Bivalent bendamustine and melphalan derivatives as anticancer agents

Article abstract of DOI:10.1016/j.ejmech.2011.02.008

The alkylating agents bendamustine and melphalan are currently used in the treatment of various tumoral diseases. In order to increase their antitumor potency and tumor selectivity both compounds were integrated in structure-activity relationship studies

Full text of DOI:10.1016/j.ejmech.2011.02.008

European Journal of Medicinal Chemistry 46 (2011) 1604e1615
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Bivalent bendamustine and melphalan derivatives as anticancer agents
,b,
Ana Maria Scutaru a, Maxi Wenzel a, Ronald Gust a
*
a Institute of Pharmacy, Freie Universität Berlin, Königin Luise Str. 2þ4, 14195 Berlin, Germany
b Institute of Pharmacy, Department of Pharmaceutical Chemistry, University of Innsbruck, Innrain 52a, A-6020 Innsbruck, Austria
a r t i c l e i n f o
a b s t r a c t
Article history:
The alkylating agents bendamustine and melphalan are currently used in the treatment of various
tumoral diseases. In order to increase their antitumor potency and tumor selectivity both compounds
were integrated in structureeactivity relationship studies including new drug carrier systems. Here we
describe the synthesis and the cytotoxicity of new bivalent bendamustine and melphalan derivatives.
Two molecules each esterified with N-(2-hydroxyethyl)maleimide were connected by diamines with
various chain lengths (n ¼ 6, 7, 8, 12). It was supposed that these conjugates (5aed, 10aed, 11aed) cause
cytotoxic effects preferred as bivalent drug. Indeed the cytotoxicity of the new compounds increased
compared to bendamustine and melphalan as determined in concentration-dependent in vitro assays
using the human MCF-7 and MDA-MB-231 breast cancer cell lines.
Received 10 August 2010
Received in revised form
1 February 2011
Accepted 4 February 2011
Available online 3 March 2011
Keywords:
Bendamustine
Melphalan
Dimer
Ó 2011 Elsevier Masson SAS. All rights reserved.
Drug targeting
Breast cancer
1. Introduction
myeloma (bone-marrow cancer), breast cancer and chronic
myelogenous leukemia [17,18].
Bendamustine firstly synthesized by Ozegowski and Krebs [1] is
a very promising drug with applicability in the treatment of various
tumoral diseases such as non-Hodgkin’s lymphoma and multiple
myeloma [2e9].
Further clinical trials illustrated some remarkable results in the
treatment of chronic lymphocytic leukemia (CLL) [10e12] and thus,
in March 2008, bendamustine was also approved in the USA for this
indication [13]. Furthermore, bendamustine has a promising future
in the treatment of breast cancer as it has been shown by recent
studies [14e16].
The design of bendamustine and melphalan is based on the
knowledge that the use of alkylating agents in the treatment of
cancer is restricted due to the diversity of the side effects (for
example: nephrotoxicity, nausea and vomiting). Thus melphalan
was obtained by combining the N-lost moiety with the amino acid
phenylalanine whereas in bendamustine the N-lost group is bound
to a purine-like heterocycle [19e22]. This drug design should
guarantee a higher accumulation of the compounds in tumor cells
compared to normal cells, because of a higher biomolecule traf-
ficking in fast growing (tumor) cells.
Its structure can be divided into three parts: a bis(2-chloroethyl)-
amino (N-lost) group providing the alkylating properties, a benz-
imidazole core and a carboxylic acid side chain mediating water
solubility. This structure assigned it to the class of nitrogen mustard
agents comparable to melphalan, which is used in the treatment of
cancer, including ovarian cancer, malignant melanoma, multiple
Since the discovery of melphalan in the late 1950s, many efforts
have been made to synthesize derivatives of melphalan with the
purpose of diminishing the side effects of the drug [22e24]. A
suitable concept might be the tumor targeting using natural and
synthetic polymers as carrier as already demonstrated for chlor-
ambucil by Kratz et al. [25,26]. They combined the anticancer agent
in a first step with a maleimide spacer and bound it in a second step
selectively to thiol groups present in the carrier protein, the Human
Serum Albumin (HSA) [27,28]. These HSA conjugates accumulated
in the tumors due to the EPR (enhanced permeability and reten-
tion) effect. The same conjugates were built in situ with endoge-
nous HSA, if the drug-maleimide is administered iv. This pro-drug
(drug-carrier) design resulted in an enhanced antitumor activity
accompanied by a better compatibility due to lower unwanted toxic
side effects.
Abbreviations: DADC, 1,12-diaminododecane; DAHEPT, 1,7-diaminoheptane;
DAH, 1,6-diaminohexane; DAOCT, 1,8-diaminooctane; EPR (enhanced permeability
and retention) effect, the property by which macromolecules build up in the tumor
due to the special properties of the cells (e.g., poor lymphatic drainage); FCS, fetal
calf serum; PBS, phosphate-buffered saline; DCC, N,N-dicyclohexylcarbodiimide;
DMAP, (4-dimethyl-aminopyridine).
* Corresponding author. Institute of Pharmacy, Department of Pharmaceutical
Chemistry, University of Innsbruck, Innrain 52a, A-6020 Innsbruck, Austria.
Tel.: þ43 512 507 5245; fax: þ43 512 507 2940.
E-mail address: ronald.gust@uibk.ac.at (R. Gust).
0223-5234/$ e see front matter Ó 2011 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2011.02.008

A.M. Scutaru et al. / European Journal of Medicinal Chemistry 46 (2011) 1604e1615
1605
In our approach bendamustine and melphalan were attached to
small synthetic dendrimers. For this purpose, both drugs esterified
with N-(2-hydroxyethyl)maleimide (analogously to the investiga-
tions of Kratz et al.) were bound via the maleimide moiety to
1,3,5-tris(3-aminopropyl)benzene (G0) and its G1 analog 3,5-bis-
(3-aminopropyl)-N-(3-{3,5-bis[3-{3,5-bis(3-aminopropyl)benzoy-
lamino}propyl]phenyl}propyl)benzamide. Compounds G0 and G1
have seen to be suitable synthetic carriers in previous studies with
platinum complexes [29,30].
The study of the drug-dendrimer conjugates includes also
comparable investigations of melphalan and its maleimide ester
with the free and Boc-protected amino groups (compounds 8 and 9,
see Scheme 2). In accordance to Bergel and Stock [23] the cytotoxic
effects were unaffected by the esterification of the COOH group but
strongly decreased by protection of the NH2 group. It should be
mentioned that these structural modifications did not reflect on the
reactivity of the N-lost group of melphalan. The hydrolysis rates of 8
and 9 were comparable to melphalan.
crosslinking achieved by the N-lost group plays a subordinate role
in the antitumor activity of bendamustine.
In order to further optimize the in vitro cytotoxicity of bend-
amustine and melphalan, we decided to use the concept of bivalent
drugs. The positive results of amino-terminated dendrimers and
the increased cellular accumulation of diaminoalkyl connected
platinum drugs [31] induced us to link two N-(2-hydroxyethyl)-
maleimide-drug conjugates to diamines with variable chain
lengths (1,6-diaminohexane (DAH), 1,7-diaminoheptane (DAHEPT),
1,8-diaminooctane (DAOCT) and 1,12-diaminododecane (DADC))
(see Schemes 1 and 2). The synthesis and the activity against the
MCF-7 and MDA-MB-231 mammary carcinoma cell lines are
described in this paper.
2. Results and discussion
2.1. Chemistry
In contrast, esterification of bendamustine with N-(2-hydroxy-
ethyl)maleimide (4a, Scheme 1) strongly improved the hydrolytic
stability of the N-lost moiety and consequently the cytotoxicity in
cell culture experiments [30]. Although very active in vivo, bend-
amustine was inactive in vitro because of its fast hydrolysis under
cell culture conditions. Therefore, it is assumed, that the DNA
Two different strategies were evaluated for the synthesis of the
dimers 5aed (bendamustine, Scheme 1), 10aed (melphalan Boc-
protected, Scheme 2) and 11aed (melphalan, Scheme 2).
A first approach included a two steps route: 1) the reaction of
N-(2-hydroxyethyl)maleimide with the diamine through a Michael-
like addition; the maleimide being used as an electrophilic olefin and
O
O
O
OH
H
N
OH
a
H2N
+
OH
NH2
N
N
N
n
N
H
n
O
O
O
2 a:
1
n= 6
3 a: n= 6
b:
b: n= 7
n= 7
c:
n= 8
c: n= 8
d:
d: n= 12
n= 12
Cl
Cl
O
N
N
O
N
b, 24h
N
O
N
N
O
+ 1
N
OH
Cl
O
Cl
4a
4
b
3 a, b, c, d
+
Cl
Cl
N
Cl
O
N
O
N
N
H
N
N
O
N
N
N
O
n
O
N
H
Cl
O
O
O
5 a: n= 6
b:
n= 7
c: n= 8
d: n= 12
Scheme 1. Synthesis of the bivalent bendamustine conjugates. Reagents and conditions: a) chloroform, reflux; b) DCC, DMAP, CH2Cl2, RT.

