Exploring synthetic avenues for the effective synthesis of selenium- and tellurium-containing multifunctional redox agents

Article abstract of DOI:10.1039/b907831b

Various human illnesses, including several types of cancer and infectious diseases, are related to changes in the cellular redox homeostasis. During the last decade, several approaches have been explored which employ such disturbed redox balances for the

Full text of DOI:10.1039/b907831b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Exploring synthetic avenues for the effective synthesis of selenium- and
tellurium-containing multifunctional redox agents†
Susanne Mecklenburg,a Saad Shaaban,a Lalla A. Ba,a Torsten Burkholz,a Thomas Schneider,a Britta Diesel,b
Alexandra K. Kiemer,b Anne Ro¨seler,c Katja Becker,c Jo¨rg Reichrath,d Alexandra Stark,d Wolfgang Tilgen,d
Muhammad Abbas,e Ludger A. Wessjohann,e Florenz Sasse f and Claus Jacob*a
Received 20th April 2009, Accepted 3rd August 2009
First published as an Advance Article on the web 11th September 2009
DOI: 10.1039/b907831b
Various human illnesses, including several types of cancer and infectious diseases, are related to changes
in the cellular redox homeostasis. During the last decade, several approaches have been explored which
employ such disturbed redox balances for the benefit of therapy. Compounds able to modulate the
intracellular redox state of cells have been developed, which effectively, yet also selectively, appear to kill
cancer cells and a range of pathogenic microorganisms. Among the various agents employed, certain
redox catalysts have shown considerable promise since they are non-toxic on their own yet develop an
effective, often selective cytotoxicity in the presence of the ‘correct’ intracellular redox partners.
Aminoalkylation, amide coupling and multicomponent reactions are suitable synthetic methods to
generate a vast number of such multifunctional catalysts, which are chemically diverse and, depending
on their structure, exhibit various interesting biological activities.
increased activities of pro-oxidant enzymes, down-regulation of
1. Introduction
antioxidant enzymes and inflammation around a tumour site
All living organisms need to maintain a healthy intracellular redox
with production of Reactive Oxygen Species (ROS).‡ Strategies
balance in order to survive and to proliferate.1,2 Not surprisingly,
the maintenance of a healthy redox balance has been a matter of
intense investigation during the last fifty years, and there have been
exploiting OS in and against such cancer cells involve compounds
able to generate ROS (e.g. As2O3), superoxide dismutase (SOD)
inhibitors (e.g. 2-methoxyestradiol) and, perhaps most interest-
numerous attempts to interfere with the redox state of the cell, for
ingly, the employment of catalysts able to convert OS in cancer
cells into a lethal cocktail of cytotoxic chemical species.7,8 At the
same time, many unpleasant parasitic microorganisms, such as
dermatophytes, plasmodia and bacteria, possess an intracellular
redox network that differs significantly from the one found in the
human cell.9 In principle, such organisms may also be targeted
with redox modulating agents.
instance by administering antioxidants to counteract Oxidative
Stress (OS). More recently, it has become apparent that certain
imbalances in the intracellular redox state are not just detrimental,
but may actually be exploited for therapeutic purposes. Similar
to cancer-cell-specific receptors or protein expression profiles,
unusual intracellular redox states may be used to ‘single out’ and
kill target cells.3
A range of studies conducted during the last ten to fifteen years
have demonstrated that many different types of cancer cells prolif-
erate under conditions of OS, among them lung, kidney, prostate
and skin cancer cells.4,5 The disturbed redox balance in those cells
may have several causes, including an increased metabolic activity,
OS is a ‘multi-stressor event’ which includes various ROS
∑
(e.g. O2 -, H2O2), but also a range of nitrogen species and
adventitious metal ions (e.g. iron and copper).1,2 High efficiency
and selectivity of compounds may therefore largely depend on
their ability to ‘recognize’ the biochemical signature of OS in
the cell, i.e. to respond to two or more of the ingredients of
OS, and not just to one. Several attempts have been made to
develop multifunctional agents able to recognize various ingredi-
ents of OS.10,11 Unfortunately, the synthesis of such compounds
encounters increasing difficulties when moving from just one or
two to three or more redox sites – in particular if a combination
of quinone, macrocycle, selenium and/or tellurium is involved.
To resolve this obstacle, various synthetic avenues have been
explored in order to synthesize compounds with three or even four
OS-related functions. Depending on their structure and reactivity,
several interesting lead compounds have now been identified
aDivision of Bioorganic Chemistry, School of Pharmacy, Saarland Uni-
versity, PO Box 151150, D-66123, Saarbruecken, Germany. E-mail:
c.jacob@mx.uni-saarland.de; Fax: +49 (0)681 302 3464; Tel: +49 (0)681
302 3129
bDivision of Pharmaceutical Biology, School of Pharmacy, Saarland Univer-
sity, PO Box 151150, D-66123, Saarbruecken, Germany
cDepartment of Nutritional Biochemistry, Interdisciplinary Research Center,
Giessen University, Heinrich-Buff-Ring 26-32, D-35392, Giessen, Germany
dDepartment of Dermatology, School of Medicine, Saarland University,
D-66421, Homburg (Saar), Germany
eLeibniz Institute of Plant Biochemistry, Department of Bioorganic Chem-
istry, Weinberg 3, D-06120, Halle (Saale), Germany
f
Helmholtz-Zentrum fu¨r Infektionsforschung GmbH, Inhoffenstraße 7,
D-38124, Braunschweig, Germany
† Electronic supplementary information (ESI) available: Synthesis of
compounds; SK-Mel-5 and HL-60 cell culture studies; activity assays
against Plasmodium falciparum and Trichophyton rubrum; 58-cell line
screening. See DOI: 10.1039/b907831b
‡ By contrast, cells found in hypoxic regions of tumours are often more
reducing than normal cells, which is exploited in bioreductive drug therapy,
e.g. by using Mitomycin C (MMC).6 OS and bioreductive environments
are not necessarily mutually exclusive.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4753–4762 | 4753
©

View Article Online
which exhibit activity against cancer cells, dermatophytes and
Plasmodium falciparum.
