Triazene drug metabolites. Part 17: Synthesis and plasma hydrolysis of acyloxymethyl carbamate derivatives of antitumour triazenes

Article abstract of DOI:10.1016/S0968-0896(00)00100-0

A series of 3-acyloxymethyloxycarbonyl-1-aryl-3-methyltriazenes 5 was synthesised by the sequential reaction of 1-aryl-3-methyltriazenes with (i) chloromethyl chloroformate, (ii) NaI in dry acetone, and (iii) either the silver carboxylate or the carboxylic acids in the presence of silver carbonate. The hydrolysis of these compounds was studied in pH 7.7 isotonic phosphate buffer and in human plasma. Triazene acyloxycarbamates demonstrated their ability to act as substrates for plasma enzymes. For compound 5f, a pH-rate profile was obtained which showed the hydrolysis to involve acid-base catalysis. The reaction is also buffer catalysed. Thus, at pH 7.7, pH-independent, base-catalysed and buffer-catalysed processes all contribute to the hydrolysis reaction. The sensitivity of the hydrolysis reaction to various structural parameters in the substrates indicates that hydrolysis occurs at the ester rather than the carbamate functionality. In plasma, the rates of hydrolysis correlate with partition coefficients, the most lipophilic compounds being the most stable. An aspirin derivative suffers two consecutive enzymatic reactions, the scission of the aspirin acetyl group being followed by the scission of the acyloxy ester group. These results indicate that triazene acyloxymethyl carbamates are prodrugs of the antitumour monomethyltriazenes. They combine chemical stability with a rapid enzymatic hydrolysis, and are consequently good candidates for further prodrug development. Moreover, this type of derivative allowed the synthesis of mutual prodrugs, associating the antitumour monomethyltriazenes with anti-inflammatory NSAIDs as well as with the anticancer agent butyric acid. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(00)00100-0

Bioorganic & Medicinal Chemistry 8 (2000) 1719±1725
Triazene Drug Metabolites.
Part 17: Synthesis and Plasma Hydrolysis of Acyloxymethyl
Carbamate Derivatives of Antitumour Triazenes
Emõlia Carvalho,a,* Ana Paula Francisco,a Jim Ileyb and Eduarda Rosaa,*
a
Â
Centro de Estudos de CieÃncias FarmaceÃuticas (CECF), Faculdade de Farmacia, Universidade de Lisboa,
Avenida das ForcËas Armadas, 1600 Lisbon, Portugal
bChemistry Department, The Open University, Milton Keynes, MK7 6AA, UK
Received 4 December 1999; accepted 15 March 2000
AbstractÐA series of 3-acyloxymethyloxycarbonyl-1-aryl-3-methyltriazenes 5 was synthesised by the sequential reaction of 1-aryl-
3-methyltriazenes with (i) chloromethyl chloroformate, (ii) NaI in dry acetone, and (iii) either the silver carboxylate or the car-
boxylic acids in the presence of silver carbonate. The hydrolysis of these compounds was studied in pH 7.7 isotonic phosphate
bu€er and in human plasma. Triazene acyloxycarbamates demonstrated their ability to act as substrates for plasma enzymes. For
compound 5f, a pH±rate pro®le was obtained which showed the hydrolysis to involve acid±base catalysis. The reaction is also bu€er
catalysed. Thus, at pH 7.7, pH-independent, base-catalysed and bu€er-catalysed processes all contribute to the hydrolysis reaction.
The sensitivity of the hydrolysis reaction to various structural parameters in the substrates indicates that hydrolysis occurs at the
ester rather than the carbamate functionality. In plasma, the rates of hydrolysis correlate with partition coecients, the most lipo-
philic compounds being the most stable. An aspirin derivative su€ers two consecutive enzymatic reactions, the scission of the
aspirin acetyl group being followed by the scission of the acyloxy ester group. These results indicate that triazene acyloxymethyl
carbamates are prodrugs of the antitumour monomethyltriazenes. They combine chemical stability with a rapid enzymatic hydro-
lysis, and are consequently good candidates for further prodrug development. Moreover, this type of derivative allowed the
synthesis of mutual prodrugs, associating the antitumour monomethyltriazenes with anti-in¯ammatory NSAIDs as well as with the
anticancer agent butyric acid. # 2000 Elsevier Science Ltd. All rights reserved.
Introduction
of 3 and 4 have been synthesised, such as esters,5 ethers,6
thioethers,7 and acyl derivatives. 8,9 None of these possess
ideal prodrug characteristics. More recently, an
imidazotetrazine Ð temozolomide Ð has been found to
be active against several tumours in phase I clinical
studies. Its activity is thought to be due to the genera-
tion of an alkylating monomethyltriazene by reaction
with water.10 Anticancer triazenes containing amino
acid carriers have also been studied.11,12 One such com-
pound, an N-acetylalanyl derivative, displayed chemical
stability in isotonic phosphate bu€er yet rapid hydro-
lysis to the monomethyltriazene in plasma, indicating it
to be a promising candidate for prodrug development.