1606
A.M. Scutaru et al. / European Journal of Medicinal Chemistry 46 (2011) 1604e1615
Cl
Cl
O
O
HO
N
N
a
NH2
HN
O
O
N
+
O
Cl
Cl
O
OH
OH
6
7
1
b
O
O
Cl
H2N
O
O
Cl
HN
O
O
O
O
O
N
c
N
N
N
8
Cl
Cl
O
9
d
+ 2a-d
Cl
Cl
N
Cl
N
Cl
O
O
O
NHX
N
O
H
O
XHN
N
HN
N
O
O
n
O
X= Boc 10 a: n= 6
b: n= 7
c:
n= 8
d: n= 12
e
11 a:
X= H·HCl
n= 6
b: n= 7
c:
n= 8
d:
n= 12
Scheme 2. Synthesis of the bivalent melphalan derivatives. Reagents and conditions: a) Boc2O, TEA, methanol, RT, 1 h; b) DCC, DMAP, CH2Cl2, RT, 24 h; c) aqueous HCl, THF, RT, 24 h;
d) chloroform, reflux; e) aqueous HCl, THF, RT.
the diamine as a nucleophile [32], 2) a Steglich reaction, involving the
free acid group of the drug, the bismaleimide-adducts and typical
reagents like DCC (N,N-dicyclohexylcarbodiimide) and catalytic
amounts of DMAP (4-dimethylaminopyridine).
The second approach implicated at first the Steglich esterifica-
tion of drug and spacer, followed by the addition of the diamines
linker to the double bond at the maleimide moiety.
Taking the reaction conditions into consideration and the need
for high yields, the first strategy was chosen for the formation of the
bendamustine derivatives and the second one for the synthesis of
the melphalan conjugates.
The route of preparing the dimers of bendamustine is depicted
in Scheme 1. N-(2-hydroxyethyl)maleimide (1) was refluxed in
chloroform with the respective diamine (2aed) to give the prod-
ucts 3aed. The addition to the maleimides results in asymmetric
centers and the building of isomers which were not separated.
Bendamustine hydrochloride was transformed into the free base 4,
activated by addition of DCC followed by conversion into the
desired compounds 5aed after addition of the correspondent bis-
maleimide-adducts 3aed and DMAP.
Melphalanemaleimide derivatives were synthesized from
L-melphalan (6) after Boc-protection and reaction with the spacer

A.M. Scutaru et al. / European Journal of Medicinal Chemistry 46 (2011) 1604e1615
1607
reference substances) possessed comparable activity in both cell
lines. The time response curves of MCF-7 and MDA-MB-231 cells
are presented in supporting information. For discussion the
(maximal) effect against MCF-7 cells after an incubation time of
72 h are presented in Figs. 1e4.
140
120
100
80
0.5 µM
1 µM
5 µM
Cisplatin influenced the cell growth in a concentration-depen-
dent manner and caused 40% growth inhibition at the highest
60
applied concentration (5
Bendamustine (4) was nearly inactive at all tested concentra-
tions (1, 5 and 10 M, see Fig. 2), while melphalan (6) caused
M.
mM) (Fig. 1).
40
20
m
cytostatic effects (T/Ccorr ¼ 20%, Fig. 3) at 10
m
0
The esterification of bendamustine (4) with the N-(2-hydrox-
yethyl)maleimide giving compound 4a increased the activity highly
-20
-40
-60
-80
(Fig. 2). Already 5 mM of 4a caused cytostatic effects. All bivalent
bendamustine derivatives (5aed) were more active than the parent
compound 4 but not as active as 4a. The dimers 5b and 5c caused
similar antiproliferative effects at 10 mM and were slightly more
active than 5a and 5d. A clear preference of a chain length could not
Cisplatin
Fig. 1. In vitro cytotoxicity of cisplatin at 72 h in the MCF-7 cell line. For detailed
information on the investigation of the T/Ccorr [%] and
section.
be detected from these results.
s [%] values see Experimental
The antiproliferative effects of -melphalan (6) depended on the
L
derivatization of the amino group as well as the carboxyl group. The
Boc-protected compound 7 proved to be inactive in both cell lines
[30]. The esterification with N-(2-hydroxyethyl)maleimide not only
compensated this inactivation but led to compound 8 with
increased cytotoxicity (Fig. 3). Interestingly, the connection of two
molecules by a diaminoalkyl spacer (10aed) drastically reduced the
antiproliferative activity (T/Ccorr values of about 50% were achieved
as depicted in Scheme 2 and described in the literature [33,34]. The
resulting ester 8 (Boc-protected) was purified by column chroma-
tography and treatment with aqueous HCl to yield 9.
Heating of 8 with the respective diamines 2aed in chloroform
gave the Boc-protected dimers 10aed which were transformed into
the final compounds 11aed by treatment with aqueous HCl
(Scheme 2). No attempts were made to separate 10aed and 11aed
into optically pure compounds.
at 10 mM). The effect of 10b and 10c were somewhat more active
than 10a and 10d (see Fig. 3).
The cleavage of the Boc groups in 10aed highly increased the
cytotoxicity of the bivalent drugs (Fig. 3). Compound 11a was
All final compounds and those used in the biological tests were
characterized by their fully assigned 1H and 13C NMR spectra and
molecular ion peaks in the mass spectra. The conjugates were
soluble in water and in buffered cell culture media as judged by
visual examination.
cytostatic at 10
The most active compound 11d was tested in lower concentrations
(Fig. 4) and caused cytotoxic effects even at 1 M.
mM, while 11bc were even cytocidal at 5 and 10 mM.
m
From the results depicted in Figs. 2e4 (and also from the data
included in the supporting information) it is obvious that the
concept of bivalent drugs was successful. The activity of melphalan
was strongly increased and in the case of bendamustine in vitro
active bendamustine compounds were obtained.
2.2. Antitumor activity
The antiproliferative activity of the synthesized compounds was
determined utilizing human MCF-7 and MDA-MB-231 breast
cancer cell lines. From the graphs antiproliferative and cytostatic
effects (80% > T/Ccorr > 0%) as well as cytocidal effects (
< 0%) can be deduced.
s
¼ T/Ccorr
2.3. Discussion
Cisplatin, bendamustine (4) and melphalan (6) were used as
references. The new synthesized compounds (as well as the
Melphalan and bendamustine are alkylating agents and
currently used in the treatment of various tumoral diseases. The
1 µM
5 µM
10 µM
140
120
100
80
60
40
20
0
-20
-40
-60
4
4a
5a
5b
5c
5d
-80
Fig. 2. In vitro cytotoxicity of bendamustine (4), its maleimide (4a) and bivalent derivatives (5aed) at 72 h in the MCF-7 cell line. For detailed information on the investigation of the
T/Ccorr [%] and [%] values see Experimental section.
s

1608
A.M. Scutaru et al. / European Journal of Medicinal Chemistry 46 (2011) 1604e1615
140
120
100
80
1 µM
5 µM
10 µM
60
40
20
0
-20
-40
-60
-80
6
8
10a
10b
10c 10d
11a
11b
11c
Fig. 3. In vitro cytotoxicity of melphalan (6), its Boc-protected maleimide derivative (8) as well as its Boc-protected bivalent derivatives (10aed) and of the compounds 11aec at 72 h
in the MCF-7 cell line. For detailed information on the investigation of the T/Ccorr [%] and
s
[%] values see Experimental section.
mode of action of alkylating drugs included a selective DNA-
interaction and the formation of interstrand cross links [35e37].
The N-lost moiety forms an aziridinium ion which attacks the
nucleobases. Therefore, it is necessary to transfer the intact alky-
lator into the cells and prevent hydrolysis in the cell medium.
Bendamustine and melphalan are non-stable in aqueous solu-
tions including the media used for cell culture experiments
[38e40]. After hydrolysis into the dihydroxyl species the cytotox-
icity is lost. Nevertheless, melphalan was distinctly more active
than bendamustine against MCF-7 and MDA-MB-231 cells (see Figs.
2 and 3 and supporting information) which point to different
accumulation kinetics into the cells. Melphalan might be trans-
ferred through the cell membrane e.g. by amino acid carrier
systems prior to hydrolysis. This would explain the loss of activity
after Boc-protection, which prevent the recognition by the trans-
porter. Active transport of melphalan into cells was confirmed e. g.
for the ASC-like amino acid system at low concentrations and the
amino acid transport system L at higher concentrations [41].
With the objective to increase uptake and accumulation of
melphalan in tumor cells a series of dipeptide derivatives were
synthesized [23,24]. Other approaches focused on the combina-
tions with carriers for instance to increase the uptake in the central
nervous system for the therapy of intracerebral tumors (for further
information about pharmaceutics, combinations, Co-drugs and
prodrugs of melphalan see Wickström et al. [22]).
A new melphalan derivative represents the L-melphalanyl-p-L-
fluorophenylalanine ethyl ester (J1) which was designed for the
treatment of neuroblastoma, the most common and deadly tumor
of childhood [42]. J1 is rapidly incorporated into the cytoplasm
followed by intracellular ester hydrolysis resulting in a release of
melphalan. The enzymes responsible for the activation have to be
identified to be aminopeptidases. This concept could also be real-
ized using ester derivatives which can be cleaved due to the low
lysosomal pH or intracellular esterases.
Because up to now a carrier system for the uptake of bend-
amustine into tumor cells could not be confirmed we bound
bendamustine to low molecular weight dendrimers of the G0 and
G1 type [43] (Fig. 5) to achieve endocytotic or macropinocytotic
accumulation in tumor cells as already demonstrated for platinum
complexes [29]. The same modification was done with melphalan.
In the first part of our structure-activity relationship (SAR) study
[30] we investigated the stability of the dendrimer bound drugs in
aqueous solution as well as the cytotoxicity against MCF-7 and
MDA-MB-231 cells in relation to the free drugs and the drug-mal-
eimide conjugates.
Data obtained from the stability studies clearly documented that
the esterification of bendamustine decreased the degradation in
aqueous solution. The same experiments done with melphalan did
not indicate higher stability. Under physiological conditions (PBS,
pH 7.4) the bendamustine-maleimide ester showed a half-life of
10h. Interestingly, the maleimide derivative of melphalan under-
went in this experiment a complete hydrolysis. Binding of the
maleimide esters to diaminoalkanes did not change the hydrolysis
profile. For instance bendamustine release by ester cleavage and
hydrolysis processes were detected in comparable amounts also for
the dimeric compounds.
The dimeric bendamustine derivatives were more active than
the dendrimer derivatives (under consideration of the number of
bound drug molecules) but not as active as the monomeric mal-
eimide ester 4a. This is an indication for the participation of
a secondary effect of the maleimide moiety in the case of 4a. As
demonstrated by Kratz et al., maleimide derivatives can effectively
bind to HSA which act as a carrier molecule but also prevent
140
0.5 µM
1 µM
2.5 µM
120
100
80
60
40
20
0
-20
11d
-40
-60
-80
Fig. 4. In vitro cytotoxicity of the bivalent melphalan derivative 11d at 72 h in the MCF-
7 cell line. For detailed information on the investigation of the T/Ccorr [%] and
values see Experimental Section.
s [%]