2. Results
2.1 Synthesis of agents combining two or more redox-active and
metal binding sites
While there is a strong demand for multifunctional redox agents
containing selenium or tellurium, synthetic avenues leading to
such agents are often marred by difficulties and low yields. We
have therefore explored five different avenues for the synthesis of
agents combining two, three or even four redox/metal binding
sites in one molecule.
2.1.1 Nucleophilic substitution by a redox-active chalcogen
moiety at a quinone core structure. We have previously reported
a method which relies on the direct attack of nucleophilic se-
lenolate or tellurolate compounds at suitable bromoquinones.10,11
This approach has enabled us to generate several compounds
combining two redox centres relevant to OS, albeit in rather low
yields (Fig. 1). In order to increase yields, the original method
was improved by employing an heterogeneous solvent system
(water and ethyl acetate (1:1)) and a phase transfer catalyst (PTC,
tricaprylmethylammonium chloride). Under these conditions, the
reaction proceeds in a more controlled manner, appears to avoid
side-reactions and results in significantly higher yields. In the
case of compound 1, for instance, yields of 78–98% have been
obtained with the two-phase-system, compared to a yield of just
9% in the original method. Related compounds, such as 2 and 3,
could also be obtained by this two-phase method and in similar
yields. While this one-step synthetic method is attractive because
of its comparable simplicity, its scope appears to be limited by the
chemical diversity and number of compounds achievable.
Fig. 1 Schematic overview of the synthetic method employing a bro-
moquinone and various chalcogen-based nucleophiles. A selection of
compounds synthesized according to this method is shown. Experimental
details are provided in the text and the ESI.†
lation is a Michael addition reaction targeting quinones, and
not bromoquinones. Secondly, the reactivity is controlled by the
amine and not by the chalcogen (yet see below). This method
was employed successfully to synthesize compounds combining a
naphthoquinone with one or two sulfur or selenium moieties in
rather good yields (20–35%). In the case of compounds 4 and 5,
two redox-centres – i.e. one quinone and a sulfur or selenium atom
– are present, while compound 6 even combines three redox sites in
one molecule. Although this method shows considerable promise
for the synthesis of multifunctional sulfur and selenium agents, it
could not (yet) be used for the synthesis of the corresponding tel-
lurium analogues. One may speculate that the long reaction times
of three days, combined with the oxidizing conditions required to
re-oxidize the semi- or hydroquinone to quinone (i.e. presence of
air), may adversely affect the tellurium moiety.
2.1.2 Aminoalkylation. To overcome these limitations,
aminoalkylation of the quinone by various chalcogen-containing
amines has been explored (Fig. 2, 4–6).12,13 This reaction proceeds
via nucleophilic attack of the amine-based nucleophile at the
quinone. Mechanistically it resembles to the method described in
section 2.1.1, with two significant differences: Firstly, aminoalky-
2.1.3 Reductive amination. Alternative methods for ‘cou-
pling together’ individual functional groups under mild conditions
have therefore been explored. Reductive amination involves the
reaction of amine and aldehyde to imine with subsequent reduc-
tion to amine (Fig. 3). This approach appears to be particularly
Fig. 2 The aminoalkylation method. A selection of compounds synthesized according to this method is shown. Experimental details are provided in the
text and the ESI.†
4754 | Org. Biomol. Chem., 2009, 7, 4753–4762
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Fig. 3 Reductive amination as a synthetic route for agents incorporating nitrogen-based macrocycles.
suitable for molecules combining redox and metal binding sites,
such as compound 7.14 Reductive amination takes advantage of
the nucleophilic character of amines, which react efficiently with
aldehydes under mild conditions and may also form a site for metal
binding. Several new compounds containing one or two selenium
atoms and a metal binding site (e.g. compounds 8 and 9) were
synthesized by this method, with yields around 30%.
quinone redox centre with two chalcogen centres, while compound
21 contains two redox centres (quinone and sulfur) and one metal
binding site.
Amide coupling works well for sulfur, selenium and tellurium
and provides compounds in rather good yields of around 50 to
60% (Fig. 4). The preparation of the individual building blocks
and the coupling reaction itself can be performed under rather
mild conditions. It is also possible to ‘switch’ the acid and amine
functions, i.e. to generate amine-bearing quinones and chalcogen-
containing acids. This method is suitable for the creation of a small
library of agents, but is limited by the number of building blocks
which can be assembled at one time.
2.1.4 Amide coupling of appropriate building blocks. Reduc-
tive amination demonstrates that it is possible to employ coupling
of two or more pre-formed ‘building blocks’ for the synthesis of
higher functionalized agents. The ‘coupling’ approach is generally
rather promising: By incorporating redox and metal binding
centres into different building blocks, it is possible to synthesize the
individual blocks separately, even under harsh conditions. They
may then be ‘mixed and matched’ to generate a range of target
molecules (e.g. a small ‘library’), which can be assembled under
mild reaction conditions. Importantly, coupling does not have to
rely on the chemistry of the redox or metal binding sites themselves,
hence avoiding the problems discussed in section 2.1.1.
Alternative coupling methods have therefore been considered.