For many years dacarbazine 1 (Fig. 1) has been one of the
most widely used drugs to treat malignant melanoma. The
incidence of this disease has increased recently and is
expected to continue in the near future.1
The biological action of dacarbazine and, in general, the
anticancer 1-aryl-3,3-dimethyltriazenes (2, in Fig. 2) is a
consequence of their capacity to alkylate DNA. These
compounds su€er metabolic oxidation by cytochrome
P450 enzymes to give hydroxymethyltriazenes 3, which,
by loss of formaldehyde, generate the cytotoxic mono-
methyltriazenes 4. These are known alkylating agents,
capable of methylating DNA and RNA (Fig. 2).2,3 With
the objective of ®nding suitable prodrugs that by-pass
the need for oxidative metabolism,4 various derivatives
As part of our ongoing work directed toward the
development of triazene prodrugs, we have extended
our investigations to encompass the acyloxymethyl car-
bamates 5 (Fig. 3). The acyloxymethyl carbamate func-
tionality has been proposed as a promoiety for primary
and secondary amines. Such compounds su€er esterase-
catalysed hydrolysis of the ester function, producing an
*Corresponding author. Tel.: +351-217946477; fax: +351-21794670;
e-mail: merosa@€.ul.pt
0968-0896/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00100-0

1720
E. Carvalho et al. / Bioorg. Med. Chem. 8 (2000) 1719±1725
carbonyl group. Compounds 5 display the typical tri-
1
azene N-methyl singlet in the H NMR spectra, ranging
from d 3.48 to 3.53 ppm. Compounds 5a±e and 5g±i
exhibit an OCH2O methylene singlet ranging from 5.99
to 6.27 ppm. In compounds 5f and 5j, the OCH2O pro-
tons are diastereotopic and the signal for these appears
as two doublets with J=5.6 Hz.
Figure 1. Structure of dacarbazine.
Kinetic studies
The hydrolysis of the triazene carbamate prodrugs 5 to
the corresponding 1-aryl-3-methyltriazenes was studied
both in isotonic phosphate bu€er and in 80% human
plasma containing 20% isotonic phosphate bu€er. The
hydrolysis of 5f was also studied across the pH range
1±9 in aqueous bu€ers using several bu€er concentrations
at each pH value. Under these reaction conditions the
monomethyltriazenes are also unstable and hydrolyse
further to the corresponding anilines. These reactions
were easily monitored by HPLC, following the loss of
starting material.
Half-lives for the hydrolysis of triazenes 5 in isotonic
phosphate bu€er are given in Table 1. Inspection of
Table 1 shows that they decompose at physiological pH
with half-lives ranging from 80 min to 7 days. The most
reactive of these, 5f, was studied in more detail, and
pseudo-®rst-order rate constants for its hydrolysis were
determined in aqueous bu€ers (using several bu€er
concentrations at each pH) and in HCl solutions. The
values of kobs obtained are dependent upon both the pH
of the solution and the bu€er concentration. Using the
intercepts of plots of kobs versus bu€er concentration,
together with kobs values determined in HCl solutions, a
pH±rate pro®le was constructed (Fig. 4). This shows
that there are regions of acid- and base-catalysis as well
as a pH-independent region between pH 2 and 5 across
which the compound is most stable. The bimolecular
rate constants for bu€er catalysis, kB, were determined
using the slopes of plots of kobs versus bu€er concentration
(data not shown). Figure 5 shows the correlation between
Figure 2. Oxidative demethylation pathway for 1-aryl-3,3-dimethyl-
triazenes.
unstable hydroxymethyl carbamate intermediate that
liberates the amine drug and a carboxylic acid.13,14 To
assess the potential of compounds 5 as bioreversible
prodrugs of the monomethyltriazenes 4, we report the
synthesis of some novel 3-(acyloxymethyloxycarbonyl)-
1-aryl-3-methyltriazenes 5a±j and an examination of
their stability both in isotonic pH 7.7 phosphate bu€er
and plasma. The compounds synthesised use as carriers
simple non-toxic carboxylic acids like acetic acid and N-
acetylalanine, or other drugs containing a carboxylic
acid group, forming mutual prodrugs.
0
kB and pKa (the pKa values corrected for ionic strength15)
which is linear (r2=0.95) and gives a Brùnsted b value
of 0.5. The pH±rate pro®le, together with the Brùnsted
plot for 5f, reveals that at pH 7.7 the hydrolysis reaction
comprises three components: pH-independent, base-
catalysed and bu€er-catalysed reactions. The major
component appears to be the base-catalysed reaction.
Thus, conclusions about the factors a€ecting the relative
magnitudes of the t1/2 values in phosphate bu€er for the
other compounds in Table 1 must be made with caution.
Nevertheless, certain features are apparent. Comparison
of 5a±d reveals that the rate of reaction is only very
slightly in¯uenced by the nature of the substituent -X in
the triazene aryl group, the more reactive compounds
being those containing electron-attracting substituents.