A.M. Scutaru et al. / European Journal of Medicinal Chemistry 46 (2011) 1604e1615
1609
XHN
XHN
NHX
XHN
O
O
NH
XHN
NH
XHN
NHX
NHX
G0: X = H·HCl
XHN
HN
O
G1: X = H·CF3COOH
Fig. 5. Structure of G0 and G1 basic dendrimers.
hydrolysis reactions. Hopwood and Stock showed that, by coupling
nitrogen mustard to proteins or semi-synthetic polymers, the rate
of hydrolysis can be slowed down and, in some cases, this proce-
dure led to a better biological activity [44]. In the first SAR study
[30] fast interactions of the compounds 4a and 9 with HSA were
observed.
The derivatization of melphalan to bivalent drugs was more
efficient than in the case of bendamustine. Compound 5b as
dimeric bendamustine derivative was cytostatic (T/Ccorr < 20%),
whereas 5a, 5c and 5d exhibited antiproliferative effects
(20% < T/Ccorr < 80%). In contrast all dimeric melphalan derivatives
were cytocidally active (T/Ccorr < 0%).
It should be emphasized that all bivalent derivatives of melp-
halan and bendamustine exist as a mixture of isomers, therefore it
can be argued that higher cytostatic effects may be expected for
some of the optically pure compounds. A separation of the optically
pure compounds is planned in a subsequent study.
This finding confirms the positive results of other attempts to
optimize the pharmacological effects of antitumor drugs, e.g. plat-
inum complexes. Farrell and co-workers have successfully synthe-
sizedmultinuclearplatinumcomplexeswithbridginglinkers leading
to circumvention of cisplatin resistance. These compounds were
found to possess a different mode of action [45]. In our group we
could demonstrate a higher accumulation of platinum complexes in
MCF-7 cells due to an active transport into the tumor cells [46].
Unfortunately, the bivalent bendamustine and melphalan
derivatives are not suitable to study the uptake into tumor cells,
because of a missing tracer. Therefore, we will label in a forth-
coming study the derivatives in such a way that either high reso-
lution continuum source atomic absorption or fluorescence
analysis can be used for pharmacological studies.
In continuation of this study, wewill quantify the cellular uptake and
the mode of action of the anticancer agents, e.g. the DNA-binding.
4. Material and methods
4.1. General
1H and 13C NMR spectra were taken on a Bruker Avance/DPX
400 MHz with TMS as internal standard. Melting points (uncor-
rected) weremeasured with a Büchi Melting Point B-545. All column
chromatography purifications were done using silica gel 60H (230
mesh ASTM, Merck). TLC was performed on silica gel 60 GF254 plates
(Merck). EI spectrawere recorded on a Thermo-Fisions VG Auto Spec
(70 eV). The ESI-TOF spectra were measured on an Agilent 6210 ESI-
TOF, Agilent Technologies, Santa Clara, CA, USA. Solvent flow rate
was adjusted to 4 mL/min, Spray voltage set to 4 kV. Drying gas flow
rate was set to 15 psi (1 bar). All other parameters were adjusted for
a maximum abundance of the relative [M þ H]þ.
4.2. Synthesis
N-tert-Butoxycarbonyl-4-[bis(2-chloroethyl)amino]-
anine (7) was prepared according to literature procedures [33].
Bendamustine hydrochloride was a gift from Bendalis. -Melphalan
hydrochloride was synthesized as described by Bergel and Stock
[47] starting from -phenylalanine. All other starting materials were
L-phenylal-
L
L
purchased from commercial sources and were used without puri-
fication. Solvents were dried under standard conditions.
4.2.1. General procedure for the synthesis of the bismaleimide-
adducts 3aed (mixture of optical isomers)
3. Conclusions
N-(2-Hydroxyethyl)maleimide (1) and the diamine (2aed) were
dissolved in 50 mL of chloroform and refluxed for 24 h. The solvent
was removed in vacuo. The residue was chromatographed on
a silica gel column to afford the desired product.
This work focused on the development of bivalent bendamustine
and melphalan derivatives with increased cytotoxicity. The in vitro
assay in MCF-7 and MDA-MB-231 breast cancer cell lines showed an
important increase of the antitumor activity of bendamustine
conjugates compared to the free drug. With exception of the dia-
minododecane-derivative, all bivalent compounds exhibited cyto-
static properties in both cell lines. The analogous melphalan
derivatives were much more active and reached cytocidal effects at
4.2.1.1. N,N0-Bis[1-(2-hydroxyethyl)pyrrolidine-2,5-dione-3-yl]-1,6-
diaminohexane (3a). N-(2-Hydroxyethyl)maleimide (2.12 g, 15
mmol), 1,6-diaminohexane (581 mg, 5 mmol), column chroma-
tography with chloroform/methanol ¼ 5:1. Yield: 1.6 g (80%) as an
yellow gum: Rf ¼ 0.22 (chloroform/methanol 5:1). 1H NMR (CDCl3):
the highest used concentrations (5 and 10
and C12 spacer were about 10 fold higher cytotoxic than melphalan.
mM). Dimers with C7, C8
d
¼ 1.37 (s, 4H, CH2DAH), 1.50 (quin, 3J(H,H) ¼ 6.49 Hz, 4H, CH2DAH),
2.54e2.63 (m, 3H, 1 CH2), 2.67e2.73 (m, 3H, 1 CH2), 2.93e2.97 (m,