The amide coupling method is based on amide bond formation
between acid and amine, a reaction used extensively in peptide
synthesis and renowned for its wide applicability and mild
conditions (Fig. 4). This method was used to synthesize a range
of compounds, some of which are shown in Figs. 4a–c. In the
first step, chalcogen-containing amines15 and quinone-containing
carboxylic acid building blocks were obtained: It is possible, for
instance, to attach a carbon chain to the quinone via oxidative
decarboxylation of an acid, such as glutaric acid (Fig. 4a).16
Similarly, primary amines and thiols may be attached to quinones
to form secondary amines and sulfides, respectively (Figs. 4b
and 4c). In the second step, amide coupling can be carried out
under mild conditions to yield compounds combining two or three
‘active’ groups. Compounds 14 and 15, for instance, combine one
2.1.5 Multicomponent Passerini and Ugi reactions. Although
amide coupling may be limited by the fact that just two blocks can
be combined at a time (more if equivalent sites are present), similar
building blocks can be used in multicomponent reactions, such as
the three-component Passerini and four-component Ugi reactions
(Fig. 5). The Passerini reaction combines an aldehyde, an acid and
an isonitrile to an a-acyloxy amide, and the Ugi reaction combines
an aldehyde, an acid, an isonitrile and an amine to amide-
bonded a-aminoacyl amide structures.17–20 Although the synthesis
of appropriate quinone- and chalcogen-bearing building blocks
is not always straightforward, a range of such compounds have
now been reported in the literature. Of particular importance is
the tellurium-containing isonitrile 23.20 By assigning the tellurium
moiety to the isonitrile building block, the remaining building
blocks, which are chemically more accessible, can be designed
at will, for instance to incorporate further exciting functional
groups. The building blocks required were synthesized in sufficient
yield and purity, with analytical data according to literature.20
The various building blocks were used to synthesize a series of
representative compounds 26–29, whose chemical structures are
shown in Fig. 5. All of these compounds were produced under
mild conditions (H2O as solvent, room temperature) and, from
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4753–4762 | 4755
©

View Article Online
Fig. 4 The two-step amide coupling method, which follows the ‘building block’ approach, whereby individual components carrying different functional
groups are synthesized first and then assembled in a final, straightforward step. This method employs ‘linkers’, which attach to the quinone via a carbon,
nitrogen or sulfur atom (Fig. 4a, 4b and 4c, respectively). A selection of compounds synthesized according to this method is shown. Experimental details
are provided in the text and the ESI.†
4756 | Org. Biomol. Chem., 2009, 7, 4753–4762
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Fig. 5 Schematic overview of the Passerini and Ugi multicomponent reactions. These reactions expand the concept of ‘building blocks’ and allow a
wide range ‘mix and match’ approach of readily available, functionalized aldehyde, acid, isonitrile and – in the case of the Ugi reaction – amine building
blocks. Only a few selected examples of compounds synthesized by this method are shown, with experimental details provided in the text and the ESI.†
the perspective of synthetic selenium and tellurium chemistry, in
good yields (54–81%).
Institute of Health (NIH) in Bethesda, MD, USA. This screen
includes 58 cell lines grouped into leukaemia, non-small-cell lung
cancer, colon cancer, CNS cancer, melanoma, ovarian cancer,
renal cancer, prostate cancer and breast cancer cell lines.21,22 In
this screen, compounds 18 (see ESI for screening data†), 27
and 28 exhibited a significant cytotoxicity against cancer cell
lines and were subsequently selected for 5-dose testing and LC50
determination. The LC50 values obtained at the NCI were in the
range 5–10 mM. It should be noted that compounds 27 and 28
feature a direct attachment of the selenium to the quinone, in
contrast to the structurally similar, yet less active 26, where the
selenium atom is located further away from the quinone moiety.
2.2 Activity screens in cancer cell lines
Ultimately, it is important to show that the chemistry devel-
oped leads to ‘useful’ molecules. A selection of the structurally
most interesting compounds was therefore evaluated for bio-
logical activity. Since previous studies have linked combined
quinone/chalcogen compounds with (selective) cytotoxicity, the
main focus of these studies was on cancer cells and infectious
microbes. Due to the fact that selenium compounds are on
occasion considered as antioxidants, some of the compounds
were also tested for antioxidant activity. It should be emphasized
from the outset that the biological data presented here is of
an early-stage, preliminary nature, and requires more in-depth
pharmacological studies in the future.
2.2.2 Activity against human SK-Mel-5 melanoma cells. The
initial activity screen indicated that compounds combining
quinone and chalcogen redox centres may be active against certain
types of cancer cells. To investigate this activity further, human
SK-Mel-5 melanoma cells were used as model. This cell line
was chosen for several reasons. Skin cancer is rather ‘easy’ to
target with drugs, either systemically or by topical applications.
Furthermore, the initial screens (section 2.2.1) have already shown
2.2.1 Activity in cancer cell line screen. In order to gain
an initial insight into the ‘usefulness’ of multifunctional agents,
a broad-spectrum, one-dose (10 mM) cell culture screen was
performed at the National Cancer Institute (NCI) of the National
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4753–4762 | 4757
©

View Article Online
some activity of quinone and/or chalcogen compounds against
melanoma cell lines. And finally, multifunctional selenium agents
appear to respond well to ROS present in skin cells, as studied
previously in a different, antioxidant context in skin fibroblast
cells.14,23
The SK-Mel-5 cell culture studies have been designed to measure
toxicity as well as to evaluate a possible pro- or antioxidative
response in the presence of ROS. As a result, compounds can
be classified overall as either (a) inactive, (b) generally toxic or
(c) toxic in response to elevated levels of ROS. Compounds com-
bining sulfur and quinones, and compounds containing either just
quinones or just sulfur were mostly inactive at concentrations up to
200 mM, which was the highest concentration tested. In contrast,
compounds containing tellurium, including compounds 12, 15
and 18 (see section 2.2.1) were generally toxic at concentrations of
5 mM, which was the lowest concentration tested.