However, compared to the hydrolysis of acyltriazenes8,11
this e€ect is very small. In contrast, comparison of
compounds 5b,e,g reveals the carboxylic acid moiety to
have a much greater in¯uence on reactivity,16 the relative
di€erences in rates paralleling the known di€erences in
reactivity for the hydrolysis of ethyl esters.17 These re¯ect
the electronic and steric e€ects on the rate. Indeed, the
Results and Discussion
Chemical synthesis
Synthesis of the desired compounds 5 proceeds via the 1-
aryl-3-chloromethyloxycarbonyl-3-methyltriazenes 6 and
the corresponding iodomethyl derivatives 7 (Fig. 3). The
triazenes 6 themselves are eciently synthesised by
reaction of 1-aryl-3-methyltriazenes with chloromethyl
chloroformate. Exchange of the Cl atom using NaI
yields the corresponding 1-aryl-3-iodomethyloxycarbo-
nyl-3-methyltriazenes 7, which react in acetone in the
presence of a tetrabutylammonium salt with either a
silver carboxylate or its equivalent a carboxylic acid/
silver carbonate. The nature of the solvent appears
important for this latter reaction; in the synthesis of 5e,
when using tetrahydrofuran, we isolated the anhydride
product of competitive attack of the nucleophile at the

E. Carvalho et al. / Bioorg. Med. Chem. 8 (2000) 1719±1725
1721
Figure 3. Synthesis and structures of the triazenes 5. Reagents: (i) ClCH2OCOCl; (ii) NaI; (iii) RCO2Ag or RCO2H/Ag2CO3.
Table 1. Half-lives, t1/2, in both pH 7.7 isotonic phosphate bu€er and
80% human plasma at 37 ꢀC and log P values for the triazenes 5a
around the ester carbonyl.16 The salicylate derivatives
5h,i have similar reactivity to 5g, re¯ecting their similar
structures. These observations help to identify which of
the two functional groupsÐthe ester or the triazene
carbamateÐis hydrolysed ®rst. In general, esters are
more reactive toward hydrolysis than carbamates.
However, we have observed previously that N-acyl-
triazenes have a reactivity similar to that of esters.8,9
Therefore, triazene derived carbamates may be antici-
pated to display enhanced reactivity over simple carba-
mates. Even so, the sensitivity of the reaction to the acyl
moiety and insensitivity to the structure of the triazene
moiety implies that these compounds su€er ester
hydrolysis. Moreover, the Brùnsted b value is typical of
ester hydrolysis.16,17
t1/2 (min)
Log Pexp
Log Pcalc
pH 7.7 bu€er
Plasma
5a
5b
5c
5d
5e
5f
5g
5h
5i
2.08Æ0.02
3.44Æ0.09
3.64Æ0.05
3.28Æ0.06
4.51Æ0.25
2.73Æ0.01
Ð
2.29Æ0.61
3.23Æ0.61
3.37Æ0.63
2.81Æ0.60
4.40Æ0.61
2.76Æ0.66
5.25Æ0.61
4.37Æ0.61
5.29Æ0.62
7.34Æ0.6
358Æ7
388Æ19
505Æ35
428Æ21
664Æ79
81Æ2
5Æ0.2
7Æ0.07
12Æ1.1
17Æ1.5
15Æ0.3
12Æ0.7
47Æ3.8
1Æ0.07c
34Æ0.7
125Æ6.25
>7500b
5622Æ56
6386Æ650
9870Æ1300
Ð
Ð
Ð
5j
aValues represent the means of at least 3 determinations ÆS.D.
bEstimated value. Due to the insolubility of the compound, results are
not precise.
Blood serum and plasma contain a range of enzymes that
catalyse the hydrolysis of esters and amides.18 Conse-
quently, we were interested in examining the eciency of
the hydrolysis of the acyloxymethyloxycarbonyltriazenes
5 to the cytotoxic monomethyltriazenes 4 in human
plasma. The results in Table 1 clearly show that all
compounds are substrates for the plasma enzymes.
Indeed, most liberate the monomethyltriazenes in a few
cHalf-life for the cleavage of the acetyl group in the aspirin moiety.
increased rate of 5f over 5b despite increased steric hin-
drance is almost certainly due to the known electron-
withdrawing ability of an NHAc group to enhance the
reactivity of an adjacent carbonyl group.12 Moreover, the
greater stability of 5j results from the steric hindrance

1722
E. Carvalho et al. / Bioorg. Med. Chem. 8 (2000) 1719±1725
Figure 6. Time course for the formation and decay of intermediates in
the plasma mediated cleavage of 5h: *, 5h; &, 5i; ~, 4; *, aniline
(X=COOEt).
Figure 4. pH±Rate pro®le for the hydrolysis of 5f; points are experi-
mental, line is best ®t using rate constants for the acid-, base- and
5
1
1
uncatalysed hydrolyses of 4Â10
M
s
1, 42.3 M 1 s and 1.55Â
6
10
s
1, respectively.
between prodrug stability (log t1/2) and lipophilicity (log
P) (eq 1). Elsewhere, it has been found that substrate
lipophilicity is one of the main features that in¯uences
esterase activity, decreasing enzyme activity being asso-
ciated with increasing lipophilicity.21 However, bulky
substituents in the substrate also in¯uence enzyme
activity. In the present case, the most lipophilic sub-
strates are also the most sterically hindered and these
e€ects cannot be separated.
minutes. The rather short half-life observed for the aspirin
derivative, 5h, refers to the very fast enzymatic hydrolysis
of the acetyl group which leads to liberation of the salicylic
acid derivative 5i (Fig. 6). As the reaction proceeds, the
formation of both monomethyltriazene and aniline is
also observed. However, due to the binding of 4 to plasma
proteins19,20 its quantitative measurement in serum proved
dicult and complete mass balance was not possible.
