Dopamine- and tyramine-based derivatives of triazenes: Activation by tyrosinase and implications for prodrug design

Article abstract of DOI:10.1016/j.ejmech.2009.03.025

A range of triazene derivatives were synthesized and investigated as prodrug candidates for melanocyte-directed enzyme prodrug therapy (MDEPT). The prodrugs contained a tyramine or dopamine promoiety required for tyrosinase activation and this was joined

Full text of DOI:10.1016/j.ejmech.2009.03.025

European Journal of Medicinal Chemistry 44 (2009) 3228–3234
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Dopamine- and tyramine-based derivatives of triazenes:
Activation by tyrosinase and implications for prodrug design
,
b
b
a
M. Jesus Perry a, Eduarda Mendes a , Ana Luısa Simplıcio , Ana Coelho , Ricardo V. Soares ,
*
´
´
Jim Iley c, Rui Moreira a, Ana Paula Francisco a
a
´
iMedUL, Faculdade de Farmacia da Universidade de Lisboa, Av. Prof. Gama Pinto, 1649-003 Lisboa, Portugal
b ITQB/IBET, Av. Repu´blica-Quinta do Marqueˆs, 2781-901 Oeiras, Portugal
c Department of Chemistry and Analytical Sciences, The Open University, Walton Hall, Milton Keynes MK7 6AA, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
A range of triazene derivatives were synthesized and investigated as prodrug candidates for melanocyte-
directed enzyme prodrug therapy (MDEPT). The prodrugs contained a tyramine or dopamine promoiety
required for tyrosinase activation and this was joined via a urea functional group to the cytotoxic tri-
azene. The stability of each of the prodrugs in phosphate buffer, human plasma and in mushroom
tyrosinase is discussed. The identification of the main peak formed after the tyrosinase reaction was
attempted by LC–MS and the conversion of prodrug to the quinone was confirmed.
Received 25 November 2008
Received in revised form
16 March 2009
Accepted 20 March 2009
Available online 27 March 2009
Ó 2009 Elsevier Masson SAS. All rights reserved.
Keywords:
Prodrugs
Triazenes
Tyrosinase
Melanoma
1. Introduction
a cyclization-drug release mechanism [6]. Jordan and co-workers
have synthesized several prodrugs 1 of nitrogen mustards as the
Disseminated melanoma represents one of the most treatment-
resistant of all human malignancies. Drug resistance is not only
acquired following drug therapy, but may be also present in
untreated lesions. Consequently, adjuvant therapy, immunotherapy
or chemotherapy, provide only modest improvements in outcome.
Among new cancer cases, melanoma is continuing to rise in inci-
dence at a rate exceeding all other cancers [1–3].
Melanocyte-directed enzyme prodrug therapy (MDEPT) has
been suggested as a possible selective strategy towards the treat-
ment of malignant melanoma [4]. MDEPT offers a highly selective
triggering mechanism for drug delivery because of the unique
occurrence of tyrosinase enzyme expression in melanocytes and
melanoma cells. When melanocytes become malignant, the genes
expressing tyrosinase become up-regulated, resulting in a marked
increase in the tyrosinase levels within the cancerous cells [5]. A
number of tyrosinase-dependent prodrug strategies have been
investigated for the treatment of melanoma, in which tyrosinase
mediates the release of an anticancer agent from prodrugs via
cytotoxic agent connected via a carbamate, urea or thiourea group
to the tyrosinase sensitive unit [4,7,8]. These prodrugs act as
tyrosinase substrates and release the mustard drug after enzymatic
oxidation [7–9].
Dacarbazine (DTIC), 2, remains the single FDA-approved
chemotherapeutic agent in common use for metastatic melanoma.
Among all the chemotherapeutic agents tested against this malig-
nancy, DTIC is the most active single agent and thus considered the
reference drug [10,11]. Related 1-aryl-3,3-dimethyltriazenes, 3 (see
Scheme 1), are also documented as anticancer drugs, but with
reduced selectivity for tumour cells [12,13]. The biological action of
DTIC and the dimethyltriazenes, 3, is a consequence of their
capacity to alkylate DNA and RNA via the corresponding mono-
methyltriazene 4 following metabolic oxidation by cytochrome
P450 enzymes (Scheme 1) [12,14,15]. Thus, improvements in the
treatment of melanoma may be attained by the development of
a prodrug strategy that can specifically deliver the ultimate meth-
ylating agent to malignant cells. Since monomethyltriazenes are as
good leaving groups as alcohols [16], we envisaged that coupling
them to tyrosinase substrates would identify suitable prodrugs 5
capable of regenerating the alkylating metabolite 4 via the pathway
depicted in Scheme 2. We now report the synthesis of triazene
* Corresponding author. Tel.: þ+351 217946413 fax: þ351 217946470.
