Self-immolative anthracycline prodrugs for suicide gene therapy

Article abstract of DOI:10.1021/jm980696v

Four novel potential prodrugs derived from daunorubicin (8, 10) and doxorubicin (12, 14) were designed and synthesized. They are self-immolative prodrugs for suicide gene therapy activation by the enzyme carboxypeptidase G2 (CPG2) subsequently releasing t

Full text of DOI:10.1021/jm980696v

J . Med. Chem. 1999, 42, 2485-2489
2485
Self-Im m ola tive An th r a cyclin e P r od r u gs for Su icid e Gen e Th er a p y
Ion Niculescu-Duvaz, Dan Niculescu-Duvaz, Frank Friedlos, Robert Spooner, J anet Martin, Richard Marais, and
Caroline J . Springer*
CRC Centre for Cancer Therapeutics, Institute of Cancer Research, 15 Cotswold Road, Sutton, Surrey SM2 5NG, U.K.
Received December 14, 1998
Four novel potential prodrugs derived from daunorubicin (8, 10) and doxorubicin (12, 14) were
designed and synthesized. They are self-immolative prodrugs for suicide gene therapy activation
by the enzyme carboxypeptidase G2 (CPG2) subsequently releasing the corresponding
anthracyclines, by a 1,6-elimination mechanism. A mammary carcinoma cell line (MDA MB
3
61) was engineered to express CPG2 intracellularly (CPG2*) or extracellularly, tethered to
the outer cell membrane (stCPG2(Q)3). The prodrugs derived from doxorubicin showed prodrug/
drug cytotoxicity differentials of 21-fold (compound 12) and 23-fold (compound 14). Prodrug
1
2 underwent an 11-fold activation when assayed in the cell line expressing externally surface-
tethered CPG2.
In tr od u ction
icity and myelosuppression. To overcome these toxicities
several “self-immolative” and “non-self-immolative” an-
thracycline prodrugs, obtained mainly by derivatization
of the amine functionality of the daunosamine, have
been described for activation by a wide range of enzymes
Gene-directed enzyme prodrug therapy (GDEPT)1,2
and virally directed enzyme prodrug therapy (VDEPT)
are two-step treatments for solid tumors. In the first
step, gene delivery technology ideally leads to the
expression of a prodrug-activating enzyme only in cells
in the tumor. During the second step, a nontoxic prodrug
is administered which is converted by the activating
3
1
2-30
in ADEPT strategies.
However, CPG2 cannot acti-
vate any of the reported prodrugs. Furthermore, direct
addition of L-glutamyl residues to doxorubicin or dauno-
rubicin does not generate a molecule which is a sub-
strate for CPG2. We therefore describe the synthesis of
novel prodrugs from doxorubicin and daunorubicin,
based on a similar strategy, which include self-immo-
lative linkers designed to reduce their intrinsic toxicity.
These produgs, for use in suicide gene therapy, can be
activated by CPG2, releasing the active drugs by a 1,6-
4
enzyme to a potent cytotoxic agent. We have previously
described a GDEPT system in which the bacterial
enzyme carboxypeptidase G2 (CPG2) is used to activate
5
6
prodrugs. CPG2 catalyzes the scission of amidic,
urethanic, or ureidic7,8 bonds between a benzene nucleus
and L-glutamic acid. This is a versatile system, because
2
CPG2 can be expressed either intracellularly or extra-
3
1-33
elimination mechanism.
cellularly anchored to the outer membrane of tumor
9
cells. The potential advantages of extracellular expres-
Ch em istr y
sion are two-fold. First, it should, theoretically, give an
improved bystander effect because the drug will be
generated in the interstitial spaces within the tumor,
rather than inside the cells as with an intracellularly
expressed activating enzyme. Second, the prodrug need
not enter cells to become activated, and therefore non-
cell-permeable prodrugs can be used. Thus, prodrugs
which release drugs with intracellular targets may be
rendered nontoxic by preventing their entry into cells.
Upon activation, a potent and cell-permeable active
moiety is released. This has already been demonstrated
Linkers cleavable by CPG2 were incorporated into the
prodrugs. The linkers 3a (Z ) NH) and 3b (Z ) O)
(Scheme 1) were previously synthesized using a five-
5
step and a three-step procedure, respectively. Here, we
describe a simple one-pot procedure which allows the
concise synthesis of 3a and 3b by direct coupling of
diallyl L-glutamylisocyanate, 2, with 4-amino- or 4-hy-
droxybenzylic alcohol, 1a and 1b, respectively, under
controlled conditions. The condensation of the 4-ami-
nobenzylic alcohol is catalyzed by triethylamine. t-BuOK
and 18-crown-6 ether were used for the condensation
of the 4-hydroxybenzylic alcohol (Scheme 1).
9
to be beneficial for prodrug-impermeable tumor cells.
However, the potential for increased toxicity due to the
diffusion of the active drug away from the tumor is a
potential disadvantage, although this could happen also
to active drugs from an intracellularly expressed en-
zyme.
The anthracyclines, such as doxorubicin and dauno-
rubicin, are antitumor drugs with the widest spectrum
of activity in human tumors.10,11 Their therapeutic
efficacy is limited by toxic side effects, mainly cardiotox-
The linkers 3a and 3b were activated as 4-nitrophenyl
5
carbonates 4a and 4b, respectively, and reacted with
daunorubicin or doxorubicin in DMF (see Scheme 2).