Á
n ˆ 9; r2 ˆ 0:89
ꢀ1†
There appears to be a super®cial correlation of the
plasma hydrolysis rates with those in pH 7.7 phosphate
bu€er despite the ca. 50-fold increase in the hydrolysis
rates in plasma. For example, in the series 5a±d elec-
tron-withdrawing groups in the aryl ring of the triazene
moiety exert a slight rate enhancement; for 5b,e,j,
increasing steric bulk near to the ester carbonyl decrea-
ses the rate, the sterically hindered 5j being the most
stable compound studied; and, for 5b,g the aroyl ester is
less reactive than its alkanoyl counterpart. However, the
enhanced chemical reactivity of 5f is suppressed in
plasma and, moreover, there is no apparent relationship
between the reactivities of 5b, 5e and 5f. Indeed, a plot
of t1/2 (plasma) versus t1/2 (bu€er) (not shown) reveals
no obvious correlation between the two.
log t1=2 ˆ 0:25log P ‡ 0:21
Conclusions
Dimethyltriazenes such as dacarbazine require meta-
bolic demethylation to the corresponding cytotoxic
monomethyltriazene by cytochrome P450 enzymes for
them to display anticancer activity.4 Indeed, there
appears to be a direct correlation between the species-
dependent extent of metabolism and anticancer activity,
with humans exhibiting only a low level of metabo-
lism.4 Moreover, the peak plasma concentration of the
active monomethyl metabolite is only ca. 0.6% of the
overall drug dose.4 Consequently, the current study
demonstrates that the acyloxymethyl carbamate deriva-
tives provide an alternative prodrug system for the
The most striking correlation that encompasses all of
the compounds studied (except 5i, vide supra) is
Figure 5. Brùnsted plot for the bu€er-catalysed hydrolysis of 5f.

E. Carvalho et al. / Bioorg. Med. Chem. 8 (2000) 1719±1725
1723
anticancer triazenes, one that by-passes the need for
oxidative demethylation and allows for extensive and
rapid transformation to the active metabolite. In addi-
tion, and in contrast to the dimethyltriazenes, a suitable
choice of the acyloxy moiety in the carbamate prodrugs
enables the half-life to be modi®ed by between one and
two orders of magnitude. The acyloxymethyl carbamate
derivatives of the present study exhibit much better
prodrug characteristicsÐchemical stability combined
with enzymatic activationÐthan many of the systems
studied previously. Thus, hydroxymethyltriazenes are
too unstable in both aqueous media and plasma,24 the
derived ethers are stable chemically but themselves
require oxidative metabolism to exert biological activ-
ity,25 the corresponding thioethers appear not to be
chemically or enzymically activated7 and simple acyl-
triazenes are stable chemically and enzymically.9 Only
the aminoacyltriazenes have similar properties though
they tend to have 4-fold larger plasma half-lives.11
However, one of the bene®ts of the current strategy over
the corresponding aminoacyl prodrug approach is that
it enables the development of mutual prodrugs. For
example, compounds 5h±j are mutual prodrugs of anti-
in¯ammatory agents which are often administered in
conjuction with antitumour alkylating agents. Of more
interest is compound 5e, which is a mutual prodrug of
butyric acid, a reported non-toxic natural product with
antitumour activity.22 Such prodrugs allow the possibi-
lity of simultaneously delivering two anticancer agents
that function by di€erent antitumour mechanisms. In
this work the problem of speci®city was not addressed.
Nonetheless, it is possible to conceive that this strategy
could include a promoiety that would confer site-selec-
tive cleavage. We are currently pursuing this strategy
and also trying to extend the range of mutual prodrugs.
General procedure for the synthesis of 1-aryl-3-chlorome-
thyloxycarbonyl-3-methyltriazenes (6). Chloromethyl
chloroformate (5.8 mmol) was added slowly to a cold
solution of the 1-aryl-3-methyltriazene (4.83 mmol), either
in pyridine (10 mL) (X=CN, COOEt) or in dichloro-
methane (5 mL) containing triethylamine (5.8mmol)
(X=Br, Me). For the reactions in pyridine, the solution
was stirred at room temperature, then poured onto ice
and the resultant precipitate was dried. For the reac-
tions in dichloromethane, the solution was evaporated
under reduced pressure. The resultant solid from either
method was recrystallised from diethyl ether/petroleum
ether.
3-Chloromethyloxycarbonyl-1-(4-cyanophenyl)-3-methyl-
triazene (6a). Yield 84%; mp 117±8 ꢀC; IR (KBr) 2224,
1
1761, 1136 cm 1; H NMR d 3.49 (3H, s, NCH3); 5.98
(2H, s, OCH2Cl); 7.76 (4H, s, Ar) ppm.
3-Chloromethyloxycarbonyl-1-(4-ethoxycarbonylphenyl)-
3-methyltriazene (6b). Yield 86%; mp 86±7 ꢀC; IR (KBr)
1719, 1741, 1278 cm 1; 1H NMR d 1.4 (3H, t, CH3); 3.5
(3H, s, NCH3); 4.35 (2H, q, CH2); 5.95 (2H, s, OCH2Cl);
7.65±7.68 and 8.05±8.08 (4H, AA0BB0, J=10 Ar) ppm.
