Schiff bases derived from p-aminobenzyl alcohol as trigger groups for pH-dependent prodrug activation

Article abstract of DOI:10.1016/j.tetlet.2010.06.055

A number of novel acid-sensitive Schiff bases derived from p-aminobenzyl alcohol and various benzaldehyde derivatives were synthesized and were subsequently shown to trigger benzyl elimination reactions. The kinetics of acid-catalyzed hydrolysis at pH 5.0 as well as stability at pH 7.4 were studied using fluorogenic model compounds. Two fluoro-substituted Schiff bases showed efficient hydrolysis at pH 5.0 combined with a long-term stability at pH 7.4 and are considered suitable candidates for the development of anticancer prodrugs.

Full text of DOI:10.1016/j.tetlet.2010.06.055

Tetrahedron Letters 51 (2010) 4371–4374
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Schiff bases derived from p-aminobenzyl alcohol as trigger groups
for pH-dependent prodrug activation
Ivonne A. Müller a, Felix Kratz a, Manfred Jung b, André Warnecke a,
*
a Tumor Biology Center, Breisacher Str. 117, 79106 Freiburg, Germany
b Institute of Pharmaceutical Sciences, University of Freiburg, Albertstr. 25, 79104 Freiburg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
A number of novel acid-sensitive Schiff bases derived from p-aminobenzyl alcohol and various benzalde-
hyde derivatives were synthesized and were subsequently shown to trigger benzyl elimination reactions.
The kinetics of acid-catalyzed hydrolysis at pH 5.0 as well as stability at pH 7.4 were studied using fluo-
rogenic model compounds. Two fluoro-substituted Schiff bases showed efficient hydrolysis at pH 5.0
combined with a long-term stability at pH 7.4 and are considered suitable candidates for the develop-
ment of anticancer prodrugs.
Received 7 May 2010
Revised 8 June 2010
Accepted 11 June 2010
Available online 16 June 2010
Keywords:
Ó 2010 Elsevier Ltd. All rights reserved.
Prodrug activation
Acid-sensitive
Benzyl elimination
Schiff base
Coupling low-molecular weight cytotoxic drugs to suitable car-
riers (macromolecules, antibodies, or receptor-specific ligands) is a
promising strategy to overcome the drawbacks associated with
conventional anticancer chemotherapy.1 For a controlled and
site-specific release of the carrier-bound anticancer drug, the
incorporation of acid-sensitive bonds (e.g., acyl hydrazone, acetal)
between the drug and its carrier is an efficient and a versatile ap-
proach.2 Since the cellular uptake of targeted prodrugs generally
occurs via the endocytic pathway, these bonds are cleaved due to
the significant shift from pH 7.2–7.4 in the blood or extracellular
spaces to pH 4.0–6.5 in the various intracellular compartments
such as endosomes or lysosomes. In the past, acyl hydrazone bonds
have been successfully used for the development of acid-sensitive
macromolecular prodrugs of doxorubicin.3 Most drugs, however,
do not provide an appropriate carbonyl or hydrazine moiety to
form acyl hydrazone bonds. Therefore we set out to develop a more
versatile method for a pH-specific prodrug activation that should
be applicable to a large number of drugs. By incorporating a self-
immolative linker, drugs with nucleophilic groups such as OH,
NH2, or SH can be chemically bound via transiently stable carbon-
ate, carbamate, or O,S-thiocarbonate bonds. A pH-dependent trig-
gering event causes a chemical breakdown of the linker with a
subsequent release of the drug. This modular strategy comprising
a trigger, a self-immolative linker, and the drug is often referred
to as the double prodrug concept.4
A suitable self-immolative linker is the p-aminobenzyloxycar-
bonyl (PABC) system (Scheme 1) first described by Katzenellenbo-
gen and co-workers5 Upon activation, the PABC linker reacts in a
O
E
O
Z
X= masked amino group
e.g. N3, NO2, R-CO-NH
X
stable
E= effector
1. de-masking of X
Z= NH, O, S
O
E
O
Z
H2N
2.1,6-benzyl elimination
(spontaneous)
OH
CO2
HZ
+
E
+
H2N
* Corresponding author. Tel.: +49 761 206 2175; fax: +49 761 206 2174.
E-mail address: warnecke@tumorbio.uni-freiburg.de (A. Warnecke).
Scheme 1. The p-aminobenzyloxycarbonyl (PABC) system.
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.06.055