The more interesting compounds – exhibiting a differentiated
toxicity in response to H2O2 – included a range of multifunctional
selenium compounds incorporating quinone and macrocycle
moieties. Noteworthy in particular are compounds 8 and 21
followed by compounds 7 and 11. Compounds 7 and 8, as well as
21, combine redox catalysis with metal binding properties. These
compounds appear to respond well to the presence of OS in form
of H2O2. This effect is best exemplified by compound 21, which
combines a quinone, sulfur and macrocycle group in one molecule.
The results obtained for this compound are shown in Fig. 6: In the
absence of H2O2, 21 exhibits little toxicity against SK-Mel-5 cells
when used in concentrations up to 25 mM (97% cell survival). In the
presence of H2O2, however, cell survival is reduced dramatically
to 38%. This effect is not additive, but due to a synergy of 21 and
H2O2. The latter enhances the cytotoxicity of 21 by about 2–3-fold.
metal binding site(s) appears to be particularly effective (see also
section 2.2.4): Compared to 21 and 8, compounds 7 and 11 were
less active – although they still exhibited a significant synergistic
effect between compound and H2O2.
It should be mentioned that the activity of compounds depends
on the cell type used. In HL-60 human leukaemia cells,24 for
instance, some of the compounds tested showed a lower toxicity
rather than enhanced toxicity in the presence of H2O2, which was
also lower than the toxicity of H2O2 on its own. These compounds
include 17, but also compound 7, which in SK-Mel-5 cells is more
toxic in the presence of H2O2. Such results underline the differences
seen in the 58-cell screen. They also confirm a certain antioxidant
effect of 7, which has been observed previously in skin fibroblasts.14
Issues surrounding such pro- and antioxidant effects of catalysis
are explained in section 3.
Considered together, the results obtained in the cancer cell
culture screen and SK-Mel-5 cell line studies point towards a range
of different activities. Worth considering are responses towards the
presence of OS, either in form of an enhanced cytotoxicity (8 and
21) or as possible antioxidants (7). It is therefore sensible to expand
the scope of biological assays for redox modulators beyond cancer
cell lines and to consider other diseases linked to unusual redox
states or events, such as infectious diseases.
2.2.3 Activity against Plasmodium falciparum. The apicom-
plexan parasite Plasmodium falciparum is a unicellular organism
multiplying – after an initial liver stage – in human erythrocytes.
Plasmodium falciparum is the causative agent of tropical malaria
and responsible for approximately 350 million cases of malaria
and around one million human deaths each year. Novel drugs,
vaccines, and insecticides, as well as deeper insights into parasite
biology are essential to counteract increasing drug resistance.25 As
demonstrated by glucose-6-phosphate dehydrogenase deficiency,
a genetic disorder protecting from malaria, the parasite host-
cell unit is very sensitive to disturbances in its redox balance. This
feature is actually shared by many rapidly growing and dividing
cells. In Plasmodium, the lack of catalase and a typical glutathione
peroxidase certainly add to this sensitivity.26 This also implies that
the redox metabolism of this organism is an excellent target for
anti-malarial drug development.26,27
Since both quinones and chalcogens are known to act as
redox modulators able to generate ROS, the toxicity of our
multifunctional agents against Plasmodium falciparum was studied
in vitro. Most of the compounds exhibited activity in the lower
micromolar range, which was not studied in more detail (Table 1).
Two compounds combining a naphthoquinone with a selenium or
tellurium moiety, however, showed excellent anti-parasitic activity
in the nanomolar range (compounds 11 and 12). The difference
in IC50 values between the selenium (225 nM) and the tellurium
compound (216 nM) was found to be minor, which is equally
surprising and important, since tellurium compounds are usually
more active when compared to their selenium analogues. In
contrast, structurally similar compounds 6, 17 and 18 were not
particularly active in this assay (IC50 > 2 mM).
Fig. 6 Cytotoxic effects of compound 21 against SK-Mel-5 cells in the
absence (solid line) and presence (dashed line) of H2O2. The presence of
H2O2 enhances the toxicity of 21 about 2–3-fold. Please note that the results
are normalized to 100% survival for cells both in the absence and presence
of H2O2. n = 3.
Similar results were obtained for compound 8. At concentra-
tions up to 50 mM, 8 is virtually non-toxic against SK-Mel-5 cells
in the absence of H2O2 (91% survival at 25 mM), yet shows a
2–3-fold increased toxicity in the presence of H2O2 (65% survival
at 25 mM). Interestingly, the combination of redox centre(s) with
2.2.4 Activity against Trichophyton rubrum. Finally, a selec-
tion of compounds was evaluated for possible activity against the
dermatophyte Trichophyton rubrum, a parasitic fungus affecting
humans.28 Fungi are, on occasion, sensitive to ROS and, at the
same time, require considerable amounts of trace metal ions
4758 | Org. Biomol. Chem., 2009, 7, 4753–4762
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Table 1 Activity of selected compounds against Plasmodium falciparum
compound 7 also shows good antioxidant activity, especially on
human skin cells, which may provide an interesting lead for a
combined antifungal/protective activity on skin (see section 3).
in vitro
Compound
IC50 on P. falciparum
4
>2 mM
>2 mM
>2 mM
>2 mM
>2 mM
225 nM
216 nM
>2 mM
>2 mM
>2 mM
>2 mM
>2 mM
3. Discussion
5
6
3.1 Comparison of the synthetic avenues
7
8
11
12
14
15
17
18
21
A range of synthetic methods have been explored which allow the
synthesis of highly functionalized redox and metal binding agents.