E-mail address: ermendes@ff.ul.pt (E. Mendes).
0223-5234/$ – see front matter Ó 2009 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2009.03.025

M.J. Perry et al. / European Journal of Medicinal Chemistry 44 (2009) 3228–3234
3229
P450
-HCHO
H3C
N
Ar
H3C
N
Ar
H3C
N
Ar
N
N
N
N
N
H
N
CH3
CH2OH
4
3
DNA
+
N
H3C
N
Ar
+
H3C DNA
N2
+
H2N Ar
N
H3C
N
N
H
Scheme 1. Metabolic pathway for dimethyltriazenes.
derivatives 5a–g (Scheme 3), and their evaluation as substrates for
2.2. Stability in phosphate buffer and human plasma
tyrosinase.
The hydrolysis of triazene derivatives, 5a–g, to the corre-
sponding 1-aryl-3-methyltriazenes, 4, was studied both in isotonic
phosphate buffer (PBS, containing 20% of dimethylsulfoxide) and in
80% human plasma, at 37 ꢀC. These reactions were followed by
HPLC by monitoring both the loss of starting material and the
formation of products. The kinetics of the hydrolysis of each pro-
drug in both media followed pseudo-first-order kinetics over four
to five half-lives.
H
CH3
N
N
X
H
N
N
Cl
Y
N
H3C
N
N
H2N
HO
O
Cl
1 X = N, O; Y = O, S
2
Half-lives for the hydrolysis in PBS are presented in Table 1.
Inspection of Table 1 shows that these compounds are generally
very stable in this medium; no measurable decomposition of the
tyramine-based derivatives 5a–c over 15 days or of the dopamine
derivative 5g over 2 days could be detected, while the tyramine
derivative 5d (half-life 15 days) and the dopamine derivatives 5e,f
(half-lives 15 h and 20 h, respectively) hydrolyzed only very slowly.
It seems, however, that the prodrugs containing dopamine are the
more reactive.
2. Results and discussion
2.1. Synthesis
We were interested in examining the stability of the prodrugs in
human plasma because it contains a range of enzymes that could
potentially catalyze the hydrolysis of such urea derivatives. The
dopamine-based triazene derivatives 5e–g are hydrolyzed to the
corresponding 1-aryl-3-methyltriazene 4 with half-lives ranging
from 6 h to 15 h, implying that they are hydrolyzed in human
plasma only ca. 1.5 times faster than in pH 7.4 buffer. In contrast, the
tyramine-based derivatives 5a,c,d are stable in human plasma for 2
days, while 5b is hydrolyzed with a half-life of 71 h. These results
Initial attempts to synthesize compounds 5 involved the prep-
aration of ethyl carbamates, 6a, from the corresponding mono-
methyltriazenes 4, according to the published procedure
(Scheme 3) [17]. However, carbamates 6a proved to be unreactive
towards tyramine and dopamine in the presence of a wide range of
bases. Consequently, we synthesized the corresponding 4-nitro-
phenyl carbamate derivatives 6b and gratifyingly reaction of these
with tyramine and dopamine in the presence of triethylamine gave
the desired compounds 5a–g in good yields.
CH3
H
CH3
H
N
N
N
N
N
N
..
N
N
Y
O
O
O
O
tyrosinase
X
X
HO
5
7
X
X
HO
HO
HO
HO
hydrolysis
CO2
+
+
N
H
NH
N
+
N
N
N
H
N
N
O
CH3
CH3
4
Scheme 2. Activation pathway for tyramine-based triazene prodrugs triggered by tyrosinase.

3230
M.J. Perry et al. / European Journal of Medicinal Chemistry 44 (2009) 3228–3234
Me
N
NHMe
N
N
OR
N
N
(i)
O
X
X
6a, R = Et
4
6b, R = 4-C6H4NO2
(ii)
Me
N
H
N
N
Y
N
O
X
OH
5a X = CH3CO Y = H
5b X = CO2Et Y = H
5e X = CH3CO Y = OH
5f X = CO2Et Y = OH
5g X = Me Y = OH
5c X = CN Y = H
5d X = Me Y = H
Scheme 3. Synthesis of triazene prodrugs 5. (i) Ethyl or 4-nitrophenyl chloroformate, pyridine, CH2Cl2; (ii) tyramine or dopamine, triethylamine, THF.
indicate that tyramine substrates are less prone to plasma hydro-
lysis when compared with their dopamine counterparts.