1610
A.M. Scutaru et al. / European Journal of Medicinal Chemistry 46 (2011) 1604e1615
Cl
O
6
H
N
1
11
9
13
1
2
9
9
8
HN
O
N
N
10
N
5
N
O
1
9
12
7
N
O
O
2
Cl
3
O
HN
4
OH
Chart 3. (Supporting scheme for NMR data of compound 5b).
Chart 1. (Supporting scheme for NMR data of compound 3a).
2H, 2 CH), 3.66e3.74 (m, 4H, CH2Nsuccinimide), 3.77e3.81 (m,
4H þ 2H, CH2OH þ CH) (Chart 1).
4.2.2.1. N,N0-Bis[(2-(2,5-dioxopyrrolidin-1-yl)ethyl-4-(5-(bis(2-
chloroethyl)amino)-1-methyl-1H-benzo[d]imidazol-2-yl)butanoate)-
3-yl]-1,6-diaminohexane (5a). Bendamustine as free base (700 mg,
1.95 mmol), 3a (394 mg, 0.97 mmol), 80 mL of dry dichloro-
methane, DCC (450 mg, 2.18 mmol), 80 mL of dry dichloromethane,
column chromatography with chloroform/methanol ¼ 20:1 and
10:1. Yield: 250 mg (23.73%) as a brown oil: Rf ¼ 0.4 (dichloro-
4.2.1.2. N,N0-Bis[1-(2-hydroxyethyl)pyrrolidine-2,5-dione-3-yl]-1,7-dia-
minoheptane (3b). N-(2-Hydroxyethyl)maleimide (2.12 g, 15 mmol),
1,7-diaminoheptane (651 mg, 5 mmol), column chromatography with
chloroform/methanol ¼ 10:1. Yield: 1.64 g (80%) as an orange gum:
Rf ¼ 0.22 (chloroform/methanol 10:1). 1H NMR (CDCl3):
d
¼ 1.34 (s,
methane/methanol 10:1). 1H NMR (CDCl3):
d
¼ 1.24 (s, 4H, 1 CH2),
6H, CH2 DAHEPT), 1.52 (m, 4H, CH2 DAHEPT), 2.55e2.74 (m, 4H þ 2H,
CH2NH þ CH2succinimide), 2.93e2.99 (m, 2H, CH2succinimide), 3.72 (m, 4H,
CH2Nsuccinimide), 3.76e3.82 (m, 4H þ 2H, CH2OH þ CH).
1.32 (s, 4H, 2 CH2), 2.15 (m, 4H, 3 CH2), 2.43e2.56 (m, 8H, 4 CH2),
2.62e2.66 (m, 2H, 5 CH2), 2.90 (m, 4H þ 2H, 5 þ 6 CH2), 3.60e3.65
(m, 11H, 7 þ 8 þ 9 CH2), 3.69e3.71 (m, 17H, 7 þ 8 þ 9 CH2), 4.22 (m,
4H, 10 CH2), 6.76 (dd, 3J(H,H) ¼ 1.73 Hz, 8.64 Hz, 2H, 11 CH), 7.05 (d,
3J(H,H) ¼ 2.56 Hz, 2H, 12 CH), 7.17 (d, 3J(H,H) ¼ 8.79 Hz, 2H, 13 CH).
4.2.1.3. N,N0-Bis[1-(2-hydroxyethyl)pyrrolidine-2,5-dione-3-yl]-1,8-dia-
minooctane (3c). N-(2-Hydroxyethyl)maleimide (2.12 g, 15 mmol),
1,8-diaminooctane (721.3 mg, 5 mmol), column chromatography with
chloroform/methanol ¼ 20:1. Yield: 1.6 g (80%) as a yellow solid:
Rf ¼ 0.25 (chloroform/methanol 10:1); mp: 124e125 ꢀC. 1H NMR
13C NMR (CDCl3):
d
¼ 22.53 (3 CH2), 26.18 (6 CH2), 26.35 (1 CH2),
29.67 (2 CH2), 32.85 (CH3), 32.96 (4 þ 5 CH2), 37.90 (9 CH2), 40.8
(7 þ 8 CH2), 47.51 (4 CH2), 54.78 (9 CH), 60.86 (10 CH2), 103.01 (12
CH), 109.81 (11 CH), 110.83 (13 CH), 128.82 (CNCH3), 142.75 (CN¼C),
143.00 (Caromatic), 154.36 (N¼C), 172.98 (COO), 173.62 (COsuccinimide),
175.05 (COsuccinimide). MS (ESI): calcd. for C50H69N10O8Cl4 1079.40,
found 1079.3947 (Chart 2).
(CDCl3):
d
¼ 1.31 (s, 8H, CH2 DAOCT), 1.50 (m, 4H, CH2 DAOCT), 2.17 (s, 4H,
NH þ OH), 2.54e2.63 (m, 4H þ 2H, CH2NH þ CH2succinimide),
2.94e3.04 (m, 2H, CH2succinimide), 3.71 (m, 4H, CH2Nsuccinimide),
3.75e3.81 (m, 4H þ 2H, CH2OH þ CH).
4.2.1.4. N,N0-Bis[1-(2-hydroxyethyl)pyrrolidine-2,5-dione-3-yl]-1,12-
diaminododecane (3d). N-(2-Hydroxyethyl)maleimide (1.27 g,
9 mmol), 1,12-diaminododecane (601.1 mg, 3 mmol), column chro-
matography with dichloromethane/methanol ¼ 10:1. Yield: 700 mg
(48.3%) as a white solid: Rf ¼ 0.33 (dichloromethane/methanol 10:1);
4.2.2.2. N,N0-Bis[(2-(2,5-dioxopyrrolidin-1-yl)ethyl-4-(5-(bis(2-
chloroethyl)amino)-1-methyl-1H-benzo[d]imidazol-2-yl)butanoate)-
3-yl]-1,7-diaminoheptane (5b). Bendamustine as free base (335 mg,
0.93 mmol), 3b (193 mg, 0.46 mmol), 25 mL of dry dichloromethane,
DCC (211.6 mg, 1.025 mmol), 25 mL of dry dichloromethane,
column chromatography with chloroform/methanol ¼ 20:1. Yield:
370 mg (72.5%) as a brown oil: Rf ¼ 0.38 (chloroform/methanol
mp: 170 ꢀC. 1H NMR (CDCl3):
d
¼ 1.27 (s, 16H, CH2DADC), 1.49 (quin, 3J
(H,H) ¼ 7.14 Hz, 4H, CH2DADC), 2.13 (s, br, 4H, NH þ OH), 2.54e2.72 (m,
4H þ 2H, 1 CH2), 2.93 (d, 3J(H,H) ¼ 8.28 Hz, 1H, 2 CH), 2.97 (d, 3J
(H,H) ¼ 8.29 Hz, 1H, 2 CH), 3.69e3.73 (m, 4H, CH2Nsuccinimide),
3.77e3.81 (m, 4H þ 2H, CH2OH þ CH).
20:1). 1H NMR (CDCl3):
d
¼ 1.28 (s, 6H,1 CH2),1.45 (s, 4H, 2 CH2), 2.15
(quin, 3J(H,H) ¼ 7.15 Hz, 4H, 3 CH2), 2.45 (m, 4H, 4 CH2), 2.50e2.57
(m, 4H, 5 CH2), 2.62 (m, 2H, 6 CH2), 2.89 (m, 2H þ 4H, 6 þ 7 CH2),
3.61 (m, 8H, 8 CH2), 3.69e3.76 (m, 20H, 9 CH2), 4.22 (m, 4H,10 CH2),
6.77 (dd, 3J(H,H) ¼ 2.03 Hz, 8.76 Hz, 2H, 11 CH), 7.05 (d, 3J
4.2.2. General procedure for the synthesis of the bendamustine
spacer diamines 5aed (mixture of optical isomers)
(H,H) ¼ 1.73 Hz, 2H,12 CH), 7.16 (d, 3J(H,H) ¼ 8.75 Hz, 2H,13 CH). 13
C
Bendamustine as free base, DMAP (tip of a spatula) and the
bismaleimide-adducts (3aed) were dissolved in dry dichloro-
methane at room temperature. DCC, dissolved in dry dichloro-
methane, was added dropwise to this solution within 1 h, and the
mixture was stirred for additional 24 h. The solution was filtered
and evaporated in vacuo. The residue was chromatographed on
a silica gel column to afford the desired product.
NMR (CDCl3):
d
¼ 22.22 (3 CH2), 26.12 (7 CH2), 26.98 (1 CH2), 29.18
(CH2), 29.67 (2 CH2), 29.95 (CH3), 32.95 (6 CH2), 36.00 (4 CH2), 37.91
(9 CH2), 40.82 (8 þ 9 CH2), 47.6 (5 CH2), 56.18 (9 CH), 60.88 (10 CH2),
102.4 (12 CH), 110.03 (11 CH), 110.96 (13 CH), 129.00 (CNCH3), 142.12
(CN¼C), 143.02 (Caromatic), 154.18 (N¼C), 172.96 (COO), 175.05
(COsuccinimide), 177.50 (COsuccinimide). MS (ESI): calcd. for
C51H71N10O8Cl4 1093.4186, found 1093.4158; C51H71N10O8Cl42þ
546.2139, found 546.2118 (Chart 3).
O
5
6
O
Cl
13
8
11
9
Cl
9
HN
13
11
HN
9
7
9
9
N
N
N
10
10
N
8
5
N
N
O
O
1
1
12
9
9
6
N
O
7
O
N
12
4
2
2
Cl
3
Cl
O
3
NH
O
NH
4
4
Chart 2. (Supporting scheme for NMR data of compound 5a).
Chart 4. (Supporting scheme for NMR data of compound 5c).

