On the role of copper and iron in DNA cleavage by ochratoxin A. Structure-activity relationships in metal binding and copper-mediated DNA cleavage

Article abstract of DOI:10.1139/v98-088

Ochratoxin A (OTA, 1: X = Cl) is a fungal carcinogen that facilitates single-strand DNA cleavage and DNA adduction when metabolically activated. To determine if redox-active transition metals induce OTA-mediated DNA damage, we have examined the toxin's ab

Full text of DOI:10.1139/v98-088

907
On the role of copper and iron in DNA cleavage
by ochratoxin A. Structure–activity relationships
in metal binding and copper-mediated DNA
cleavage
Jason A. Ardus, Ivan G. Gillman, and Richard A. Manderville
Abstract: Ochratoxin A (OTA, 1: X = Cl) is a fungal carcinogen that facilitates single-strand DNA cleavage and DNA
adduction when metabolically activated. To determine if redox-active transition metals induce OTA-mediated DNA damage,
we have examined the toxin’s ability to bind Cu(II) and Fe(III) in aqueous media and facilitate DNA cleavage in their
presence using agarose gel electrophoresis and supercoiled plasmid DNA. Using fluorescence spectroscopy, 1 was found to
bind Cu(II) readily at physiological pH, while acidic conditions (pH 2.6) were employed to study Fe(III) binding due to the
formation of Fe-oxide precipitates at higher pH values. Structure–activity relationships employing synthetic derivatives of 1
implied that 1 binds both Cu(II) and Fe(III) by its phenolic oxygen, while the carboxylic acid of its phenylalanine moiety
binds Cu(II), but does not appear to play a role in Fe(III) coordination at pH 2.6. In terms of metal-mediated DNA cleavage,
no role for 1 could be detected in Fe-induced DNA strand scission. With Cu(II), DNA cleavage by the 1:1 copper-bound
complex of 1 could only be initiated by addition of a suitable reducing agent (sodium ascorbate). However, 1 was found to
facilitate DNA cleavage by the Cu(II) complex of 1,10-phenanthroline (Cu(OP)2); a prototypical Cu-mediated nuclease
system that cleaves DNA upon activation by an external reducing agent. Structure–activity relationships employing analogs
lacking the chlorine atom, ochratoxin B (2: X = H), and the lactone (12), indicated that the chlorine atom is essential for
activity of the OTA in potentiating DNA cleavage by Cu(OP)2. The implications of our findings to the genotoxic properties of
1 are discussed.
Key words: ochratoxin, DNA cleavage, copper, iron, 1,10-phenanthroline.
Résumé : L’ochratoxine A (OTA, 1; X = Cl) est un cancérigène de champignon qui, lorsqu’il est métaboliquement activé,
facilite le clivage de l’ADN monobrin et l’adduction de l’ADN. Dans le but de déterminer si les métaux de transition actifs en
oxydoréduction induisent des dommages à l’ADN sous l’influence de l’OTA, on a utilisé l’électrophorèse sur gel d’agarose et
un plasmide d’ADN surtorsadé pour examiner l’habilité de la toxine à se lier au Cu(II) et au Fe(III) en milieu aqueux et à
faciliter le clivage de l’ADN en leur présence. En utilisant la spectroscopie de fluorescence, on a trouvé que 1 se lie facilement
au Cu(II) à pH physiologique alors que, à cause de la formation de précipités d’oxyde de Fe à des valeurs de pH plus élevées,
des conditions acides sont nécessaires pour étudier la liaison avec le Fe(III). Des relations structure–activité développées à
l’aide de dérivés synthétiques de 1 impliquent que 1 se lie au Cu(II) ainsi qu’au Fe(III) par son oxygène phénolique; par
ailleurs, l’acide carboxylique de sa partie phénylalanine se lie au Cu(II), mais ne semble pas jouer de rôle dans la coordination
du Fe(III) à un pH de 2,6. En termes de clivage de l’ADN sous l’influence des métaux, on n’a détecté aucun rôle pour 1 dans
la scission de l’ADN catalysé par le Fe(III). Avec le Cu(II), le clivage de l’ADN par le complexe 1:1 de cuivre lié à 1 ne peut
être initié que par l’addition d’un agent réducteur approprié (ascorbate de sodium). Toutefois, on a trouvé que 1 facilite le
clivage de l’ADN par le complexe du Cu(II) avec la 1,10-phénanthroline, (Cu(OP)2), un prototype de système de nucléase
catalysée par le Cu qui clive l’ADN lors d’une activation par un agent réducteur externe. Les relations structure–activité
utilisant des analogues ne comportant pas d’atome de chlore, ochratoxine B (2, X = H) et la lactone (12), indiquent que
l’atome de chlore est essentiel pour l’activité de l’OTA pour potentialiser le clivage de l’ADN par le Cu(OP)2. On discute des
implications de nos observations en relation avec les propriétés génotoxiques de 1.
Mots clés : ochratoxine, clivage d’ADN, cuivre, fer, 1,10-phénanthroline.
[Traduit par la rédaction]
This paper is dedicated to Professor Erwin Buncel, my mentor, who was the first to stimulate my interest in physical organic chemistry.
Received November 17, 1997.
J.A. Ardus, I.G. Gillman, and R.A. Manderville.1 Department of Chemistry, Wake Forest University, Winston-Salem, NC 27109–7486,
U.S.A.