These Km values compare favourably with those reported for tyra-
mine (0.51 mM) and dopamine (2.2 mM). The Vmax values corre-
spond to turnover numbers of 0.6 sꢂ1 and 6 sꢂ1 for 5c and 5f,
respectively, which are rather smaller than the 25.9 sꢂ1 and 439 sꢂ1
observed for tyramine and dopamine substrates [18], though the
relative sizes between the tyramine and dopamine derivatives are
similar. This observation is consistent with the report that tyrosinase
accepts mono-, di- and trihydroxyphenols as substrates but displays
the highest turnover rates for dihydroxyphenols [19,20]. The rates of
decomposition of tyramine derivatives 5e–g are not significantly
affected by the substituents in the triazene moiety, presumably
reflecting the significant distance between these substituents and
the catechol moiety. Interestingly, although the half-lives of the
tyramine derivatives 5a–d are little affected by the substituent in the
triazene moiety also, there is a correlation between log t1/2 values
2.3. Triazene derivatives as a substrate for tyrosinase
The ability of triazene derivatives 5 to act as substrates of tyros-
inase was studied by HPLC by determining the half-lives for activa-
tion for each compound in the presence of the enzyme (Table 1).
HPLC analysis of reaction mixtures showed that both tyramine, 5a–
d, and dopamine derivatives, 5e–g, are rapidly consumed with half-
lives ranging from 6 to 18 min, thus indicating that theyare excellent
tyrosinase substrates. These triazene prodrugs have shorter half-
lives in the presence of tyrosinase than the urea-linked N-mustard
prodrugs described by Knaggs et al. [9]. For compound 5c
Km ¼ 0.31 ꢁ0.08 mM and Vmax ¼ 0.030 ꢁ 0.005 mM/min, and for
compound 5f Km ¼ 0.71 ꢁ0.4 mM and Vmax0.31 ꢁ0.14 mM/min.
and the Hammett sp values yielding a
r
value of ꢂ0.2 (r2 ¼ 0.99,
Fig. 1). The origin of this rather small effect is not clear.
Inspection of the HPLC traces of the reaction mixtures of the
tyrosinase-catalysed reactions revealed that triazenes 5 were
Table 1
Triazene prodrug 5 half-lives, t1/2, in both PBS buffer (20% DMSO) and 80% human
plasma at 37 ꢀC, their log P values, and their ability to act as tyrosinase substrates
(100 U/mL).
1.3
CH3
H
5d
N
N
N
N
O
Y
X
1.2
HO
5a
5
log Pexp
t1/2
80% human Tyrosinaseb
Compound
X
Y
pH 7.4
buffera (h) plasmaa (h) (min)
1.1
5b
c
c
c
d
5a
5b
5c
5d
5e
5f
Ac
CO2Et
CN
Me
Ac
H
H
H
H
3.25 ꢁ 0.07
3.96 ꢁ 0.09
2.89 ꢁ 0.06
13.0
12.9
11.8
18.2
9.9
5c
71.5
d
d
3.90 ꢁ 0.06 364
OH 3.24 ꢁ 0.13
15.1
9.30
14.7
5.66
CO2Et OH 3.78 ꢁ 0.09 20.8
6.1
8.4
1.0
d
5g
Me
OH
4.17 ꢁ 0.03
-0.2
0
0.2
0.4
0.6
0.8
a
Substrate concentration 10ꢂ5 M.
Substrate concentration 5 ꢃ10ꢂ5 M.
Stable for 15 days.
σp
b
c
Fig. 1. Hammett plot of the half-lives for the decomposition of tyramine-based tri-
azenes 5a–d against sp values for the triazene substituents.
d
Stable for 2 days.

M.J. Perry et al. / European Journal of Medicinal Chemistry 44 (2009) 3228–3234
3231
Table 2
(Table 2). For this reason ammonium formate was added to the
aqueous eluent. To achieve adequate HPLC separations, we also
required the pH of the eluent to be kept low; therefore the aqueous
reaction solutions were acidified to pH 3.5 using formic acid.
We were able to identify the main HPLC peak formed through
activation of triazenes 5b,c and e by tyrosinase. In the case of the
tyramine derivatives, 5b,c, the mass of the metabolite was 14 units
higher than the mass of the substrate (as exemplified for 5c in
Fig. 2). This mass difference can be ascribed to the conversion of 5b,
and 5c to the corresponding ortho-quinone 7 (Scheme 2). In the
case of the dopamine derivative, 5e, chromatographic separation
proved more difficult since the product peak overlapped with the
substrate, as expected for a quinone product with similar structure
and polarity to the substrate. Separation proved possible, however,
APCI MS of the triazene substrates 5b,c,e and their corresponding metabolites
produced by tyrosinase activation.