A.M. Scutaru et al. / European Journal of Medicinal Chemistry 46 (2011) 1604e1615
1611
4.2.2.3. N,N0-Bis[(2-(2,5-dioxopyrrolidin-1-yl)ethyl-4-(5-(bis(2-
chloroethyl)amino)-1-methyl-1H-benzo[d]imidazol-2-yl)butanoate)-
3-yl]-1,8-diaminooctane (5c). Bendamustine as free base (335 mg,
0.93 mmol), 3c (200 mg, 0.46 mmol), 25 mL of dry dichloromethane,
DCC (211.6 mg,1.025 mmol), 25 mL of dry dichloromethane, column
chromatography with chloroform/methanol ¼ 20:1 and chloro-
form/methanol ¼ 10:1. Yield: 180 mg (35%) as a brown oil: Rf ¼ 0.36
30 mL of dry dichloromethane at room temperature. DCC (176.5 mg,
0.85 mmol), dissolved in 25 mL of dry dichloromethane, was added
dropwise to this solution within 1 h, and the solution was then
stirred for additional 20 h. The solution was filtered and evaporated
in vacuo. The red residue was chromatographed on a silica gel
column (hexane/ethyl acetate ¼ 2:1) to afford the desired product.
Yield: 137 mg (32%) as a yellow solid: Rf ¼ 0.3 (hexane/ethyl acetate
(chloroform/methanol 10:1).1H NMR (CDCl3):
d
¼ 1.21 (s, 8H,1 CH2),
2:1); mp: 101 ꢀC. 1H NMR (CDCl3):
d
¼ 1.29 (s, 9H, CH3), 2.62 (m, 1H,
1.39 (s, 4H, 2 CH2), 2.07 (quin, 3J(H,H) ¼ 7.18 Hz, 4H, 3 CH2), 2.39 (m,
4H, 4 CH2), 2.44e2.51 (m, 4H, 5 CH2), 2.55e2.61 (m, 2H, 6 CH2), 2.82
(m, 2H þ 4H, 6 þ 7 CH2), 3.55 (m, 8H, 8 CH2), 3.62e3.69 (m, 20H, 9
CH2), 4.14 (m, 4H,10 CH2), 6.70 (dd, 3J(H,H) ¼ 2.15 Hz, 8.77 Hz, 2H,11
CH), 6.99 (d, 3J(H,H) ¼ 2.01 Hz, 2H, 12 CH), 7.10 (d, 3J(H,H) ¼ 8.75 Hz,
CH2Ar), 2.74 (dd, 3J(H,H) ¼ 4.46 Hz, 9.44 Hz, 1H, CH2Ar), 3.62 (t, 3J
(H,H) ¼ 5.28 Hz, 2H, CH2 maleimide), 3.66 (s, 8H, CH2CH2Cl), 4.09 (m,
2H, CH2O), 4.22 (m,1H, CH), 6.63 (d, 3J(H,H) ¼ 8.52 Hz, 2H, ArH), 7.02
(d, 2H þ 2H, ArH þ Hmaleimide), 7.17 (d, 3J(H,H) ¼ 8.01 Hz,1H, NH). 13
C
NMR (CDCl3):
d
¼ 28.31 (CH3), 36.64 (CH2Ar), 36.98 (CH2Nmaleimide),
2H,13 CH).13C NMR (CDCl3):
d
¼ 22.23 (3 CH2), 26.25 (7 CH2), 27.06 (1
40.46 (CH2NAr þ CH2Cl), 54.49 (CH), 62.11 (CH2O), 79.84 (C(CH3)3),
112.09 (Caromatic), 124.8 (Caromatic), 130.56 (Caromatic), 134.25 (CHma-
leimide), 145.04 (Caromatic), 155.07 (COOBoc), 170.29 (COmaleimide),
171.75 (COO). MS (EI, 80 eV, 300 ꢀC); m/z (%): 527.2 (5.93) [Mþ ꢁ 1],
404.1 (0.36) [MꢁC6H6NO2]þ, 230.3 (100) C15H14NCl2.
CH2), 29.31 (CH2), 29.87 (2 CH2), 32.99 (CH3 þ 6 CH2), 36.12 (4 CH2),
37.86 (9 CH2), 40.83 (8 þ 9 CH2), 47.65 (5 CH2), 56.26 (9 CH), 60.81 (10
CH2),102.84 (12 CH),109.90 (11 CH),110.88 (13 CH),129.28 (CNCH3),
142.80 (CN¼C þ Caromatic), 154.30 (N¼C), 172.95 (COO), 175.07
(COsuccinimide), 177.65 (COsuccinimide). MS (ESI): calcd. for
C52H73N10O8Cl4 1107.4342, found 1107.4327, C52H73N10O8Cl42þ
554.2208, found 554.2196 (Chart 4).
4.2.4. O-{4-[Bis(2-chloroethyl)amino]-L-phenylalanyl}-2-hydroxy-
ethylmaleimide hydrochloride (9)
8 (350 mg, 0.66 mmol) was dissolved in 6 mL of THF. To this
solution was added 250
mixture was stirred for 20 h at room temperature. The solvent was
removed under reduced pressure. The remained oil was dissolved
in methanol and the product was precipitated with diethyl ether,
filtered and dried, to afford 245 mg (80%) as a colorless solid; mp:
4.2.2.4. N,N0-Bis[(2-(2,5-dioxopyrrolidin-1-yl)ethyl-4-(5-(bis(2-
chloroethyl)amino)-1-methyl-1H-benzo[d]imidazol-2-yl)butanoate)-
3-yl]-1,12-diaminododecane (5d). Bendamustine as free base
(660 mg, 1.86 mmol), 3d (410 mg, 0.85 mmol), 50 mL of dry di-
chloromethane, DCC (422 mg, 2.05 mmol), 50 mL of dry dichloro-
methane, column chromatography with dichloromethane/
methanol ¼ 20:1. Yield: 560 mg (56.7%) as a crimson solid: Rf ¼ 0.46
(dichloromethane/methanol 20:1); mp: 322 ꢀC. 1H NMR (CDCl3):
m
L of hydrochloric acid (w ¼ 25%) and the
130 ꢀC. 1H NMR (CD3OD):
d
¼ 3.00 (dd, 3J(H,H) ¼ 7.89 Hz, 14.59 Hz,
1H, CH2Ar), 3.14 (dd, 3J(H,H) ¼ 5.22 Hz, 14.60 Hz, 1H, CH2Ar), 3.69
(m, 4H, CH2Cl), 3.78 (m, 4H þ 2H, CH2CH2Cl þ CH2maleimide), 4.18 (m,
1H, CH), 4.38 (m, 1H, CH), 6.77 (d, 3J(H,H) ¼ 8.81 Hz, 2H, ArH), 6.86
(s, 2H, Hmaleimide), 7.11 (d, 3J(H,H) ¼ 8.75 Hz, 2H, ArH); 13C NMR
d
¼ 1.25 (s, 16H,1 CH2), 1.48 (quin, 3J(H,H) ¼ 7.02 Hz, 4H, 2 CH2), 2.15
(t, 3J(H,H) ¼ 7.00 Hz, 4H, 3 CH2), 2.46 (t, 3J(H,H) ¼ 6.84 Hz, 4H, 4 CH2),
2.50e2.60 (m, 4H, 5 CH2), 2.63e2.70 (m, 2H, 6 CH2), 2.89e2.96 (m,
2H þ 4H, 6 þ 7 CH2), 3.64 (t, 3J(H,H) ¼ 6.8 Hz, 8H, 8 CH2), 3.72e3.78
(m, 20H, 9 CH2), 4.22 (quin, 3J(H,H) ¼ 5.54 Hz, 4H, 10 CH2), 6.79 (dd,
3J(H,H) ¼ 2.3 Hz, 8.81 Hz, 2H, 11 CH), 7.08 (d, 3J(H,H) ¼ 2.22 Hz, 2H,
12 CH), 7.19 (d, 3J(H,H) ¼ 8.81 Hz 2H, 13 CH). 13C NMR (DMSO-d6):
(CD3OD):
d
¼ 37.55 (CH2Nmaleimide), 39.67 (CH2Ar), 41.07 (CH2N),
42.6 (CH2Cl), 53.7 (CH), 60.14 (CH2O), 119.5 (Caromatic), 132.50 (Car-
omatic), 132.54 (Caromatic), 132.7 (Caromatic), 135.43 (CHmaleimide),
135.68 (Caromatic), 170.33 (COmaleimide), 172.3 (COO). MS (ESI): calcd.
for C19H24N3O4Clþ2 428.1138, found 428.1152.
d
¼ 24.34 (3 CH2), 25.23 (7 CH2), 26.18 (CH2), 28.48 (CH2), 28.82 (1
CH2), 29.87 (2 CH2), 32.49 (CH3), 33.24 (4 þ 6 CH2), 37.16 (9 CH2),
41.18 (8 þ 9 CH2), 47.38 (5 CH2), 52.99 (9 CH2), 60.27 (10 CH2), 110.68
(12 CH), 117.62 (11 þ 13 CH), 127.62 (CNCH3), 143.27 (CN¼C), 146.05
(Caromatic), 156.56 (N¼C), 171.88 (COO), 172.29 (COsuccinimide), 173.75
(COsuccinimide). MS (ESI): calcd. for C56H81N10O8Cl4 1163.4963, found
1163.4892 ( Chart 5).
4.2.5. General procedure for the synthesis of the Boc-protected
L
-melphalan-spacer-diamines 10aed (mixture of optical isomers)
8 and the respective diamine were dissolved in chloroform and
40 h refluxed. The solvent was removed in vacuo and the oily
residue was purified by column chromatography (silica gel).
4.2.5.1. N,N0-Bis[(2-(2,5-dioxopyrrolidin-1-yl)ethyl-3-(4-(bis(2-chloro-
ethyl)amino)phenyl)-2-(tert-butoxycarbonylamino)propanoate)-3-yl]-
1,6-diaminohexane(10a). 8(255.8 mg, 0.48mmol),1,6-diaminohexane
(25.6 mg, 0.22 mmol), 10 mL of chloroform, column chromatography
4.2.3. O-{N-tert-Butoxycarbonyl-4-[bis(2-chloroethyl)amino]-L-
phenylalanyl}-2-hydroxyethylmaleimide (8)
7 (330 mg, 0.81 mmol), DMAP (tip of a spatula) and N-
(2-hydroxyethyl)maleimide (441 mg, 3.1 mmol) were dissolved in
O
H
N
7
9
Cl
3
7
8
N
O
8
N
1
Cl
3
O
Cl
5
12
5
O
11
2
O
2
N
1
H
HN
O
8
4
N
13
3
HN
N
O
O
NH
N
9
6
10
10
9
7
O
4
N
6
9
Cl
O
2
Chart 5. (Supporting scheme for NMR data of compound 5d).
Chart 6. (Supporting scheme for NMR data of compound 10a).