1
Author to whom correspondence may be addressed. Telephone: (336)-758–5513. Fax: (336)-758–4698. E-mail: manderra@wfu.edu
Can. J. Chem. 76: 907–918 (1998)
© 1998 NRC Canada

908
Can. J. Chem. Vol. 76, 1998
Scheme 1. Synthetic route used for the preparation of the dihydroisocoumarine.
agent (sodium ascorbate). However, more importantly to the
possible mechanism of OTA genotoxicity, the toxin was found
to facilitate DNA cleavage by the Cu(II) complex of 1,10-
phenanthroline (Cu(OP)2), a well-known footprinting reagent
(15) that requires an external reducing agent to convert Cu(II)
→ Cu(I). Structure–activity relationships employing synthetic
derivatives of 1 that lack the chlorine atom, OTB (2, X = H),
and the lactone functionality indicated that these components
(especially the chlorine atom) facilitate the ability of 1 to po-
tentiate DNA cleavage by Cu(OP)2. This contrasting behavior
for 1 in Cu-mediated DNA cleavage is discussed through con-
sideration of the toxin’s chelating versus redox properties.
Ochratoxin A (OTA, 1: X = Cl) is a fungal carcinogen pro-
duced by species of Aspergillus and Penicillium (1). Its occur-
rence in feed is widespread (2) and ingestion has been linked
to the disease Balkan endemic nephropathy in which patients
suffer from urinary tract tumors (3). The toxin consists of a
phenolic dihydroisocoumarin, amide linked to L-phenylalan-
ine, and has been shown to promote DNA adduction (4) and
single-strand DNA cleavage (5); properties that establish a ba-
sis for its genotoxic action.
Results
Synthesis of structural analogs
To derive structure–activity relationships for metal binding
and DNA cleavage by ochratoxin A (OTA, 1: X = Cl), we
prepared analogs that lack the following: the chlorine atom,
OTB, 2: X = H; the dihydroisocoumarine (lactone), (12); and
the phenylalanine moiety, (±)-mellein (7a) (16). Derivatives of
2 containing an altered phenylalanine side-chain, i.e., the
methyl ester (9a) and the amide (9b), were also prepared.
Compounds containing the lactone ring system of 1 were
synthesized according to methods outlined in Scheme 1. Using
procedures described by Snieckus and co-workers (17), o-an-
isic acid (2-methoxybenzoic acid, 3a) was converted into (±)-
mellein (7a) (16), while 2-methoxyisophthalic acid (3b)
yielded the carboxylic acid derivative 7b (17).
To attach the L-β-phenylalanine side-chain moiety, the phe-
nolic groups were initially protected as the acetate and then
reacted with the requisite phenylalanine using a dicyclohexyl-
carbodiimide (DCC)/1-hydroxybenzotriazole (HOBT) peptide
coupling procedure (18), as outlined in Scheme 2. The coupled
products, 9a, 9b, and 11, were isolated as the free phenols as
the acetate protecting groups were lost in the work-up. Hy-
drolysis of the methyl ester derivatives (9a and 11) into the
corresponding free acids (2 and 12) was carried out under basic
conditions. These strategies provided mixtures of stereoisom-
ers for 2, 9a, and 9b that were not separated.
Although the mechanisms underlying the genotoxicity of
OTA are uncertain, in vivo studies have established that the
chlorine atom (6) and lactone (7) are essential and that activa-
tion may be mediated by cytochrome P-450 (8) with sub-
sequent formation of activated oxygen species, such as
superoxide (9) and (or) hydroxyl radicals (10, 11). The ability
of 1 to bind Fe(III) has also been implicated as a possible
source of reactive oxygen species that promote lipid peroxida-
tion (12), and may facilitate DNA cleavage (10). Reports by
Hecht et al. (13) and Li and Trush (14) that certain phenols and
hydroquinones damage DNA by a copper-redox mechanism
implies that Cu(II) ions may also play a role in OTA carcino-
genicity.
Since OTA activation appears to be oxidative (3–11) and
may involve participation of redox-active transition metals
(9–14), we examined the toxin’s ability to bind Cu(II) and
Fe(III) in aqueous solutions and facilitate DNA cleavage in
their presence using supercoiled plasmid DNA and agarose gel
electrophoresis.
In the present article, we report that, although no role for 1
in Fe-mediated DNA cleavage could be detected, the toxin was
found to bind Cu(II) to form a stable 1:1 complex that initiates
DNA strand scission upon activation by a suitable reducing
© 1998 NRC Canada

909
Ardus et al.
Scheme 2. Synthetic route used for attachment of the phenylalanine moiety.
Fig. 1. The UV–absorption spectrum of 1 in ethanol (0.175 mM) in
the absence (bold trace) and presence of 1 equiv. of Cu(OAc)2.
Inset: spectrum of Cu(II) OTA complex at 1.75 mM highlighting
the absorbance at 608 nm.
complexation between 1 (m/z 403) and Cu(II) (m/z 63) leads to
the loss of two protons. This observation strongly implies that
the ligands bearing the two acidic groups of 1 (COOH and
phenolic oxygen) participate in copper coordination. The OTA
also forms a 1:1 complex with FeCl3 that is pink in alcoholic
solution (10). Generation of the complex in methanol followed
by analysis by ES– showed the presence of a single iron com-
plex with [M – H]– = 563. This mass corresponds to a complex
with molecular formula FeC20H18NO6Cl4 and indicates that
FeCl3 (m/z = 161) binding in methanol does not lead to depro-
tonation of 1. Table 1 contains UV–absorption data in ethanol
for the protonated and deprotonated forms of 1, 2, and 12,
along with their corresponding 1:1 Cu(II) and Fe(III) com-
plexes.
The ability of 1 to bind Cu(II) and Fe(III) was determined
in aqueous buffer at 25°C using fluorescence spectroscopy.
