Structure activity analysis of the pro-apoptotic, antitumor effect of nitrostyrene adducts and related compounds

Article abstract of DOI:10.1016/S0006-2952(02)01618-0

In the present study, we outlined the part of the molecule mediating the prominent pro-apoptotic effect of the Michael adduct of ascorbic acid with p-chloro-nitrostyrene, a new synthetic phosphatase inhibitor. The nitrostyrene (NS) moiety was identified as the structure essential for apoptosis induction. NS and its ascorbic acid adducts displayed LC₅₀ values of 10-25μM with no significant reduction of potency in okadaic acid resistant cells overexpressing the MDR1 P-glycoprotein. Induction of apoptosis by NS derivatives and the protein phosphatase 2A inhibitor cantharidic acid was proven by the analysis of caspase-3 activation and subsequent fragmentation of DNA. Further structure activity analysis revealed the necessity of the nitro group at the β-position of the side chain. The pro-apoptotic potential of adducts of NS with pyrimidine- or pyridine-derivatives varied between NS and a progressive reduction in potency up to a nearly complete loss of cytotoxicity. Substitutions at the benzene core of NS suggested a prominent enhancement of toxicity only by substitutions at the 2- or 3-position. Heterocyclic aromatics can substitute for the benzene ring of NS albeit with a 2-3-fold reduced potency. In conclusion, nitrostyrene was identified as the core structure mediating the pro-apoptotic effect of a new synthetic phosphatase inhibitor. Further studies defined a nitrovinyl side chain attached to an aromatic ring as the pharmacophore structure of a new group of pro-apoptotic agents. These observations present the basis for the development of a new group of anticancer drugs.

Full text of DOI:10.1016/S0006-2952(02)01618-0

Biochemical Pharmacology 65 (2003) 603±610
Structure activity analysis of the pro-apoptotic, antitumor effect
of nitrostyrene adducts and related compounds
Sylvia Kaapa,b, Iris Quentina, Dereje Tamirub, Mohammed Shaheenb, Kurt Egerb,
a,*
È
Hans Jurgen Steinfelder
a
È
È
Institute of Pharmacology & Toxicology, University of Gottingen, Robert-Koch-Str.40, D-37075 Go ttingen, Germany
bInstitute of Pharmacy, University of Leipzig, Liebigstr.18, D-04103 Leipzig, Germany
Received 16 October 2002; accepted 21 November 2002
Abstract
In the present study, we outlined the part of the molecule mediating the prominent pro-apoptotic effect of the Michael adduct of ascorbic
acid with p-chloro-nitrostyrene, a new synthetic phosphatase inhibitor. The nitrostyrene (NS) moiety was identi®ed as the structure
essential for apoptosis induction. NS and its ascorbic acid adducts displayed LC values of 10±25 mM with no signi®cant reduction of
50
potency in okadaic acid resistant cells overexpressing the MDR1 P-glycoprotein. Induction of apoptosis by NS derivatives and the protein
phosphatase 2A inhibitor cantharidic acid was proven by the analysis of caspase-3 activation and subsequent fragmentation of DNA.
Further structure activity analysis revealed the necessity of the nitro group at the b-position of the side chain. The pro-apoptotic potential
of adducts of NS with pyrimidine- or pyridine-derivatives varied between NS and a progressive reduction in potency up to a nearly
complete loss of cytotoxicity. Substitutions at the benzene core of NS suggested a prominent enhancement of toxicity only by substitutions
at the 2- or 3-position. Heterocyclic aromatics can substitute for the benzene ring of NS albeit with a 2±3-fold reduced potency. In
conclusion, nitrostyrene was identi®ed as the core structure mediating the pro-apoptotic effect of a new synthetic phosphatase inhibitor.
Further studies de®ned a nitrovinyl side chain attached to an aromatic ring as the pharmacophore structure of a new group of pro-apoptotic
agents. These observations present the basis for the development of a new group of anticancer drugs.
# 2002 Elsevier Science Inc. All rights reserved.
Keywords: Nitrostyrene derivatives; Drug-induced apoptosis; Phosphatase inhibitors; Caspase activation; DNA fragmentation; Multidrug-resistance
1. Introduction
apoptosis induction in animal and human tumor cell lines
[3±5]. These inhibitors include the shell®sh toxins okadaic
acid and calyculin A and the blister beetle toxin cantharidin
or some of its derivatives [6,7]. All these drugs are fairly
selective inhibitors of Ser/Thr phosphatases 1 and 2A
[8±11]. With the exception of calyculin A they are also
at least 10-fold more potent inhibitors of PP2A than of PP1.