1612
A.M. Scutaru et al. / European Journal of Medicinal Chemistry 46 (2011) 1604e1615
Cl
as a brown solid: Rf ¼ 0.22 (dichloromethane/acetone 1:1); mp:
110 ꢀC. 1H NMR (CDCl3):
1.49 (s, 4H, 2 CH2), 2.58 (m, 4H, 3 CH2), 2.69 (m, 2H, 4 CH2), 2.86e3.03
(m, 2H þ 4H, 4 þ 5 CH2), 3.62 (t, 3J(H,H) ¼ 6.33 Hz, 8H, 6 CH2), 3.69 (t,
3J(H,H) ¼ 6.50 Hz, 8H, 7 CH2), 3.76 (m, 6H, 7 CH2), 4.24 (m, 2H, 9 CH2),
4.35 (m, 2H, 9 CH2), 4.45 (m, 2H, 9 CH2), 4.94 (m, 2H, NH), 6.60 (d, 3J
6
d
¼ 1.30 (s, 8H, 1 CH2), 1.40 (s, 18H, Boc CH3),
5
N
9
O
10
4
H
N
H
N
2
Cl
1
7
3
(H,H) ¼ 8.57 Hz, 4H, 10 CH), 7.00 (d, 3J(H,H) ¼ 8.05 Hz, 4H, 11 CH). 13
C
N
O
3
7
O
NMR (CDCl3):
d
¼ 26.74 (CH2), 28.33 (CH2 þ CH3), 29.27 (CH2), 31.74
8
O
8
(CH2), 36.8 (4 CH2), 37.66 (5 þ 8 CH2), 40.45 (6 þ 7 CH2), 47.84 (3 CH2),
53.48 (9 CH2), 54.49 (8 CH2), 61.26 (9 CH2), 79.86 (C),112.08 (Caromatic),
124.84 (Caromatic), 130.56 (Caromatic), 145.07 (Caromatic), 155.16 (CO),
171.82 (CO). MS (ESI): calcd. for C56H83N8O12Cl4 1201.4862, found
1201.4890 (Chart 8).
HN
O
O
Chart 7. (Supporting scheme for NMR data of compound 10b).
with chloroform/methanol ¼ 20:1. Yield: 165 mg (64%) as a yellow oil:
4.2.5.4. N,N0-Bis[(2-(2,5-dioxopyrrolidin-1-yl)ethyl-3-(4-(bis(2-
chloroethyl)amino)phenyl)-2-(tert-butoxycarbonylamino)prop-
anoate)-3-yl]-1,12-diaminododecane (10d). 8 (150 mg, 0.28 mmol),
1,12-diaminododecane (25.8 mg, 0.13 mmol), 5 mL of chloroform,
column chromatography with dichloromethane/acetone ¼ 6:1 and
3:1. Yield: 114 mg (70%) as a yellow oil: Rf ¼ 0.64 (dichloromethane/
Rf ¼ 0.45 (chloroform/methanol 20:1). 1H NMR (CDCl3):
¼ 1.25 (s, 4H,
d
1 CH2), 1.49 (s, 22H, 2 CH2), 2.46e2.70 (m, 4H þ 2H, 3 CH2), 2.87e3.03
(m, 2H þ 4H, 3 þ 4 CH2), 3.62 (t, 3J(H,H) ¼ 6.29 Hz, 8H, 5 CH2), 3.69 (t, 3J
(H,H) ¼ 6.35 Hz, 8H, 6CH2), 3.76(m, 6H, 7CH2), 4.24(m, 2H, 8CH2), 4.33
(m, 2H, 8 CH2), 4.45 (m, 2H, 8 CH2), 4.94 (m, 2H, NH), 6.60 (d, 3J
(H,H) ¼ 8.63 Hz, 4H, 9 CH), 7.00 (d, 3J(H,H) ¼ 7.88 Hz, 4H, 10 CH). 13
C
acetone 3:1). 1H NMR (CDCl3):
d
¼ 1.39 (s, 16H, 1 CH2), 1.33 (s, 18H, 2
NMR (CDCl3):
d
¼ 26.89 (1 CH2), 28.33 (CH3), 29.69 (2 CH2), 33.94 (3
CH2), 1.41 (quin, 3J(H,H) ¼ 7.04 Hz, 4H, 3 CH2), 2.40e2.64 (m,
4H þ 2H, 4 CH2), 2.80e2.96 (m, 2H þ 4H, 5 þ 4 CH2), 3.54 (t, 3J
(H,H) ¼ 6.27 Hz, 8H, 6 CH2), 3.61 (t, 3J(H,H) ¼ 6.47 Hz, 8H, 7 CH2),
3.69 (m, 6H, 8 CH2), 4.16 (m, 2H, 9 CH2), 4.26 (m, 2H, 9 CH2), 4.37 (m,
2H, 9 CH2), 4.89 (m, 2H, NH), 6.54 (d, 3J(H,H) ¼ 8.69 Hz, 4H, 10 CH),
CH2), 37.63 (4 þ 7 CH2), 40.46 (5 þ 6 CH2), 47.62 (3 CH2), 53.48 (8 CH2),
56.14 (7 CH2), 61.29 (8 CH2), 79.86 (C), 112.09 (9 CH), 124.88 (Caromatic),
130.56 (Caromatic), 145.08 (Caromatic), 155.15 (CO), 171.81 (CO). MS (ESI):
calcd. for C54H79N8O12Cl4 1173.45422, found 1173.4615 (Chart 6).
6.93 (d, 3J(H,H) ¼ 7.26 Hz, 4H, 11 CH). 13C NMR (CDCl3):
d
¼ 26.8
4.2.5.2. N,N0-Bis[(2-(2,5-dioxopyrrolidin-1-yl)ethyl-3-(4-(bis(2-chloro-
ethyl)amino)phenyl)-2-(tert-butoxycarbonylamino)propanoate)-3-yl]-
1,7-diaminoheptane (10b). 8 (485 mg, 0.91 mmol), 1,7-dia-
minoheptane (60 mg, 0.45 mmol), 25 mL of chloroform, column
chromatography, successivewithdichloromethane/acetone¼ 6:1, 3:1,
1:1. Yield: 270 mg (49.72%) as a brown solid: Rf ¼ 0.22 (dichloro-
(CH2), 28.6 (CH3), 29.27 (1 CH2), 31.74 (3 CH2), 33.44 (4 CH2), 37.66
(5 þ 8 CH2), 41.35 (6 þ 7 CH2), 47.10 (4 CH2), 54.98 (9 CH2), 58.49 (8
CH2), 62.26 (9 CH2), 80.01 (C), 112.08 (10 CH), 125.84 (Caromatic),
130.47 (Caromatic), 145.87 (Caromatic), 156.16 (CO), 172.82 (CO). MS
(ESI): calcd for C60H91N8O12Cl4 1257.54963, found 1257.5461
(Chart 9).
methane/acetone 1:1); mp: 120 ꢀC. 1H NMR (CDCl3):
d
¼ 1.32 (s, 6H, 1
CH2),1.40 (s,18H, Boc CH3),1.49 (s, 4H, 2 CH2), 2.49e2.70 (m, 4H þ 2H,
3 CH2), 2.86e3.03 (m, 2H þ 4H, 3 þ 4 CH2), 3.62 (t, 3J(H,H) ¼ 6.33 Hz,
8H, 5 CH2), 3.69 (t, 3J(H,H) ¼ 6.49 Hz, 8H, 6 CH2), 3.76 (m, 6H, 7 CH2),
4.21 (m, 2H, 8 CH2), 4.33 (m, 2H, 8 CH2), 4.45 (m, 2H, 8 CH2), 4.94 (m,
2H, NH), 6.60 (d,3J(H,H) ¼ 8.62Hz, 4H, 9 CH), 7.00 (d, 3J(H,H) ¼ 8.24 Hz,
4.2.6. General procedure for the synthesis of the deprotected L-
melphalan-spacer-diamines 11aed (mixture of optical isomers)
10aed were respectively dissolved in THF. To this solution
hydrochloric acid (w ¼ 25%) was added and the mixture was stirred
for 3 h at room temperature. The reaction was continuously
monitored by TLC. The solvent was then removed under reduced
pressure to afford the deprotected product.
4H,10 CH).13C NMR (CDCl3):
d
¼ 28.33 (CH2 þ CH3), 29.27 (CH2), 36.76
(3 CH2), 37.71 (4 þ 7 CH2), 40.46 (5 þ 6 CH2), 47.68 (3 CH2), 53.48 (8
CH2), 54.49 (7 CH2), 61.22 (8 CH2), 79.86 (C), 112.09 (Caromatic), 124.88
(Caromatic), 130.56 (Caromatic), 145.07 (Caromatic), 155.2 (CO), 171.83 (CO).
MS (ESI): calcd. for C55H81N8O12Cl4 1187.4705, found 1187.4730
(Chart 7).
4.2.5.3. N,N0-Bis[(2-(2,5-dioxopyrrolidin-1-yl)ethyl-3-(4-(bis(2-chloro-
ethyl)amino)phenyl)-2-(tert-butoxycarbonylamino)propanoate)-3-yl]-
1,8-diaminooctane (10c). 8 (485 mg, 0.91 mmol), 1,8-diaminooctane
(66 mg, 0.45 mmol), 25 mL of chloroform, column chromatography
with dichloromethane/acetone ¼ 3:1 and 1:1. Yield: 360 mg (65.45%)
1
4
3
HN
4
8
4
NH
O
O
O
2
H
8
N
O
N
3
O
8
NH
O
9
N
8
9
5
O
Cl
Cl
4
1
O
9
6
O
O
HN
9
5
11
6
10
2
O
N
7
O
Cl
HN
N
10
11
7
Cl
Chart 8. (Supporting scheme for NMR data of compound 10c).
Chart 9. (Supporting scheme for NMR data of compound 10d).