Addition of Cu(II) and Fe(III) quenched the fluorescence of
the toxin (10), which permitted determination of apparent
equilibrium binding constants (log K*). That the observed
fluorescence quenching was due to metal binding was con-
firmed by performing the metal titration in the presence of
excess EDTA. This metal chelator inhibited the fluorescence
quenching if added prior to addition of the metal, and restored
the fluorescence of OTA if added after the titration.
Copper and iron binding affinity
Figure 3a shows fluorescence quenching of the emission
spectrum (λem = 441 nm) of 1 by Cu(OAc)2 in 4-mor-
pholineethanesulfonic acid (MES) buffer (pH 6.0), while the
accompanying Scatchard analysis (20) is shown in Fig. 3b.
From the negative slope in Fig. 3b, an apparent equilibrium
binding constant of log K* = 5.35 M–1 for formation of the 1:1
Cu(II) OTA complex was determined. For Fe(III) binding by
1 (log K* = 3.34 M–1), aqueous HCl KCl buffer (pH 2.6) was
utilized due to the formation of Fe-oxide precipitates at higher
pH values.
In ethanol, 1 exhibits an absorption spectrum with λmax at 215
(ε = 28 000 M–1 cm–1) and 333 nm (ε = 5500 M–1 cm–1) (19).
Addition of 1 equiv. of Cu(OAc)2 yielded a pale green solution
with λmax at 222 (ε = 20 911 M–1 cm–1), 258 (sh), 367 (ε =
6420), and 608 nm (ε = 55) (Fig. 1). Figure 2 shows the elec-
trospray mass spectrum of the 1:1 mixture (in methanol) using
negative ionization (ES–). The copper complex was found to
have [M – H]– = 463, which corresponds to a complex with
molecular formula CuC20H16NO6Cl and suggests that
© 1998 NRC Canada

910
Can. J. Chem. Vol. 76, 1998
Fig. 2. Electrospray mass spectrum of the Cu(II) OTA complex in methanol (m/z = 462.9). Also present is free 1 (m/z = 402.1).
Table 1. The UV–absorption data in ethanol for ochratoxin A (OTA, 1), OTB (2), and 12 in the
absence (Free) and presence of base, and with 1 equiv. of Fe(III) and Cu(II).a
Compound
Free
+ Baseb
+ Fe(III)c
+ Cu(II)d
OTA (1, X = Cl)
215 (28 000)
333 (5500)
218 (32 000)
320 (6900)
214 (17 500)
312 (3600)
222 (20 300)
383 (11 000)
230 (10 400)
371 (10 200)
222 (19 900)
342 (5050)
224, 342, 483
222 (21 000), 258 (sh),
367 (6420), 608 (55)
235 (11 050), 331 (sh)
356 (6050), 620 (60)
250 (19 600)
OTB (2, X = H)
241, 340, 508
215, 311
12
322 (3700), 584 (130)
a Absorbance data are given in nm; extinction coefficients are in M–1 cm–1
b +2 equiv. of NaOH.
.
c +1 equiv. of FeCl3.
d +1 equiv. of Cu(OAc)2.
Apparent equilibrium binding constants (log K*) for forma-
tion of 1:1 Cu(II) and Fe(III) complexes of 1 and our synthe-
sized derivatives are listed in Table 2. Values in parentheses
for Cu(II) binding represent the copper binding affinity of the
anionic forms of the various derivatives (log K1:1). For com-
parison to our synthesized sample of 2, metal binding by the
natural stereoisomer of 2 (purchased from Sigma) was deter-
mined, and as anticipated, the binding constants of the pur-
chased 2 and our synthesized sample were nearly identical.
The binding affinity of 1 for Zn(II) in 4-morpholinepropane-
sulfonic acid (MOPS) buffer (pH 7.4) was also measured to
provide insight into Cu(I) binding, as both Zn(II) and Cu(I) are
diamagnetic d10-electron metals that favor a tetrahedral bind-
ing arrangement (21). Also presented are pKa values for 1, 2,
7a, and 12. Values for 1 (7, 22) and 2 (7, 23) are from the
literature, while pKas for 7a and 12 were measured at 25°C in
0.1 M aqueous buffer solutions using the recommended spec-
trophotometric procedure (24).
DNA cleavage
The ability of 1 to effect DNA cleavage in the presence of Fe
and Cu was determined using supercoiled plasmid DNA and
agarose gel electrophoresis. In our initial experiments, at-
tempts were made to establish the toxin’s ability to facilitate
DNA cleavage in the presence of Fe. However, utilizing both
Fe(III) and Fe(II) in the presence of excess 1 under various
conditions (sodium ascorbate and HOOH as additives), no de-
finitive role for 1 was determined, as control lanes in the gels
for free Fe were equally efficient in promoting single-strand
DNA cleavage (data not shown).
For DNA cleavage by Cu(II) (Cu(OAc)2, 20 µM) admix-
ture of 1 had very little effect on the extent of single-strand
DNA cleavage (over a 15 h incubation period at 37°C in
MOPS buffer, pH 7.4). Thus, the ability of the Cu(II) OTA
complex to facilitate DNA cleavage in the presence of sodium
ascorbate was examined. For these experiments, 20 µM Cu(II)
was premixed with 60 µM 1 prior to addition of plasmid DNA
© 1998 NRC Canada

911
Ardus et al.