4372
I. A. Müller et al. / Tetrahedron Letters 51 (2010) 4371–4374
rapid and well-understood 1,6-benzyl elimination6 and has there-
fore been employedfor thedesign ofvarious anticancerprodrugs7 or
even more complex self-immolative structures, such as oligomers,8
polymers,9 and dendrimers.10 Self-immolation of the PABC linker is
achieved by liberatingthe aromaticamino group from a masked pre-
cursor. In previous work, it was shown that the aniline of the PABC
system could be effectively masked as an azide, a nitro group, or
by acylation. De-masking was accomplished by reduction,8,10 enzy-
matically11 or chemically via the Staudinger reaction.12
Here we present novel PABC-based trigger groups that are acti-
vated by decreasing the pH. Since hydrolysis of Schiff bases (imi-
nes) occurs under mild acidic conditions, we assumed that this
reaction would be suitable for de-masking the PABC’s amino group.
However, it had to be shown that the formation of a Schiff base of-
fers an effective masking, i.e., prevents the PABC linker from benzyl
elimination at neutral pH. As a model compound we synthesized
the Schiff base derived from p-aminobenzyl alcohol and benzalde-
hyde that was subsequently reacted with 7-isocyanato-4-methyl-
coumarin (Scheme 2).
OH
H2N
H
OH
CHO
5 % AcOH
N
benzene reflux
1a
The fluorophore 7-amino-4-methyl-coumarin (AMC) thereby
serves as a model for an amino-functionalized drug. In contrast
to the acylated form, free AMC shows a strong blue fluorescence
(ex. 390 nm; em. 460 nm) and concentrations can be determined
conveniently in the micromolar or even nanomolar range.13 The re-
lease kinetics of the model compound (1b) was investigated at pH
5.0 (20 mM acetate buffer) and pH 7.4 (20 mM phosphate buffer)
by measuring the increase in fluorescence that directly correlates
with the liberation of AMC. All experiments were performed at
CH3
O
CH3
O
OCN
O
O
H
N
O
N
H
O
DBTL, THF, rt
1b
Scheme 2. Preparation of fluorogenic model compound 1b.
low concentrations (10
lM) to minimize potential catalytic or
inhibitory effects of the cleavage products.
Table 1
Yields of 1a/b–12a/b and half-lives of model compounds 1b–12b in buffer at pH 5.0 and 7.4 at 37 °C
CH3
CH3
O
H
O
N
H
O
O
O
O
H2N
R
N
1b-12b
fluorescent
Entry
R =
Yield xa (%)
Yield xb (%)
t1/2
t1/2
pH 5.0a (h)
pH 7.4b (h)
2a/b
3a/b
1a/b
4a/b
5a/b
77
81
77
73
71
87
95
66
84
76
0.28
0.33
0.57
0.45
0.62
26
13
8
NMe
2
OMe
H
14
36
F
Cl
O
6a/b
34
37
0.47
9
7a/b
8a/b
90
67
85
74
1.02
1.35
36
65
CF
3
NO2
Cl
Cl
9a/b
36
66
75
74
63
82
0.83
0.55
0.57
50
39
89
Cl
F
10a/b
11a/b
F
F
F
F
F
F
F
F
12a/b
84
87
2.88
365
F
F
a
Acetate buffer, 20 mM.
Phosphate buffer, 20 mM.
b

