.25 mg/kg VCR ip. We chose
the oral route of administration in order to maintain,
as far as possible, plasma levels long enough to fully
restore the sensitivity of resistant cells.12 When tested
alone, the compounds were devoid of both toxicity and
antitumor activity at 100 mg/kg po (T/C ) 91-116%).
Table 5 gives the increase in VCR antitumor activity
induced by the compounds, expressed as T/V (defined
in the Experimental Section). Some compounds signifi-
cantly increased the survival time of mice when admin-
istered in association with VCR with respect to mice
treated by VCR alone, the more active ones being 3, 4,
9, 25, and 26. For these derivatives, the sensitization
was complete, since the T/C obtained in P388/VCR-
bearing mice with the association was similar to that
obtained in mice bearing the sensitive P388 leukemia
with VCR alone.
a,d,eSee the corresponding footnotes for Table 1.
Chart 3) does not dramatically change the affinities
(compare 28 vs 3, 30 vs 6, 31 vs 7, 32 vs 8, and 34 vs
10, respectively).
(2) Part B: The shorter spacers (m ) 0) afford
compounds with lower affinities for the calcium channel
(compare 8 vs 16, 3 vs 19, and 6 vs 11 and 12).
(3) Part C: Modulation of this aromatic part provided
the greatest variations in affinities (from 5 nM for 4 to
3700 nM for 17). When part C is a benzhydryl group,
presence of halogens increased affinity (compare 4 vs
5). Benzo (21) or dibenzosuberane groups (3) are less
favorable than a naphthyl (7) group. In the case of a
dibenzosuberane-like structure, the nature of the bridge
becomes a very sensitive parameter. The affinity
decreased as a function of the nature of this bridge: 3
> 8 > 10 . 17. In the case of 11 and 12, which are two
enantiomers, the (-)-form (12) is slightly less active
than the (+)-form (11).
In order to study the relationship between calcium
channel affinity and reversal of MDR, all the compounds
were tested for the inhibition of [3H]D888 binding to
L-type channel of rat skeletal muscle cell. Table 5 lists
the values of Ki obtained in this assay. A large range
of affinities were measured, from 5 nM (4) to 3700 nM
(17). The lack of relationship between calcium channel
affinity and in vitro reversal of resistance is illustrated
in Figure 2.
Biologica l Resu lts. The study of the MDR-revers-
ing properties of these new derivatives was first carried
out in vitro on P388/VCR-20 cells, a murine leukemia
cell line whose resistance was induced by vincristine,
and KB-A1 cells, a human epidermoid carcinoma cell
line whose resistance was induced by adriamycin. The
compounds were tested at four concentrations (0.5-5
µM) in association with VCR (P388/VCR-20 cells) or
adriamycin (KB-A1 cells). The cytotoxicity due to the
modulator alone was systematically measured under the
same experimental conditions (four doubling times).
Figure 1a shows a typical dose-response curve obtained
in KB-A1 cells with VRP and CsA used as reference
compounds and compounds 2 and 9 taken as examples.
As already shown,11,12 VRP was moderately active at
these relatively low concentrations, and CsA was mark-
edly active only at 5 µM, a toxic concentration (57% of
cell viability).
The mechanism, at the molecular level, of the inhibi-
tion of Pgp function remains unknown. However, the
modulators are likely to interact directly with Pgp, as
shown, for example, by the covalent binding of photo-
affinity substrates.14 The verapamil analog LU-49888
was shown to bind both to Pgp15 and the R1-subunit of
L-type calcium channel,16 suggesting the existence of
homologous sites within these two proteins. We can
thus hypothesize similar interactions for our com-
pounds, some being more specific for the calcium chan-
nel and others for Pgp, which could explain the lack of
This approach allowed the determination of the most
active, noncytotoxic concentration for each modulator.

4102 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20
Ta ble 3. Target Compounds: Piperazinopurines
Dhainaut et al.
a,bSee the corresponding footnotes for Table 1.
Sch em e 1. Compounds 3-27 (Table 1)a
Sch em e 2. Compounds 28-35 (Table 2)a
a
Reagents and reaction conditions: (a) 2 equiv of R1NH2 in
EtOH at room temperature and for 30 min at 50 °C; (b) 10 equiv
of 4,4-diethoxypiperidine for 4 h at 100 °C; (c) EtOH, HCl at 50
°C for 2 h; (d) 2 equiv of NaBH(OAc)3 in CH2Cl2 at room
temperature for 5 h. X, Y, R, R1, R2, and m have the same meaning
as mentioned in Table 1.
a
Reagents and reaction conditions: (a) 2 equiv of 4,4-dieth-
oxypiperidine in EtOH for 1 h at room temperature; (b) 10 equiv
of allylamine in EtOH, autoclave for 40 h at 100 °C; (c) EtOH,
HCl at 50 °C for 2 h; (d) 2 equiv of NaBH(OAc)3 in CH2Cl2 at room
relationship between affinity for the calcium channel
and reversal of MDR.
temperature for 5 h. R and
mentioned in Table 2.
m have the same meaning as
Con clu sion s
The present work deals with the synthesis and
evaluation, in vitro and in vivo, of a series of agents
modulating multidrug resistance derived from S9788.
During the phase 1 clinical study of S9788, electrocar-
diographic followup showed a dose-limiting side effect.2,3
Because of the affinity of S9788 for the calcium channel,
we systematically measured the ability of the com-
pounds to inhibit [3H]D888 binding.
the cardiovascular side effects of S9788 and these new
analogs are now in progress in animal models.
Exp er im en ta l Section
Biologica l P r oced u r es. Cell Cu ltu r e a n d Cytotoxicity
Assa y. The reversal agents were solubilized at 10-2 M in
DMSO (except CsA solubilized in absolute ethanol) and diluted
directly in culture medium. The maximum concentration of
DMSO used (0.5%) was not cytotoxic and had no effect on
chemosensitivity. Verapamil (VRP) was obtained from Sigma
Co. Vincristine (VCR) and adriamycin (ADR) were obtained
from Roger Bellon (Neuilly, France) and were dissolved in
water, and CsA was a generous gift of Sandoz Co.
KB-A1, a human epidermoid carcinoma, was kindly pro-
vided by Dr. Gottesman (Bethesda, MD) and is 340-fold more
resistant to ADR with respect to the sensitive cell line KB-3-
1. P388/VCR-20 was obtained by culturing P388/VCR cells
in culture medium containing 20 nM VCR17 and was 69-fold
more resistant to VCR with respect to the corresponding
sensitive cell line P388.