Fig. 3. (a) Emission spectrum (λex = 380 nm, λem = 440 nm) of
fluorescence quenching of 1 (10 µM) at increasing Cu(II)(OAc)2
concentrations in 10 mM MES (pH 6.0) and 100 mM NaClO4
buffer solution at 25°C, and (b) Scatchard plot of the fluorescence
quenching of part a.
facilitate the production of Cu(I)(OP)2, the species necessary
for DNA cleavage (15). Figure 5 shows that DNA cleavage by
Cu(OP)2 was potentiated by OTA in a concentration dependent
fashion. Additional experiments showed that the reaction was
completely inhibited by the enzyme catalase, which mimicked
the findings observed by Sigman and co-workers (15) for
Cu(OP)2 cleavage in the presence of thiol. However, unlike
thiol-potentiation, the cleavage reaction was also partially in-
hibited by superoxide dismutase (SOD) (ca. 70%, data not
shown), which suggests that the reduction of Cu(II)(OP)2 is
also mediated by O2– (15).
Since in vivo studies on OTA genotoxicity suggest a role
for the chlorine atom and lactone (6, 7, 11), we also examined
Cu(OP)2-mediated DNA cleavage in the presence of 2 and 12
(Scheme 2). Figure 6 shows that under conditions where 1 ef-
fectively potentiates DNA cleavage by Cu(OP)2 (lane 4), the
derivative 12 (lane 5) was only slightly less efficient, while 2
(lane 6) failed to induce DNA cleavage above background
(lane 2). To test if differences in the phenolic pKas for these
species (Table 2) could account for these results, the experi-
ment was repeated in 2-(cyclohexylamino)-ethanesulfonic
acid (CHES) buffer at pH 8.6. Interestingly, no change in the
extent of cleavage was detected for all species (data not
shown), indicating that the presence of the phenoxy anion does
not explain the differences noted in the gel.
The ability of 1 to induce DNA cleavage in the presence of
Fe(III) prebound to the metal chelator EDTA (25) was also
examined. At 20 µM Fe(III) EDTA, additions of 1 up to
250 µM, followed by 30 min incubation time at 37°C, failed
to induce DNA cleavage (data not shown). This result was
contrasted by our findings with Cu(OP)2 and by the reaction
of Fe(III) EDTA (20 µM) with sodium ascorbate (50 µM),
which almost facilitated the complete conversion of Form I
into Form II DNA after the 30 min incubation time.
Discussion
One of the major goals in the field of toxicology is to elucidate
the mechanisms by which xenobiotics produce detrimental
biological responses in an organism (26). This knowledge can
be used to identify chemical structures that predispose a mole-
cule to a sequence of events that leads to toxicity. In the present
study, we have evaluated the role played by Fe(III) and Cu(II)
ions in DNA cleavage by the fungal carcinogen ochratoxin A
(OTA, 1). Since it had been shown that an Fe(III) complex of
1 produces hydroxyl radicals to an extent greater than free iron
(10), it was anticipated that 1 would facilitate Fe-mediated
DNA cleavage. That 1 possesses a para-chlorophenolic group
also suggested that copper ions may promote OTA-induced
DNA damage, based on analogy to Cu-mediated DNA cleav-
age by hydroquinones (13, 14).
Initial studies focused on the ability of the toxin and several
structural analogs (Schemes 1 and 2) to bind Fe(III) and Cu(II)
using fluorescence spectroscopy (Table 2). Both metals were
found to quench the fluorescence of the toxin, which permitted
determination of apparent equilibrium binding constants (log
K*) for formation of the 1:1 complexes. Of particular rele-
vance to the biological properties of these compounds was the
ability to monitor Cu(II) binding at physiological pH, while
acidic conditions (pH 2.6) were employed for Fe(III) binding
due to the formation of Fe-oxide precipitates at higher
and sodium ascorbate. Under these conditions, approximately
99% of Cu(II) is bound by the toxin, determined from log K*
= 6.26 M–1 for formation of the 1:1 Cu(II) OTA complex at
pH 7.4. Figure 4 shows a representative gel highlighting that
the Cu(II) OTA complex is capable of redox cycling and thus
inducing DNA cleavage in the presence of sodium ascorbate.
Additional experiments showed the cleavage to be completely
inhibited by the enzyme catalase (1000 units/mL) and sodium
azide (100 mM), while the hydroxyl radical scavengers, di-
methyl sulfoxide (1 M) and tert-butanol (1 M), provided only
partial protection (data not shown). These results indicate a
requirement for HOOH and favor the involvement of a cop-
per-oxo species over a freely diffusible hydroxyl radical (15).
To further explore Cu-mediated DNA cleavage by 1, its
ability to induce DNA cleavage by Cu(II) prebound to 2 equiv.
of 1,10-phenanthroline (OP) was examined. The OP ligand
binds Cu(II) to form a Cu(OP)2 complex that is a prototypical
copper nuclease system and facilitates DNA cleavage in the
presence of an external reducing agent (15). As reported by
Sigman (15), reduction of Cu(II)(OP)2 to Cu(I)(OP)2 followed
by HOOH oxidation produces a copper-oxo species that
cleaves DNA through H1′-atom abstraction of the deoxyribose
sugar. The utility of Cu(OP)2 here was to test if 1 could
© 1998 NRC Canada

912
Can. J. Chem. Vol. 76, 1998
Table 2. Equilibrium binding data for ochratoxin A (OTA, 1) and structural analogs in aqueous
buffered solutions at 25°C.
pKa
Cu(II)e
Fe(III)i
Zn(II)j
log K*
3.92
f
Compound
COOH/PhOH
log K* (log K1:1
)
log K*
OTA (1, X = Cl)
OTB (2, X = H)
9a (R = OMe)
9b (R = NH2)
12
(±)-Mellein (7a)
a Taken from refs. 7 and 22.
b Taken from refs. 7 and 23.