While all of them are, in principle, able to generate molecules
combining redox and/or metal binding sites, there are obvious
differences between them.
The most basic approach employs direct ‘coupling’ of the
chalcogen (sulfur, selenium, tellurium) to the quinone as part of
a simple, one-step procedure. The chalcogen serves as nucleophile
and the resulting product features a quinone–chalcogen bond, a
structural feature that appears to be particularly active in some
of the biological assays. This coupling method can be employed
for the synthesis of a range of mono- and di-substituted quinones,
including benzo- and naphthoquinones. Good yields (75% and
more) may be obtained when the procedure is carried out in a
water–ethyl acetate mixture in the presence of a phase transfer
catalyst. Nonetheless, this method is ultimately limited in scope by
the availability of suitable and reactive quinones and chalcogen-
derivatives. It is also questionable if higher substitutions, such
as three- or four-substituted quinones, may be accessible by this
method or stable in aqueous media, especially in the case of
tellurium. Although the method appears to work well for simple
compounds, the attraction of one-step synthesis is lost once rather
complex starting materials have to be synthesized first.
Nonetheless, nucleophilic attack at the quinone provides an
excellent entry point for the synthesis of bifunctional agents,
especially when more reactive and stable amines are used. Amines
avoid some of the problems encountered with selenium and
tellurium nucleophiles, and the aminoalkylation method (using
quinones instead of bromoquinones) works rather well for sulfur
and selenium compounds 4–6. It appears to fail, however, for the
corresponding tellurium analogues, possibly due to the instability
of 3-(phenyltelluryl)propane-1-amine and/or the product under
the reaction conditions employed.
for normal growth. The use of compounds with a combined
ROS generating and metal sequestering activity was therefore
envisaged as an alternative to current antifungal treatment. The
latter includes the drug ketoconazole, which is taken systemically
or topically over a time course of several weeks.
In order to test the compounds in a ‘real-life situation’,
Trichophyton rubrum samples were obtained from patients treated
at the Dermatology Unit at Saarland University Hospital. As a
consequence, each dermatophyte sample differed as to the exact
nature of the fungus. The results obtained as part of the antifungal
tests were therefore to a certain degree dependent on each patient
and their individual dermatophyte. Nonetheless, the results overall
point towards a reasonable antifungal activity of compounds 7 and
9 when used at concentrations of 100 mM (or above). Fig. 7 shows
a representative concentration-dependent activity of 7 against
the fungus. It is worth noting that the benchmark compound
ketoconazole is active in these assays at similar concentrations
(188.2 mM). The other compounds tested against Trichophyton
rubrum were inactive at concentrations up to 10 mM. Interestingly,
The aminoalkylation method has therefore been developed
further. By initially ‘separating’ the quinone from the chalcogen
and macrocycle chemistry, it has been possible to synthesize appro-
priate building blocks which in a final step are coupled together
under mild conditions. Employing such coupling methods, it is
possible to take full advantage of the specific quinone, chalcogen
and macrocycle chemistries – which must be kept separately
at the beginning to avoid the kind of complications arising
once these groups encounter each other (see above). Reductive
amination, a rather straightforward assembly method which
couples a (functionalized) aldehyde to a (functionalized) amine,
has been used successfully to synthesize compounds containing
macrocycles, such as 8 and 9. Nonetheless, the scope of the method
is rather limited.
Fig. 7 Activity of compound 7 against Trichophyton rubrum. Fungal
growth towards the plate containing 7 is inhibited (grey bars). In contrast,
the fungus grows readily towards the plate containing the DMSO control,
as measured by the (rapidly decreasing) distance between fungus and plate
(white bars). Please note that this compound was able to inhibit growth
(i.e. maintain the distance between the fungus and the point of application)
for at least ten days. Students t-test, n = 8, * P < 0.05, *** P < 0.001.
Experimental details are provided in the text and the ESI.†
By using the amide coupling approach, it has been possi-
ble to successfully synthesize various selenium and tellurium
compounds, including compounds 11, 12 and 21, which show
interesting biological activities. This method works rather well
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4753–4762 | 4759
©

View Article Online
with a linker containing an amine (for attachment to the quinone)
and acid (for amide coupling), and it has also been possible to
employ a linker containing a carbon or thiol instead of an amine
group. The use of a thiol allows a two-fold ‘functionalization’ of the
quinone: Compound 21, for instance, combines three important
groups in one molecule, namely a quinone redox-centre, a sulfur
redox-centre and a cyclen metal binding site. At the same time, 21
also highlights the limitations of this method: Although further
groups of interest may be introduced – such as a selenium atom in
the linker – more diverse or higher functionalization may become
difficult. This is due to the fact that generally just two building
blocks are assembled via amide coupling, and additional groups
would have to be introduced via more complex building blocks.
Ultimately, the issue of generating highly functionalized molecules
would simply shift from the synthesis of the final product to the
synthesis of the individual building blocks.