The protected prodrugs 7, 9, 11, and 13 were purified
by a two-step preparative TLC procedure and depro-
tected with tetrakis(triphenylphosphine)Pd(0) and an
34
allyl scavenger (morpholine, sodium toluenesulfinate )
to prodrugs 8, 10, 12, and 14. The desalting of prodrugs
1
2 and 14 was achieved using a weak acidic ion
exchanger (IRC 50). The final purification of these
prodrugs was performed by semipreparative HPLC.
*
To whom reprint requests should be addressed. Tel: +44 181-
7
22 4214. Fax: +44 181-643 6902.
1
0.1021/jm980696v CCC: $18.00 © 1999 American Chemical Society
Published on Web 06/11/1999

2
486 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 13
Brief Articles
Sch em e 1a
a
(1a) Z ) NH, THF, NEt3; (1b) Z ) O, THF, t-BuOK, 18-crown-6; (2) 4-nitrophenyl chloroformate, NEt3.
Sch em e 2a
a
(3) DMF, NEt3; (4) Pd(0), morpholine; (5) Pd(0), sodium toluenesulfinate or morpholine.
Ta ble 1. Half-Lives (t1/2) of Doxorubicin and Daunorubicin Prodrugs in the Presence or Absence of CPG2 and Their Cytotoxicitya
to MDA MB 361 Cells Expressing Control LacZ, CPG2 Internally (CPG2*), or CPG2 Tethered to the Outer Surface of the Cell
Membrane (stCPG2(Q)3)
cytotoxicity in MDA MB 361 cell line expressing:
t1/2 (min)
(apparent)
prodrug in the
LacZ
differential
between
prodrug/drug
CPG2*
differential
between prodrug,
LacZ/CPG2*
stCPG2(Q)3
prodrug in
the absence
compound,
IC50
(nM)
compound,
IC50
compound,
differential
IC50
between prodrug,
LacZ/stCPG2(Q)3
b
(nM)b
(nM)b
drug
Doxo
of enzyme presence of enzyme
12 ( 2
256 ( 48
279 ( 66
17 ( 6
112 ( 57
115 ( 57
na
21.3
23.3
na
6.6
6.8
33 ( 12
47 ( 26
na
18 ( 1
24 ( 6
na
10.7
3.0
na
3.9
0.7
1
1
2
4
948
1702
42.0
267.4
5.4
1.4
na
0.9
0.3
204 ( 26
36 ( 13
92 ( 22
21 ( 5
Dauno
8
1187
1482
38.7
132.4
123 ( 66
405 ( 304
29 ( 9
177 ( 144
1
0
a
Cytotoxicity differentials describe either the ratio between prodrug:drug in the control cell line expressing LacZ (LacZ column), the
ratio for a given prodrug between the cell line expressing CPG2* compared to that expressing LacZ (CPG2* column), or the ratio for a
given prodrug between the stCPG2(Q)3 line compared to the LacZ line (stCPG2(Q)3 column). b Data are nM ( SEM; na, not applicable;
see Enzymic Activity and Biological Evaluation for the analysis of variance of Table 1 data.
En zym ic Activity a n d Biologica l Eva lu a tion
of daunorubicin and doxorubicin reduced their cyto-
toxicity was determined by comparing their IC50 to that
of 8 and 10 or 12 and 14, respectively, in control MDA
MB 361 (LacZ) cells. Two-tailed analysis of variance
showed significant differences (P < 0.0001) between the
cytotoxicity of doxorubicin compared to 12 (21-fold) and
14 (23-fold) and differences (P < 0.085) between dauno-
rubicin, 8, and 10 (7-fold) (Table 1). A P value of <0.1
was selected as indicating significance. The assumptions
The ability of 8, 10, 12, and 14 to act as substrates
for CPG2 was determined by comparing their half-lives
in phosphate-buffered saline (pH 7.4, 37 °C) in the
presence or absence of CPG2 (50 mU). All four prodrugs
were similarly stable, but their half-lives were shortened
in the presence of CPG2 indicating that they are
substrates, 12 and 8 being more rapidly hydrolyzed than
1
4 and 10 (Table 1). The extent to which derivatization

Brief Articles
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 13 2487
), before evaporating
again. A yellow oil resulted (10.25 g); 2.7 g of the obtained
product was submitted to purification by preparative HPLC
of the analysis of variance were shown to be correct in
that the residuals were normally distributed about zero
with equal variance. The ability of 8, 10, 12, and 14 to
act as prodrugs was determined by comparing their IC50
in control MDA MB 361 (LacZ) cells with that in cells
expressing CPG2 intracellularly (CPG2*) or extracel-
lularly (stCPG2(Q)3). There were highly significant
differences (P < 0.0001) between the cytotoxicities
experienced by all three cell lines following exposure to
doxorubicin, 12, and 14. Compound 12 gave the higher
cytotoxicity differential in both cell lines, commensurate
with its more rapid digestion by CPG2 (Table 1). Cells
expressing CPG2 extracellularly showed approximately
twice the activation of both 12 and 14, compared with
those expressing CPG2 intracellularly, probably reflect-
ing greater access of the prodrug to the enzyme.