Considering the in vitro MDR reversal, the majority
of the 36 compounds tested in this work were at least
as active as S9788 and markedly more active than
verapamil. No relationship between affinity for calcium
channel and MDR reversal in vitro was observed. The
compounds were tested in vivo on the P388/VCR murine
leukemia in association with vincristine. The five more
active compounds (3, 4, 9, 25, 26) totally restored the
sensitivity of this tumor to VCR when administered po
for 4 days. Interestingly, they have different affinities
for the phenylalkylamine binding site of the calcium
channel, ranging from 5 to 560 nM. Investigations of
These cell lines were previously characterized with respect
to Pgp overexpression and cross-resistance to MDR drugs.13,17

Purines as Modulators of MDR
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20 4103
Ta ble 4. Starting Materials
Sch em e 3. Compounds 36-38 (Table 3)a
a
Reagents and reaction conditions: (a) 2 equiv of amine in
EtOH for 1 h at room temperature and for 30 min at 50 °C; (b) 2
equiv of amine in EtOH for 24 h at 150 °C in an autoclave; (c) 6.5
equiv of allylamine in nBuOH for 72 h at 150 °C in an autoclave.
R1 and R2 have the same meaning as mentioned in Table 3.
Cells were cultivated in RPMI 1640 medium supplemented
with 10% fetal calf serum (Gibco), 2 mM L-glutamine, 100
units/mL penicillin, 100 µg/mL streptomycin, and 10 mM
HEPES buffer (pH ) 7.4). The cytotoxicity was measured by
the microculture tetrazolium assay essentially as previously
described.1 Cells were exposed for four doubling times (48 h
for P388 cells, 96 h for KB cells) to both cytotoxic drug (nine
graded concentrations in triplicate) and the reversal agent.
Results are expressed as IC50 values, the drug concentration
inhibiting by 50% the absorbance with respect to untreated
cells.
The activity of the reversal agents is expressed as fold
reversion (FR ) IC50(cytotoxic alone)/IC50(cytotoxic + modula-
tor)), calculated for each concentration of modulator agent. To
monitor the effect of the modulator agents on cell viability,
systematic controls were performed under the same conditions,
and the results are expressed as percent viability (absorbance
cells grown in the presence of modulator/absorbance control
cells) ×100.
ADR Up ta k e Stu d ies. KB-A1 cells (5 × 105/mL) were
incubated with 50 µM ADR at 37 °C for 5 h with 2.5-10 µM
of the tested compounds. For each sample the ADR fluores-
cence of 10 000 cells was analyzed, at 4 °C, on an ATC3000
flow cytometer (Bruker, Wissembourg, France) using an argon
laser (Spectra-Physics, Les Ulis, France) emitting 600 mW at
488 nm. Results are expressed as the increase of the mean of
ADR fluorescence of treated cells compared with the mean of
ADR fluorescence of untreated cells.
An titu m or Activity. The parental sensitive P388 murine
leukemia and P388/VCR, the subline resistant to VCR, were
provided by the NCI (Frederick, MD). Groups of 8-10
B6D2F1 female mice (Iffa Credo, Lyon, France) weighing 18-
20 g were transplanted ip with 106 cells on day 0. The drugs
were administrated po daily on days 1-4. The modulator
agents were formulated as a suspension in 2.5% Tween 80/
H2O (v/v) and administered 30-60 min before VCR. Antitu-
mor activity was expressed in terms of T/V ) (median survival
time of modulator + VCR-treated group)/(median survival time
of VCR-treated group).
Bin d in g to th e Ca lciu m Ch a n n els. Rat skeletal muscle
microsomes (100 µg/mL) prepared as described18 were incu-
bated in 20 mM Mops/Na+ (pH 7.4) containing 1 mM EDTA
and 0.1% bovin serum albumin. After a 60 min incubation at
20 °C in the presence of 1 nM [3H]-(-)-D888 with or without
tested compounds, the samples were filtered under vacuum
on GF/C filters pretreated with 0.1% poly(ethylenimine) and
counted by liquid scintillation. Nonspecific binding was
determined in a parallel incubation containing an excess of
10 µM (-)-D888. All binding data were analyzed using
Lundon 2 (Lundon Software Inc., Cleveland).
a
C, H, N analyses were within 0.4% of theoretical values for
Gen er a l Meth od s for P r ep a r a tion of Com p ou n d s. All
nonaqueous reactions were carried out under an atmosphere
of nitrogen. Flash column chromatography was carried out
the formulae given, except for 43: N calcd, 30.95; found, 30.54.
All compounds exhibited NMR consistent with assigned structures.
b
Synthesis described in the Experimental Section.

4104 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20
Sch em e 4. Compounds 39 and 73 (Table 4)a
Dhainaut et al.
with the reported 2-methoxy homolog.24 Phthalic anhydride,
trimethoxybenzoic acid, and sodium acetate were heated for
3 h at 240 °C. The resulting (trimethoxybenzylidene)phthalide
(162 °C, 68%) was hydrogenated for 24 h at 150 °C and 150
kg of H2 pressure to give the 2-(methoxyphenethyl)benzoic acid
(166 °C, 33%). The latter was cyclized by means of P2O5 in
xylene under reflux for 1 h to give the corresponding tri-
methoxydibenzosuberone (oil, 31%). The ketone was reduced
in alcohol with LiAlH4 in Et2O for 2 h under reflux (mp 57-
60 °C, 50%). The alcohol was chlorinated with HCl gas in CH2-
Cl2 at 10 °C for 2 h (131 °C, 100%). The chloride was
substituted with AgCN in CH3CN for 3 h under reflux to give
the corresponding nitrile (mp 111 °C, 42%) which was reduced
by means of BMS in THF under reflux for 3 h to give the title
product (oil, 58%): IR (neat) 2850 cm-1 (NH2); 1H NMR (CDCl3,
200 MHz) δ 7.7 and 7.1 (2m, 4H, CHd), 6.4 (s, 1H, CHd), 4.35
(dd, 1H, CH-CH2-NH2), 3.8 (3s, 9H, 3 × OCH3), 3.3 (m, 2H,
CH2-CH2), 3.1 (dd, 2H, CH-CH2-NH2), 2.8 (m, 2H, CH2-CH2),
1.4 (bd, 2H,-NH2, exchangeable for D2O).
a
Reagents and reaction conditions: 1.5 equiv of P(C6H5)3, 2
equiv of allyllic alcohol, 1.5 equiv of DEAD in THF overnight at
room temperature.
Sch em e 5. Compounds 64 and 65 (Table 4)a
(-)-5-Am in o -8-c h lo r o -10,10-d io x o -11-m e t h y l-5,11-
d ih yd r od iben zo[c,f][1,2]th ia zep in e: prepared by analogy
with the reported (+)-isomer20 and using (+)-dibenzoyltartaric
acid.
F or th e P r ep a r a tion of Com p ou n d s 39-74 (Ta ble 4).
2,6-Dichloropurine is a commercial product. 4,6-Dichloro-1H-
pyrazolo[3,4-d]pyrimidine10 and 2,4-dichloro-1H-pyrrolo[2,3-
d]pyrimidine9 were prepared as reported.