A.M. Scutaru et al. / European Journal of Medicinal Chemistry 46 (2011) 1604e1615
1613
O
O
O
O
H
N
H
7
N
6
5
3
6
7
N
N
3
O
Cl
O
Cl
Cl
3
1
1
10
10
5
8
O
O
2
12
9
H2N
13
11
H2N
2
N
HN
N
4
4
8
11
HN
9
12
Cl
Chart 12. (Supporting scheme for NMR data of compound 11c).
Chart 10. (Supporting scheme for NMR data of compound 11a).
CH), 130.70 (13 CH), 130.81 (Caromatic), 142.40 (Caromatic), 169.76
(COO), 172.28 (COsuccinimide), 172.89 (COsuccinimide). MS (ESI): calcd.
for C45H65N8O8Cl4 987.3652, found 987.3879 (Chart 11).
4.2.6.1. N,N0-Bis[(2-(2,5-dioxopyrrolidin-1-yl)ethyl-3-(4-(bis(2-
chloroethyl)amino)phenyl)-2-aminopropanoate)-3-yl]-1,6-diamino-
hexane dihydrochloride (11a). 10a (550 mg, 0.46 mmol), 40 mL of
THF, 5.12 mL of hydrochloric acid. Yield: 450 mg (92%) as green oil
4.2.6.3. N,N0-Bis[(2-(2,5-dioxopyrrolidin-1-yl)ethyl-3-(4-(bis(2-chloro-
ethyl)amino)phenyl)-2-aminopropanoate)-3-yl]-1,8-diaminooctane
dihydrochloride (11c). 10c (330 mg, 0.27 mmol), 25 mL of THF, 3 mL of
hydrochloric acid. Yield: 279 mg (95%) as a brown solid after lyophi-
after lyophilization from water. 1H NMR (CD3OD):
d
¼ 1.51 (s, br, 4H,
1 CH2), 1.79 (m, 4H, 2 CH2), 2.97 (dd, 3J(H,H) ¼ 5.43 Hz, 17.97 Hz, 2H,
3 CH2), 3.07e3.12 (m, 4H, 3 þ 4 CH2), 3.17e3.25 (m, 2H þ 4H, 4 þ 5
CH2), 3.64e3.71 (m, 2H þ 4Hþ8H, 6 þ 7þ8 CH2), 3.83 (t, 3J
(H,H) ¼ 6.68 Hz, 8H, 9 CH2), 4.19 (t, 3J(H,H) ¼ 5.43 Hz, 1H, 11 CH),
4.29 (m, 1H, 11 CH), 4.40 (m, 2H, 10 CH2), 4.49 (m, 2H, 10 CH2), 6.96
(d, 3J(H,H) ¼ 8.38 Hz, 4H, 12 CH), 7.24 (d, 3J(H,H) ¼ 8.46 Hz, 4H, 13
lization from water; mp: 129 ꢀC. 1H NMR (CD3OD):
d
¼ 1.43 (s, br, 8H, 1
CH2), 1.75 (m, 4H, 2 CH2), 2.91e3.24 (m, 4H þ 4H þ 2H, 3 þ 4 CH2),
3.36 (m, 2H, 4 CH2), 3.67 (m, 8H þ 6H þ 2H, 5 þ 6 þ 7 CH2), 3.83 (m,
8H þ 2H, 8 þ 9 CH2), 4.40 (m, 2H,10 CH2), 4.47 (m, 2H,10 CH2), 6.78 (d,
3J(H,H) ¼ 8.55 Hz, 4H, 11 CH), 7.16 (d, 3J(H,H) ¼ 8.55 Hz, 4H, 12 CH). 13
C
CH). 13C NMR (CD3OD):
d
¼ 25.52 (1 CH2), 31.86 (2 CH2), 35.07 (3
NMR (CD3OD):
d
¼ 25.81 (1 CH2), 28.23 (CH2), 31.85 (2 CH2), 35.05 (3
CH2), 38.67 (4 þ 7 CH2), 41.32 (8 þ 9 CH2), 46.24 (5 CH2), 53.82 (11
CH), 55.37 (6 CH), 57.74 (10 CH2), 116.70 (13 CH), 130.81 (12 CH),
130.93 (Caromatic), 141.54 (Caromatic), 169.71 (COO), 172.74 (COsuccini-
mide), 172.89 (COsuccinimide). MS (ESI): calcd. for C44H63N8O8Cl4
973.3495, found 973.3498 (Chart 10).
CH2), 38.72 (6 CH2), 38.85 (4 CH2), 41.29 (5 þ 8 CH2), 46.47 (3 CH2),
53.79 (9 CH), 55.41 (7 CH), 57.71 (10 CH2), 116.66 (11 CH), 130.83 (12
CH), 130.95 (Caromatic), 141.59 (Caromatic), 169.73 (COO), 172.27 (COsuc-
cinimide), 172.93 (COsuccinimide). MS (ESI): calcd. for C46H67N8O8Cl4
1001.3809, found 1001.4197 (Chart 12).
4.2.6.2. N,N0-Bis[(2-(2,5-dioxopyrrolidin-1-yl)ethyl-3-(4-(bis(2-
chloroethyl)amino)phenyl)-2-aminopropanoate)-3-yl]-1,7-diamino-
heptane dihydrochloride (11b). 10b (290 mg, 0.24 mmol), 20 mL of
THF, 2.67 mL of hydrochloric acid. Yield: 240 mg (93%) as a brown
solid after lyophilization from water; mp: 213 ꢀC. 1H NMR (CD3OD):
4.2.6.4. N,N0-Bis[(2-(2,5-dioxopyrrolidin-1-yl)ethyl-3-(4-(bis(2-
chloroethyl)amino)phenyl)-2-aminopropanoate)-3-yl]-1,12-diamino-
dodecane dihydrochloride (11d). 10d (280 mg, 0.22 mmol), 20 mL of
THF, 2.43 mL of hydrochloric acid. Yield: 245 mg (98%) as a brown
solid after lyophilization from water; mp: 194e196 ꢀC. 1H NMR
d
¼ 1.46 (s, br, 6H, 1 CH2), 1.76 (m, 4H, 2 CH2), 2.95 (dd, 3J
(CD3OD):
d
¼ 1.32 (s, 16H, 1 CH2), 1.77 (m, 4H, 2 CH2), 3.05 (m, 2H, 3
(H,H) ¼ 5.59 Hz, 17.98 Hz, 2H, 3 CH2), 3.04e3.12 (m, 2H, 3 CH2),
3.15e3.25 (m, 4H þ 4H, 4 þ 5 CH2), 3.64e3.70 (m, 2H þ 4H þ 8H,
6 þ 7 þ 8 CH2), 3.83 (t, 3J(H,H) ¼ 6.65 Hz, 8H, 9 CH2), 4.17 (m, 1H, 11
CH), 4.27 (m, 1H, 11 CH), 4.40 (m, 2H, 10 CH2), 4.47 (m, 2H, 10 CH2),
6.86 (d, 3J(H,H) ¼ 8.59 Hz, 4H, 12 CH), 7.20 (d, 3J(H,H) ¼ 8.61 Hz, 4H,
1
13 CH). 13C NMR (CD3OD):
d
¼ 25.65 (1 CH2), 27.86 (CH2), 31.87 (2
4
CH2), 35.07 (3 CH2), 37.65 (7 CH2), 39.00 (4 CH2), 41.34 (8 þ 9 CH2),
2
HN
46.45 (5 CH2), 53.82 (11 CH), 54.89 (6 CH), 57.76 (10 CH2), 115.83 (12
8
3
O
O
NH
H
6
N
N
O
5
7
10
NH2
N
7
O
Cl
5
1
3
O
10
11
8
O
O
12
13
11
H2N
O
2
N
4
13
NH
9
Cl
12
N
6
9
Cl
Cl
Chart 11. (Supporting scheme for NMR data of compound 11b).
Chart 13. (Supporting scheme for NMR data of compound 11d).