Statistically significant differences did not exist between
the IC50 of daunomycin, 8, and 10 in the three cell lines,
although the highest differential of 4-fold seen with 8
in stCPG2(Q)3-expressing cells is consistent with the
result seen with its doxorubicin congener (12) and,
considering that 8 is shown to be a substrate for CPG2,
probably reflects a genuine prodrug effect.
with water (2 × 20 mL), and dried (MgSO
4
(
2 2
CH Cl :EtOAc, 1:1) which yielded 1.52 g (63%) of pure 3a :
-
1
ν
max cm (film) 3354 (NH-, OH, broad), 1737 (CdO, ester),
1
1
659 (CdO, urea); H NMR δ
H
1.85-1.93 (m, 1H, CH
CH(NH)-), 2.46 (t, 2H, CH
CO , J ) 5.4 Hz), 4.26-4.35 (m, 1H, -CH(NH)-), 4.40 (d, 2H,
2
CH-
(
NH)-), 2.00-2.05 (m, 1H, -CH
2
2
-
2
CH
.61 (d, 2H, CH
CH dallyl), 5.85-5.94 (m, 2H, CHdallyl), 6.56 (d, 1H, NH-G,
J ) 8.0 Hz), 7.17 (d, 2H, H2+6, J ) 8.5 Hz), 7.32 (d, 2H, H3+5),
2
-Ph, J ) 5.6 Hz), 4.55 (d, 2H, CH
2
O-allyl, J ) 5.3 Hz),
4
2
O-allyl), 4.98 (t, 1H, OH), 5.17-5.37 (m, 4H,
2
+
+
8
1
.52 (s, 1H, NH-Ph); MS m/z 399 (M + 23, 35), 377 (M + 1,
+
00), 359 (M - H
2 24 2 6
O, 34). Anal. (C19H N O ) C, H, N.
Dip r op -2-en yl N-[(4-{H yd r oxym et h yl}p h en oxy)ca r -
bon yl]-L-glu ta m a te (3b). 4-Hydroxybenzyl alcohol (1.00 g,
8.10 mmol) was dissolved in THF (dry) (10 mL) and 18-crown-6
ether (0.20 mL) and potassium tert-butoxide (0.16 g, 1.4 mmol)
were added. To this mixture was added a clear solution of
5
diallyl L-glutamylisocyanate (2) (8.20 mmol) in 50 mL of THF,
dropwise at room temperature with vigorous stirring. The
reaction was monitored by IR (disappearance of the νNCO 2252
-1
cm peak) and stopped by addition of AcOH (0.5 mL). The
reaction mixture was filtered and the solvent removed under
vacuum. The residue was dissolved in EtOAc (20 mL), and the
solution was washed sequentially with aqueous 4% NaOH (20
mL) and water (2 × 20 mL), dried, and evaporated to dryness.
Compound 3b (1.62 g, 53%) resulted and was purified by
preparative HPLC (eluent cyclohexane:EtOAc, 1:1) which
In conclusion, we show that self-immolative prodrugs
can be synthesized, which release cytotoxic anthracy-
clines when cleaved by CPG2. We show that doxorubi-
cin-based prodrugs are the more deactivated and yield
higher cytotoxicity differentials. Prodrugs based on
linker 3a are more rapidly cleaved by CPG2 and give
higher cytotoxicity differentials, with the highest dif-
ferential seen in cells expressing CPG2 extracellularly.
-1
resulted in a clear oil (0.80 g, 26.5%): νmax cm (film) 3343
1
(NH-, OH, broad), 1734 (CdO, ester);
(m, 1H, CH CH(NH)-), 2.05-2.09 (m, 1H, -CH
2.50-2.56 (m, 2H, CH CO ), 3.96-4.05 (m, 1H, -CH(NH)-),
, J ) 5.7 Hz), 4.58 (d, 2H, CH O-allyl, J )
.3 Hz), 4.63 (d, 2H, CH O-allyl), 5.16 (t, 1H, OH), 5.19-5.38
dallyl), 5.95-5.99 (m, 2H, CHdallyl), 7.04 (d, 2H,
3+5, J ) 8.5 Hz), 7.32 (d, 2H, H2+6), 8.24 (d, 1H, NH-G, J )
H NMR δH 1.90-2.01
2
2
CH(NH)-),
2
2
4
5
.48 (d, 2H, PhCH
2
2
2
(m, 4H, CH
2
H
7
Exp er im en ta l Section
+
+
.8 Hz); MS m/z 400 (M + 23, 82), 378 (M + 1, 3); mass
Na) calcd 400.1372, found 400.1376. Anal. (C19
) H, N; C: calcd, 60.47; found, 60.01.
(C
19
H
7
23NO
7
23
H -
All starting materials, reagents, and anhydrous solvents
were from Aldrich, unless otherwise stated. The diallyl
glutamate was obtained from Sigma. Daunorubicin‚HCl and
doxorubicin‚HCl were obtained from Meheco, China. Kieselgel
NO
N-[4-(Dip r op en -2-yl-L-glu ta m ylca r bon yla m in o)ben zyl-
oxyca r bon yl]d a u n or u bicin (7). To a solution of 60 mg (0.10
mmol) of daunorubicin hydrochloride in 4 mL and DMF were
added 54 mg (0.10 mmol) of diallyl 4-[[(4′-nitrophenyloxycar-
6
0 was used in gravity columns (Art. 9385 and 15111, Merck).