1-(4-Bromophenyl)-3-chloromethyloxycarbonyl-3-methyl-
triazene (6c). Yield 82%; mp 101±3 ꢀC; IR (KBr) 1735,
1479, 1124 cm 1; 1H NMR d 3.5 (3H, s, NCH3); 6.0 (2H,
s, OCH2Cl); 7.6 (4H, s, Ar) ppm.
3-Chloromethyloxycarbonyl-3-methyl-1-(4-tolyl)triazene
(6d). Yield 86%; mp 78±81 ꢀC; IR (KBr) 1736, 1474, 1123
cm 1; 1H NMR d 2.45 (3H, s, CH3); 3.57 (3H, s, NCH3);
6.0 (2H, s, OCH2Cl); 7.27±7.73 (4H, AA0BB0, J=8, Ar)
ppm.
General procedure for the synthesis of 1-aryl-3-iodome-
thyloxycarbonyl-3-methyltriazenes (7). NaI (1.67 mmol)
was added to a solution of the 1-aryl-3-chloromethyl-
oxycarbonyl-3-methyltriazenes in dry acetone. The
mixture was stirred at room temperature until comple-
tion of the reaction (ca. 18 to 24 h) as monitored by the
disappearance of the CH2Cl, and appearance of the
Experimental
Melting points were determined in a Ko¯er camera
Bock-Monoscop ``M'' and are uncorrected. IR spectra
were recorded using a Perkin Elmer 1310 spectro-
photometer. 1H NMR spectra were recorded using
either a Jeol PMX-60 or a Jeol JNM-EX 400 spectro-
meter. Mass spectra were recorded using a VG Mass
Lab 20-250 spectrometer. High-performance liquid
chromatography (HPLC) was performed using a system
consisting of a Merck-Hitachi lachrom L-7110 pump, a
Spectra-Physics UV spectrachrom 100 detector, a Spec-
tra-Physics SP4270 integrator, a Merck-Hitachi AS-
2000 autosampler and a Lichrospher1 100 RP-8 (5 mm)
column. Elemental analyses were obtained from Medac
Ltd., Brunel Science Park, Engle®eld Green, Egham,
Surrey, UK.
1
CH2I, peak in the H NMR. The solvent was removed
under reduced pressure, the residue redissolved in di-
chloromethane, washed with saturated sodium bisul®te
solution, dried and evaporated under reduced pressure.
The resultant solid was used immediately without further
puri®cation.
General procedure for the synthesis of 3-acyloxymethyl-
oxycarbonyl-1-aryl-3-methyltriazenes (5). The appro-
priate carboxylic acid silver salt (8.42 mmol), prepared
by reaction of the sodium salt with silver nitrate, and
tetrabutylammonium bromide (8.42 mmol) were added
to a solution of 7 (8.42 mmol) in dry acetone (5 mL). The
mixture was re¯uxed for ca. 10 h, ®ltered and evapo-
rated under reduced pressure. The residue was subjected
to chromatography (silica gel 60 H 77331, Merck) using
petroleum ether/diethyl ether. The desired triazenes 5
were recrystallised from petroleum ether/diethyl ether.
For compounds 5f,i,j, the carboxylic acid and one
equivalent of silver carbonate were used in place of the
Plasma was derived from di€erent healthy individuals,
pooled, and kept at 18 ꢀC until required. Ibuprofen,
aspirin and salicylic acid were purchased commercially.
All other chemicals were reagent grade except those used
in kinetic and HPLC studies, which were analytical or
LiChrosolv (Merck) grade, respectively. 1-Aryl-3-methyl-
triazenes 4 were synthesised by previously described
methods.23

1724
E. Carvalho et al. / Bioorg. Med. Chem. 8 (2000) 1719±1725
silver carboxylate. In one case 5e we attempted the
above reaction in dry THF but isolated the anhydride
(ꢀmax 1736, 1709, 1697, 1272, 1051 cm 1; dH 1.03 (3H, t,
J=5Hz, Me), 1.27 (3H, t, J=6Hz, Me), 1.7 (2H, m,
CH2), 2.93 (2H, t, J=6Hz, CH2), 3.5 (3H, s, N-Me), 4.40
(2H, q, J=6 Hz, OCH2), 7.6±8.23 (4H, AA0BB0,
J=10 Hz, Ar)). We did not investigate further the e€ect
of solvent, but used dry acetone throughout.
J=8.55 Hz, Ar) ppm; EIMS 394 (M+), 349 (M EtO), 177
(ArN+2 ), 149 (Ar+), 86. Anal. calcd for CHNO: C, 51.78;
H, 5.62; N, 14.20. Found: C, 51.8; H, 5.7; N, 14.2.
3-(Benzoyloxymethyloxycarbonyl)-1-(4-ethoxycarbonyl-
phenyl) - 3 - methyltriazene (5g). Yield 49%; mp 108±
10 ꢀC; IR (KBr) 1742, 1740, 1718, 1272 cm 1; H NMR
1
d 1.42 (3H, t, J=7.1 Hz, Me), 3.53 (3H, s, NMe), 4.40
(2H, q, J=6.8 Hz, OCH2), 6.27 (2H, s, OCH2O), 7.48
(2H, t, J=6.1 Hz, ArR), 7.60 (1H, t, J=7.3 Hz, ArR),
7.68±8.11 (4H, AA0BB0, J=8.5 Hz, ArX), 8.14 (2H, d,
J=5.9 Hz, ArR) ppm; EIMS 385 (M+), 340 (M EtO),
177 (ArN+2 ), 149 (Ar+), 105 (PhCO+). Anal. calcd for
CHNO: C, 59.22; H, 4.97; N, 10.90. Found: C, 59.1; H,
4.9; N, 11.0.