9-Allyl-2,6-d ich lor op u r in e (39) a n d 7-Allyl-2,6-d ich lo-
r op u r in e (73). Diethyl azodicarboxylate (67.9 g, 0.39 mol)
was added dropwise over 30 min at room temperature to a
stirred solution of 2,6-dichloropurine (50 g, 0.26 mol), tri-
phenylphosphine (102.3 g, 0.39 mol), and allylic alcohol (35.4
mL, 0.52 mol) in THF (1 L). The temperature increased
to 50 °C. The mixture was stirred overnight at room temper-
ature and evaporated to dryness. The crude residue was
poured into Et2O, and the insoluble part was filtered off.
The filtrate was concentrated, and the residue was purified
by flash chromatography eluting with CH2Cl2-acetone (95:5)
to provide 43.7 g (73%) of 39: mp 73-74 °C; IR (Nujol) 1620-
a
Reagents and reaction conditions: (a) excess of allylamine for
4 h at 120 °C; (b) 1.1 equiv of NaNO2 in H2O-AcOH for 45 min at
50 °C; (c) 3.4 equiv of Na2S2O4 in H2O for 15 min at 80 °C; (d)
NaNO2 in H2O, HCl (4 N) up to pH 5; (e) C6H5POCl2 in excess for
24 h at 160 °C.
with Amicon silica gel (230-400 mesh) under medium nitrogen
pressure. 1H NMR were recorded on either a Bruker AC200
or a Bruker AM400 at 200 and 400 MHz, respectively. 13C
NMR were measured at 50.29 or 75.43 MHz. Chemical shifts
are expressed in ppm downfield from internal tetramethylsi-
lane. Significant 1H NMR data are reported in the following
order: multiplicity (s, singlet; d, doublet; t, triplet; q, quartet;
m, multiplet), number of protons. IR spectra were recorded
on a Bruker IFS48 infrared spectrometer. Elemental analyses
were carried out on a Carlo ERBA autoanalyzer. All melting
points were determinated on a MEL TEMP capillary ap-
paratus and are uncorrected.
1
1593 cm-1 (CdC, CdN); H NMR (CDCl3, 200 MHz) δ 8.1 (s,
1H, NdCH-N), 6.2-5.9 (m, 1H, CH2-CHdCH2), 5.5-5.3 (m,
2H, CHdCH2), 4.85 (m, 2H, CH2-CHd). Anal. (C8H6Cl2N4)
C, H, N, Cl.
Further eluting provided 11 g (18.5%) of 73: mp 102-103
°C; IR (Nujol) 1640 cm-1 (CdC); 1H NMR (CDCl3, 200 MHz) δ
8.3 (s, 1H, N)CH-N), 6.2-6.0 (m, 1H, CH2-CH)CH2), 5.45-
5.2 (m, 2H, CHdCH2), 5.15 (d, 2H, CH2-CHd). Anal. (C8H6-
Cl2N4) C, H, N, Cl.
Sta r tin g Ma ter ia ls: F or th e P r ep a r a tion of Com -
pou n ds 3-38 (Tables 1-3). 2,2-Diphenylethylamine, 1-naph-
thalenemethylamine, and 1-[bis(4-fluorophenyl)methyl]pipera-
zine are commercially available. (10,11-Dihydro-5H-dibenzo-
[a,d]cyclohepten-5-yl)methylamine, (8-chloro-10,10-dioxo-11-
methyl-5,11-dihydrodibenzo[c,f][1,2]thiazepin-5-yl)meth-
ylamine, (6,11-dihydrodibenz[b,e]oxepin-11-yl)methylamine,
(2-methoxy-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5-yl)-
methylamine, (5H-dibenzo[a,d]cyclohepten-5-yl)methylamine,
(6-oxo-5,11-dih ydr o-6H -diben z[b,e]a zepin -11-yl)m et h yl-
amine, 5-amino-10,11-dihydro-5H-dibenzo[a,d]cycloheptene,
and 2,2,2-triphenylethylamine have been described in a pre-
ceding paper.1 [Bis(4-fluorophenyl)methyl]amine,19 (+)-5-
amino-8-chloro-10,10-dioxo-11-methyl-5,11-dihydrodibenzo-
[c,f][1,2]thiazepine,20 11-amino-6,11-dihydrodibenz[b,e]oxepine,21
and 1-(aminomethyl)benzosuberane22 were prepared by re-
ported procedures. 5-Chloro-10,10-dioxo-11-propyl-5,11-
dihydrodibenzo[c,f][1,2]thiazepine23 was substituted with AgCN
in CH3CN for 2 h under reflux to give the corresponding nitrile
(mp 102-103 °C, 46%), which was reduced by BMS in THF
for 2 h under reflux to give 5-(aminomethyl)-10,10-dioxo-11-
propyl-5,11-dihydrodibenzo[c,f][1,2]thiazepine (oil, 70%): IR
(neat) 3383-3323 (NH2), 1329-1149 cm-1 (SO2); 1H NMR
(CDCl3, 200 MHz) δ 7.9 (d, 1H, CHd), 7.45-7.2 (m, 7H, CHd),
4.1 (t, 1H, CH2-CH-C), 3.8-3.45 (m, 2H, N-CH2-CH2), 3.3 (d,
2H, CH-CH2-NH2), 1.7 (m, 2H, CH2-CH2-CH3), 1.1 (bd, 2H,
NH2, exchangeable for D2O), 0.85 (t, 3H, CH2-CH3). Anal.
(C17H20N2O2S) C, H, N, S.
1-Allyl-4,6-d ich lor op yr a zolo[3,4-d ]p yr im id in e (43). A
representative procedure is as follows for the conversion of 47
(90%). 4,6-Dichloropyrazolo[3,4-d]pyrimidine10 (38.1 g, 0.20
mol), triphenylphosphine (76 g, 0.3 mol), and allylic alcohol
(29 mL, 0.42 mol) were stirred in THF while diethyl azodi-
carboxylate (48 mL, 0.3 mol) was added dropwise over 30
min at room temperature. The solvent was concentrated, and
the crude residue was dissolved in Et2O. The insoluble
materials were filtered off, and the resulting solution was
evaporated under vacuum. The residue was purified by flash
chromatography eluting with CH2Cl2 to provide 43 which
crystallized on standing (11.4 g, 24.9%): mp 60-63 °C; IR
(Nujol) 1645, 1589 cm-1 (CdN, CdC); 1H NMR (CDCl3, 200
MHz) δ 8.15 (s, 1H, NdCH), 6.2-5.9 (m, 1H, CH2-CHdCH2),
5.3 (m, 2H, CHdCH2), 5.05 (d, 2H, N-CH2). Anal. (C8H6Cl2N4)
C, H, N, Cl.