TLC was performed on precoated sheets of Kieselgel 60 F254
Art. 5735, Merck). Semipreparative HPLC was performed on
ThermoQuest equipment using an Apex ODS-WP, column (25
0.8 cm) from J ones Chromatography, using an isocratic
5
bonyl)oxymethylphenyl]carbamoyl]-L-glutamate, 4a , and 70
(
µL (0.50 mmol) of triethylamine with stirring at room tem-
perature. The reaction was complete after 3.5 h. The reaction
mixture was evaporated under vacuum, EtOAc (5 × 5 mL) and
×
mobile phase of 35 mM ammonium acetate buffer, pH 5.0/
MeOH, 30:70. Melting points were determined on a Kofler hot-
stage melting point apparatus (Reichert Thermovar) and are
uncorrected. Low-resolution FAB mass spectra were performed
on a VG-2AB-SE double-focusing magnetic sector mass spec-
trometer (Fisons Instruments, Warrington, Manchester, U.K.),
operating at a resolution of 1000. High-resolution accurate
mass spectra were determined on the same system, but with
the resolution set to 8 000-10 000. Reported spectra are by
FAB unless otherwise stated. NMR spectra were determined
CH
2
Cl
2
(2 × 5 mL) were added, and the solution was evapo-
rated again. A dark-red solid resulted which was purified by
preparative TLC using first THF as eluent (5-6-cm migration)
followed by EtOAc:cyclohexane (2:1) (for another 10 cm). After
extraction of the compound in THF, 73 mg (74.1%) of a red
-1
solid resulted: mp 142-4 °C; νmax cm (film) 3358 (NH, broad),
1
1
3
2
736 (CdO, ester), 1717 (CdO, ketone); H NMR δ
H, (CH 5′, J ) 6.4 Hz), 2.15 (m, 2H, H ), 2.26 (s, 3H, (CH
.43 (m, 2H, CH CO A), 2.88 (s, 2H, H ,), 3.28-3.30 (m, 1H,
4′), 3.99 (s, 3H, OCH ), 4.27-4.35 (m, 1H, -CH(NH)-), 4.54
O-allyl, J ) 5.3 Hz), 4.59 (d, 2H, CH O-allyl), 4.86
-Ph), 4.96 (t, 1H, H10), 5.13-5.34 (m, 4H, CH
H
1.12 (d,
3
)
9
3
)14),
2
2
7
H
3
in Me
2 6
SO-d on a Brucker AC250 spectrometer (250 MHz) at
3
0 °C (303 K) unless otherwise stated. For proton assignment
(d, 2H, CH
(s, 2H, CH
2
2
a numbered formula is given in Scheme 2. The assignment of
protons is partial for the prodrugs. IR spectra (film) were
recorded on a Perkin-Elmer 1720X FTIR spectrometer. El-
emental analyses were determined by Butterworth Laborato-
ries Ltd. (Teddington, Middlesex, U.K.) and are within 0.4%
of theory except when stated.
2
2
d
allyl), 5.82-5.99 (m, 2H, CHdallyl), 6.58 (d, 1H, NH-G, J )
8.1 Hz), 6.84 (d, 1H, NH-dauno, J ) 8.0 Hz), 7.17 (d, 2H, H3′′+5′′
J ) 8.3 Hz), 7.32 (d, 2H, H2′′+6′′), 7.63-7.67 (m, 1H, H ), 7.91-
7.95 (m, 2H, H +H ), 8.58 (s, 1H, NH-Ph); MS m/z 930 (M
+ 1, 2), 952 (M + 23, 62); mass (C47
52.3116, found 952.3140.
The same procedure was used to obtain N-[4-(d ip r op en -
2-yl-L-glu ta m ylca r bon yloxy)ben zyloxyca r bon yl]d a u n o-
,
3
+
2
4
+
51 3
H N O17Na) calcd
9
Dip r op -2-en yl N-[(4-{Hyd r oxym eth yl}p h en yl)ca r ba m -
oyl]-L-glu ta m a te (3a ). To a solution of diallyl glutamyl
isocyanate (2) (25.0 mmol) in 100 mL of THF were added
5
5
-1
4
-aminobenzyl alcohol (3.0 g, 24.3 mmol) and triethylamine
3.41 mL, 24.3 mmol) in 20 mL of toluene, dropwise, over 10
min, at room temperature. The reaction was complete within
5 min. The reaction mixture was filtered and evaporated to
dryness; the residue was dissolved in 20 mL of EtOAc, washed
r u bicin (9) from 4b: yield 92%; mp 99-101 °C; νmax cm
(
(film) 3349 (NH, broad), 1736 (CdO, ester), 1719 (CdO,
1
ketone); H NMR δ
2.00 (m, 1H, CH CH(NH)-), 2.15 (m, 2H, H
(CH 14), 2.47-2.50 (m, 2H, CH CO A), 2.96 (s, 2H, H
H
1.13 (d, 3H, (CH
3
)
5′, J ) 6.4 Hz), 1.90-
), 2.26 (s, 3H,
). 3.26-
1
2
9
3
)
2
2
7

2
488 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 13
Brief Articles
3
.29 (m 1H, H4′), 3.99 (s, 3H, OCH
3
), 4.15-4.20 (m, 1H,
O-allyl, J ) 5.2 Hz), 4.60 (d, 2H,
-Ph+H10), 5.18-5.34 (m, 4H,
2H, H
7
), 3.99 (s, 3H, OCH
3
), 4.94 (s, 3H, CH
2
-Ph+H10), 6.85
-CH(NH)-), 4.55 (d, 2H, CH
2
(d, 1H, NH-doxo, J ) 7.9 Hz), 7.05 (d, 2H, H3′′+5′′, J ) 8.3 Hz),
CH
CH
2
O-allyl), 4.95 (s, 3H, CH
2
7.31 (d, 2H, H2′′+6′′), 7.63-7.67 (m, 2H, H
3
+NH-G), 7.90-7.93
+
-
2
dallyl), 5.82-5.99 (m, 2H, CHdallyl), 6.84 (d, 1H, NH-
(m, 2H, H
- H, 58); mass (C41
2
+H
4
); MS m/z 889 (M + 23, 25), ESI, m/z 865 (M
dauno, J ) 8.0 Hz), 7.05 (d, 2H, H3′′+5′′, J ) 8.2 Hz), 7.32 (d,
2
8
42 2
H N O19Na) calcd 889.2279, found 889.2287;
H, H2′′+6′′), 7.64-7.68 (m, 1H, H
.25 (d, 1H, NH-G, J ) 8.0 Hz); MS m/z 953 (M + 23, 100);
18Na) calcd 953.2956, found 953.2940.