Compound
5b
5c
5e
355.0a
400.6b
Substrate molecular ion
322.7a
368.4b
336.6a
382.2b
415.7b
383.6a
429.3b
Metabolite molecular ion
398.6b
a
Molecular ion.
Formic acid adduct.
b
rapidly converted to a metabolite that did not correspond either to
the appropriate monomethyltriazene 4 or its aniline decomposition
product [12]. To gain some insight to the structure of this metab-
olite, an LC–MS study of the reaction mixtures was performed. It
was not possible to obtain efficient ionization of the substrates or
products by electrospray ionization (positive or negative modes)
using the chromatographic conditions used for kinetic experi-
ments. However, using the APCI source we were able to detect
formic acid adducts of the substrates 5 and their corresponding
metabolites; in some cases we could also detect the molecular ions
using
a gradient program and lower formate concentration.
Formation of the ortho-quinone 7 corresponding to 5e was
confirmed by LC–MS since the metabolite peak, which eluted
0.5 min earlier than the substrate, gave a formic acid adduct ion of 2
mass units less than the substrate (Table 2). Overall, these results
indicate that although the prodrugs are substrates for tyrosinase,
the release of the cytotoxic monomethyltriazene 4 does not occur
Fig. 2. TIC (A) and DAD (B) chromatograms of a solution of triazene 5c after enzymatic reaction with tyrosinase; the substrate is detected at 6.8 min and the product at 5.3 min.
Mass spectra for the product (C) and the triazene substrate (D) peaks.

3232
M.J. Perry et al. / European Journal of Medicinal Chemistry 44 (2009) 3228–3234
under the conditions used. This suggests that subsequent cycliza-
tion activation of the ortho-quinone 7 (Scheme 2) is not an efficient
process. Our results are consistent with recent studies that show
that in urea or carbamate derivatives of dopamine the nitrogen
atom is not sufficiently nucleophilic to effect the necessary cycli-
zation that results in release of the antitumour agent [21,22].
J ¼ 8.0 Hz, COAr); 13C NMR (100 MHz, CDCl3):
d 27.29 (CH3CO), 29.18
(NCH3), 35.21 (CH2Ar), 42.56 (NCH2), 115.66–136.41 (CAr), 152.37
(CNN), 154.16 (COH), 156.18 (CO), 197.67 (CH3CO). Anal. Calcd for
C18H20N4O3: C, 63.5; H, 5.9; N, 16.5. Found: C, 63.5; H, 5.9; N, 16.4.
4.2.3. 3-[2-(4-Hydroxyphenyl)ethylaminocarbonyl]-3-methyl-1-(4-
ethoxycarbonylphenyl)triazene (5b)
3. Conclusions
Yield 68%; m.p. 158–160 ꢀC; nmax/cmꢂ1 3415, 3280, 2962, 1708,
1676; 1H NMR (CDCl3):
d
1.42 (3H, t, J ¼ 7.0 Hz, CH3CH2), 2.84 (2H, t,
In conclusion, the synthesis and analysis of a range of prodrugs
derived from tyramine and dopamine, of potential use within
MDEPT, have been evaluated. These compounds display excellent
stability both in pH 7.4 buffer and in human plasma and simulta-
neously are good substrates for tyrosinase. The tyrosinase-cata-
lysed reaction is very efficient in liberating an oxidized metabolite,
but our HPLC and LC–MS data demonstrate that release of a cyto-
toxic monomethyltriazene does not occur for the urea-linked
derivatives 5. The present work demonstrates that drug release,
which requires an activation process involving an intramolecular
Michael addition reaction, requires a much better nucleophile than
the urea nitrogen atom. Future work will therefore focus on the
design of derivatives that have improved cyclization capacity while
retaining excellent stability in aqueous solution. Such an approach
shall also take account of the fact that dopamine derivatives are
more rapidly metabolized by tyrosinase.
J ¼ 6.8 Hz, CH2Ar), 3.46 (3H, s, NCH3), 3.65 (2H, q, J ¼ 6.4 Hz,
NHCH2CH2Ar), 4.40 (2H, q, J ¼ 7.0 Hz, CH3CH2), 6.14 (1H, s, OH), 6.52
(1H, br t, J ¼ 5.2 Hz, NH), 6.85–7.10 (4H, AA0BB0, J ¼ 8.4 Hz, HOAr),
7.41–8.09 (4H, AA0BB0, J ¼ 8.8 Hz, NArCO); 13C NMR (100 MHz,
CDCl3):
d 14.33 (CH3CH2), 28.61 (NCH3), 34.98 (CH2Ar), 41.68
(NCH2), 61.33 (CH3CH2), 115.69–130.65 (CAr), 152.36 (CNN), 154.36
(COH), 154.87 (CO), 166.48 (COCH2CH3). Anal. Calcd for C19H22N4O4:
C, 61.6; H, 6.0; N, 15.1. Found: C, 61.5; H, 5.8; N, 14.9.