Fostriecin is an experimental anticancer drug [12,13]
which as phosphatase inhibitor displays a much higher
selectivity for PP2A than any of the other agents [14]. A
problem in clinical trials with fostriecin was its limited
availability [15] which might now be solved since its
complete synthesis has been reported recently [16]. The
limited availability of natural inhibitors urged for the
investigation of potential synthetic inhibitors of Ser/Thr-
phosphatases. Such studies were successful with Michael
adducts of ascorbic acid analogues with a,b-unsaturated
carbonyl compounds, especially with acrolein [17]. These
Protein phosphatase inhibitors have been shown in
various reports to be cytotoxic and display an antitumor
potential [1,2] which apparently is the consequence of
* Corresponding author. Tel.: ‡49-551-395777; fax: ‡49-551-399652.
E-mail address: hsteinfe@med.uni-goettingen.de (H.J. Steinfelder).
Abbreviations: CA, cantharidic acid; PP, protein phosphatase; MDR,
multidrug-resistance; MBA, methylene blue assay; AEBSF, 4-(2-ami-
noethyl)benzenesulfonyl ¯uoride; DEVD-AFC, DEVD-7-amino-4-tri¯uor-
omethylcoumarin; NS, trans-b-nitrostyrene; NVF, nitrovinylfuran; NVI,
nitrovinylindole; DNS, dinitrostyrenes; 3-NS, 3-nitrostyrene; CiA, trans-
cinnamic acid; HCiA, hydrocinnamic acid; Cl-NS, 4-chloro-nitrostyrene;
NS-Asc, ascorbic acid adduct of nitrostyrene; MeO, 4-methoxy; TriMeO,
3,4,5-trimethoxy; DAP, 3-(2-nitro-1-phenyl-ethyl)-2,6-diamino-pyridine;
TAP, 5-(2-nitro-1-phenyl-ethyl)-2,4,6-triamino-pyrimidine; DAHP, 5-(2-
nitro-1-phenylethyl)-2,6-diamino-pyrimidin-4-ol; P-AHP, 5-(2-nitro-1-
phenylethyl)-2-pyridin-2-yl-6-amino-pyrimidin-4-ol.
0006-2952/02/$ ± see front matter # 2002 Elsevier Science Inc. All rights reserved.
doi:10.1016/S0006-2952(02)01618-0

604
S.Kaap et al./ Biochemical Pharmacology 65 (2003) 603±610
ascorbic acid adducts which as spiro compounds possess a
spiroacetal moiety comparable to those present in the
natural compounds okadaic acid, tautomycin and caly-
culin A could be shown to be fairly selective inhibitors of
PP1 [17]. The therapeutic antitumor options of Ser/Thr
phosphatase inhibitors, especially PP2A inhibitors, moti-
vated us to test three of these synthetic phosphatase
inhibitors for a potential pro-apoptotic effect [18]. Herein
we observed that the two most potent PP1 inhibitors,
ascorbic acid adducts with acrolein, had no effect on
PP2A activity and no pro-apoptotic potential in the
micromolar range while the Michael adduct of ascorbic
acid with 4-chloro-nitrostyrene was a potent PP2A inhi-
bitor and also a potent inducer of apoptosis at micromolar
concentrations.
Our present study is based on these observations. Here
we investigated which part of the molecule of this new
agent is necessary for the pro-apoptotic effect and which
further modi®cations might either enhance or reduce this
effect. First, the nitrostyrene moiety was conclusively
identi®ed as the essential pro-apoptotic core structure.
Then, we studied substitutions at the NS benzene ring
and substituted the NS core by related heterocyclic aro-
matic structures, e.g. in nitrovinylfuran or in nitrovinylin-
dole, to compare the pro-apoptotic potency of these
derivatives with that of NS. Finally, we tested whether
Michael adduct formation of NS with compounds other
than ascorbic acid affected the cytotoxicity of the NS core.
18 hr at room temperature and the adduct was puri®ed by
column chromatography.
The adducts of nitrostyrene: 3-(2-nitro-1-phenyl-ethyl)-
2,6-diamino-pyridine-NS, 5-(2-nitro-1-phenyl-ethyl)-2,4,
6-triamino-pyrimidine-NS, 5-(2-nitro-1-phenylethyl)-2,6-
diamino-pyrimidin-4-ol and 5-(2-nitro-1-phenylethyl)-2-
pyridin-2-yl-6-amino-pyrimidin-4-ol were generated as
described in [20] by Michael addition of the respective
substituted heterocycles with nitrostyrene in an ethyl acetate
water (1:1) mixture.
2.3. Cell culture
Rat pituitary GH3 tumor cells were cultured in DMEM
medium supplemented with 8% fetal calf serum at 378 in a
humidi®ed atmosphere of 95% air and 5% CO2.
2.4. Viability testingÐmethylene blue assay
The day before treatment 15,000 cells per well were
plated in the inner 60 wells of 96-well plates. Treatment
with the respective agents was done in triplicate in com-
plete DMEM medium (8% fetal calf serum) for 24 hr.