4.2/7.0a
4.2/7.8b
—
5.35 (6.40)
4.42 (6.24)
3.73 (5.52)g
3.65 (5.45)
5.17 (6.69)h
2.55 (6.95)
3.34
3.49
3.24
3.28
3.71
3.93
—
7.5c
10.4d
c Determined by UV-vis spectroscopy in 0.1 M phosphate buffers.
d Determined by UV-vis spectroscopy in 0.1 M carbonate buffers containing 20 vol.% 1,4-dioxane.
e Determined by fluorescence spectroscopy in 10 mM MES buffer (pH 6.0) containing 100 mM NaClO4.
f log K1:1 = [Cu(II) OTA]/[Cu(II)][OTA2–]; determined from log K* by accounting for the percent anionic forms at
pH 6.0.
g Phenolic pKa of 9a and 9b has been taken as 7.8 to estimate log K1:1
h pKa of COOH has been taken as 4.2 to estimate log K1:1
.
.
i Determined by fluorescence spectroscopy in 10 mM KCl HCl (pH 2.6), 100 mM NaClO4.
j Determined by fluorescence spectroscopy in 10 mM MOPS buffer (pH 7.4), 100 mM NaClO4.
Fig. 5. The OTA–potentiation of Form I DNA cleavage by
Cu(II)(OP)2. Cleavage was carried out at 37°C for 30 min in
10 mM MOPS (pH 7.4), 50 mM NaCl buffer solution. Lane 1:
DNA alone; lane 2: 20 µM Cu(II)(OP)2; lane 3: 20 µM Cu(II)(OP)2
+ 10 µM sodium ascorbate; lane 4–7: 20 µM Cu(II)(OP)2 + 50,
100, 150, and 250 µM 1, respectively.
Fig. 4. The DNA cleavage by Cu(II) OTA in the presence of
sodium ascorbate. Cleavage was carried out at 37°C for 30 min in
10 mM MOPS (pH 7.4), 50 mM NaCl buffer solution. Lane 1:
DNA alone; lane 2: 20 µM Cu(OAc)2; lane 3: 20 Cu(II) OTA; lane
4–7: 20 µM Cu(II) OTA + 10, 20, 50, and 250 µM sodium
ascorbate, respectively.
pH values. This finding indicates that 1 would be more effi-
cient at binding Cu(II) than Fe(III) ions in biological systems.
Although we have not unequivocally determined the exact
Cu(II) binding mode for 1, that formation of the Cu(II) OTA
complex in methanol (Fig. 2) was accompanied by the loss of
two protons suggests strongly that the two acidic groups of 1
(COOH and phenol) participate in Cu(II) coordination. Sup-
port for participation of the phenolic group stems from the fact
that at pH 6, log K* = 5.35 M–1 for 1, while for 2 log K* = 4.42
M–1 (Table 2). Clearly the binding affinity is sensitive to ioni-
zation of the phenolic group. Participation of the COOH group
was supported by the finding that the OTB derivatives 9a and
9b (Scheme 2), which possess the methyl ester and amide
As demonstrated by Bredenkamp et al. (27), 1 adopts a β
phenylalanine group respectively, showed diminished Cu(II)
binding affinity when compared to the free acid 2. The results
presented in Table 2 also indicate that the dihydroisocoumarin
moiety (lactone) of 1 is not important for formation of the 1:1
Cu OTA complex. This hypothesis stems for the fact that the
derivative 12, which lacks the lactone moiety, has a greater
affinity for Cu(II) than 1 (compare log K1:1 values in Table 2).
For both the phenolic oxygen and the carboxylic acid of the
phenylalanine moiety of 1 to participate in Cu(II) binding, it
also appears that the amide nitrogen coordinates Cu(II). This
hypothesis was derived from inspection of molecular models
and from consideration of the conformation of the free toxin
derived from single-crystal X-ray structural studies (27).
form (crystals obtained from benzene) conformation that
© 1998 NRC Canada

913
Ardus et al.
Fig. 6. Structure–activity relationships in OTA–potentiation of
Form I DNA cleavage by Cu(II)(OP)2. Cleavage was carried out at
37°C for 30 min in 10 mM MOPS (pH 7.4), 50 mM NaCl buffer
solution. Lane 1: DNA alone; lane 2: 20 µM Cu(II)(OP)2; lane 3:
20 µM Cu(II)(OP)2 + 10 µM sodium ascorbate; lane 4: 20 µM
Cu(II)(OP)2 + 50 µM OTA; lane 5: 20 µM Cu(II)(OP)2 + 50 µM
12; lane 6: 20 µM Cu(II)(OP)2 + 50 µM 2.
(13) and Li and Trush (14) on the ability of hydroquinones to
induce copper-mediated DNA cleavage. They found that
30 min incubation time at 37°C was more than sufficient to
cleave the DNA through reduction of Cu(II) to Cu(I) with
concomitant oxidation of the hydroquinone to the respective
benzoquinone (13, 14). The inability of 1 to induce Cu-medi-
ated DNA cleavage in the absence of a reducing agent, how-
ever, can be rationalized by consideration of the toxin’s
chelating properties. Unlike ortho- or para-hydroquinone,
which are not expected to coordinate Cu(II) with high binding
affinity, 1 binds Cu(II) very efficiently (Table 2). If binding
affinity of the OTA for Zn(II) (log K = 3.92 M–1 at pH 7.4) is
representative of the toxin’s affinity for Cu(I) (both are d10-
electron metals that favor a tetrahedral binding arrangement
(21)) then the Cu(II) complex is significantly more stable than
the putative Cu(I) complex (log (KCu/KZn) at pH 7.4 =
6.26/3.92 = 1.6). This finding would explain the inability of 1
to facilitate Cu-mediated DNA damage in the absence of a
reducing agent, and is consistent with the fact that Cu(II) is
much less efficient in reduction when it is bound to protein
ligands than when it is free in solution (31). Studies by Mar-
quardt and co-workers (11a) on hydroxyl radical production
by 1 in bacteria also show that additions of divalent cations
such as Fe(II) and Cu(II) ions fail to enhance hydroxyl radical
production, which appears consistent with the findings pre-
sented here on the inability of 1 to cleave DNA in the presence
of these metal ions free in solution. Furthermore, Cu(II) OTA
mediated DNA cleavage in the presence of sodium ascorbate
is probably irrelevant to OTA genotoxicity, since antioxidants
inhibit the toxicity of 1 (9).