47: mp 64-68 °C; IR (Nujol) 1643 (CdC), 1589 cm-1 (CdN);
1H NMR (DMSO-d6, 200 MHz) δ 7.1 (d, 1H, N-CHd), 6.7 (d,
1H, N-CHdCH), 6.0 (m, 1H, CH2-CHd), 5.2-4.9 (m, 2H,
dCH2), 4.85 (d, 2H, CH2). Anal. (C9H7Cl2N3) C, H, N, Cl.
9-Allyl-2-ch lor o-6-(p r op yla m in o)p u r in e (58). A repre-
sentative procedure is as follows for the conversions of 39 f
40 (72.2%), 43 f 44 (96%), 47 f 48 (92%), 43 f 51 (71%), 54
f 55 (90%), 54 f 61 (80.2%). A solution of 39 (40 g, 0.175
mol) and n-propylamine (20.7 g, 0.35 mol) in EtOH (700 mL)
was stirred for 1 h at room temperature and then for 30 min
at 50 °C. EtOH was evaporated in vacuo and the residue
treated with CH2Cl2 and water. The organic layer was
separated and dried (MgSO4), and the solvent was evaporated
(2,3,4-Tr im et h oxy-10,11-d ih yd r o-5H -d ib en zo[a ,d ]cy-
cloh ep ten -5-yl)m eth yla m in e. This was made by analogy

Purines as Modulators of MDR
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20 4105
Ta ble 5. Pharmacological Properties of the Target Compounds
in vitro reversal fold reversionb (% of viability)c
calcium channel affinitya
in vivo reversal T/Vd
(optimal dose po, mg/kg)
compd
Ki [3H]D888, nM
P388/VCR-20
KB-A1
2
89
58
ND
46
5
630
87
42 ( 7 (93 ( 3)
4 (91)
260 (65)
50 (98)
59 (95)
78 (93)
75 (70)
53 (80)
84 (95)
236 (114)
93 (111)
124 (146)
57 (135)
30 (74)
189 ( 28 (111 ( 4)
14 (79 ( 6)
9 (82)
1.19 (200)
1.22 (50)
ND
VRP
CsA
3
4
5
6
7
8
9
171 (104)
121 (120)
278 (115)
238 (121)
236 (110)
406 (108)
160 (108)
208 (110)
102 (60)
68 (64)
120 (132)
75 (145)
136 (86)
44 (109)
499 (126)
723 (86)
83 (111)
370 (127)
272 (122)
147 (86)
152 (103)
209 (107)
171 (112)
156 (127)
264 (84)
210 (95)
49 (106)
214 (141)
113 (106)
578 (110)
200 (91)
189 (102)
142 (119)
6 (73)
1.42 (100)
1.49 (100)
1.30 (100)
1.13 (100)
1.33 (100)
1.36 (100)
1.54 (100)
1.21 (50)
1.27 (100)
1.11 (100)
1.31 (100)
1.21 (100)
1.16 (100)
ND
440
120
12
170
650
1000
270
490
99
550
3700
200
770
410
62
420
400
48
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
57 (96)
108 (96)
37 (123)
71 (106)
108 (126)
15 (111)
27 (119)
78 (123)
56 (98)
1.13 (100)
1.24 (100)
ND
1.30 (100)
1.20 (50)
1.30 (100)
1.18 (100)
1.31 (100)
1.38 (50)
1.57 (100)
1.17 (100)
1.19 (100)
ND
1.03 (100)
1.31 (100)
1.19 (100)
1.06 (100)
1.17 (100)
ND
75 (90)
51 (102)
70 (103)
129 (98)
13 (83)
288 (112)
36 (108)
70 (89)
35 (104)
190 (91)
133 (131)
193 (119)
24 (120)
13 (98)
84
560
860
62
ND
170
1600
93
110
90
620
ND
ND
ND
ND
ND
ND
24 (99)
3 (97)
9 (103)
5 (104)
a
b
Ki inhibition constant of [3H]D888 binding to calcium channel of membranes from rat skeletal muscle. Ratio of IC50(cytotoxic alone
(VCR for P388/VCR-20, ADR for KB-A1 cells))/IC50(cytotoxic + modulator) (1 µM in association with VCR or 2.5 µM in association with
d
ADR). c Percent of cell viability remaining after incuation with the modulator alone. Ratio of the survival time of mice treated with
VCR (0.25 mg/kg ip) + modulator given at the optimal dose po/survival time of mice treated with VCR alone. Modulators were adminstered
60 min before VCR, on days 1-4 after tumor implantation. ND: not done.
to provide 37.4 g (84.9%) of 58 as yellowish gray crystals: mp
117 °C; IR (Nujol) 3271 (NH), 1618 cm-1 (CdC, CdN); 1H NMR
(DMSO-d6, 200 MHz) δ 8.3 (m, 1H, NH, exchangeable for D2O),
8.1 (s, 1H, NdCH-N), 6.2-6.0 (m, 1H, CH2-CHdCH2), 5.25-
5.0 (m, 2H, CHdCH2), 4.75 (d, 2H, N-CH2-CHd), 3.4 (m, t after
exchange, 2H, NH-CH2-CH2), 1.6 (m, 2H, CH2-CH2-CH3), 0.85
(t, 3H, CH3). Anal. (C11H14ClN5) C, H, N, Cl.
(250 mL) was stirred at 50 °C for 2 h. EtOH was evaporated
in vacuo, the crude residue was treated with CH2Cl2 and 10%
Na2CO3 solution, and the organic layer was separated and
washed with water. After drying, the solvent was evaporated
in vacuo to provide 22 g (90.6%) of yellowish gray crystalline
residue 60: mp 122 °C; IR (Nujol) 3271-3207 (NH), 1712
(CdO), 1628-1600 cm-1 (CdC, CdN); 1H NMR (DMSO-d6,
200 MHz) δ 7.75 (s, 1H, NdCH-N), 7.45 (m, 1H, NH,
exchangeable for D2O), 6.2-6.0 (m, 1H, CH2-CHdCH2), 5.3-
5.0 (m, 2H, CHdCH2), 4.65 (d, 2H, N-CH2-CHd), 4.0 (t, 4H,
CH2-N-CH2), 3.4 (m, 2H, NH-CH2-), 2.35 (t, 4H, CH2-O-CH2),
1.6 (m, 2H, CH2-CH2-CH3), 0.9 (t, 3H, CH3). Anal. (C16H22N6O)
C, H, N.
3-Allyl-5,7-d ih yd r oxy-1,2,3-t r ia zolo[4,5-d ]p yr im id in e
(64). 6-Chlorouracil (25.5 g, 0.174 mol) and allylamine (200
mL) were stirred for 4 h at 120 °C. After cooling, the mixture
was filtered, and the solid was washed with Et2O and dissolved
in 150 mL of NaOH (2 N). The solid residue was filtered off,
and the solution was cautiously acidified with concentrated
AcOH (pH ∼6) and then cooled. The precipitate was filtered
and washed with water and Et2O to provide 25 g (85.9%) of
6-(allylamino)uracil: mp 245-250 °C (lit.25 mp 254-255 °C).