Thereafter the medium was aspirated and 100 mL of ®xa-
tive added (10% formaldehyde in 0.9% NaCl (v/v)) for at
least 30 min at room temperature. The ®xative was de-
canted and cells were stained with 100 mL 1% methylene
blue in 0.01 M sodium-borate buffer (w/v; pH8.5) for
another 30 min [21]. After decanting the dye solution the
wells were washed four times with 0.01 M sodium-borate
buffer (pH8.5). To elute the dye from the cells 100 mL 50%
ethanol in 0.1 M HCl (v/v) were added to each well and the
plates incubated for 10 min at room temperature. After
shaking the plates, the dye content was determined by
measuring absorption at 650 nm in a microplate reader
(Molecular Dynamics). The OD of untreated cells was set
to 100% and the values of the treated samples referred to
this control value.
2. Materials and methods
2.1. Materials
Trans-b-nitrostyrene, 3-nitrostyrene and all three dini-
trostyrenes were purchased from Lancaster. Trans-cinnamic
acid and hydrocinnamic acid were from Sigma-Aldrich. The
caspase-3 substrate DEVD-AFC was obtained from
Bachem.
2.5. Analysis of DNA fragmentation
2.2. Synthesis of nitrostyrene derivatives and related
agents
In order to study apoptotic fragmentation of DNA 1
million cells were plated per well of a 6-well plate and
treated 24 hr later with the respective agent for another 24 hr.
DNA fragments were isolated according to the column
isolation protocol described in [22]. After neutral lysis of
the recovered cells with SDS, proteins and genomic DNA
were precipitated with a CsCl solution and the DNA frag-
ments in the supernatant isolated by the use of GFX columns
(Pharmacia). DNA was separated by agarose gel electro-
phoresis on a 1.6% NuSieve gel (Biozyme). Ten microliters
of sample were mixed with 3 mL loading buffer containing
orange G and loaded on the gels (Gibco Minigel
6 cm  8 cm). Gels were run for about 35 min at 90 V,
stained with ethidiumbromide and documented with the
Biodoc system (Biometra) after transillumination at 312 nm.
4-Chloro-, 4-methoxy-, 3,4,5-trimethoxy-trans-b-nitro-
styrene, nitrovinylindole and nitrovinylfuran were synthe-
sized via Henry condensation reaction [19] of nitromethane
with the respective corresponding aldehydes. For NVF
potassium hydroxide was used as catalyst, for all other
agents ethylene diammonium diacetate was the catalyst.
The reactions were performed under soni®cation for 3 hr at
228. The synthesis of the ascorbic acid adducts is described
in [18]. The dodecanoic acid-4-(2-nitroethenyl)phenyl ester
of NS was synthesized from 4-hydroxy-trans-b-nitrostyrene
and dodecanoic acid dissolved in dichloromethane contain-
ing 4-pyrrolidin-1-yl-pyridine as catalyst. After addition of
N,N0-dicyclohexylcarbodiimide the mixture was stirred for

S.Kaap et al./ Biochemical Pharmacology 65 (2003) 603±610
605
2.6. Fluorometric assay of caspase-3 like activity
as 100%. Graphs were designed and curves were ®tted
with SigmaPlot (SPSS Science). Values presented are
For the measurements of caspase-3 like activity 1 mil-
lion cells were plated per well of a 6-well plate and
treatment started after 24 hr. After the indicated periods
of time cells were scraped off the plates while still being in
the medium. Cells and medium were transferred to 2 mL
vials. Cells were separated from the medium by centrifu-
gation at 500 g for 10 min at 48. The supernatants were
sucked off and cells lysed on ice for 15 min in 125 mL of
buffer (pH7.4) consisting of 10 mM HEPES (w/v), 0.1%
NP-40 (w/v), 0.01 mM Digitonin and 1 mM AEBSF. After
centrifugation at 14,000 g for 10 min at 48 the supernatant
was transferred to a new vial, frozen in liquid nitrogen and
stored until further use at À208.
means Æ SE (N ˆ 3). The LC values calculated from
50
the graphs represent the concentration reducing viability
by 50%.
3. Results
Fig. 1 gives an overview of the agents used in the present
study to gain insights in the structural requirements for the
strong pro-apoptotic effect of nitrostyrene derivatives.