To explore the role played by Cu(II) ions in OTA-induced
DNA cleavage in more detail, we also examined the ability of
1 to potentiate DNA cleavage by the Cu(II) complex of 1,10-
phenanthroline (Cu(OP)2). As described by Sigman (15), the
nuclease activity of Cu(OP)2 proceeds by an obligatory, or-
dered mechanism, whereby the cupric complex Cu(II)(OP)2 is
reduced to the cuprous complex, Cu(I)(OP)2, which is then
oxidized by HOOH to generate the copper-oxo species directly
responsible for strand-scission. Since the reaction is inde-
pendent of the reductant used to potentiate it (15), we used the
Cu(OP)2 nuclease system to gain insight into redox properties
of the OTA in DNA cleavage. Interestingly, 1 was found to
potentiate DNA cleavage by Cu(OP)2 (Fig. 5), presumably by
reducing Cu(II)(OP)2 → Cu(I)(OP)2 with the concomitant oxi-
dation of 1. Although the exact nature of OTA oxidation is
uncertain, it is informative that the nonchlorinated derivative
2 failed to induce DNA cleavage by Cu(OP)2. The derivative
12, which lacks the lactone of 1, showed only slightly dimin-
ished activity (Fig. 6), which indicates that the chlorine atom
is more important than the lactone for OTA potentiation of
Cu(OP)2-mediated DNA cleavage.
maximizes hydrogen bonding interactions between the amide,
phenolic group, and lactone carbonyl. The alternative α form
was not detected. Although the fully protonated form of 1 is
not important for Cu(II) coordination, the orientation of the
phenylalanine moiety in the β form appears to facilitate coor-
dination to Cu(II) by the phenolic oxygen and carboxylic acid
of the phenylalanine. This conformation also predicts that the
amide nitrogen participates in Cu(II) binding, which is consis-
tent with previous studies on Cu(II) binding by the amide
group of peptides (28), and Cu(II) phenylalanine complexes
(29).
For Fe(III) binding the data in Table 2 indicates that all
derivatives possess similar binding affinity at pH 2.6. These
results are consistent with the mass spectral data obtained in
methanol, which indicated that FeCl3 binding by 1 is not ac-
companied by proton removal from the toxin. This suggests
that FeCl3 forms a coordinate bond with the phenolic oxygen.
Consistent with this hypothesis, 7a, which lacks both the elec-
tron-withdrawing chlorine atom and amide group (pKa = 10.4),
possesses the greatest affinity for FeCl3 at pH 2.6 (Table 2). In
contrast to Cu(II) binding, the derivatives 9a and 9b do not
show significant differences in binding from 2. This result ap-
pears to indicate that the carboxylic acid does not participate
in Fe(III) binding at pH 2.6 and may suggest that these com-
pounds bind Fe(III) by the phenolic oxygen (7, 10) and an
ortho-carbonyl, either from the lactone (10) or the amide group
(7, 30).
To define the role played by Fe(III) and Cu(II) in OTA-in-
duced DNA cleavage, we examined the ability of 1 to facilitate
DNA strand-scission in the presence of free iron and copper
ions. Utilizing both Fe(III) and Fe(II) ions, no definitive role
for the toxin could be identified. This result was surprising,
given reports that 1 coordinates Fe(III), facilitates lipid peroxi-
dation (12), and produces the hydroxyl radical via the Fenton
reaction (10). However, results by Marquardt and co-workers
(11b) on production of the hydroxyl radical in a microsomal-
NADPH preparation, showed conclusively that addition of 1
and Fe(III) produces the hydroxyl radical in a lower yield than
NADPH and Fe(III) alone. Our in vitro results are in accord
with this finding and reaffirm that coordination of 1 to Fe(III)
is not responsible for OTA genotoxicity.
To provide a rationale for the ability of 1 to induce DNA
cleavage by Cu(OP)2, it is informative to consider oxidation
of phenols, where the oxidation takes place according to the
routes presented in Scheme 3 (32). Under aprotic conditions,
oxidation of phenol (I) produces the cation radical (II), which
deprotonates to the neutral radical (III). For solutions contain-
ing the phenolate ion (V), oxidation to the phenoxyl radical
(III) proceeds readily and the resulting phenoxyl radical (III)
may be further oxidized to the phenoxonium ion (IV), or cou-
ple to yield a variety of products (32).
For Cu-mediated DNA cleavage, 1 was found to form a 1:1
Cu(II) OTA complex that induced DNA strand-scission only
upon activation by a suitable reducing agent (Fig. 4). This re-
sult is in sharp contrast to the reports by Hecht and co-workers
© 1998 NRC Canada

914
Can. J. Chem. Vol. 76, 1998
Scheme 3. Oxidation pathway for common phenols.