4.2.4. 3-[2-(4-Hydroxyphenyl)ethylaminocarbonyl]-3-methyl-1-(4-
cyanophenyl)triazene (5c)
Yield 55%; m.p. 156–158 ꢀC; nmax/cmꢂ1 3221, 3411, 3018, 2966,
2224, 1678, 1613; 1H NMR (CDCl3):
d
2.83 (2H, t, J ¼ 6.8 Hz, CH2Ar),
3.45 (3H, s, NCH3), 3.59–3.36 (2H, q, J ¼ 4.0 Hz, CH2CH2Ar), 6.56 (1H,
br t, NHCH2), 6.72 (1H, s, ArOH), 6.79–7.08 (4H, AA0BB0, J ¼ 4.8 Hz,
ArH), 7.47–7.74 (4H, AA0BB0, J ¼ 8.4 Hz, ArH); 13C NMR (100 MHz,
CDCl3):
d 27.38 (NCH3), 33.49 (CH2Ar), 40.29 (CH2CH2Ar), 109.80
4. Experimental
(CCN), 114.24–117.46 (CAr), 121.14 (CN), 128.46 (CAr), 131.94 (CAr),
150.73 (CNN); 152.61 (COH); 154.13 (CO). Anal. Calcd for
C17H17N5O2: C, 63.2; H, 5.3; N, 21.7. Found: C, 63.2; H, 5.5; N, 21.6.
4.1. General information
Melting points were determined using a Kofler camera Bock
Monoscop M and are uncorrected. Elemental analyses were per-
4.2.5. 3-[2-(4-Hydroxyphenyl)ethylaminocarbonyl]-3-methyl-1-(4-
tolyl)triazene (5d)
´
Yield 53%; m.p. 196–197 ꢀC; nmax/cmꢂ1 3363, 3027, 2923, 1678;
formed by Instituto Tecnologico e Nuclear, Portugal. FTIR spectra
were recorded as KBr discs using a Nicolet Impact 400 spectro-
photometer. 1H and 13C NMR spectra were obtained in CDCl3 or
DMSO solutions using a Bruker AM 400 WB. Coupling constants, J,
are expressed in Hz. Chemicals were purchased from Sigma–
Aldrich Chemical Company (U.K.) and B.D.H. Merck Chemical
Company (U.K.). All chemicals and solvents were of reagent grade,
except buffer substances and HPLC solvents, which were analytical
grade and LiChrosolvÒ grade, respectively. Column chromatog-
raphy was performed using silica gel 60 mesh 70–230 (Merck).
1H NMR (CDCl3):
d
2.41 (3H, s, CH3Ar), 2.85 (2H, t, J ¼ 6.8 Hz, CH2Ar),
3.45 (3H, s, NCH3), 3.64 (2H, q, J ¼ 6.2 Hz, CH2CH2Ar), 5.64 (1H, s,
OH), 6.55 (1H, br t, NH), 6.85–7.12 (4H, AA0BB0, J ¼ 7.6 Hz, ArMe),
7.24–7.34 (4H, AA0BB0, J ¼ 7.8 Hz, ArCN); 13C NMR (100 MHz, CDCl3):
d
21.23 (CH3Ar), 28.65 (NCH3), 35.30 (CH2Ar), 42.46 (NCH2), 115.66–
138.39 (CAr), 146.91 (CNN), 154.54 (COH), 156.17 (CO). Anal.
C17H20N4O2: C, 65.4; H, 6.5; N, 17.9. Found: C, 65.3; H, 6.3; N, 17.7.
4.2.6. 3-[2-(3,4-Dihydroxyphenyl)ethylaminocarbonyl]-3-methyl-
1-(4-acetylphenyl)triazene (5e)
4.2. Synthesis
Yield 45%; m.p. 158–160 ꢀC; nmax/cmꢂ1 3560, 3500, 3400,
1720, 1650; 1H NMR (DMSO-d6):
d 2.65 (3H, s, CH3CO), 2.77 (2H,
4.2.1. General procedures for the synthesis of prodrugs 5a–g
A dichloromethane solution (5 mL) containing the appropriate
monomethyltriazene (1.31 mmol) was added to a solution of 4-
nitrophenyl chloroformate (1.58 mmol) and pyridine (1.31 mmol,
t, J ¼ 7.2 Hz, CH2Ar), 3.45 (3H, s, NCH3), 3.56 (2H, t, J ¼ 7 Hz,
NHCH2), 6.58–6.74 (3H, m, Ar), 7.65–8.09 (4H, AA0BB0, J ¼ 7.0 Hz,
ArAc); 13C NMR (100 MHz, DMSO-d6):
d 27.29 (CH3), 29.18
(NCH3), 35.49 (CH2Ar), 42.61 (NCH2), 116.02–136.41 (CAr), 144.09
(COH), 145.64 (COH), 152.37 (CNN), 154.14 (CO), 197.68(CO). Anal.