3
), 7.91-7.94 (m, 2H, H
2
+H
4
),
purity (HPLC) 91.8% (280 nm); 94.6% (495 nm).
+
Deter m in a tion of Ha lf-Lives. The half-lives of the pro-
drugs were determined by HPLC in 100 µM solutions in
phosphate-buffered saline at 37 °C in the presence or absence
of 50 mU of CPG2. Samples of solutions (10 µM) were injected
at timed intervals onto a reverse-phase HPLC column (Par-
tisphere ODS, 5 mm, 4.6 × 250 mm), eluted isocratically with
75% methanol:15% water:10% 10 mM NaH PO (pH 5.0) at 1
mass (C47
50 2
H N O
N-[4-(Dip r op en -2-yl-L-glu ta m ylca r ba m oyla m in o)ben -
zyloxyca r bon yl]d oxor u bicin (11): yield 91%; mp 145-6 °C;
1
H NMR δ
H, CH
8.2 Hz), 3.45 (m, 1H, H4′), 3.99 (s, 3H, OCH
m, 6H, (CH O-allyl), 4.86 (s, 2H, CH -Ph), 4.96 (t,
H, H10), 5.19-5.36 (m, 4H, CH dallyl), 5.88-6.00 (m, 2H,
H
1.12 (d, 3H, (CH
CH(NH)-), 2.18 (d, 2H, H
3
)
5′, J ) 6.4 Hz), 1.89-2.03 (m,
), 2.44 (t, 2H, CH CO A, J
), 4.53-4.56
2
)
(
2
9
2
2
3
2
4
)
14+CH
2
2
mL‚min-1 and the eluate continuously monitored at 500 nm.
2
1
2
The chemical half-lives were determined from the slope of
semilog plots. In the presence of the enzyme, straight-line plots
were obtained on linear axes, and the apparent half-life was
calculated as the time it took for the concentration to fall to
one-half its starting value.
CHdallyl), 6.59 (d, 1H, NH-G, J ) 9.4 Hz), 6.84 (d, 1H, NH-
doxo, J ) 8.0 Hz), 7.18 (d, 2H, H3′′+5′′, J ) 8.4 Hz), 7.32 (d, 2H,
H
2′′+6′′), 7.63-7.67 (m, 1H, H
3
), 7.91-7.95 (m, 2H, H
2
+
4
+H ), 8.58
+
(
s, 1H, NH-Ph); MS m/z 947 (M + 1, 5), 969 (M + 23, 32);
mass (C47 18Na) calcd 968.3065, found 968.3100.
51 3
H N O
P r ep a r a tion of Histid in e-Ta gged CP G2*. Polyhistidine-
tagged CPG2* was expressed in Sf9 insect cells, and the
N-[4-(Dip r op en -2-yl-L-glu t a m ylca r b on yloxy)b en zyl-
oxyca r bon yl]d oxor u bicin (13): yield 84%; mp 114-6 °C;
protein was purified by nickle-agarose (Quingen) affinity
-
1
1
ν
max cm (film) 3371 (NH, OH, broad), 1728 (CdO, ester); H
chromatography, as described.2
NMR δ
CH CH(NH)-), 2.47-2.52 (m, 2H, CH
OCH ), 4.49-4.61 (m, 6H, (CH
CH -Ph+H10), 5.18-5.37 (m, 4H, CH
H
1.13 (d, 3H, (CH
3
)
5′, J ) 6.4 Hz), 1.76-2.14 (m, 2H,
CO A), 3.99 (s, 3H,
O-allyl), 4.95 (sb, 3H,
dallyl), 5.88-6.00 (m,
Biologica l Meth od s: Cytotoxicity Assa ys. MDA MB 361
cells stably expressing either surface-tethered CPG2 (stCPG2-
2
2
2
3
2
)
14+ CH
2
(Q)3) or an intracellularly located CPG2 (CPG2*) were used;
2
2
9
control cells expressed â-galactosidase (LacZ).
The compounds were dissolved in DMSO at 10 mM im-
mediately prior to treatment, and the cytotoxicity assays were
performed as described, except that after the treatment cells
were harvested and reseeded in quadruplicate in 96-well plates
2
2
H, CHdallyl), 6.85 (d, 1H, NH-doxo, J ) 9.0 Hz), 7.05 (d,
H, H3′′+5′′, J ) 8.4 Hz), 7.32 (d, 2H, H2′′+6′′), 7.65-7.68 (m, 1H,
H
3
), 7.91-7.93 (m, 2H, H
2
+H
Hz); MS m/z 929 (M + 1, 25); mass (C47
69.2905, found 969.2940.