First, we con®rmed in another tumor cell line our previous
results obtained in HIT hamster B-cells [18], in which we
described a new group of pro-apoptotic agents also inhibit-
ing PP2A. In the present study we used rat pituitary GH3
tumor cells to initially evaluate the viability loss induced
by the Michael adduct of ascorbic acid with 4-chloro-
nitrostyrene (Fig. 1A), the most potent substance in the
previous study. In addition to the parental rat pituitary GH3
cells we included a GH3 subclone being at least 10-fold less
sensitive to the pro-apoptotic effect of the phosphatase
inhibitor okadaic acid. The resistant subclone (GH3-
OA100) is overexpressing the P-glycoprotein mediating
the MDR1 multidrug-resistance phenotype and displays a
comparable 10-fold lower sensitivity towards the anti-
cancer drugs doxorubicin and actinomycin D [23]. Both
cell lines were affected to the same extent by the Cl-NS
After thawing the samples on ice 100 mL were pipetted
into a well of a 96-well plate. To each sample were added
150 mL assay buffer (lysis buffer plus 4 mM NaEDTA and
15 mM DTT). The buffer contained the substrate DEVD-
AFC (®nal concentration: 25 mM). The enzymatic reaction
was run at 378. The liberated amount of AFC was repeat-
edly measured in a ¯uorometer (Canberra Packard,
Ex ˆ 425 nm; Em ˆ 530 nm). From the linear part of
the increase in ¯uorescence the enzymatic activity per
mg protein was calculated and the activation by the
respective treatments is presented relative to the activity
of the untreated controls. In parallel the protein content was
measured in a 10 mL sample with the biscinchoninic acid
reaction (Pierce).
derivative with an LC of about 25 mM (Fig. 2). The
50
nitrostyrene adduct of ascorbic acid lacking the chloro-
substitution appeared to be equally effective. Interestingly,
NS itself was the most potent cytotoxic agent
(lc50 ꢀ 15 mM). None of these agents displayed a loss
in effectiveness as a consequence of the MDR1 phenotype
2.7. Data analysis and presentation
Data for individual agents are given as percentage of
viability while setting the viability in the absence of agents
Fig. 1. Overview of tested agents (A) Michael adduct of ascorbic acid with p-chloro-nitrostyrene, a new synthetic phosphatase inhibitor [18], (B) ring
substituted nitrostyrenes, (C) agents used to evaluate the role of the nitro group at the b-position (3-NS; cinnamic acid; hydrocinnamic acid), (D) pyrimidine
and pyridine derivatives, (E) related aromatic heterocycles (NVF: nitrovinylfurane; NVI: nitrovinylindole).

606
S.Kaap et al./ Biochemical Pharmacology 65 (2003) 603±610
Fig. 2. Effect of Cl-NS-Asc, NS-Asc and NS on viability of parental GH3 rat pituitary tumors cells (open columns) and GH3-OA100 cells (filled columns), a
subclone that is resistant to 100 nM okadaic acid. This line is at least 10-fold less sensitive to okadaic acid, doxorubicin and actinomycin D than the parental
cells.
in the GH3-OA100 cells. Thus, the nitrostyrene moiety of
the new agent apparently was suf®cient to promote cell
death in these tumor cells and its effect was not hampered
by the MDR1 phenotype.
Apoptosis speci®c DNA fragmentation is the result of
the preceding activation of caspases, especially down-
stream effector caspases like caspase-3. Thus, the time-
dependent activation of caspase-3 was tested in a ¯uoro-
metric assay using the substrate DEVD-AFC. The effect of
NS and its ascorbic acid adduct was compared to that of the
phosphatase inhibitor cantharidic acid at the concentration
of 25 mM. All three agents resulted in an approximately 40-
fold increase of caspase-3 activation after 18 hr (Fig. 4).
After 15 hr NS and its ascorbic acid adduct had already
reached a comparable level while activity in cantharidic
acid treated cells was still about 25% below that observed
after 18 hr. This observation suggested a faster progression
of apoptosis induced by NS and its derivative compared to
cantharidic acid (see also Fig. 8B).
Next, we con®rmed that NS induced cell death occurred
via apoptosis. As a marker of apoptosis we studied the
induction of DNA fragmentation by NS, its 4-chloro-
substituted derivative and the respective Michael adducts
of ascorbic acid in parental GH3 cells (Fig. 3). While both
chloro-derivatives resulted in a distinct DNA ladder at
25 mM (Fig. 3A), NS itself as well as its ascorbic acid
adduct (Fig. 3B) appeared to induce DNA fragmentation
even at 10 mM. With all treatments the most prominent
ladder was observed at 25 mM. When compared to their
respective adducts both nitrostyrenes displayed ladders at
25 mM that were most prominent at the bands representing
lower molecular weight DNA fragments. This observation
suggests that apoptosis was induced by the nitrostyrenes
themselves and progression in this case was more rapid
than with the respective adducts of ascorbic acid.
Ascorbic acid itself revealed no toxicity up to 1 mM
(data not shown). Taken together, these data had de®nitely
identi®ed nitrostyrene as the core structure required for the
pro-apoptotic activity of the ascorbic acid adducts. There-
fore, all further experiments were intended to refer the
Fig. 3. DNA fragmentation induced by treatment of GH3 cells with (A) NS-Asc or NS or (B) Cl-NS and Cl-NS-Asc.

S.Kaap et al./ Biochemical Pharmacology 65 (2003) 603±610
607
Fig. 4. Activation of DEVDase (caspase-3 like) activity by NS-Asc, NS or
cantharidic acid (CA) after 15 and 18 hr.
effects observed with related agents to those achieved by
NS itself. Our intention was to further elucidate structural
modi®cations that might either enhance or decrease the
pro-apoptotic activity of the NS core.