A mixture of 6-(allylamino)uracil (20 g, 0.12 mol) with NaNO2
(10 g, 0.14 mol) in H2O (200mL) was heated at 50 °C. AcOH
was added to achieve pH 4.5, the mixture was stirred for 45
min at 50 °C and cooled, and the resulting precipitate was
filtered and washed with H2O and Et2O to provide 12 g (51%)
of 6-(allylamino)-5-nitrosouracil: mp 218 °C.
9-Allyl-2-(4,4-d ieth oxyp ip er id in -1-yl)-6-(p r op yla m in o)-
p u r in e (59). A representative procedure is as follows for the
conversions of 40 f 41 (88%), 44 f 45 (93.3%), 48 f 49 (30%),
51 f 52 (82.6%), 55 f 56 (87%), 61 f 62 (68.5%), 66 f 67
(60%). 58 (25 g, 0.099 mol) and 4,4-diethoxypiperidine (175
g, 1.01 mol) were stirred for 4 h at 100 °C. Excess 4,4-
diethoxypiperidine was evaporated in vacuo. The crude
residue was treated with CH2Cl2 and water. The organic layer
was dried and evaporated in vacuo, and the residue was
purified by flash chromatography eluting with CH2Cl2-acetone
(90:10) to provide 31.6 g (82.1%) of 59 as white crystals: mp
1
117 °C; H NMR (DMSO-d6, 200 MHz) δ 7.7 (s, 1H, NdCH-
N), 7.3 (m, 1H, NH, exchangeable for D2O), 6.0 (m, 1H, CH2-
CHdCH2), 5.2-5.0 (m, 2H, CHdCH2), 4.6 (d, 2H, N-CH2-
CHd), 3.7 (m, 4H, NH-CH2-, CH-N-CH), 3.45 (m, 6H, 2 ×
O-CH2-CH3, CH-N-CH), 1.8 (m, 6H, CH2-CH2-CH3, 2 × CH2-
CH2-N), 1.1 (t, 6H, 2 × CH3-CH2-O), 0.85 (t, 3H, CH2-CH2-
CH3). Anal. (C20H32N6O2) C, H, N.
9-Allyl-2-(4-oxop ip e r id in -1-yl)-6-(p r op yla m in o)p u -
r in e (60). A representative procedure is as follows for the
conversions of 41 f 42 (98%), 45 f 46 (90%), 49 f 50 (89%),
52 f 53 (88%), 56 f 57 (98%), 62 f 63 (91%), 67 f 68 (97.5%).
A solution of 59 (30 g, 0.077 mol) in EtOH (250 mL) and HCl
This compound was dissolved in water (20 mL) at 80 °C,
and Na2S2O4 (3 g, 0.015 mol) was added portionwise with

4106 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20
Dhainaut et al.
of 3-allyl-5,7-dihydroxy-1,2,3-triazolo[4,5-d]pyrimidine (64):
mp 281 °C; IR (Nujol) 3400-2500 (OH + H2O), 1741-1700
(CdN), 1630 cm-1 (CdC); 1H NMR (DMSO-d6, 200 MHz) δ
12.5-12.0 (m, 1H, OH, exchangeable for D2O), 11.25 (m, 1H,
OH, exchangeable for D2O), 6.0 (m, 1H, CHdCH2), 5.3-4.9
(m, 4H, CH2-CHdCH2); 13C NMR (DMSO-d6, 75.7 MHz) δ 48.8,
118.3, 124.2, 131.6, 141.7, 150.6, 156.2. Anal. (C7H7N5O2) C,
H, N.
a
3-Allyl-5,7-dich lor o-1,2,3-tr iazolo[4,5-d]pyr im idin e (65).
A mixture of 64 (8.7 g, 0.045 mol) and phenylphosphine
dichloride (25 mL) was stirred for 24 h at 160 °C. This mixture
was cooled, triturated with crushed ice, and partitioned
between water and CH2Cl2. The organic layer was separated,
washed with water, dried (MgSO4), and concentrated. The
crude material was purified by flash chromatography, eluting
with CH2Cl2, and then triturated with isopropyl ether to
provide 5.7 g (55%) of 65: mp 59 °C; IR (Nujol) 1647 (CdC),
1581 cm-1 (CdN); 1H NMR (DMSO-d6, 200 MHz) 6.2-6.0 (m,
1H, CHdCH2), 5.4-5.1 (m, 4H, CH2-CHdCH2). Anal. (C7H5-
Cl2N5) C, H, N, Cl.
b
9-Allyl-2-ch lor o-6-(4,4-d iet h oxyp ip er id in -1-yl)p u r in e
(69). A solution of 4,4-diethoxypiperidine (80.7 g, 0.46 mol)
and 9-allyl-2,6-dichloropurine (39) (53.5 g, 0.23 mol) in EtOH
(1.3 L) was stirred for 1 h at room temperature and then
concentrated. The crude material was partitioned between
CH2Cl2 and water; the organic layer was separated, washed
in water, dried (MgSO4), and concentrated to provide 71.5 g
of 69 (85%) as an oil: IR (neat) 1645 cm-1 (CdC); 1H NMR
(CDCl3, 200 MHz) δ 7.65 (s, 1H, NdCH-N), 6.15-5.9 (m, 1H,
CHdCH2), 5.35 (m, 2H, CHdCH2), 4.75 (d, 2H, N-CH2-CHd),
4.6-3.8 (m, 4H, CH2-N-CH2), 3.5 (q, 4H, 2 × OCH2), 1.9 (m,
4H, CH2-CH-N-CH-CH2), 1.25 (t, 6H, 2 × CH2-CH3). Anal.
(C17H24ClN5O2) C, H, N, Cl.
9-Allyl-2-(a llyla m in o)-6-(4,4-d ieth oxyp ip er id in -1-yl)p u -
r in e (70). A solution of 69 (1 g, 0.0027 mol), allylamine (2
mL, 0.027 mol), KI (1 crystal), and EtOH (6 mL) was heated
in an autoclave for 40 h at 100 °C. The mixture was
concentrated, the resulting residue was treated with CH2Cl2
and water, and the organic layer was separated, dried (Mg-
SO4), and concentrated. The crude product was purified by
flash chromatography eluting with CH2Cl2-acetone (95:5) to
provide 0.72 g (69%) of 70 as an oil: IR (neat) 3363 (NH),
1645-1593 cm-1 (CdC, CdN); 1H NMR (DMSO-d6, 200 MHz)
δ 7.7 (s, 1H, NdCH-N), 6.5 (t, 1H, NH, exchangeable for D2O),
6.15-5.7 (m, 2H, 2 × CHdCH2), 5.2-4.95 (m, 4H, 2 × dCH2),
4.6 (d, 2H, N-CH2-CHd), 4.15 (m, 4H, CH2-N-CH2), 3.9 (t, 2H,
NH-CH2-CHd), 3.45 (q, 4H, 2 × OCH2), 1.7 (m, 4H, CH2-CH2-
N-CH2-CH2), 1.1 (t, 6H, 2 × CH3-CH2-O). Anal. (C20H30N6O2)
C, H, N.