N-[4-(L-Glu ta m ylca r bon yla m in o)ben zyloxyca r bon yl]-
d a u n or u bicin (8). Compound 7 (65 mg, 0.07 mmol) was
dissolved in 5 mL of CH Cl at room temperature, and 8.1 mg
0.007 mmol) of tetrakis(triphenylphosphine)Pd(0) and 12 µL
0.14 mmol) of morpholine were added with stirring. After 3 h
4
), 8.58 (s, 1H, NH-G, J ) 7.8
2
+
50 2
H N O
19Na) calcd
9
3
at ∼2 × 10 cells/well. When the control plates has reached
confluence, the cells were fixed and stained with sulfo-
rhodamine-B.35 The absorbance at 590 nm was determined,
and the results are expressed as percentage of control growth,
with IC50 (concentration required to give half-maximal cyto-
toxicity) values being determined by interpolation. Each
determination was performed at least three times.
2
2
(
(
a red solid precipitated. Filtration, washing (2-3 mL of CH
Cl ), and drying gave 26 mg (40%) of a red solid: mp 176-8
2
-
2
1
°
(
(
C; H NMR δ
CH 14), 3.99 (s, 3H, OCH
s, 2H, CH -Ph), 4.96 (t, 1H, H10), 6.45 (d, 1H, NH-G, J ) 6.7
H
1.13 (d, 3H, (CH
3
)5′, J ) 6.5 Hz), 2.26 (s, 3H,
)
3
), 4.16 (q, 1H, H5′, J ) 6.1 Hz), 4.84
Ack n ow led gm en t. These studies were supported by
The Cancer Research Campaign and the Institute of
Cancer Research. The authors are grateful to Mr. L.
Griggs for analytical HPLC, to Ms. Colleen McGahan
for performing the analysis of variance, and to Dr. M.
Cocksedge for the mass spectra. We are also indebted
to Prof. P. Workman and Prof. K. Harrap for support.
3
2
Hz), 6.74 (d, 1H, NH-dauno, J ) 7.9 Hz), 7.15 (d, 2H, H3′′+5′′
J ) 8.5 Hz), 7.32 (d, 2H, H2′′+6′′), 7.63-7.67 (m, 1H, H +NH-
), 8.86 (s, 1H, NH-Ph); MS m/z
17Na) calcd 872.2485, found
,
3
G), 7.90-7.93 (m, 2H, H
2
+H
4
+
8
8
72 (M + 23, 70); mass (C41H N O
43 3
72.2490; purity (HPLC) 98.7% (280 nm), 100% (495 nm).
The following compounds were obtained according to the
same route.
N-[4-(L-Glu ta m ylca r bon yloxy)ben zyloxyca r bon yl]d a u -
Refer en ces
1
n or u bicin (10): yield 57%; mp 135-7 °C; H NMR δ
H
1.13
14), 2.97 (s, 2H,
-Ph+H10), 6.86 (d,
(
(
(
(
1) Bridgewater, G.; Springer, C. J .; Knox, R.; Minton, N.; Michael,
P.; Collins, M. Expression of the bacterial nitroreductase enzyme
in mammalian cells renders them selectively sensitive to killing
by the prodrug CB1954. Eur. J . Cancer 1995, 31A, 2362-2370.
2) Marais, R.; Spooner, R. A.; Light, Y.; Martin, J .; Springer, C. J .
Gene-directed enzyme prodrug therapy with a mustard prodrug/
carboxypeptidase G2 combination. Cancer Res. 1996, 56, 4735-
(
H
1
7
(
d, 3H, (CH
3
)
5′, J ) 6.4 Hz), 2.26 (s, 3H, (CH
3
)
7
), 3.99 (s, 3H, OCH
3
), 4.94 (s, 3H, CH
2
H, NH-dauno, J ) 7.9 Hz), 7.05 (d, 2H, H3′′+5′′, J ) 8.3 Hz),
.31 (d, 2H, H2′′+6′′), 7.63-7.67 (m, 2H, H
3
+NH-G), 7.90-7.93
+
m, 2H, H
2
+H
4
); MS m/z 873 (M + 23, 100); mass (C41
H
42
N
2
O
18
-
Na) calcd 873.2330, found 873.2350; purity (HPLC) 94.9% (280
nm), 96.9% (495 nm).
4
742.
3) Huber, B. A.; Richards, C. A.; Krenitsky, T. A. Retroviral-
mediated gene therapy for the treatment of hepatocellular
carcinoma: An innovative approach for cancer therapy. Proc.
Natl. Acad. Sci. U.S.A. 1991, 88, 8039-8043.
4) Niculescu-Duvaz, I.; Spooner, R. A.; Marais, R.; Springer, C. J .
S. Gene-directed enzyme prodrug therapy. Bioconjugate Chem.
1998, 9, 4-22.