Fig. 5. Cytotoxic potency of pyrimidine and pyridine adducts of NS
compared to that of NS itself. Cells were incubated for 24 hr with the
indicated concentrations of NS and the respective adducts with 2,4,6-
triamino-pyrimidine (TAP-NS [X ˆ N; R1,2,3 ˆ NH2]), 2,4-diamino-6-
hydroxy-pyrimidine (DAHP-NS [X ˆ N; R1,3 ˆ NH2; R2 ˆ OH]),
2-pyridinyl-4-amino-6-hydroxy-pyrimidine (P-AHP-NS [X ˆ N; R1 ˆ
pyridin-2-yl; R2 ˆ OH; R3 ˆ NH2]) and 2,6-diamino-pyridine DAP-NS
(X ˆ C; R1,2 ˆ NH2; R3 ˆ H).
In the following structure activity analysis we used
viability tests to evaluate the consequences of chemical
modi®cations on the potency of NS. First, we tested
whether the formation of adducts with aromatic hetero-
cycles (Fig. 1D) instead of ascorbic acid affects the cyto-
toxic potency of the NS core (Fig. 5). The adduct with
2,4,6-triamino-pyrimidine was as potent (lc50 ꢀ 15 mM)
and as effective as NS itself, the adduct with 2,4-diamino-
6-hydroxy-pyrimidine was equally effective but 2-fold less
potent (lc50 ꢀ 30 mM) while the adduct with 2-pyridinyl-
4-amino-6-hydroxy-pyrimidine was distinctly less potent
(lc50 ꢀ 75 mM). Finally, the adduct with 2,6-diamino-
pyridine displayed only a remote cytotoxic effect (approxi-
mately 20% viability loss) at 100 mM.
Next, we compared NS with the related aromatic
heterocyclic compounds nitrovinylfuran and nitrovinylin-
dole. Both agents clearly displayed cytotoxic properties
(Fig. 6) but were less potent than NS by 2±3-fold
(NVI: lc50 ꢀ 30 mM; NVF: lc50 ꢀ 45 mM). Comparative
analysis of NVF and its adducts with 2,4,6-triamino-pyr-
imidine, 2,4-diamino-6-hydroxy-pyrimidine and 2,6-dia-
mino-pyridine revealed an order of potency (data not
shown) resembling the one observed with the respective
NS adducts.
Since all of the tested agents contained a nitro group at
the b-position of the side chain we tested whether its
dislocation to the 3-position of the benzene ring (3-NS)
or a replacement by the bioisosteric carboxyl group at the
b-position either in cinnamic or hydrocinnamic acid affects
the pro-apoptotic property (Fig. 1C). All three agents were
Fig. 6. Substitution of the benzene ring by the aromatic heterocycles furan
and indole. Comparison of the 24 hr concentration-dependent cytotoxicity
of nitrovinylfuran (NVF: R ˆ 2-furyl) and nitrovinylindole (NVI: R ˆ 3-
indolyl) with that of nitrostyrene (NS: R ˆ phenyl).

608
S.Kaap et al./ Biochemical Pharmacology 65 (2003) 603±610
Fig. 7. Modifications in the NS core structure. (A) Role of the nitro group at the b-position. The nitro group was either transferred to the 3-position (3-NS) or
substituted by a carboxy group (cinnamic and hydrocinnamic acid). (B) Effects of ring substitutions by methoxy groups or dodecanoic acid on NS toxicity.
without cytotoxic effect (Fig. 7A) up to 50 mM (3-NS) or
even 100 mM (cinnamic or hydrocinnamic acid). Thus,
apparently the nitro group at the b-position is an absolute
requirement for the observed pro-apoptotic effect.
Thereafter, we investigated whether substitutions at the
benzene ring might affect the pro-apoptotic potency of NS.
Data with 4-halogenated NS suggested a comparable
potency (F-NS, Cl-NS) (data not shown). In addition we
tested substitutions at position 4 with a methoxy group or
even larger substituents like dodecanoic acid (Fig. 1B).
The potency of 4-MeO-NS was comparable to that of NS
(lc50 ꢀ 15±20 mM) while 3,4,5-tri-MeO-NS appeared to
be more potent (lc50 ꢀ 10 mM) suggesting substitutions at
positions other than at the 4-position might have a greater
impact on the pro-apoptotic potency (Fig. 7B). Finally, the
introduction of a larger substitution at the 4-position like
Fig. 8. Comparison of the pro-apoptotic effect of NS and 2-, 3- and 4-dinitrostyrene (2-,3-, 4-DNS). (A) Viability testing after a 24 hr treatment. (B) Caspase-
3 activation after 6 hr treatment with NS, its DNS derivatives and the known PP2A inhibitor cantharidic acid. Agents were added at 10 mM (open columns)
and 25 mM (filled columns).