3-(Acetoxymethyloxycarbonyl)-1-(4-cyanophenyl)-3-meth-
yltriazene (5a). Yield 38%; mp 107 ꢀC; IR (KBr) 2229,
1755, 1739, 1225 cm 1; 1H NMR d 2.12 (3H, s, Ac), 3.51
(3H, s, NMe), 5.99 (2H, s, OCH2O), 7.74 (4H, s, Ar) ppm;
EIMS 276 (M+), 160 (ArN3HCH3+), 130 (ArN+2 ), 102
(Ar+), 73, 43. Anal. calcd for CHNO: C, 52.17; H, 4.35;
N, 20.29. Found: C, 52.4; H, 4.2; N, 20.4.
3-(2-Acetoxybenzoyloxymethyloxycarbonyl)-1-(4-ethoxy-
carbonylphenyl)-3-methyltriazene (5h). Yield 46%; mp
3-(Acetoxymethyloxycarbonyl)-1-(4-ethoxycarbonylphe-
nyl)-3-methyltriazene (5b). Yield 65%; mp 84±6 ꢀC; IR
124±5 ꢀC; IR (KBr) 1745 (br), 1717 (br), 1606, 1276,
1
1
(KBr) 1740 br, 1717, 1275 cm 1; H NMR d 1.4 (3H, t,
1206 cm
;
1H NMR d 1.42 (3H, t, J=7.1 Hz, Me),
J=5 Hz, Me), 2.1 (3H, s, Ac), 3.5 (3H, s, NMe), 4.35
(2H, q, J=5 Hz, OCH2), 6.0 (2H, s, OCH2O), 7.65±7.7
and 8.05±8.1 (4H, AA0BB0, J=10 Hz, Ar) ppm; EIMS
323 (M+), 277 (M EtOH), 177 (ArN+2 ), 149 (Ar+), 43.
Anal. calcd for CHNO: C, 52.01; H, 5.30; N, 12.99.
Found: C, 51.9; H, 5.1; N, 12.7.
2.39 (3H, s, Ac), 3.52 (3H, s, NMe), 4.41 (2H, q,
J=7.3 Hz, OCH2), 6.21 (2H, s, OCH2O), 7.12 (1H, d,
J=8.3 Hz, ArR), 7.35 (1H, t, J=7.1 Hz, ArR), 7.62
(1H, t, J=5.9 Hz, ArR), 7.68±7.70 and 8.10±8.13 (4H,
AA0BB0, J=8.3 Hz, ArX), 8.12 (1H, d, J=7.8 Hz,
ArR) ppm; EIMS 443 (M+), 398 (M+ EtO), 234
(ArXN3CO+), 177 (ArXN2+), 149 (ArX). Anal. calcd for
CHNO: C, 56.88; H, 4.77; N, 9.48. Found: C, 56.8; H,
4.8; N, 9.4.
3-(Acetoxymethyloxycarbonyl)-1-(4-bromophenyl)-3-me-
thyltriazene (5c). Yield 43%; mp 88±9ꢀC; IR (KBr) 1764,
1
1737 (br), 1138cm 1; H NMR d 2.16 (3H, s, Ac), 3.48
(3H, s, NMe), 5.99 (2H, s, OCH2O), 7.52±7.57 (AA0BB0,
J=8.8Hz, Ar) ppm; EIMS 329/331 (M+), 185/183
(ArN+2 ), 157/155 (Ar+), 43. Anal. calcd for CHNO: C,
40.02; H, 3.66; N, 12.73. Found: C, 39.4; H, 3.7; N,
12.85.
1-(4-Ethoxycarbonylphenyl)-3-(2-hydroxybenzoyloxyme-
thyloxycarbonyl)-3-methyltriazene (5i). Yield 29%; mp
106±7 ꢀC; IR (KBr) 3284, 1740, 1719, 1688, 1278cm 1; 1H
NMR d 1.41 (3H, t, J=7.1Hz, Me), 3.53 (3H, s, NMe),
4.39 (2H, q, J=7.1 Hz, OCH2), 6.27 (2H, s, OCH2O), 6.91
(1H, t, J=7.6 Hz, 5-H ArR), 7.0 (1H, d, J=8.4 Hz, 3-H
ArR), 7.5 (1H, m, 4-H ArR), 7.66±7.69 and 8.09±8.12 (4H,
AA0BB0, J=8.5, ArX), 7.92 (1H, d, J=8.1 Hz, 6-H ArR),
10.39 (1H, s, OH) ppm; EIMS 401 (M+), 177 (ArXN2+),
149 (ArX), 121 (ArRCO+). Anal. calcd for CHNO: C,
56.86; H, 4.77; N, 10.47. Found: C, 56.5; H, 5.0; N,
10.3.