I. A. Müller et al. / Tetrahedron Letters 51 (2010) 4371–4374
4373
0
-1
-2
-3
-4
-5
At pH 5.0 we found a rapid release of AMC following a pseudo
first-order kinetics with t1/2 = 34 min. The hydrolysis of the Schiff
base is the rate-determining step because the subsequent 1,6-ben-
zyl elimination occurs rather rapidly (t1/2 = 16 s at pH 5.0, see Sup-
plementary data). In contrast, 1b proved to be significantly more
stable at pH 7.4 showing a half-life of 8 h. For the design of acid-
sensitive anticancer prodrugs, however, a long-term stability at
neutral pH (t1/2 >48 h) is required. Thus we were interested in both
which factors influence the kinetic behavior of 1b and how the sta-
bility at neutral pH could be further improved.
To clarify these issues, we synthesized a small library of Schiff
bases using benzaldehyde derivatives with different substituents
(2a–12a). Reaction with 7-isocyanato-4-methylcoumarin accord-
ing to Scheme 2 afforded the model compounds 2b–12b of which
the release kinetics were studied as described above (Table 1). We
11b
12b
osb
y = -0.8891x + 2.8414
k
y = -0.8748x + 1.9739
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
pH
found that the imine bond is generally cleaved rapidly (t1/2
=
17À173 min) at pH 5.0 whereas at pH 7.4 our model compounds
proved to be significantly more stable (t1/2 = 8À365 h). At both
pH values we observed a marked influence of the substituents of
the benzaldehyde component on the rate of cleavage with the gen-
eral finding that electron-withdrawing groups have a stabilizing
effect. In order to evaluate whether this effect can be described
by Hammett’s equation, we studied the linear free-energy relation-
ships (log(k/k0) vs rp) for the para-substituted compounds at both
pH values (Fig. 1).
For pH 5.0, the plot shows a linear relationship in good approx-
imation albeit with two exceptions: the data points for the substit-
uents p-Ac and p-NMe2. An explanation for the deviation of the
dimethylamino substituent might be that this group is partially
Figure 2. pH dependence of first-order rate constants for 11b and 12b (acetate
buffer: pH 4.5–6, phosphate buffer: pH 6.5–7.4, 37 °C).
relationship was not found. The impact of substituents on the sta-
bility, however, was generally more pronounced than at pH 5.
The stability of halogen-substituted Schiff bases at pH 7.4 could
be further improved by introducing additional substituents as
shown with the di-, tetra-, and pentafluoro- as well as the trichlo-
robenzaldehyde derivatives 9b–12b. Especially the 2,3,5,6-tetra-
fluoro- and pentafluorobenzaldehyde derivatives (11b, 12b) have
proven to be considerably stable at neutral pH making them inter-
esting candidates for medical applications. Therefore we studied
the release behavior of both compounds more extensively over
the physiologically relevant pH range (pH 4.5–7.4, Fig. 2).
Both compounds show a constant decrease in their rate of
hydrolysis over the observed pH range. Between pH and log kobs
a linear relation was found in good approximation. This should
be expected for a reaction with specific acid catalysis, that is, a
reaction with an acid–base pre-equilibrium with only the proton-
ated species (iminium) undergoing the next (rate-determining)
reaction steps as shown for the hydrolysis of imines in previous
mechanistic studies.14 For a reaction with specific acid catalysis
the first-order rate constant kobs can be described as a function of
pH: kobs = k0 + kcatÁ[H+] with k0 as the rate constant for the uncata-
lyzed reaction. For k0 ꢀ kcatÁ[H+] the slope of the graph should
reach 1 in theory whereas for 11b and 12b a slope of ꢁ0.9 was cal-
culated. This means only a slight deviation from ideal acid cataly-
sis. Moreover, the linear behavior suggests that the rate-
determining step does not change within the examined pH range.
Compounds 11b and 12b are the two model compounds with the
most efficient pH response and were thus selected for the develop-
ment of acid-sensitive prodrugs.
Although imine hydrolysis is a rather complex multistep reac-
tion15 and our analysis of the kinetic data does not take into account
all effects that may result from general acid/base catalysis,14 we feel
that the general trend of pH-dependent stability is apparent within
the spectrum of molecules that were studied.
In summary, we have developed a number of novel acid-sensitive
Schiff bases that are capable of triggering benzyl elimination reac-
tions. We showed that both acid sensitivity and stability at neutral
pH can be adjusted and fine-tuned by varying the substituents of
the benzaldehyde component. The Schiff base trigger groups, espe-
cially those derived from fluorine-substituted benzaldehydes, form
the basis for developing prodrugs with anticancer drugs or self-
immolative polymers that are degraded in an acidic environment.
protonated at acidic pH whereas the
tonated form. However, we have no explanation for the significant
deviation of the acetyl substituent. The reaction constant ( value)
rp value relates to the depro-
q
was calculated to be À0.57 thus indicating a moderate influence of
substituents at pH 5.0. In contrast, for pH 7.4, a linear free-energy
0.4
p-NMe2
0.2
p-OMe
p-F
p-Ac
0
p-Cl
/k
0
-0.2
-0.4
p-H
p-NO2
0.8
p-CF3
-0.8
-0.6
-0.4
-0.2
0
σ p
0.2
0.4
0.6
0.2
0
p-H
p-OMe
p-Ac
-0.2
-0.4
-0.6
-0.8
-1
p-F
p-NMe2
-0.6
p-Cl
p-CF3
p-NO2
0.8
-0.8
-0.4
-0.2
0
σ p
0.2
0.4
0.6
Acknowledgment
Figure 1. Linear free-energy relationships for the hydrolysis of para-substituted
model compounds 1b–8b at pH 5 (top) and pH 7.4 (bottom). rp values were
obtained from the literature.16
The financial support of the Deutsche Forschungsgemeinschaft
(DFG) is gratefully acknowledged.