Calcd for C18H20N4O4: C, 60.7; H, 5.7; N, 15.7. Found: C, 60.5; H,
5.3; N, 15.6.
106 mL). After stirring for 24 h at room temperature, the reaction
mixture was filtered and concentrated in vacuo. The crude product
so-obtained was purified by column chromatography and used
immediately. The pure carbamate (0.425 mmol) was dissolved in
THF (10 mL) and tyramine (or dopamine) (0.863 mmol) and trie-
4.2.7. 3-[2-(3,4-Dihydroxyphenyl)ethylaminocarbonyl]-3-methyl-
1-(4-ethoxycarbonylphenyl)triazene (5f)
thylamine (1.3 mmol, 180 mL) were added. After 48 h at room
temperature, the solution was concentrated in vacuo and the crude
product purified by column chromatography (ethylic ether-
petroleum ether, 7:3).
Yield 56%; m.p. 173–174 ꢀC; nmax/cmꢂ1 3410, 1708, 1602; 1H
NMR (DMSO-d6):
d
1.35 (3H, t, J ¼ 7.2 Hz, CH3CH2O), 2.66 (2H,
t, J ¼ 8.0 Hz, CH2Ar), 3.35 (5H, br s, NCH3 and NCH2), 4.34 (2H,
q, J ¼ 7.2 Hz, OCH2), 6.48–6.65 (3H, m, Ar), 7.82–8.04 (4H, AA0BB0,
J ¼ 8.0 Hz, ArCO2Et), 7.91 (1H, br t, J ¼ 8.0 Hz, NH), 8.70 (1H, s, OH),
4.2.2. 3-[2-(4-Hydroxyphenyl)ethylaminocarbonyl]-3-methyl-1-(4-
acetylphenyl)triazene (5a)
8.81 (1H, s, OH); 13C NMR (100 MHz, DMSO-d6):
d 14.65 (CH3),
Yield 50%; m.p.178–180 ꢀC; nmax/cmꢂ13409, 3362,1748,1672; 1H
29.18 (NCH3), 35.50 (CH2Ar), 42.64 (NCH2), 61.33 (OCH2), 116.01–
130.69 (CAr), 144.09 (COH), 145.63 (COH), 152.48 (CNN), 154.12
(CO), 165.76 (CO). Anal. Calcd for C17H20N4O3: C, 62.2; H, 6.1; N,
15.2. Found: C, 62.2; H, 5.9; N, 15.1.
NMR (CDCl3):
d
2.67 (3H, s, CH3CO), 2.87 (2H, t, J ¼ 8.0 Hz, CH2Ar),
3.48 (3H, s, NCH3), 3.67 (2H, q, J ¼ 8.0 Hz, NHCH2CH2), 5.47 (1H, s,
OH), 6.85–7.13 (4H, AA0BB0, J ¼ 8.0 Hz, OHAr), 7.44–8.03 (4H, AA0BB0,

M.J. Perry et al. / European Journal of Medicinal Chemistry 44 (2009) 3228–3234
3233
4.2.8. 3-[2-(3,4-Dihydroxyphenyl)ethylaminocarbonyl]-3-methyl-
4.7. Tyrosinase degradation studies
1-(4-tolyl)triazene (5g)
Yield 26%; m.p. 151–153 ꢀC; nmax/cmꢂ1 3500, 3317, 3408, 1683;
Mushroom tyrosinase (Sigma) was used at a concentration of
100 U/mL in a solution comprising phosphate buffer (pH 7.4,
2.4 mL) and DMSO (0.6 mL) at 37 ꢀC. A solution of the prodrug
1H NMR (DMSO-d6):
d
2.35 (3H, s, CH3Ar), 2.65 (2H, t, J ¼ 8.0 Hz,
NCH2CH2Ar), 3.32 (3H, s, NCH3), 3.37 (2H, m, NCH2CH2Ar), 6.48–
6.67 (3H, m, Ar), 7.28–7.61 (4H, AA0BB0, J ¼ 8.0 Hz, Ar-triazene), 7.73
(1H, br t, J ¼ 5.6 Hz, NH), 8.71 (1H, OH), 8.82 (1H, OH); 13C NMR
under investigation (15 m
L of 10ꢂ2 M solution) was added and the
mixture incubated with gentle stirring at 37 ꢀC. Aliquots were
collected at selected time intervals and quenched by addition of an
organic solvent (acetonitrile). The disappearance of the prodrug,
with concomitant appearance of metabolites, was followed by
HPLC for evaluation of the hydrolysis kinetics.