3-(Acetoxymethyloxycarbonyl)-3-methyl-1-(4-tolyl)tria-
zene (5d). Yield 44%; mp 52 ꢀC; IR (KBr) 1769, 1739,
1370, 1133 cm 1; 1H NMR d 2.16 (3H, s, Ac), 2.39 (3H, s,
ArMe), 3.48 (3H, s, NMe), 5.99 (2H, s, OCH2O), 7.25±
7.56 (AA0BB0, J=8.3 Hz, Ar) ppm; EIMS 265 (M+), 119
(ArN+2 ), 91 (Ar+), 43. Anal. calcd for CHNO: C, 54.33;
H, 5.70; N, 15.84. Found: C, 54.6; H, 5.7; N, 15.8.
1 - (4 - Ethoxycarbonylphenyl) - 3 - (2 - (4 - isobutylphenyl)-
propanoyloxymethyloxycarbonyl)-3-methyltriazene (5j).
Yield 60%; mp 59±61 ꢀC; IR (KBr) 1746 (br), 1718,
3-(Butanoyloxymethyloxycarbonyl)-1-(4-ethoxycarbonyl-
phenyl)-3-methyltriazene (5e). Yield 48%; mp 62 ꢀC; IR
(KBr) 1741 br, 1721, 1275 cm 1; 1H NMR d 0.95 (3H, t,
J=5 Hz, Me), 1.4 (3H, t, J=5 Hz, Me), 1.6±1.7 (2H, m,
CHR2 ), 2.35 (2H, t, J=8 Hz, COCH2), 3.45 (3H, s, NMe),
4.35 (2H, q, J=5 Hz, OCH2), 6.0 (2H, s, OCH2O), 7.6±
7.65 and 8.05±8.1 (4H, AA0BB0, J=10 Hz, Ar) ppm;
EIMS 351 (M+), 306 (M EtO), 177 (ArN+2 ), 149 (Ar+),
71, 43. Anal. calcd for CHNO: C, 54.70; H, 6.02; N,
11.95. Found: C, 54.8; H, 6.0; N, 11.8.
1604, 1274 cm 1; H NMR d 0.88 (6H, d, J=6.8, Pr),
1.43 (3H, t, J=7.1 Hz, Me), 1.53 (3H, d, J=7.3 Hz, Me),
1.78 (1H, m, CHiPr), 2.39 (2H, d, J=7.3 Hz, CH2ArR),
3.45 (3H, s, NMe), 3.78 (1H, q, J=6.83 Hz, CHCO), 4.41
(2H, q, J=6.8 Hz, OCH2), 5.97±6.01 (2H, 2d, J=5.62
Hz, OCH2O), 7.03±7.05 and 7.18±7.20 (4H, AA0BB0,
J=8.3 Hz, ArR), 7.65±7.67 and 8.11±8.13 (4H, AA0BB0,
J=8.55 Hz, ArX) ppm; EIMS 469 (M+), 177 (ArXN+2 ),
161 (Me2CHArCHCH3+), 149 (ArX+), 43. Anal. calcd
for CHNO: C, 63.95; H, 6.65; N, 8.95. Found: C, 63.7;
H, 6.7; N, 8.9.
1
i
3-(2-(N-acetylamino)propanoyloxymethyloxycarbonyl)-
1-(4-ethoxycarbonylphenyl)-3-methyltriazene (5f). Yield
24%; mp 127±9 ꢀ C; IR (KBr) 3295, 1743, 1718, 1656,
1276 cm 1; 1H NMR d 1.42 (3H, t, J=7 Hz, Me), 1.45 (3H,
d, J=7.3 Hz, a-Me), 2.04 (3H, s, Ac), 3.54 (3H, s, NMe),
4.41 (2H, q, J=7.1 Hz, OCH2), 4.66±4.72 (1H, m, a-CH),
6.02±6.09 (2H, 2Âd, J=5.62 Hz, OCH2O), 5.97 (1H, d,
J=6.8 Hz, NH), 7.67±7.69 and 8.11±8.13 (4H, AA0BB0,
Hydrolysis studies
Compounds 5a±j were incubated at 37ꢀC in 0.066 M pH
7.7 isotonic phosphate bu€er. Compound 5f was also stu-
died at 37ꢀC in bu€er solutions of di€erent concentration

E. Carvalho et al. / Bioorg. Med. Chem. 8 (2000) 1719±1725
1725
and pH. The ionic strength of these solutions was main-
References
tained at 0.2 M using NaClO4. Samples of these reaction
mixtures were analysed directly by HPLC. The 3-acyl-
oxymethyloxycarbonyl-1-aryl-3-methyltriazenes were
also incubated at 37 ꢀC in 80% human plasma diluted
with 0.066 M pH 7.4 isotonic phosphate bu€er (the ®nal
pH of this mixture was 7.7). At appropriate intervals,
samples of the plasma reactions were withdrawn, dilu-
ted with acetonitrile (0.4 mL) and centrifuged at 13,000
rpm for 5 min. The clear supernatant was analysed for
substrate by HPLC. Quantitative analysis of the pro-
drugs and of their hydrolysis products was achieved by
using a calibration curve constructed using standards
chromatographed under the same conditions. For very
fast reactions, e.g., those for 5h, samples were with-
drawn every 10 s and frozen at 18 ꢀC prior to analysis.