S.Kaap et al./ Biochemical Pharmacology 65 (2003) 603±610
609
the ester with dodecanoic acid reduced the pro-apoptotic
potency by 3±4-fold (lc50 ꢀ 50 mM).
These agents retain the inhibitory activity of the mother
compound towards Ser/Thr phosphatases and also display
prominent pro-apoptotic potencies which might be useful
in new anticancer regimens. Interestingly, some of these
derivatives appear to display a selective cytostatic effect
towards tumor subgroups, e.g. colon cancer cells [27,29]
which are usually considered to be fairly insensitive to
most commonly used anticancer drugs.
We have described easily available synthetic agents
inhibiting PP2A activity which are also potent inducers
of apoptosis. We identi®ed the core structure for apoptosis
induction by the most potent of these agents [18]. This
agent was a Michael adduct of 4-chloro-nitrostyrene with
ascorbic acid. Its NS moiety was proven to be essential for
apoptosis induction.
Apoptosis by NS was shown by the time- and concentra-
tion-dependent increase of caspase-3 activity and subse-
quent DNA fragmentation. NS and its ascorbic acid adduct
were effective in the same concentration range (10±25 mM)
as cantharidic acid, a well known inhibitor of PP2A and
inducer of apoptosis [6,7]. Interestingly, caspase activation
data suggest a distinct faster progression of apoptosis by NS
derivatives when compared to cantharidic acid.
Based on these observations a more detailed structure
activity analysis was performed. Within the NS core
structure the nitro group was required at the b-position
and could not be replaced by a carboxyl group. The
benzene ring on the other hand could be replaced by
aromatic heterocycles like furan or indole. This modi®ca-
tion retained the pro-apoptotic ef®cacy suggesting an
important contribution of the nitro vinyl structure or the
respective saturated nitro ethyl side chain attached to an
aromatic ring system. Since such a replacement resulted in
a reduction of the pro-apoptotic potency, nitrostyrene as
the most potent of these agents appears the basic structure
for further developments. Substitution at the benzene ring,
especially at positions 2 and 3 by a second nitro group or
positions 3,4,5 by methoxy groups enhanced the pro-
apoptotic potency by about 2±3-fold. Thus, this initial
structure activity analysis provided many details to be kept
in mind when the nitrostyrene core is considered as the
basis for the development of derivatives that might be
useful in new anticancer regimens.
In summary, in the present paper we have characterized
nitrostyrene as the structural core underlying the pro-
apoptotic effect of a recently identi®ed synthetic inhibitor
of PP2A-like activity [18]. For this fast and extensive pro-
apoptotic activity a nitro vinyl side chain attached to an
aromatic ring could be identi®ed as the underlying phar-
macophore structure. Further studies should concentrate on
pharmacokinetic aspects, e.g. water solubility, stability at
neutral or slightly acid pHwhich is found in tumors, the
form of drug application and ®nally tissue distribution.
Pharmacodynamic aspects of further developments should
concentrate on the aspects of speci®cally targeting the
nitrostyrene core to tumor cells, reducing its toxicity to
Since the nitro group appeared to be necessary and on
the other hand substitutions at the benzene ring might
increase the potency of NS we compared the potency of
three different b-dinitrostyrenes (DNS) having an addi-
tional substitution of a nitro group at positions 2, 3 or 4
with NS (Fig. 8A). All dinitrostyrenes were at least as
effective as the mother compound. While 4-DNS displayed
only a marginally higher potency, 3-DNS was clearly more
potent (lc50 ꢀ 10 mM) and 2-DNS was the most potent of
all tested agents (lc50 ꢀ 5 mM). In order to support this
observation we compared NS with the dinitrostyrenes in
caspase-3 activation assays (Fig. 8B). Again the order of
potency was 2-DNS ꢁ 3-DNS > 4-DNS ꢁ NS. The two
more potent dinitrostyrenes, 2- and 3-DNS, were strong
caspase activators even after 6 hr at 10 mM, while NS and
4-DNS induced caspase-3 only at 25 mM. Cantharidic acid
was again distinctly less effective at this time point sup-
porting the hypothesis of a slower progression of apoptosis
than with NS or the DNS. The observation that 2-DNS at
25 mM displayed a lower induction than at 10 mM appeared
strange but can be interpreted as a result of a very fast
progression of apoptosis at the higher concentration. Thus,
with this treatment probably cells were detected which
were already severely damaged and have lost activated
caspase in the course of secondary necrosis with its
characteristic disintegration of cell membranes.
4. Discussion
The search for new antitumor drugs has been empowered
by the continuous progress in understanding the mechan-
isms underlying drug-induced apoptosis. In addition,
mechanisms of resistance in tumors towards established
chemotherapeutic drugs are better understood and can be
taken into account when screening for agents that circum-
vent resistance by acting via different initiator mechanisms.