(100 MHz, DMSO-d6):
d 21.25 (CH3), 28.66 (NCH3), 35.59 (CH2Ar),
42.49 (NCH2), 116.01–138.39 (CAr), 144.07 (COH), 145.64 (COH),
146.90 (CNN), 154.51 (CO). Anal. Calcd for C17H20N4O3: C, 62.2; H,
6.1; N, 17.1. Found: C, 62.5; H, 6.0; N, 17.1.
4.8. LC–MS assays
4.3. HPLC analysis
Analysis was performed on a Surveyor HPLC from Thermo Fin-
negan, San Jose, C.A., USA equipped with an autosampler and
a diode array detector. The mass spectrometry system was an LCQ
ion trap mass spectrometer also from Thermo Finnegan, equipped
with an APCI source. The LC–MS system was controlled by Excalibur
version 1.3 software (Thermo Finnegan – Surveyor, San Jose, USA).
The analytical high-performance liquid chromatography (HPLC)
system comprised a Merck Hitachi L-7110 pump with an L-7400 UV
detector, a manual sample injection module equipped with a 20 mL
loop, a Merck LichrospherÒ 100 RP18 125 mm ꢃ 4.6 mm (5
mm)
column equipped with a Merck Lichrocart pre-column (Merck,
Germany). For compounds 5a–d an acetonitrile/water (45:55) iso-
cratic solvent system was used. For compounds 5e–g a gradient
elution beginning with a mobile phase of 95% water/5% acetonitrile
ramping to 95% acetonitrile/5% water over 30 min was used.
The column effluent was monitored at 300 nm for all compounds
(5a–g).
The analytical column was a Lichrospher RP-8 (125 ꢃ 4 mm, 5
mm)
from Merck with a guard column of the same material. Separations
were carried out with a flow rate of 0.7 mL/min at 50 ꢀC. Injections
were made with 20–100 m
L of the 10ꢂ4 to 10ꢂ5 M solutions of the
substrates in acetonitrile or the corresponding solutions after
enzymatic reaction.
In the case of the tyrosine derivatives, an isocratic system
comprising 45% ammonium formate (p.a. Fluka, 20 mM, pH 3.5
with formic acid), 45% acetonitrile (LC–MS grade, Reidel-de Hae¨n)
and 10% methanol (LC–MS grade, Reidel-de Hae¨n) was used. For the
dopamine derivatives, separation was carried out using a gradient
program, starting with 95% ammonium formate (2 mM, pH 3.5)/5%
acetonitrile for 7 min ramping to 35% acetonitrile over the next
10 min. Detection by UV absorption was in the range 200–400 nm.
The following conditions were used for the APCI source in
negative mode: vaporizer temperature, 400 ꢀC; discharge current,
4.4. Apparent partition coefficients
The apparent partition coefficients (log Pexp) were determined
at room temperature using 1-octanol–pH 7.4 phosphate buffer.
Each phase was mutually saturated before the experiment. The
volumes of each phase were chosen so that solute concentrations in
the aqueous phase after distribution were readily measurable. The
compounds were dissolved in 1-octanol and the 1-octanol–phos-
phate buffer mixtures were shaken for 30 min to reach an equi-
librium distribution. Each phase was analyzed separately by HPLC,
but the organic phase was diluted in acetonitrile (1:10) before HPLC
analysis. Partition coefficients, Pexp, were calculated from the ratio
of the peak area in octanol to the peak area in buffer.
5 m
A; temperature of the heated capillary 150 ꢀC. Nitrogen was used
as sheath gas and auxiliary gas with a gas flow rate of 80 and 40
arbitrary units, respectively. The flow was diverted from the mass
detector in the first 2 min of each run.
LC–MS was performed in total ion content mode (TIC) from m/z
50 to 1000.
4.5. Hydrolysis in buffer solution
Acknowledgements
Reaction mixtures were analyzed using HPLC. Usually, a 10 mL
ˆ
The authors thank the Fundaça˜o para a Ciencia e Tecnologia
aliquot of a 10ꢂ2 M stock solution of prodrug 5 in acetonitrile was
added to 10 mL of the appropriate thermostated buffer solution
(8 mL PBS and 2 mL DMSO). At regular intervals, samples of the
reaction mixture were analyzed by HPLC using the following
conditions: mobile phase, acetonitrile/water (the composition of
which depended on the compound); flow rate 1.0 mL/min.