N-[4-(L-Glu ta m ylca r bon yla m in o)ben zyloxyca r bon yl]-
1
d oxor u bicin (12): yield 50%; mp 182-4 °C; H NMR δ
H
1.12
),
(
3
d, 3H, (CH
3
)
5′, J ) 6.4 Hz), 2.18 (m, 2H, H
9
), 2.99 (s, 2H, H
), 4.57 (s, 2H, (CH 14), 4.84 (s, 2H, CH
Ph), 4.96 (t, 1H, H10), 6.46 (d, 1H, NH-G, J ) 6.9 Hz), 6.76 (d,
7
.99 (s, 3H, OCH
3
2
)
2
-
1
7
(
H, NH-doxo, J ) 8.2 Hz), 7.15 (d, 2H, H3′′+5′′, J ) 8.6 Hz),
.31 (d, 2H, H2′′+6′′), 7.63-7.67 (m, 1H, H +NH-G), 7.91-7.93
+H ), 8.80 (s, 1H, NH-Ph); MS m/z 888 (M + 23,
5); mass (C41 18Na) calcd 888.2439, found 888.2410;
(5) Niculescu-Duvaz, D.; Niculescu-Duvaz, I.; Friedlos, F.; Martin,
J .; Spooner, R.; Davies, L.; Marais, R.; Springer, C. J . Self-
immolative mustard prodrugs for suicide gene therapy. J . Med.
Chem. 1998, 41, 5297-5309.
3
+
m, 2H, H
2
4
1
43 3
H N O
(
6) Springer, C. J .; Antoniw, P.; Bagshawe, K. D.; Searle, F.; Bisset,
G. M. F.; J arman, M. Novel prodrugs which are activated to
cytotoxic alkylating agents by carboxypeptidase G2. J . Med.
Chem. 1990, 33, 677-681.
7) Dowell, R.; Springer, C. J .; Davies, D. H.; Hadley, E. M.; Burke,
P. J .; Boyle, F. T.; Melton, R. G.; Connors, T. A.; Blakey, D. C.;
Mauger, A. B. New mustards prodrugs for antibody-directed
enzyme prodrug therapy: alternative for the amide link. J . Med.
Chem. 1996, 39, 1100-1105.
purity (HPLC): 90.1% (280 nm); 94.2% (495 nm).
The same procedure as described above was used to prepare
N-[4-(L-glu ta m ylca r bon yloxy)ben zylm eth yloxyca r bon yl]-
d oxor u bicin (14), with the difference that sodium toluene-
sulfinate (22.6 mg, 0.13 mmol) instead of morpholine was
(
1
added to compound 13 (40 mg, 0.04 mmol): yield 22%; H NMR
δ
H
1.12 (d, 3H, (CH
3
9
)5′, J ) 6.4 Hz), 2.18 (m, 2H, H ), 2.98 (s,

Brief Articles
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 13 2489
(
8) Springer, C. J .; Dowell, R. L.; Burke, P. J .; Hadley, E.; Davies,
D. H.; Blakey, D. C.; Melton, R. G.; Niculescu-Duvaz, I. Opti-
mization of alkylating prodrugs derived from phenol and aniline
(23) Svensson, H. P.; Vrudhula, V. M.; Emswiler, J . E.; MacMaster,
J . F.; Cosand, W. L.; Senter, P. D.; Wallace, P. M. In vitro and
in vivo activities of a doxorubicin prodrug in combination with
monoclonal antibody â-lactamase conjugates. Cancer Res. 1995,
55, 2357-2365.
(24) Vrudhula, V. M.; Svensson, H. P.; Senter, P. D. Cephalosporin
derivatives of doxorubicin as prodrugs for activation by mono-
clonal antibody-â-lactamase conjugates. J . Med. Chem. 1995, 38,
1380-1385.
(25) Meyer, D. L.; Law, K. L.; Payne, J . K.; Mikolajczyk, S. D.;
Zarrinmayeh, H.; J ungheim, L. N.; Kling, J . K.; Shepherd, T.
A.; Starling, J . J . Site-specific prodrug activation by antibody-
â-lactamase conjugates: Preclinical investigation of the efficacy
and toxicity of doxorubicin delivered by antibody directed
catalysis. Bioconjugate Chem. 1995, 6, 440-446.
mustards:
a new clinical candidate prodrug (ZD2767) for
ADEPT. J . Med. Chem. 1995, 38, 5051-5065.
(
9) Marais, R.; Spooner, R. A.; Stribbling, S. M.; Light, Y.; Martin,
J .; Springer, C. J . A cell surface tethered enzyme improves
efficiency in gene-directed enzyme prodrug therapy. Nature
Biotechnol. 1997, 15, 1373-1377.
(
10) Weiss, R. B.; Savory, G.; Clagget-Carr, K.; Russo, M.; Leyland-
J ones, B. Anthracyclines analogues: past, present and future.
Cancer Chemother. Pharmacol. 1986, 18, 185-197.
11) Grever, R.; Chabner, B. A. Cancer drug discovery and develop-
ment, 5th ed.; Lipincott-Raven: New York, 1997; Vol. I.
12) Senter, P. D. Activation of prodrugs by antibody-enzyme conju-
gates: A new approach to cancer therapy. FASEB J . 1990, 4,
(
(
(
26) Senter, P. D.; Svensson, H. P.; Schreiber, G. J .; Rodriguez, G.
L.; Vrudhula, V. M. Poly(ethylene glycol)-doxorubicin conjugates
containing â-lactamase-sensitive linkers. Bioconjugate Chem.
1
88-193.
(
13) Andrianomenjanahary, S.; Dong, X.; Florent, J . C.; Gaudel, G.;
Gesson, J . P.; J acquesy, J . C.; Koch, M.; Michel, S.; Mondon,
M.; Monneret, C.; Petit, P.; Renoux, B.; Tillequin, F. Synthesis
of novel targeted pro-drugs of anthracyclines potentially acti-
vated by a monoclonal antibody galactosidase conjugate (Part
1
995, 6, 389-394.