Insight from Scheme 3 suggests that single electron transfer
from 1 to Cu(II)(OP)2 would yield the OTA phenoxyl radical
(OTA ) and Cu(I)(OP)2 (eq. [1a]). For Cu(I)(OP)2 to initiate
DNA strand-scission, it must combine with hydrogen peroxide
(HOOH) to generate a copper-oxo species (15). The finding
that SOD inhibits the reaction (70%) suggests that the reduc-
tion of Cu(II)(OP)2 is also mediated by O2 −. Although oxida-
tion of Cu(I)(OP)2 by O2 can yield O2 − (15), another possible
source of superoxide is the oxidation of OTA by O2 to yield
the phenoxonium ion of OTA (OTA+), eq. [1b]. The liberated
may be attributable to the toxin’s redox properties, instead of
its ability to bind redox-active transition metals. The toxin does
not facilitate DNA cleavage in the presence of free Fe, or
Fe(III) bound to EDTA. With Cu(II), 1 forms a stable 1:1
Cu(II) OTA complex that facilitates DNA cleavage upon acti-
vation by sodium ascorbate. However, since OTA toxicity is
inhibited by antioxidants, this pathway is unlikely to play a
role in the mechanism(s) of OTA genotoxicity. A more signifi-
cant finding was the demonstrated ability of 1 to initiate DNA
cleavage by Cu(II)(OP)2. Here, 1 provides the reducing
equivalents to convert Cu(II)(OP)2 → Cu(I)(OP)2. The chlo-
rine atom was found to be critical for this activity, which is
consistent with in vivo studies on OTA genotoxicity (6). Cur-
rently, we are studying oxidation of 1 by electrochemical and
chemical means in an effort to determine the nature of the
oxidized OTA species and the potential of such a species to
induce DNA adduction.
−
O2 could assist OTA in the reduction of Cu(II)(OP)2
(eq. [1c]) and yield HOOH after disproportionation (eq. [1d]).
Oxidation of Cu(I)(OP)2 by HOOH would yield a copper-oxo
species (eq. [1e]), that initiates DNA cleavage (eq. [1f]).
[1a] OTA + Cu(II)(OP)2 → OTA + Cu(I)(OP)2
[1b] OTA + O2 → OTA+ + O2
−
[1c] Cu(II)(OP)2 + O2 − → Cu(I)(OP)2 + O2
[1d] 2O2 – + 2H+ → HOOH + O2
Experimental section
[1e] Cu(I)(OP)2 + HOOH → copper-oxo species
Materials
[1f]
Copper-oxo species + DNA → DNA cleavage
Ochratoxin A (OTA, 1: X = Cl) was purchased from Dr.
Ronald Marquardt, Department of Animal Sciences, Univer-
sity of Manitoba, and was used without further purification.
Supercoiled plasmid (Form I) DNA was a gift from Dr. Fred
W. Perrino, Department of Biochemistry, Bowman Gray
School of Medicine, Wake Forest University. The plasmid was
a derivative of pOXO4 (36) containing the dnaQ gene. The
following enzymes and reagents were obtained commercially
and used without further purification: 1,10-phenanthroline
(OP), copper acetate (Cu(OAc)2), ferric chloride (FeCl3), so-
dium perchlorate (NaClO4), 4-morpholinepropanesulfonic
acid (MOPS), 4-morpholineethanesulfonic acid (MES), 2-(cy-
clohexylamino)ethanesulfonic acid (CHES), ethylenedi-
aminetetraacetic acid (EDTA), 1,3-dicyclohexylcarbo-diimide
(DCC), 1-hydroxybenzotriazole hydrate (HOBT), L-pheny-
lalanine methyl ester hydrochloride, 5-chlorosalicylic acid,
sec-butyllithium (sec-BuLi), N,N,N′,N′-tetramethylethyle-
nediamine (TMEDA), methyl iodide (MeI), lithium diiso-
propyamide (LDA), acetaldehyde, and acetic anhydride
While we have yet to identify the products resulting from OTA
oxidation, it is informative that the intermediacy of the OTA
phenoxyl radical in the in vivo genotoxicity of 1 has been
proposed by Malaveille et al. (6). Oxidation to the putative
phenoxonium ion (OTA+) may also provide a rationale for the
toxin’s ability to facilitate DNA adduction (4). A DNA-de-
rived nucleophile could react directly with OTA+, or reaction
of OTA+ with H2O to furnish a quinone intermediate after loss
of HCl could precede DNA adduction. In this regard, chlorin-
ated phenols have been converted to the respective para-qui-
nones with CuCl2 and O2 (33) and by enzymatic oxidative
pathways (34), and quinones are Michael acceptors that have
been implicated in DNA adduct formation (35).
Conclusions
The present study implies that the genotoxic properties of 1
© 1998 NRC Canada

915
8-Hydroxy-3-methyl-3,4-dihydroisocoumarin (Mellein)
Ardus et al.
(Aldrich); ochratoxin B (2: X = H), catalase, superoxide dis-
mutase (SOD), and L-β-phenylalanine amide (Sigma).
(7a)
To a stirred solution of 5a (0.193 g, 0.873 mmol) in dry THF
(75 mL) at –78°C under argon was added lithium diisopropyl-
amide (LDA, 0.655 mL, 1.31 mmol, from 2.0 M solution in
heptane–THF–ethylbenzene). After stirring for 1 h at –78°C,
the resulting burgundy red mixture was quenched with acetal-
dehyde. The mixture was allowed to warm to room tempera-
ture, stirred for an additional 12 h, and then washed with 1 N
HCl (50 mL) and extracted with EtOAc. The organic layer was
dried over MgSO4, filtered, and the solvent removed under
reduced pressure. The resultant crude residue was then re-
fluxed in 6 N aqueous 6 N HCl (50 mL) for 4 days. The mix-
ture was filtered, washed with water, and dried under vacuum
to afford 7a as a light yellow solid. Recrystallization from
acetone gave 7a as a white solid: yield 0.080 g (52%); mp
Methods
Elemental analyses were carried out by Atlantic Microlab, Inc.