F igu r e 1. In vitro activity of compounds 2 (2) and 9 (9) on
KB-A1 cells compared with VRP (×) and CsA (4). (a) Cells
were exposed for 4 days to ADR alone or to ADR + modulator,
and the IC50 values were mesured by the microculture tetra-
zolium assay. Fold reversion is the ratio of the IC50(ADR
alone)/IC50(ADR + modulator). (b) Cells were incubated
with 50 µM ADR for 5 h with ADR alone or ADR + modula-
tor. The increase in ADR uptake is (mean of ADR fluores-
cence of treated cells) _ (mean of ADR fluorescence of untreated
cells).
9-Allyl-2-(allylam in o)-6-(4-oxopiper idin -1-yl)pu r in e (71).
A solution of 70 (24.2 g, 0.062 mol), 1 N HCl (240 mL), and
ethyl alcohol (240 mL) was heated for 2 h at 50 °C with
stirring. The solvent was evaporated and the residue parti-
tioned between CH2Cl2 and 10% aqueous Na2CO3. The organic
layer was separated, washed with water, dried over MgSO4,
and condensed to give 18 g (93%) of an oil which crystallized
on standing: mp 97 °C; IR (Nujol) 3350 (NH), 1697 (CdO),
1645 cm-1 (CdC); 1H NMR (DMSO-d6, 200 MHz) δ 7.75 (s,
1H, NdCH-N), 6.65 (t, 1H, NH, exchangeable for D2O), 6.15-
5.75 (m, 2H, 2 × CH2-CHd), 5.25-4.9 (m, 4H, 2 × CH2dCH),
4.65 (d, 2H, N-CH2-CHd), 4.4 (m, 4H, CH2-N-CH2), 3.85 (m,
2H, NH-CH2-CHd), 2.45 (m, 4H, CH2-CO-CH2). Anal.
(C16H20N6O) C, H, N.
9-Allyl-6-[4-[b is(4-flu or op h en yl)m et h yl]p ip er a zin -1-
yl]-2-ch lor op u r in e (72). A solution of 39 (3.2 g, 0.014 mol)
and [bis(4-fluorophenyl)methyl]piperazine in ethyl alcohol
(50 mL) was stirred for 1 h at room temperature. Stirring
was then continued for 1 h at 50 °C. The solvent was
evaporated in vacuo, and the residue was partitioned be-
tween CH2Cl2 and 10% aqueous Na2CO3. The organic layer
was separated, washed with water, dried (MgSO4), and
concentrated. The residual oil was purified by use of flash
chromatography eluting with CH2Cl2-acetone (80:20) to pro-
vide the crude product which was crystallized from ethyl
alcohol to give 5.4 g (80.2%) of the title product 72: mp 162
°C; IR (Nujol) 1597 cm-1 (CdC); 1H NMR (CDCl3, 200 MHz) δ
7.6 (s, 1H, NdCH-N), 7.35 (m, 4H, aromatic H), 7.0 (m, 4H,
F igu r e 2. Lack of direct relationship between calcium channel
affinity and sensitization of KB-A1 to ADR.
stirring until discoloration and precipitation. Stirring was
continued for 15 min at 80 °C, and then the mixture was cooled
to 0 °C and filtered. The crude solid obtained was suspended
in water (20 mL) containing 1.3 g of NaNO2. A 4 N HCl
solution was then added dropwise up to pH 5 at 5-10 °C.
Stirring was continued for 10 min at 10 °C and for 1 h at
room temperature. After standing on cooling, the precipitate
was filtered, washed with water, and dried to provide 1 g (30%)

Purines as Modulators of MDR
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20 4107
aromatic H), 6.0 (m, 1H, CH2-CHd), 5.25 (m, 2H, CHdCH2),
4.75 (m, 2H, N-CH2-CHd), 4.25 (m, 4H, CH2-N-CH2), 4.3 (s,
1H, N-CH), 2.45 (m, 4H, CH2-N-CH2). Anal. (C25H23ClF2N6)
C, H, N, Cl.
concentrated. The residue was crystallized from isopropyl
ether to provide 1.2 g (35%) of the title compound 37: mp 208-
209 °C; IR (Nujol) 3276-3200 (NH), 1601 cm-1 (CdN); 1H
NMR (DMSO-d6, 200 MHz) δ7.7 (s, 1H, NdCH-N), 7.45 (m,
4H, aromatic H), 7.15 (m, 4H, aromatic H), 6.15-5.8 (m, 2H,
2 × CHdCH2), 5.2-5.0 (m, 4H, 2 × CHdCH2), 4.6 (d, 2H,
N-CH2-CHd), 4.35 (s, 1H, N-CH), 4.05 (m, 2H, HN-CH2-CHd),
3.7 (m, 4H, CH2-N-CH2), 2.3 (m, 4H, CH2-N-CH2). Anal.
(C28H29F2N7) C, H, N.