(
27) Kerr, D. E.; Schreiber, G. J .; Vrudhula, V. M.; Svensson, H. P.;
Hellstrom, I.; Hellstrom, K. E.; Senter, P. D. Regressions and
cures of melanoma xenografts following treatment with mono-
clonal antibody â-lactamase conjugates in combination with
anticancer prodrugs. Cancer Res. 1995, 55, 3558-3563.
1
). Bioorg. Med. Chem. Lett. 1992, 2, 1093-1096.
(
14) Gesson, J . P.; J acquesy, J . C.; Mondon, M.; Petit, P.; Renoux,
B.; Andrianomenjanahary, S.; Dufat-Trinh Van, H.; Koch, M.;
Michel, S.; Tillequin, F.; Florent, J . C.; Monneret, C.; Bosslet,
K.; Czech, J .; Hoffmann, D. Prodrugs of anthracyclines for
chemotherapy via enzyme-monoclonal antibody conjugates. Anti-
Cancer Drug Des. 1994, 9, 409-423.
15) Azoulay, M.; Florent, J . C.; Monneret, C.; Gesson, G. P.;
J acquesy, G. P.; Tillequin, F.; Koch, M.; Bosslet, K.; Czech, J .;
Hoffman, D. Prodrugs of anthracycline antibiotics suited for
tumor-specific activation. Anti-Cancer Drug Des. 1995, 10, 441-
(
28) Rodrigues, M. L.; Presta, L. G.; Kotts, C. E.; Wirth, C.; Mordenti,
C.; Osaka, G.; Wong, W. T.; Nuijens, A.; Blackburn, B.; Carter,
P. Development of a humanized disulfide-stabilized anti-
p185HER Fv-â-lactamase fusion protein for activation of a
cephalosporin doxorubicin prodrug. Cancer Res. 1995, 55, 63-
(
(
(
7
0.
(
29) Kerr, D. E.; Senter, P. D.; Burnett, P. D.; Hirschberg, D. L.;
Hellstrom, I.; Hellstrom, K. E. Antibody-penicillin-V-amidase
conjugates kill antigen-positive tumor cells when combined with
doxorubicin phenoxyacetamide. Cancer Immunol. Immunother.
4
50.
16) Haisma, H. J .; Boven, E.; van Muijen, M.; de J ong, J .; van der
Vygh, W. J . F.; Pinedo, H. M. A monoclonal antibody-â-
glucuronidase conjugate as activator of the prodrug epirubicin-
glucuronide for the specific treatment of cancer. Br. J . Cancer
1
990, 31, 202-206.
(
30) Vrudhula, V. M.; Senter, P. D.; Fischer, K. J .; Wallace, P. M.
Prodrugs of doxorubicin and melphalan and their activation by
a monoclonal antibody-penicillin-G-amidase conjugate. J . Med.
Chem. 1993, 36, 919-923.
31) Wakselman, M. 1,4 and 1,6 eliminations from hydroxy- and
amino-substituted benzyl systems: chemical and biochemical
applications. N. J . Chim. 1983, 7, 439-447.
32) Carl, P. L.; Chakravarty, P. K.; Katzenellenbogen, J . A. A novel
connector linkage applicable in prodrug design. J . Med. Chem.
1981, 24, 479-480.
1
992, 66, 474-478.
17) Haisma, H. J .; Van Muijen, M.; Pinedo, H. M.; Boven, E.
Comparison of two anthracycline-based produgs for activation
by a monoclonal antibody-â-glucuronidase conjugate in specific
treatment of cancer. Cell Biophys. 1994, 24/25, 185-192.
18) Bosslet, K.; Czech, J .; Hoffmann, D. Tumor-selective prodrug
activation by fusion protein-mediated catalysis. Cancer Res.
(
(
(
1
994, 54, 2151-2159.
(
19) Leenders, R. G.; Scheren, H. W. Synthesis and evaluation of
novel daunomycin-phosphate, sulfate, â-glucuronide and â-glu-
coside prodrugs for application in ADEPT. Bioorg. Med. Chem.
Lett. 1995, 5, 2975-2980.
(33) Senter, P. D.; Pearce, W. E.; Greenfield, R. S. Development of a
drug-release strategy based on the reductive fragmentation of
benzyl carbamate disulfides. J . Org. Chem. 1990, 55, 2975-2978.
(34) Honda, M.; Morita, H.; Nagakura, I. Deprotection of allyl groups
with sulfinic acids and palladium catalyst. J . Org. Chem. 1997,
62, 8932-8936.
(35) Mistry, P.; Kelland, L. R.; Abel, G.; Sidhar, S.; Harrap, K. R.
The relationships between glutathione-S-transferase and cyto-
toxicity of platinum drugs and melphalan in eight human
ovarian carcinoma cell lines. Br. J . Cancer 1991, 64, 215-220.
(
(
(
20) Bakina, E.; Zheng, W.; Rosenblum, M.; Farquhar, D. Intensely
cytotoxic anthracycline prodrugs: Glucuronides. J . Med. Chem.
1
997, 40, 4013-4018.
21) Farquhar, D.; Cherif, A.; Bakina, E.; Nelson, J . A. Intensely
potent doxorubicin analogues: Structure-activity relationship.
J . Med. Chem. 1998, 41, 965-972.
22) J ungheim, L. N.; Shepherd, T. A.; Kling, J . K. Synthesis of a
cephalosporin-doxorubicin antitumor prodrug: A substrate for
an antibody-targeted enzyme. Heterocycles 1993, 35, 339-348.
J M980696V