1614
A.M. Scutaru et al. / European Journal of Medicinal Chemistry 46 (2011) 1604e1615
CH2), 3.17 (m, 2H, 3 CH2), 3.30 (s, 4H, 4 CH2), 3.38 (m, 4H, 5 CH2),
3.68 (m, 8H þ 4H, 6 þ 7 CH2), 3.87 (m, 2H, 8 CH), 4.07 (m, 8H, 9
CH2), 4.34 (m, 2H, 10 CH2), 4.43 (m, 2H, 10 CH2), 4.55 (m, 1H, 11 CH),
4.64 (m, 1H, 11 CH), 7.56 (m, 4H, 12 CH), 7.61 (m, 4H, 13 CH). 13C
Acknowledgments
The presented study was supported by grants Gu285/4-1,
Gu285/5-1 and Gu285/5-2 as well as the SFB 765 from the Deutsche
Forschungsgemeinschaft.
NMR (CD3OD):
d
¼ 27.50 (CH2), 30.15 (CH2), 30.41 (2 CH2), 33.24 (3
CH2), 36.24 (7 CH2), 36.50 (5 CH2), 40.91 (9 CH2), 42.73 (6 CH2),
48.00 (4 CH2), 53.67 (11 CH), 55.24 (8 CH), 59.17 (10 CH2), 116.26 (13
CH), 131.99 (Caromatic), 132.09 (Caromatic), 145.09 (Caromatic), 170.50
(COO), 173.67 (COsuccinimide), 174.15 (COsuccinimide). MS (ESI): calcd.
for C50H74N8O8Cl4 1057.4436, found 1057.4456 (Chart 13).
Appendix. Supplementary data
Supplementary data related to this article can be found online at
doi:10.1016/j.ejmech.2011.02.008.
References
4.3. Biological methods
[1] W. Ozegowski, D. Krebs, J. Prakt. Chem. 20 (1963) 178e186.
[2] B.D. Cheson, M.J. Rummel, J. Clin. Oncol. 27 (2009) 1492e1501.
[3] K.U. Chow, S. Boehrer, K. Geduldig, A. Krapohl, D. Hoelzer, P.S. Mitrou,
E. Weidmann, Haematologica 86 (2001) 485e493.
[4] A. Forero-Torres, M.N. Saleh, Clin. Lymphoma Myeloma 8 (2007) 13e17.
[5] S. Lonial, Clin. Lymphoma Myeloma 8 (2007) 18e23.
[6] G.L. Plosker, N.J. Carter, Drugs 68 (2008) 2645e2660.
[7] W. Ponisch, M. Rozanski, H. Goldschmidt, F.A. Hoffmann, T. Boldt,
A. Schwarzer, U. Ritter, R. Rohrberg, E. Schwalbe, J. Uhlig, T. Zehrfeld,
V. Schirmer, A. Haas, U. Kreibich, D. Niederwieser, Biochem. Pharmacol. 143
(2008) 191e200.
[8] D. Schrijvers, J.B. Vermorken, Semin. Oncol. 29 (2002) 15e18.
[9] E. Weidmann, S.Z. Kim, A. Rost, H. Schuppert, G. Seipelt, D. Hoelzer,
P.S. Mitrou, Ann. Oncol. 13 (2002) 1285e1289.
[10] M.A. Bergmann, M.E. Goebeler, M. Herold, B. Emmerich, M. Wilhelm, C. Ruelfs,
L. Boening, M.J. Hallek, Haematologica 90 (2005) 1357e1364.
[11] A. Ferrajoli, Haematologica 90 (2005) 1300A.
4.3.1. Cell culture
The human MCF-7 and MDA-MB-231 breast cancer cell lines
were obtained from the American Type Culture Collection (ATCC).
The MCF-7 breast cancer cell line originated from a 69-year-old
Caucasian woman and is a well-characterized estrogen receptor
(ER) positive control cell line (cells are positive for cytoplasmic
estrogen receptors). The human cell line MDA-MB-231 is a proto-
type for the study of hormone-independent breast cancer. Cell line
banking and quality control were performed according to the seed
stock concept reviewed by Hay [48]. Both cell lines were main-
tained in L-glutamine and sodium pyruvate containing DMEM High
Glucose (4.5 g/L) supplemented with 5% fetal calf serum (FCS,
Gibco) using 25 cm2 culture flasks in a humidified atmosphere (5%
CO2) at 37 ꢀC. The cell lines were passaged weekly after treatment
with trypsin (0.05%)/ethylenediaminetetraacetic acid (EDTA, 0.02%,
Boehringer). Mycoplasma contamination was regularly monitored
and only mycoplasma-free cultures were used.
[12] H. Koppler, J. Heymanns, A. Pandorf, R. Weide, Leuk. Lymphoma 45 (2004)
911e913.
[13] M.J. Keating, C. Bach, U. Yasothan, P. Kirkpatrick, Nat. Rev. Drug Discov. 7
(2008) 473e474.
[14] M.H.R. Eichbaum, F. Schuetz, T. Khbeis, I. Lauschner, F. Foerster, C. Sohn,
A. Schneeweiss, Anti-Cancer Drugs 18 (2007) 963e968.
[15] K. Hoffken, K. Merkle, M. Schoenfelder, G. Anger, M. Brandtner, K. Ridwelski,
S. Seeber, J. Cancer Res. Clin. Oncol. 124 (1998) 627e632.
[16] U. Reichmann, C. Bokemeyer, D. Wallwiener, M. Bamberg, J. Huober, Ann.
Oncol. 18 (2007) 1981e1984.
[17] R.L. Furner, R.K. Brown, Cancer Treat. Rep. 64 (1980) 559e574.
[18] R.B. Jones, Cancer Treat. Res. 112 (2002) 305e322.
[19] J.A.B. Balfour, K.L. Goa, Drugs 61 (2001) 631e638.
[20] S. Blumel, A. Goodrich,C. Martin,N.H. Dang, Clin. J. Oncol. Nurs. 12(2008)799e806.
[21] S.M. Konstantinov, A. Kostovski, M. Topashka-Ancheva, M. Genova, M.R. Berger,
J. Cancer Res. Clin. Oncol. 128 (2002) 271e278.
[22] M. Wickstroem, H. Loevborg, J. Gullbo, Lett. Drug Des. Discov. 3 (2006)
695e703.
[23] F. Bergel, J.A. Stock, J. Chem. Soc. (1960) 3658e3669.
[24] L. Kupczyk Subotkowska, K. Tamura, D. Pal, T. Sakaeda, T.J. Siahaan, V.J. Stella,
R.T. Borchardt, J. Drug Target. 4 (1997) 359e370.
[25] U. Beyer, T. Roth, P. Schumacher, G. Maier, A. Unold, A.W. Frahm, H.H. Fiebig,
C. Unger, F. Kratz, J. Med. Chem. 41 (1998) 2701e2708.
[26] F. Kratz, U. Beyer, T. Roth, M.T. Schuette, A. Unold, H.H. Fiebig, C. Unger, Arch.
Pharm. 331 (1998) 47e53.
[27] Y. Matsumura, H. Maeda, Cancer Res. 46 (1986) 6387e6392.
[28] H. Sinn, H.H. Schrenk, E.A. Friedrich, U. Schilling, W. Maierborst, Nucl. Med.
Biol. 17 (1990) 819e827.
[29] T. Kapp, A. Dullin, R. Gust, Bioconjug. Chem. 21 (2010) 328e337.
[30] A.M. Scutaru, M. Wenzel, H. Scheffler, G. Wolber, R. Gust, Bioconjug. Chem. 21
(2010) 1728e1743.
[31] T. Kapp, A. Dullin, R. Gust, J. Med. Chem. 49 (2006) 1182e1190.
[32] S. Okumoto, S. Yamabe, J. Org. Chem. 65 (2000) 1544e1548.
[33] I. Giraud, M. Rapp, J.C. Maurizis, J.C. Madelmont, J. Med. Chem. 45 (2002)
2116e2119.
[34] B. Neises, W. Steglich, Angew. Chem. Int. Edit. 17 (1978) 522e524.
[35] K. Bremer, W. Roth, Tumordiag. Ther. 17 (1996) 1e6.
[36] E.H.L. Chun, L. Gonzales, F.S. Lewis, J. Jones, R.J. Rutman, Cancer Res. 29 (1969)
1184e1194.
[37] I.G. Walker, B.D. Reid, Cancer Res. 31 (1971) 510e515.
[38] S.Y. Chang, D.S. Alberts, D. Farquhar, L.R. Melnick, P.D. Walson, S.E. Salmon,
J. Pharm. Sci. 67 (1978) 682e684.
[39] R. Gust, R. Krauser, Monatsh. Chem. 128 (1997) 291e299.
[40] W. Werner, G. Letsch, W. Ihn, Pharmazie 42 (1987) 272e273.
[41] B.A. Teicher, Cancer Principles and Practice of Oncology (1997) 405e418.
[42] M. Wickstroem, K. Viktorsson, L. Lundholm, R. Aesoy, H. Nygren, L. Sooman,
M. Fryknas, L.K. Vogel, R. Lewensohn, R. Larsson, J. Gullbo, Biochem. Phar-
macol. 79 (2010) 1281e1290.
4.3.2. In vitro chemosensitivity assays
Briefly, 100 m
L of a cell suspension of 7500 cells mLꢁ1 culture
medium were plated into each well of a 96-well microtiter plate and
incubated at 37 ꢀC for 3 days in a humidified atmosphere(5% CO2). By
adding an adequate volume of a stock solution of the appropriate
compound (solvent: DMF or methanol) to the medium, the desired
test concentrationwasobtained. After the properincubation time (0,
48, 72, 96, 120, 144 h) the medium was removed, and the cells were
fixed with a glutardialdehyde solution and stored under phosphate-
buffered saline (PBS) at 4 ꢀC. Cell biomass was determined by
a crystal violet staining technique described previously [49,50]. The
invitrocytotoxicitytestwas performed 3 times foreach substance. In
the diagrams, the vertical error bars on data points represent the
standard error of the mean.
The efficiencyofthe compoundsisexpressed ascorrected % T/Ccorr
values according to the following equations:
Cytostatic effect : %T=Ccorr ¼ ½ðT ꢁ C0Þ=ðC ꢁ C0Þꢂ100
Cytocidal effect : %
¼ ½ðT ꢁ C0Þ=C0ꢂ 100
(1)
s
(2)
in which T (test) and C (control) are the optical density values at
578 nm of the crystal violet extract of the cells in the wells (that is
the chromatin-bound crystal violet extracted with 70% ethanol),
and C0 is the density of the cell extract immediately before treat-
ment. A microplate reader at 590 nm (Flashscan Analytik Jena AG)
was used for the automatic estimation of the optical density of the
crystal violet extract in the wells. The calculated % T/C values can be
interpreted as follows:
T/Ccorr > 80%: no antiproliferative effect;
80% > T/Ccorr > 20%: antiproliferative effect;
20% > T/Ccorr > 0%: cytostatic effect;
[43] S. Fuchs, T. Kapp, H. Otto, T. Schoeneberg, P. Franke, R. Gust, A.D. Schlueter,
Chem.-Eur. J. 10 (2004) 1167e1192.
s
¼ T/Ccorr < 0%: cytocidal effect.
[44] W.J. Hopwood, J.A. Stock, Chem. Biol. Interact. 4 (1971) 31e39.

A.M. Scutaru et al. / European Journal of Medicinal Chemistry 46 (2011) 1604e1615
1615
[45] N. Farrell, Y. Qu, M.P. Hacker, J. Med. Chem. 33 (1990) 2179e2184.
[46] T. Kapp, S. Mueller, R. Gust, ChemMedChem 1 (2006) 560e564.
[47] F. Bergel, J.A. Stock, J. Chem. Soc. (1954) 2409e2417.
[48] R.J. Hay, Anal. Biochem. 171 (1988) 225e237.
[49] G. Bernhardt, H. Reile, H. Birnbock, T. Spruss, H. Schoenenberger, J. Cancer Res.
Clin. Oncol. 118 (1992) 35e43.
[50] H. Reile, H. Birnboeck, G. Bernhardt, T. Spruss, H. Schoenenberger, Anal. Bio-
chem. 187 (1990) 262e267.