Tar get Com pou n ds (Tables 1-3): 9-Allyl-6-(allylam in o)-
2-[4-[[(10,11-d ih yd r o-5H-d iben zo[a ,d ]cycloh ep ten -5-yl)-
m eth yl]a m in o]p ip er id in -1-yl]p u r in e (3). A representative
procedure is provided for the following conversions, and the
system of elution employed for flash chromatography is given
in parentheses: 42 f 4 (60.7%) (CH2Cl2-acetone, 80:20), 42
f 5 (59%) (CH2Cl2-MeOH, 96:4), 42 f 6 (33.9%) (CH2Cl2-
MeOH, 97:3), 42 f 7 (57%) (CH2Cl2-MeOH, 97:3), 42 f 8
(89%) (toluene-CH2Cl2-MeOH, 50:45:5), 42 f 9 (86%), 42f
10 (93.6%), 42 f 11 (80.6%) (toluene-CH2Cl2-MeOH, 50:45:
5), 42 f 12 (76%) (toluene-CH2Cl2-MeOH, 50:45:5), 46 f
13 (97.4%) (CH2Cl2-acetone, 80:20), 42 f 14 (81%) (CH2Cl2-
acetone, 70:30), 42 f 15 (55.8%) (CH2Cl2-MeOH, 95:5), 42 f
16 (43%) (CH2Cl2-acetone, 80:20), 42 f 17 (64.1%) (CH2Cl2-
MeOH, 95:5), 50 f 18 (95%) (toluene-MeOH, 99:1), 42 f 19
(66%) (CH2Cl2-acetone, 90:10), 42 f 20 (88.9%) (toluene-
MeOH, 96:4), 42 f 21 (68%) (CH2Cl2-acetone, 70:30), 53 f
22 (58%) (CH2Cl2-MeOH, 95:5), 46 f 23 (61.5%) (CH2Cl2-
acetone, 90:10), 57 f 24 (82%) (toluene-MeOH, 95:5), 60 f
25 (62%) (CH2Cl2-acetone, 70:30), 63 f 26 (34%) (CH2Cl2-
acetone, 70:30), 68 f 27 (62%) (toluene-EtOH, 97:3), 71 f
28 (68, 7%) (CH2Cl2-MeOH, 97:3), 71 f 29 (59%) (CH2Cl2-
acetone, 85:15), 71 f 30 (66%) (toluene-CH2Cl2-MeOH, 50:
45:5), 71 f 31 (78%) (toluene-CH2Cl2-MeOH, 50:45:5), 71
f 32 (84.6%) (CH2Cl2-MeOH, 95:5), 71 f 33 (64.4%), 71 f
34 (84%), 71 f 35 (85%) (CH2Cl2-acetone, 90:10). A solution
of 42 (5 g, 0.016 mol), (10,11-dihydro-5H-dibenzo[a,d]cyclo-
hepten-5-yl)methylamine (3.6 g, 0.016 mol), and sodium tri-
acetoxyborohydride (7.4 g, 0.035 mol) in CH2Cl2 (150 mL) was
stirred for 5 h at room temperature. A 10% aqueous Na2CO3
solution was added, and the organic layer was separated,
washed with water, dried over MgSO4, and concentrated. The
resulting oil was purified by flash chromatography eluting with
CH2Cl2-acetone (80:20) to provide the title product 3 (7.3 g,
87.9%): mp 142-144 °C; IR (Nujol) 3282 (NH), 1610 cm-1
Ack n ow led gm en t. Thanks are due to the staff of
Analytical Division of this institute for the elemental
analyses and spectral measurements. Thanks are also
due to Alain Lombet for the calcium channel binding
studies and Isabelle Duprat for her enthusiastic techni-
cal assistance.
Refer en ces
(1) Dhainaut, A.; Re´gnier, G.; Atassi, G.; Pierre´, A.; Le´once, S.;
Kraus-Berthier, L.; Prost, J . F. New Triazine Derivatives as
Potent Modulators of Multidrug Resistance. J . Med. Chem. 1992,
35, 2481-2496.
(2) Terret, C.; Goncalves, E.; Ruffie, P.; Sarkany, M.; Giroux, B.;
Lucas, C.; Funck-Brentano, C.; Armand, J . P. Specific Methodol-
ogy for a Phase I Trial with Unusual Side-effect: S9788. 9th
NCI-EORTC Symposium on New Drugs in Cancer Therapy,
1996; Abstract 438.
(3) Terret, C.; Le Cesne, A.; Lagarde, N.; Di Palma, M.; Goncalves,
E.; N’Dom, P.; Yataghene, Y.; Funck-Brentano, C.; N’Guyen, J .
P.; Marino, J . P.; Besse, B.; Armand, J . P.; Le Chevalier, T.;
Belpomme, D.; Misset, J . L.; D’Agay, L.; Berger, E.; Sarkany,
M.; Giroux, B. S 9788 a multidrug resistance (MDR) reversing
agent: French phase I studies with a 6 h continuous infusion
schedule in combination with chemotherapy in patients with
refractory cancer. Proc. Am. Assoc. Cancer Res. 1996, 37, 165,
Abstract #1139.
(4) El-Sherif, N. Early Afterdepolarizations and Arrhythmogenesis.
Experimental and Clinical Aspects. Arch. Mal. Coeur. 1991, 84
(2), 227-234.
(5) Franz, M. R.; Koller, B.; Woosley, R. L. Pharmacologic therapy
in cardiac arrhythmias. In Cardiovascular Pharmacology and
Therapeutics; Bramah, Singh, N., Dzan, V. J ., Vanhoutte, P. M.,
Woosley, R. L., Eds.; Churchill Livingstone Inc.: Edinburgh,
U.K., 1994; pp 595-616.
1
(CdN); H NMR (DMSO-d6, 200 MHz) δ 7.7 (s, 1H, NdCH-
N), 7.5 (s, 1H, NH, exchangeable for D2O), 7.2 (m, 2H, aromatic
H), 7.1 (m, 6H, aromatic H), 5.95 (m, 2H, 2 × CH2-CHd), 5.1
(m, 4H, 2 × CHdCH2), 4.6 (d, 2H, N-CH2-CHd), 4.5 (d, 2H,
CH2-N-CH2), 4.25 (t, 1H, NH-CH2-CH), 4.1 (s br, 2H, NH-CH2-
CHd), 3.2 (m, 4H, CH2-CH2), 3.0 (m, 2H, NH-CH2-CH), 2.9 (t,
2H, CH2-N-CH2), 2.7 (m, 1H, CH-NH), 1.8 (m, 2H, CH2-CH-
CH2), 1.2 (m, 2H, CH2-CH-CH2), 1.2 (s br, 1H, NH, exchange-
able for D2O). Anal. (C32H37N7) C, H, N.
9-Allyl-2-(a llyla m in o)-6-[4-[bis(4-flu or op h en yl)m eth yl]-
p ip er a zin -1-yl]p u r in e (36). A solution of 72 (3 g, 0.0062
mol), allylamine (3 mL, 0.04 mol), potassium iodide (0.05 g),
and n-butyl alcohol (60 mL) was heated for 72 h at 150 °C in
an autoclave. The mixture was cooled, and the solvent was
evaporated. The residue was partitioned between CH2Cl2
and water. The organic layer was separated, dried over
MgSO4, and concentrated. The oily residue was purified by
flash chromatography eluting with CH2Cl2-acetone (95:5)
to provide 0.96 g (31%) of 36: mp 170-171 °C; IR (Nujol)
3307 (NH), 1645 cm-1 (CdC); 1H NMR (DMSO-d6, 200
MHz) δ 7.7 (s, 1H, NdCH-N), 7.4 (m, 4H, aromatic H), 7.15
(m, 4H, aromatic H), 6.5 (t, 1H, NH, exchangeable for D2O),
6.15-5.8 (m, 2H, 2 × CH2-CHd), 5.25-4.9 (m, 4H, 2 ×
CHdCH2), 4.6 (d, 2H, N-CH2-CHd), 4.4 (s, 1H, N-CH), 4.25-
4.0 (m, 4H, CH2-N-CH2), 3.8 (t, 2H, NH-CH2-CHd, d after
exchange for D2O), 2.35 (m, 4H, CH2-N-CH2). Anal.