A bolder approach has therefore been investigated. The
Passerini or Ugi reactions allow the one-step assembly of three
or four building blocks, respectively. If one building block were to
carry just one function, three- to four-functional molecules may
become accessible. The individual building blocks required can
now be synthesized with comparable ease (including functional-
ized isonitriles) and one may ‘mix and match’ them according
to individual design criteria. The subsequent Passerini and Ugi
chemistry itself is fairly straightforward: It can be carried out
under mild conditions in solvents such as water or chloroform,
and yields are high when compared to the other methods evaluated
in this study. Multicomponent chemistry ultimately could be
used to generate molecules bearing four or more redox and/or
metal binding sites in one molecule, as demonstrated by the
synthesis of 27. Then again, it may not always be the method
of choice. The synthesis of the individual building blocks is fairly
time-consuming. In addition, the Passerini reaction leads to an
a-acyloxy amide, i.e. a compound containing an ester bond,
which is not particularly stable and may be cleaved chemically
or enzymatically. It is possible that such compounds are readily
modified in vivo, either for the benefitofactivation or, less desirably,
in conjunction with a loss of activity.
of (a) oxygen radical and H2O2 formation at the quinone and (b)
oxidation of cellular thiols catalyzed by selenium in the presence
of H2O2. Together, this ‘chemistry’ is likely to significantly increase
OS in cells already suffering from (mild) H2O2 stress – and
ultimately to push those cells over a critical ‘life and death’ (redox)
threshold. This explanation may also account for the differences in
toxicity observed in the 58 cell screen: Cells particularly sensitive
to compounds such as 18, 27 and 28 may be cells with elevated
intracellular levels of ROS and/or cells particularly sensitive to
OS. For most compounds, the highest activity in mammalian and
human cell culture is achieved when employed in concentrations
between 1 and 50 mM. LC50 values obtained in 5-dose testing of the
58 cell screen are around 10 mM, and the best results in SK-Mel-5
cells are also obtained in this concentration range. These values are
reasonable, yet still leave room for further improvement. It must
be pointed out that the compounds used were selected initially
to showcase the underlying synthetic chemistry. They are not
optimized regarding their pharmacokinetic profile (e.g. Lipinski’s
rules) or metabolic stability.
Some compounds, such as 7 and 17, also appear to counteract
the cytotoxic effects exerted by H2O2 on HL-60 cells. Such
compounds may be considered as ‘antioxidants’. Compound 7,
in particular, has already shown this kind of activity in cultured
mouse fibroblast cells exposed to UVA radiation stress.14,23 In
principle, the ability of compounds to catalyze the reaction
between H2O2 and thiols may result in antioxidant protection
as well as enhanced cytotoxicity. This apparent contradiction is
easily resolved: The impact of this type of peroxidase-like activity
is dependent on the concentrations of intracellular thiols and
the nature of the thiols consumed. Peroxidase-like catalysis is
antioxidative as long as sacrificial thiols (e.g. reduced glutathione,
GSH) are affected, and lethal once essential thiols (e.g. zinc finger
proteins) are consumed. Interestingly, the tellurium analogue of
compound 17, i.e. compound 18, is toxic to cells.
The notion of redox modulation by our compounds as a possible
approach towards antimicrobial therapy has been evaluated
further by studying their effects on parasitic organisms. The IC50
values for growth inhibition obtained for compounds 11 and 12
against the malarial parasite Plasmodium falciparum were around
200 nM. These values are considerably lower than those obtained
for cultured human cells, which represents an excellent starting
point for further developing this class of compounds as potential
anti-malarials which might act alone or in combination therapies.
Indeed, a number of currently used drugs, including the quinoline
and the endoperoxide anti-malarials, appear to act at least par-
tially by increasing OS. Current international drug development
strategies also include the synthesis of novel compounds with pro-
oxidant activity such as peroxidic compounds, anthraquinones,
and inhibitors of antioxidant enzymes.26,27 Redox-active enzymes
like glutathione reductase, thioredoxin reductase or peroxiredox-
ins rank among the currently most attractive drug targets against
parasitic infections. As a next step, we envisage the synthesis of
further compounds structurally related to 11 and 12 and a detailed
evaluation for anti-malarial activity in order to understand the
underlying structure–activity relationships.
In any case, the availability of a range of redox-active building
blocks, such as numerous quinone- and selenium-containing
acids, aldehydes and amines, allows the synthesis of a vast range
of chemically diverse molecules via different synthetic avenues.
Such molecules may be ‘tailor-made’ to respond to OS. Indeed,
the biological data obtained so far highlights the potential and
potential use(s) of such multifunctional redox/metal binding
compounds.
3.2 Biological activity
Overall, the various compounds synthesized as part of this study
have provided a number of interesting leads as far as possible drug
design and development are concerned. First of all, compounds
such as 8 and 21 have confirmed the feasibility of so-called
‘sensor/effector’ agents. The latter are catalytic agents able to
recognize a particular intracellular state of cells – such as OS –
and, by using this state, are able to kill these cells selectively. This
property manifests itself in the cytotoxicity of 8 and 21 against
SK-Mel-5 cells, which is enhanced significantly in the presence of
H2O2. The most likely explanation for this effect is a combination
At the same time, the results obtained with Trichophyton rubrum
point towards a possible strategy for fighting off this rather
unpleasant and resilient parasitic fungus. Although the concentra-
tions of the most active compounds 7 and 9 required for activity
4760 | Org. Biomol. Chem., 2009, 7, 4753–4762
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
ranged between 100 mM and 10 mM, i.e. were rather high, so
were benchmark ketoconazol concentrations (188.2 mM), pointing
towards a particularly resistant organism. Importantly, long-term
activity against Trichophyton rubrum is of particular interest and
7 and 9 were able to stop individual fungi for several days from
growing. Based on the assumption that compounds such as 7 and
9 may act by increasing OS and, at the same time, by depriving
the fungus from essential trace metal ions (such as zinc, copper
and iron), a topical rather than systemic drug treatment may
become possible, whereby growth of the fungus would be stopped
by deprivation of essential metal ions, and the remaining parts
may be killed by ROS. The impact of ROS would be particularly
severe, since the lack of metal ions such as copper and iron may
weaken the fungi’s antioxidant defence. At the same time, 7 is
known to act as an antioxidant in skin cells, implying a protective
role of 7 against ROS in human cells. A compound such as 7
may therefore attack Trichophyton rubrum by causing OS in this
organism, yet protect the skin cells affected by counteracting OS
in these cells. As mentioned above, pro- and antioxidant effects
are not mutually exclusive, but depend on the availability and
nature of intracellular thiols. In this respect, the fungus may lose
out against the skin cells, which have an almost unlimited supply
of GSH and trace metal ions available via the blood. Similar
effects may be associated with other compounds combining redox-
centres with metal binding sites. Although those considerations
are necessarily speculative at this time, they provide a lead worth
pursuing. It must be emphasized, however, that not all samples of
Trichophyton rubrum were affected equally. Nonetheless, the use of
patient samples, rather than cultivated, pedigreed fungi, provides
a more realistic model of real-life infections, which, of course, is
also more robust.