(Portugal) for financial support through their pluriannual funding
to the iMed.UL research unit.
References
[1] C.M. Balch, S.J. Soong, J.E. Gershenwald, J.F. Thompson, D.S. Reintgen,
N. Cascinelli, M. Urist, K.M. McMasters, M.I. Ross, J.M. Kirkwood, M.B. Atkins,
J.A. Thompson, D.G. Coit, D. Byrd, R. Desmond, Y. Zhang, P.Y. Liu, G. Lyman,
A. Morabito, J. Clin. Oncol. 19 (2001) 3622–3643.
[2] B. Kasper, V. D’Hondt, P. Vereecken, A. Awada, Crit. Rev. Oncol. Hematol. 62
(2007) 16–22.
4.6. Hydrolysis in human plasma
[3] K.S.M. Smalley, M. Herlyn, Mini Rev. Med. Chem. 6 (2006) 387–393.
[4] A.M. Jordan, T.H. Khan, H.M. Osborn, A. Photiou, P.A. Riley, Bioorg. Med. Chem.
7 (1999) 1775–1780.
[5] P.A. Riley, Pigment Cell Res. 6 (2003) 548–552.
[6] M. Rooseboom, J.N.M. Commandeur, N.P.E. Vermeulen, Pharmacol. Rev. 56
(2004) 53–102.
[7] A.M. Jordan, T.H. Khan, H. Malkin, H.M. Osborn, A. Photiou, P.A. Riley, Bioorg.
Med. Chem. 9 (2001) 1549–1558.
[8] A.M. Jordan, T.H. Khan, H. Malkin, H.M. Osborn, Bioorg. Med. Chem. 10 (2002)
2625–2633.
[9] S. Knaggs, H. Malkin, H.M. Osborn, N.A.O. Williams, P. Yaqoob, Org. Biomol.
Chem. 3 (2005) 4002–4010.
Human plasma was obtained from the pooled, heparinised
blood of healthy donors, and was frozen and stored at ꢂ70 ꢀC prior
to use. For the hydrolysis experiments, the substrates were incu-
bated at 37 ꢀC in human plasma that had been diluted to 80% (v/v)
with pH 7.4 isotonic phosphate buffer. At appropriate intervals,
aliquots of 200 mL were added to 400 mL of acetonitrile to both
quench the reaction and precipitate plasma proteins. These
samples were centrifuged and the supernatant analyzed by HPLC
for the presence of substrate and products.

3234
M.J. Perry et al. / European Journal of Medicinal Chemistry 44 (2009) 3228–3234
[10] A.A. Tarhini, S.S. Agarwala, Dermatol. Ther. 19 (2006) 19–25.
[11] L. Chin, L.A. Garraway, D.E. Fisher, Genes Dev. 20 (2006) 2149–2182.
[12] T.A. Connors, P.M. Goddard, K. Merai, W.C.J. Ross, D.E.V. Wilman, Biochem.
Pharmacol. 25 (1976) 241–246.
[13] K. Vaughan, M.F.G. Stevens, Q. Rev. 7 (1978) 377–397.
[14] J.L. Skibba, D.D. Beal, G. Ramirez, G.T. Bryan, Cancer Res. 30 (1970) 147–150.
[15] F.W. Kru¨ger, R. Preusssman, N. Niepelt, Biochem. Pharmacol. 20 (1971) 529–
533.
[17] E. Carvalho, J. Iley, M.J. Perry, E. Rosa, J. Chem. Soc., Perkin Trans. 2 (1998)
2375–2380.
[18] J.C. Espin, R. Varon, L.G. Fenoll, M.A. Gilabert, P.A. Garcia-Ruiz, J. Tudela,
F. Garcia-Canovas, Eur. J. Biochem. 267 (2000) 1270–1279.
[19] S.Y. Seo, V.K. Sharma, N. Sharma, J. Agric. Food Chem. 51 (2003) 2837–2853.
[20] H.M. Osborn, N.A. Williams, Org. Lett. 18 (2004) 3111–3113.
[21] B. Gasowska-Bajer, H. Wojtasek, Bioorg. Med. Chem. Lett. (2008). doi:10.1016/
j.bmcl.2008.04.041.
[16] Sparc, Sparc Performs Automated Reasoning in Chemistry V.4.2 (2008).http://
ibmlc2.chem.uga.edu/sparc.
[22] J. Borovansky, R. Edge, E.J. Land, S. Navaratnam, S. Pavel, C.A. Ramsden,
P.A. Riley, N.P.M. Smit, Pigment Cell Res. 19 (2006) 170–178.