(C28H29F2N7) C, H, N.
9-Allyl-6-(a llyla m in o)-2-[4-[bis(4-flu or op h en yl)m eth yl]-
p ip er a zin -1-yl]p u r in e (37). A representative procedure is
as follows for the conversion mentioned with the eluting
system in parentheses for flash chromatography: 74 f 38
(76%) (CH2Cl2-MeOH, 95:5). A solution of 40 (1.7 g, 0.0068
mol), 1-[bis(4-fluorophenyl)methyl]piperazine (3.9 g, 0.0136
mol), potassium iodide (0.05 g), and ethyl alcohol (50 mL) was
heated for 24 h at 150 °C in an autoclave. The solvent was
evaporated in vacuo, and the residue was treated with CH2-
Cl2 and a 10% aqueous Na2CO3 solution. The organic layer
was separated, washed with water, dried (MgSO4), and
(6) J aillon, P.; Simon, T. Les effets pro-arythmiques ventriculaires
des me´dicaments antiarythmiques. (Ventricular proarrhythmic
effects of antiarrhytmic drugs.) Therapie 1992, 47, 187-192.
(7) Somberg, J . C. Calcium channel blockers that prolong the QT
interval. Am. Heart J . 1985, 110, 416-421.
(8) Mitsunobu, O. The Use of Diethyl Azodicarboxylate and Tri-
phenylphosphine in Synthesis and Transformation of Natural
Products. Synthesis 1981, 1-28.
(9) Kazimierczuk, Z.; Cottam, H. B.; Revankar, G. R.; Robins, R. K.
Synthesis of 2′-Deoxytubercidin, 2′-Deoxyadenosine, and Related
2′-Deoxynucleosides via a Novel Direct Stereospecific Sodium
Salt Glycosylation Procedure. J . Am. Chem. Soc. 1984, 106,
6379-6382.
(10) Robins, R. K. Potential Purine Antagonists. IX. Further Studies
of Some 4,6-Disubstituted Pyrazolo[3,4-d]pyrimidines. J . Am.
Chem. Soc. 1957, 79, 6407-6415.
(11) Pierre´, A.; Dunn, T. A.; Kraus-Berthier, L.; Le´once, S.; Saint-
Dizier, D.; Re´gnier, G.; Dhainaut, A.; Berlion, M.; Bizzari, J . P.;
Atassi, G. In vitro and In vivo Circumvention of Multidrug
Resistance by Servier 9788, a Novel Triazinoaminopiperidine
Derivative. Invest. New Drugs 1992, 10, 137-148.
(12) Pe´rez, V.; Pierre´, A.; Le´once, S.; Anstett, M.; Atassi, G. Effect of
Duration of Exposure to S9788, Cyclosporin A or Verapamil on
Sensitivity of Multidrug Resistant Cells to Vincristine or Doxo-
rubicin. Anticancer Res. 1993, 13, 985-990.
(13) Le´once, S.; Pierre´, A.; Anstett, M.; Pe´rez, V.; Genton, A.; Bizzari,
J . P.; Atassi, G. Effects of a New Triazinoaminopiperidine
Derivative on Adriamycin Accumulation and Retention in Cells
Displaying P-glycoprotein-mediated Multidrug Resistance. Bio-
chem. Pharmacol. 1992, 44 (9), 1707-1715.
(14) Beck, W. T.; Qian, X. D. Photoaffinity substrates for P-glyco-
protein. Biochem. Pharmacol. 1992, 43, 89-93.
(15) Beck, W. T.; Qian, X. D. Binding of an optically pure photoaffinity
analog of verapamil, LU-49888, to P-glycoprotein from multi-
drug-resistant human leukemic cell lines. Cancer Res. 1990, 50,
1132-1137.
(16) Striessnig, J .; Knaus, H.; Grabner, M.; Moosburg, K.; Seitz, W.;
Leitz, H.; Glossman, H. Photoaffinity labeling of the phenyl-
alkylamine receptor of the skeletal muscle transverse-tubule
calcium channel. FEBS Lett. 1987, 212, 247-253.
(17) Pe´rez, V.; Pierre´, A.; Le´once, S.; Anstett, M.; Prost, J . F.; Atassi,
G. Caracte´risation in vitro de l’Activite´ du S9788, un Nouveau
Modulateur de la Re´sistance Multidrogue. (In vitro activity of
S9788 on MDR cell-lines.) Bull. Cancer 1993, 80, 310-325.

4108 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20
Dhainaut et al.
(18) Qar, J .; Galizzi, J . P.; Fosset, M.; Lazdinski, M. Receptors for
Diphenylbutylpiperidine Neuroleptics in Brain, Cardiac and
Smooth Muscle Membranes. Relations with Receptors for 1,4-
Dihydropyridines and Phenylalkylamines and with Calcium
Channel blockade. Eur. J . Pharmacol. 1987, 141, 261-268.
(19) Grisar, J . M.; Claxton, G. P.; Wiech, N. L.; Lucas, R. W.;
Mackenzie, R. D.; Goldstein, S. Benzhydryl and Fluorenyl
Lactamimides with Hypoglycemic, Diuretic, Blood Platelet Ag-
gregation Inhibitory and Anti-inflammatory Activities. J . Med.
Chem. 1973, 16, 885-893.
(22) Seidel, G.; Huigsen, R.; Hill, J . H. M. Medium Sized Rings. XXV.
Reaction of 1,2-Benzocycl-1-en-3-ylmethylamine with Nitric Acid.
Tetrahedron 1964, 20, 633-640.
(23) Malen, C.; Laubie, M. Nouveaux De´rive´s a` la Dibenzo(c,f)-
thiaze´pine(1,2) et leur Proce´de´ de Pre´paration. French Patent
FR 1566191; Chem. Abstr. 1970, 72, 100780j.
(24) Winthrop, S. O.; Davis, M. A.; Myers, G. S.; Gavin, J . G.; Thomas,
R.; Barber, R. New Psychotropic Agents. Derivatives of Dibenzo-
[a,d]-1,4-cycloheptadiene. J . Org. Chem. 1962, 27, 230-240.
(25) Wright, G. E.; Brown, N. C. Inhibitors of Bacillus subtilis DNA
Polymerase III. 6-Anilinouracils and 6-(Alkylamino)uracils. J .
Med. Chem. 1980, 23, 34-38.
(20) Kamoun, A.; Delalleau, B.; Deslandes, A. Tianeptine Isomer for
the Treatment of Cognitive and Memory Disorders. Eur. Patent
EP 671173; Chem. Abstr. 1995, 123, 266135z.
(21) Winter, W.; Thiel, M.; Stach, K.; Schaumann, W.; Ribbentrop,
A. Preparation of New Dibenzoxepines. Br. Patent GB 1119329;
Chem. Abstr. 1969, 70, 11583p.
J M960361I