Thus, these new agents might be effective in primarily
resistant tumor cells. The inhibition of the activity of various
protein phosphatases, especially PP1 and 2A, by natural
toxins has been related to a potential antitumor activity of
these drugs which is the consequence of apoptosis induction
[1±8,12±14]. Generally these toxins are expensive and are
also only available to a limited amount.
With the increasing insights in the therapeutic potentials
of selective phosphatase inhibitors attempts to identify,
generate and characterize new natural or synthetic phos-
phatase inhibitors have been intensi®ed [24]. In addition,
structural modi®cations have been performed in order to
extent the potential of already known and characterized
natural phosphatase inhibitors. Some recent papers have
dealt with the aspect of using and further investigating
cantharidin derivatives as anticancer drugs [25±27] and
also generating new derivatives of cantharidin [28±32].

610
S.Kaap et al./ Biochemical Pharmacology 65 (2003) 603±610
[14] Walsh AH, Cheng A, Honkanen RE. Fostriecin, an antitumor anti-
biotic with inhibitory activity against serine/threonine protein phos-
phatases types 1 (PP1) and 2A (PP2A), is highly selective for PP2A.
FEBS Lett 1997;416:230±4.
healthy tissue and ®nally evaluating its ef®cacy in tumor
lines resistant to standard anticancer drugs.
[15] de Jong RS, Mulder NH, Uges DR, Sleijfer DT, Hoppener FJ, Groen
HJ, Willemse PH, van der Graaf WT, de Vries EG. Phase I and
pharmacokinetic study of the topoisomerase II catalytic inhibitor
fostriecin. Br J Cancer 1999;79:882±7.
Acknowledgments
Thanks for synthesis of P-AHP-NS to A. Ludwig,
Leipzig and of FNS to M. Klemm, Leipzig. This work
was supported by grants from the Deutsche Forschungs-
gemeinschaft (STE 422/4-3) and the Faculty of Medicine,
[16] Boger DL, Ichikawa S, Zhong W. Total synthesis of fostriecin (CI-
920). J Am Chem Soc 2001;123:4161±7.
[17] Witt UG, Schultz JE, Dolker M, Eger K. Inhibition of protein
phosphatases by Michael adducts of ascorbic acid analogues with
alpha, beta-unsaturated carbonyl compounds. Bioorg Med Chem
2000;8:807±13.
È
University of Gottingen, to H.J. Steinfelder and the Inno-
vationskolleg Chemisches Signal und biologische Antwort
and the Fonds der Chemischen Industrie to K. Eger.
[18] Fathi AR, Krautheim A, Kaap S, Eger K, Steinfelder HJ. Michael
adducts of ascorbic acid as inhibitors of protein phosphatase 2A and
inducers of apoptosis. Bioorg Med Chem Lett 2000;10:1605±8.
[19] Luzzio FA. The Henry reaction: recent examples. Tetrahedron
2001;57:915±45.
References
[20] Taylor EC, Lin B. Exploitation of a new route to fused pyrroles:
synthesis of TNP-351, homo-MTA, and 5-arylpyrrolo[2,3-]pyrimi-
dines. Tetrahedron Lett 1999;40:4027±30.
[1] Tachibana K, Scheuer PJ, Tsukitani Y, Kikuchi H, Van Eugen D,
Clardy J, Gopichand Y, Schmitz FJ. Okadaic acid, a cytotoxic poly-
ether toxin from two marine sponges of the genus Halichondria. J Am
Chem Soc 1981;103:2469±71.
[21] Oliver MH, Harrison NK, Bishop JE, Cole PJ, Laurent GJ. A rapid and
convenient assay for counting cells cultured in microwell plates:
[2] Kato Y, Fusetani N, Matsunaga S, Hashimoto K. Calyculins, potent
antitumor metabolites from the marine sponge Discodermia calyx:
biological activities. Drugs Exp Clin Res 1988;14:723±8.
[3] Boe R, Gjertsen BT, Vintermyr OK, Houge G, Lanotte M, Doskeland
SO. The protein phosphatase inhibitor okadaic acid induces morpho-
logical changes typical of apoptosis in mammalian cells. Exp Cell Res
1991;195:237±46.
application for assessment of growth factors.
1989;92:513±8.
J Cell Sci
[22] Steinfelder HJ, Quentin I, Ritz V. A fast and sensitive technique to
study the kinetics and the concentration dependencies of DNA frag-
mentation during drug-induced apoptosis. J Pharmacol Toxicol Meth-
ods 2000;43:79±84.
[23] Ritz V, Marwitz J, Sieder S, Ziemann C, Hirsch-Ernst KI, Quentin I,
Steinfelder HJ. Contribution of mdr1b-type P-glycoprotein to okadaic
acid resistance in pituitary GH3 cells. Naunyn-Schmiedeberg's Arch
Pharmacol 1999;360:116±21.
[4] Kiguchi K, Glesne D, Chubb CH, Fujiki H, Huberman E. Differential
induction of apoptosis in human breast tumor cells by okadaic acid and
related inhibitors of protein phosphatases 1 and 2A. Cell Growth
Differ 1994;5:995±1004.