cells under OS. At the same time, results obtained for 7 support
the notion that compounds featuring redox and metal binding
sites may be able to attack fungi, such as Trichophyton rubrum,
while simultaneously acting as antioxidant. This matter requires
further investigation, especially in the context of skin disease.
Here, future compounds may feature metal exchange instead of
metal binding sites, for instance sites able to ‘exchange’ iron and
copper for zinc. This exchange would not only scavenge iron and
copper ions but also release (fungicidal) zinc ions, and hence hit the
fungus particularly hard.14 Since topical applications are possible,
evaluation of antifungal properties may soon reach the stage of
simple tests on volunteers. A more complex, yet equally important
activity is found against Plasmodium falciparum. This species was
chosen originally as an example of an organism highly sensitive
towards OS. The rather low IC50 values obtained for 11 and 12
are promising and support the notion that organisms with a weak
antioxidant defence may indeed be targeted by certain quinone-
selenium agents. These and related compounds targeting the redox
system of Plasmodium falciparum take advantage of an inherent
biochemical ‘weakness’ of this malaria pathogen and may become
useful in the future.
Ultimately, the studies presented here provide a fertile ground
for various multidisciplinary follow-on projects in the area of
synthetic chemistry, electrochemistry, biological chemistry, in vitro
biochemical studies, cell culture and drug development.
5. Experimental
5.1 Materials and general methods
Chemical reagents for the synthesis of compounds were purchased
from Sigma-Aldrich-Fluka (Darmstadt, Germany) and used with-
out further purification unless stated otherwise. THF and CHCl3
was refluxed with calcium hydride and freshly distilled before
use. Reactions under inert atmosphere were carried out under
argon (99.996%) using standard Schlenk techniques. Silica gel 60
(Macherey-Nagel, 50-200 mm) was used for column chromatogra-
phy. Unless noted otherwise, the dimensions of columns used were
2.5 cm (diameter) and 25–30 cm (height of silica gel). TLC plates
(silica gel 60 F254, 0.20 mm) were purchased from Merck.
1H NMR spectra were recorded at 500 MHz, and 13C NMR
spectra at 125 MHz on a Bruker (Rheinstetten) DRX 500 or
Avance 500 spectrometer. Chemical shifts are reported in d (ppm),
expressed relative to the solvent signal at 7.26 ppm (CDCl3,
1H NMR) and at 77.16 ppm (CDCl3, 13C NMR), as well as
3.31 ppm (CD3OD, 1H NMR) and 49.00 ppm (CD3OD, 13C
NMR). Coupling constants (J) are given in Hz.
LC-MS/MS analysis was performed using a TSQ Quantum
mass spectrometer equipped with an ESI source and a triple
quadrupole mass detector (Thermo Finnigan, San Jose, CA).
High-resolution mass spectrometry was performed on an Ac-
cela UPLC-system (Thermo-Fisher) coupled to a linear trap-
FT-Orbitrap combination (LTQ-Orbitrap), operating in positive
ionization mode.
4. Conclusions
In summary, the results obtained as part of this study provide an
overview of various synthetic strategies which can be employed
to synthesize compounds combining several redox-centres and/or
metal binding sites. These molecules are not only intriguing from a
purely synthetic point of view. They also provide several interesting
leads for possible drug development, which are not limited to
cancer research, but also appear to include lesser exposed, yet
equally important areas of pharmaceutical research, such as anti-
malarial and anti-fungal therapies.
The studies presented here are, of course, only an entry point
for considerably wider and more in-depth investigations, which
ultimately may also consider mechanistic aspects and biochemical
mode(s) of action. The evaluation of various synthetic strategies,
for instance, provides fertile ground for future synthesis. While the
multicomponent approach towards multifunctional agents may
appear to be the most promising avenue – and certainly needs
to be explored further – other methods, which may provide less
sophisticated end-products yet are ultimately faster to conduct
(e.g. amide coupling), also possess a certain attraction. The choice
of synthetic method ultimately depends on the target at hand. In
any case, the availability of several established avenues is a great
asset for future studies to build upon.
Melting points were recorded using a Digital Melting Point
Apparatus (IA9000 series, ThermoFischer Scientific, Rochford,
UK) and are given without correction.
As far as biological activity is concerned, several of the
compounds tested appear to respond well to the presence of
H2O2, and this may be exploited further for targeting tumour
A detailed description of the synthesis of individual compounds,
including extensive analytical data used for their characterization,
is provided as part of the ESI†.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4753–4762 | 4761
©

View Article Online
5.2 Biological assays
8 E. O. Hileman, J. S. Liu, M. Albitar, M. J. Keating and P. Huang,
Cancer Chemother. Pharmacol., 2004, 53, 209–219.
9 J. F. Turrens, Mol. Aspects Med., 2004, 25, 211–220.
10 G. I. Giles, N. M. Giles, C. A. Collins, K. Holt, F. H. Fry, P. A. S.
Lowden, N. J. Gutowski and C. Jacob, Chem. Commun., 2003, 2030–
2031.