4374
I. A. Müller et al. / Tetrahedron Letters 51 (2010) 4371–4374
Bioorg. Med. Chem. 2002, 10, 71–77; (d) Abu Ajaj, K.; Biniossek, M. L.; Kratz, F.
Supplementary data
Bioconjugate Chem. 2009, 20, 390–396; (e) Abu Ajaj, K.; Graeser, R.; Fichtner, I.;
Kratz, F. Cancer Chemother. Pharmacol. 2009, 64, 413–418.
Supplementary data (experimental section, characterization
data, and 1H and 13C NMR spectra of all compounds) associated
with this article can be found, in the online version, at
doi:10.1016/j.tetlet.2010.06.055.
8. Warnecke, A.; Kratz, F. J. Org. Chem. 2008, 73, 1546–1552.
9. (a) Sagi, A.; Weinstain, R.; Karton, N.; Shabat, D. J. Am. Chem. Soc. 2008, 130,
5434–5435; (b) Weinstain, R.; Sagi, A.; Karton, N.; Shabat, D. Chem. Eur. J. 2008,
14, 6857–6861.
10. de Groot, F. M.; Albrecht, C.; Koekkoek, R.; Beusker, P. H.; Scheeren, H. W.
Angew. Chem., Int. Ed. 2003, 42, 4490–4494.
11. (a) Niculescu-Duvaz, I.; Niculescu-Duvaz, D.; Friedlos, F.; Spooner, R.; Martin,
J.; Marais, R.; Springer, C. J. J. Med. Chem. 1999, 42, 2485–2489; (b) de Groot, F.
M. H. Bioorg. Med. Chem. Lett. 2002, 12, 2371–2376; (c) Dubowchik, G. M.;
Firestone, R. A. Bioorg. Med. Chem. Lett. 1998, 8, 3341–3346; (d) Jeffrey, S. C.;
Torgov, M. Y.; Andreyka, J. B.; Boddington, L.; Cerveny, C. G.; Denny, W. A.;
Gordon, K. A.; Gustin, D.; Haugen, J.; Kline, T.; Nguyen, M. T.; Senter, P. D. J.
Med. Chem. 2005, 48, 1344–1358.
References and notes
1. Kratz, F.; Müller, I. A.; Ryppa, C.; Warnecke, A. ChemMedChem 2008, 3, 20–53.
2. Kratz, F.; Beyer, U.; Schütte, M. T. Crit. Rev. Ther. Drug Carrier Syst. 1999, 16,
245–288.
3. (a) Kratz, F. Expert Opin. Invest. Drugs 2007, 16, 855–866; (b) Etrych, T.; Chytil,
P.; Jelinkova, M.; Rihova, B.; Ulbrich, K. Macromol. Biosci. 2002, 2, 43–52.
4. Bundgaard, H. Adv. Drug Delivery Rev. 1989, 3, 39–65.
12. van Brakel, R.; Vulders, R. C.; Bokdam, R. J.; Grull, H.; Robillard, M. S.
Bioconjugate Chem. 2008, 19, 714–718.
13. Linn, C. P.; Mize, P. D.; Hoke, R. A.; Quante, J. M.; Pitner, J. B. Anal. Biochem.
1992, 200, 400–404.
5. Carl, P. L.; Chakravarty, P. K.; Katzenellenbogen, J. A. J. Med. Chem. 1981, 24,
479–480.
14. Pitea, D.; Favini, G.; Grasso, D. J. Chem. Soc., Perkin Trans. 2 1974, 1595–
1600.
15. (a) Cordes, E. H.; Jencks, W. P. J. Am. Chem. Soc. 1963, 85, 2843–2848; (b)
Cordes, E. H.; Jencks, W. P. J. Am. Chem. Soc. 1962, 84, 832–837.
6. Wakselman, M. Nouv. J. Chem. 1983, 7, 439–447.
7. (a) Greenwald, R. B.; Pendri, A.; Conover, C. D.; Zhao, H.; Choe, Y. H.; Martinez,
A.; Shum, K.; Guan, S. J. Med. Chem. 1999, 42, 3657–3667; (b) Niculescu-Duvaz,
D.; Niculescu-Duvaz, I.; Friedlos, F.; Martin, J.; Spooner, R.; Davies, L.; Marais,
R.; Springer, C. J. J. Med. Chem. 1998, 41, 5297–5309; (c) Damen, E. W. P.;
Nevalainen, T. J.; van den Bergh, T. J. M.; de Groot, F. M. H.; Scheeren, H. W.
16. Chapman, N. B.; Shorter, J. Correlation Analysis in Chemistry
Advances; Plenum: New York, 1978.
– Recent