[24] McCluskey A, Sim AT, Sakoff JA. Serine±threonine protein phospha-
tase inhibitors: development of potential therapeutic strategies. J Med
Chem 2002;45:1151±75.
[5] von Zezschwitz C, Vorwerk H, Tergau F, Steinfelder HJ. Apoptosis
induction by inhibitors of Ser/Thr phosphatases 1 and 2A is associated
with transglutaminase activation in two different human epithelial
tumor lines. FEBS Lett 1997;413:147±51.
[25] Yang EB, Tang WY, Zhang K, Cheng LY, Mack PO. Norcantharidin
inhibits growth of human HepG2 cell-transplanted tumor in nude mice
and prolongs host survival. Cancer Lett 1997;117:93±8.
[26] Chen YN, Chen JC, Yin SC, Wang GS, Tsauer W, Hsu SF, Hsu SL.
Effector mechanisms of norcantharidin-induced mitotic arrest and
apoptosis in human hepatoma cells. Int J Cancer 2002;100:158±65.
[27] Peng F, Wei YQ, Tian L, Yang L, Zhao X, Lu Y, Mao YQ, Kan B, Lei
S, Wang GS, Jiang Y, Wang QR, Luo F, Zou LQ, Liu JY. Induction of
apoptosis by norcantharidin in human colorectal carcinoma cell lines:
involvement of the CD95 receptor/ligand. J Cancer Res Clin Oncol
2002;128:223±30.
[6] Morana SJ, Wolf CM, Li J, Reynolds JE, Brown MK, Eastman A. The
involvement of protein phosphatases in the activation of ICE/CED-3
protease, intracellular acidification, DNA digestion, and apoptosis. J
Biol Chem 1996;271:18263±71.
[7] Krautheim A, Brechlin P, Becker K, Winkler M, Steinfelder HJ.
Hamster pancreatic beta cell lines with altered sensitivity towards
apoptotic signalling by phosphatase inhibitors. Br J Pharmacol
2000;129:687±94.
[8] Ishihara H, Martin BL, Brautigan DL, Karaki H, Ozaki H, Kato Y,
Fusetani N, Watabe S, Hashimoto K, Uemura D, et al. Calyculin A and
okadaic acid: inhibitors of protein phosphatase activity. Biochem
Biophys Res Commun 1989;159:871±7.
[28] Ho YP, To KK, Au-Yeung SC, Wang X, Lin G, Han X. Potential new
antitumor agents from an innovative combination of demethylcanthar-
idin, a modified traditional Chinese medicine, with a platinum moiety.
J Med Chem 2001;44:2065±8.
[9] Haystead TA, Sim AT, Carling D, Honnor RC, Tsukitani Y, Cohen P,
Hardie DG. Effects of the tumor promoter okadaic acid on intra-
cellular protein phosphorylation and metabolism. Nature 1989;337:
78±81.
[29] McCluskey A, Bowyer MC, Collins E, Sim ATR, Sakoff JA, Baldwin
ML. Anhydride modified cantharidin analogues: synthesis, inhibition
of protein phosphatases 1 and 2A and anticancer activity. Bioorg Med
Chem Lett 2000;10:1687±90.
[10] Honkanen RE. Cantharidin, another natural toxin that inhibits the
activity of serine/threonine protein phosphatases types 1 and 2A.
FEBS Lett 1993;330:283±6.
[30] Sakoff JA, Ackland SP, Baldwin ML, Keane MA, McCluskey A.
Anticancer activity and protein phosphatase 1 and 2A inhibition of a
new generation of cantharidin analogues. Invest New Drugs 2002;
20:1±11.
[11] Honkanen RE, Codispoti BA, Tse K, Boynton AL. Characterization of
natural toxins with inhibitory activity against serine/threonine protein
phosphatases. Toxicon 1994;32:339±50.
[31] McCluskey A, Walkom C, Bowyer MC, Ackland SP, Gardiner E,
Sakoff JA. Cantharimides: a new class of modified cantharidin
analogues inhibiting protein phosphatases 1 and 2A. Bioorg Med
Chem Lett 2001;11:2941±6.
[12] Hotz MA, Del Bino G, Lassota P, Traganos F, Darzynkiewicz Z.
Cytostatic and cytotoxic effects of fostriecin on human promyelocytic
HL-60 and lymphocytic MOLT-4 leukemic cells. Cancer Res
1992;52:1530±5.
[32] McCluskey A, Keane MA, Walkom CC, Bowyer MC, Sim AT, Young
DJ, Sakoff JA. The first two cantharidin analogues displaying PP1
selectivity. Bioorg Med Chem Lett 2002;12:391±3.
[13] de Jong RS, de Vries EG, Mulder NH. Fostriecin: a review of the
preclinical data. Anticancer Drugs 1997;8:413±8.