11 F. H. Fry, A. L. Holme, N. M. Giles, G. I. Giles, C. Collins, K. Holt, S.
Pariagh, T. Gelbrich, M. B. Hursthouse, N. J. Gutowski and C. Jacob,
Org. Biomol. Chem., 2005, 3, 2579–2587.
12 L. Salmon-Chemin, E. Buisine, V. Yardley, S. Kohler, M. A. Debreu,
V. Landry, C. Sergheraert, S. L. Croft, R. L. Krauth-Siegel and E.
Davioud-Charvet, J. Med. Chem., 2001, 44, 548–565.
13 S. Bittner, S. Gorohovsky, O. Paz-Tal and J. Y. Becker, Amino Acids,
2002, 22, 71–93.
14 S. Mecklenburg, C. A. Collins, M. Doring, T. Burkholz, M. Abbas,
F. H. Fry, C. Pourzand and C. Jacob, Phosphorus, Sulfur Silicon Relat.
Elem., 2008, 183, 863–888.
The cell survival (cytotoxicity) studies using SK-Mel-5 and HL-
60 cell culture were performed according to the literature.29
Activity against Plasmodium falciparum was determined with a
method which has been developed recently and is now used
routinely by the authors to screen for activity.30,31 Details of these
methods, together with a detailed description of the activity assay
against Trichophyton rubrum, can be found in the ESI†. The
58 cell line screen at the National Cancer Institute (NCI) was
carried out independently of the authors following a standard
protocol, details of which can be obtained from the NCI website
http://dtp.nci.nih.gov.
15 J. E. Baldwin, R. M. Adlington and J. Robertson, Tetrahedron, 1989,
45, 909–922.
Acknowledgements
16 J. M. Anderson and J. K. Kochi, J. Am. Chem. Soc., 1970, 92, 1651.
17 L. Banfi, A. Basso, G. Guanti and R. Riva, in Multicomponent
Reactions, eds. J. Zhu and H. Bienayme´, Wiley-VCH, 2005.
18 A. Doemling and I. I. Ugi, Angew. Chem., Int. Ed., 2000, 39, 3168–3210.
19 I. Ugi, B. Werner and A. Doemling, Molecules, 2003, 8, 53–66.
20 S. Shabaan, L. A. Ba, M. Abbas, T. Burkholz, A. Denkert, A. Gohr,
L. A. Wessjohann, F. Sasse, W. Weber and C. Jacob, Chem. Commun.,
2009, 4702–4704.
The authors would like to thank the University of Saarland, the
Ministry of Economics and Science of Saarland and the Deutsche
Forschungsgemeinschaft (DFG; grant number Be 1540/10-1 to
KB) for financial support. We also acknowledge the support of the
European Union in form of the RedCat Initial Training Network
(grant agreement number 215009) and the Corena project of
the Interreg IV programme. The HRMS-measurements were
performed by R. Thomes in the Kazmaier laboratory.
21 H. J. Fingert, A. T. Pu, Z. Y. Chen, P. B. Googe, M. C. Alley and A. B.
Pardee, Cancer Res., 1988, 48, 4375–4381.
22 R. H. Shoemaker, Nat. Rev. Cancer, 2006, 6, 813–823.
23 C. A. Collins, F. H. Fry, A. L. Holme, A. Yiakouvaki, A. Al-Qenaei,
C. Pourzand and C. Jacob, Org. Biomol. Chem., 2005, 3, 1541–1546.
24 I. Kasugai and M. Yamada, Leukemia Res., 1992, 16, 173–179.
25 B. M. Greenwood, D. A. Fidock, D. E. Kyle, S. H. I. Kappe, P. L.
Alonso, F. H. Collins and P. E. Duffy, J. Clin. Invest., 2008, 118, 1266–
1276.
References
1 B. B. Halliwell, and J. M. C. Gutteridge, Free Radicals in Biology and
Medicine, Oxford University Press, 2007.
2 C. Jacob, and P. G. Winyard, Redox Signaling and Regulation in Biology
and Medicine, Wiley-VCH, 2009.
3 F. Fry and C. Jacob, Curr. Pharm. Des., 2006, 12, 4479–4499.
4 T. D. Oberley and L. W. Oberley, Histol. Histopath., 1997, 12, 525–535.
5 J. M. Mates, J. A. Segura, F. J. Alonso and J. Marquez, Arch. Toxicol.,
2008, 82, 273–299.
6 S. Danson, T. H. Ward, J. Butler and M. Ranson, Cancer Treat. Rev.,
2004, 30, 437–449.
7 H. Pelicano, D. Carney and P. Huang, Drug Resist. Updates, 2004, 7,
97–110.
26 K. Becker, L. Tilley, J. L. Vennerstrom, D. Roberts, S. Rogerson and H.
Ginsburg, Int. J. Parasitol., 2004, 34, 163–189.
27 S. Rahlfs and K. Becker, Mini-Rev. Med. Chem., 2006, 6, 163–176.
28 C. Seebacher, J.-P. Bouchara and B. Mignon, Mycopathologia, 2008,
166, 335–352.
29 T. Mosmann, J. Immunol. Methods, 1983, 65, 55–63.
30 R. E. Desjardins, C. J. Canfield, J. D. Haynes and J. D. Chulay,
Antimicrob. Agents. Chemother., 1979, 16, 710–718.
31 Q. L. Fivelman, I. S. Adagu and D. C. Warhurst, Antimicrob. Agents
Chemother., 2004, 48, 4097–4102.
4762 | Org. Biomol. Chem., 2009, 7, 4753–4762
This journal is
The Royal Society of Chemistry 2009
©