In all cases the HPLC conditions were: detector wave-
length, 290 nm; mobile phase, acetonitrile:H2O (50:50 to
80:20). All reactions, both in bu€ers and in plasma,
followed pseudo-®rst-order kinetics up to at least 4 half-
lives.
1. Cohen, G. L.; Falkson, C. I. Drugs 1998, 55, 791.
2. Connors, T. A.; Goddard, P. M.; Merai, K.; Ross, W. C. J.;
Wilman, D. E. V. Biochem. Pharmacol. 1976, 25, 241.
3. Julliard, M.; Vernin, G. Ind. Eng. Chem. Prod. Res. Rev.
1981, 20, 287.
4. Rutty, C. J.; Newell, D. R.; Vincent, R. B.; Abel, G.;
Goddard, P. M.; Harland, S. J.; Calvert, A. H. Brit. J. Cancer
1983, 48, 140.
5. Iley, J.; Moreira, R.; Rosa, E. J. Chem. Soc., Perkin Trans.
2 1987, 1503.
6. Fernandes, L.; Iley, J.; Rosa, E. J. Chem. Res.(S) 1987, 264.
7. Iley, J.; Moreira, R.; Rosa, E. J. Chem. Soc., Perkin Trans.
1 1991, 3241.
8. Iley, J.; Ruecroft, G.; Carvalho, E.; Rosa, E. J. Chem. Res.
(S) 1989, 162.
9. Carvalho, E.; Iley, J.; Rosa, E. J. Chem. Soc., Perkin Trans.
2 1993, 865.
10. Clark, A. S.; Deans, B.; Stevens, M. F. G.; Tisdale, M. J.;
Wheelhouse, R. T.; Denny, B. J.; Hartley, J. A. J. Med. Chem.
1995, 38, 1493.
11. Carvalho, E.; Iley, J.; Perry, M. J.; Rosa, E. Pharm. Res.
1998, 15, 931.
12. Carvalho, E.; Iley, J.; Perry, M. J.; Rosa, E. J. Chem. Soc.,
Perkin Trans. 2 1998, 2375.
Partition coecients
13. Alexander, J.; Cargyll, R.; Michelson, S. R.; Schwam, H.
J. Med. Chem. 1988, 31, 318.
For compounds 5a±f, partition coecients were deter-
mined in octanol±pH 7.4 phosphate bu€er at room
temperature. Each phase was mutually saturated before
the experiment. The compounds were dissolved in octa-
nol and the octanol±pH 7.4 mixtures were shaken for
30 min to reach an equilibrium distribution; each phase
was analysed separately by the HPLC method described
above. The partition coecients, Pexp, were obtained
from the ratio of the peak area in octanol to the peak
area in bu€er. For compounds 5g,h, the solubility in the
aqueous medium was insucient for an accurate deter-
mination of Pexp. Calculated values of P, expressed as
log Pcalc, were obtained using the software Chemsketch
version 3.60 from Advanced Chemistry Development
Inc.26 For compounds 5a±f, there is suciently good
agreement between log Pcalc and log Pexp for us to have
con®dence that the log Pcalc values for 5g,h appropriately
re¯ect the lipophilicity of these compounds.
14. Alexander, J.; Fromtling, R. A.; Bland, J. A.; Pelak, B. A.;
Gil®llan, E. C. J. Med. Chem. 1991, 34, 78.
15. Davies C. W. J. Chem. Soc. 1938, 2093.
16. Hammett, L. P. Physical Organic Chemistry. McGraw-
Hill: New York, 1940, p. 121.
17. Jencks, W. P.; Carriuolo, J. J. Am. Chem. Soc. 1961, 83,
1743.
18. Leitweber, F. J. Drug Metabolism Rev. 1989, 18, 379.
19. Delben, F.; Paoletti, S.; Manzini, G.; Nisi, C. J. Pharm.
Sci. 1981, 70, 892.
20. Godin, J. R. P.; Llanos, G.; Vaughan, K.; Renton, K. W.
Can. J. Physiol. Pharmacol. 1981, 59, 1239.
21. Huang, T. L.; Szekacs, A.; Uematsu, T.; Kuwano, E.;
Parkinson, A.; Hammock, B. D. Pharm. Res. 1993, 10, 639.
22. Nudelman, A.; Ruse, M.; Aviram, A.; Rabizadeh, E.; Shak-
lai, M.; Zimrah, Y.; Raphaeli, A. J. Med. Chem. 1992, 35, 687.
23. Cheng S. C.; Iley J. J. Chem. Res. (S) 1983, 320.
24. Vaughan, K.; Tang, T.; Llanos, G.; Horton, J. K.;
Simmonds, R. J.; Hickman, J. A.; Stevens, M. F. G. J. Med.
Chem. 1984, 27, 357.
25. Cameron, L. M.; LaFrance, R. J.; Hemens, C. M.;
Vaughan, K.; Rajaraman, R.; Chubb, D. C.; Goddard, P. M.
Anti-Cancer Drug Des. 1985, 1, 27.
Acknowledgements
This work was supported by JNICT and PRAXIS XXI,
project No. 2/2.1/SAU/1404/95.
26. Advanced Chemistry Development Inc., 133 Richmond
St., West Suite 605, Toronto ONMSH2L3, Canada.