High-resolution mass spectral analyses were carried out by the
Nebraska Center for Mass Spectrometry. Melting points were
taken on a Mel-Temp apparatus and are not corrected. NMR
spectra were recorded on a Varian VXR-200 spectrometer (1H,
200 MHz; 13C, 50 MHz). Chemical shifts are given in parts per
million (ppm) relative to tetramethylsilane (TMS), and cou-
pling constants (J) are reported in hertz (Hz). The following
abbreviations are used: s, singlet; d, doublet; t, triplet; q, quar-
tet; m, multiplet; br, broad. Absorption measurements were
made on a Hitachi U-2001 UV-vis spectrophotometer, and
fluorescence spectra were acquired with a Hitachi F-2000;
both instruments were equipped with a thermostated cell com-
partment. Distilled, deionized water from a Milli-Q system
was used for all aqueous solutions and manipulations. Other
solvents were purified and dried according to standard proce-
dures. Agarose gel electrophoresis was carried out in 40 mM
Tris-acetate buffer (pH 8.0), containing 5 mM EDTA. Agarose
gel loading buffer: 40 mM Tris-OAc (pH 8.0), 5 mM EDTA,
40% glycerol, 0.3% bromophenol blue.
1
38.0°C (lit. (17b) 38°C); H NMR (CDCl3), δ: 11.0 (s, 1H),
7.38 (m, 1H), 6.86 (m, 1H), 6.66 (m, 1H), 4.71 (m, 1H), 2.91
(m, 2H), 1.51 (d, 3H, J = 6.3 Hz); 13C NMR (CDCl3), δ: 162.9,
161.3, 142.1, 134.6, 119.3, 111.0, 74.2, 56.3, 36.2, 20.8.
N,N-Diethyl-2-methoxy-3-(diethylcarbamoyl)benzamide
(4b)
The benzamide 4b (17c) was prepared from 2-methoxyiso-
phthalic acid (0.850 g, 4.33 mmol), excess SOCl2 and diethyl-
amine using standard conditions. After purification by
silica-gel chromatography (elution with EtOAc), 4b was ob-
tained as a white solid: yield 1.13 g (85%); mp 82°C from
Et2O–hexane (lit. (17c) 82°C); 1H NMR (CDCl3), δ: 7.27 (m,
3H), 3.86 (s, 3H), 3.80–3.20 (br, 8H), 1.27 (t, 6H, J = 7.18 Hz),
1.06 (t, 6H, J = 7.14 Hz); 13C NMR (CDCl3), δ: 166.8, 150.6,
127.0, 42.0, 37.9, 12.8, 11.7.
Mass spectra were acquired using a Micromass Quattro II
operating in the negative ion spray mode (ES–). The system
acquired signal over a m/z range of 200–1000 at 8 s/scan with
a 0.06 step. Samples of 1 and Cu(II) OTA (0.1mg/mL) were
prepared in methanol and injected via syringe into a fixed loop
injector port (100 µL volume) interfaced to the ion spray
source. The Cu(II) OTA sample was also prepared in 10 mM
NH4OAc (pH 7.0) buffer and subjected to reverse-phase high-
pressure liquid chromatography (HPLC, Hewlett Packard
(HP) 1100 series with photodiode array detection,
190–600 nm, using a gradient of MeOH in 2 mM ammonium
acetate, pH 7.0), prior to analysis by ESI/MS. The sample of
Fe(III) OTA (0.1mg/mL) was prepared in methanol and was
analyzed by ESI/MS without prior separation by HPLC.
N,N-Diethyl-2-methoxy-3-(diethylcarbamoyl)-6-methyl-
benzamide (5b)
To a stirred mixture of 4b (0.304 g, 0.993 mmol) in dry THF
(75 mL) at –78°C under argon, was added via syringe sec-
BuLi (1.25 mL, 1.19 mmol, from 0.950 M in cyclohexanes)
and TMEDA (0.138 g, 1.19 mmol). The resulting yellow solu-
tion was allowed to stir for 1 h at –78°C, and dry methyl iodide
(0.423 g, 2.98 mmol) was added. The stirred mixture was al-
lowed to warm to room temperature and stirred for an addi-
tional 12 h. The solution was then washed with 1 N HCl
(100 mL) and extracted with EtOAc. The organic layer was
separated, washed with three 20 mL portions of brine, then
dried over MgSO4. The solution was concentrated under re-
duced pressure to afford a crude product that was purified by
silica-gel chromatography. Elution with ether (containing 1%
triethylamine) afforded 5b as a light yellow oil: yield 0.27 g
(85%); 1H NMR (CDCl3), δ: 7.13 (d, 1H, J = 7.6), 7.00 (d, 1H,
J = 7.6), 3.84 (s, 3H), 3.80–3.20 (br, 8H), 2.27 (s, 3H), 1.27 (m,
6H), 1.06 (t, 6H, J = 7.06 Hz); 13C NMR (CDCl3), δ: 167.9,
167.1, 151.3, 136.2, 127.2, 42.6, 42.3, 18.3, 13.5, 13.4, 12.3;
mass spectrum (HREI), m/z: 320.209 (C18H28N2O3 requires
320.210).
N,N-Diethyl-2-methoxybenzamide (4a)
To prepare 4a (37), 2-methoxybenzoic acid (1.00 g,
6.57 mmol) was reacted with excess thionyl chloride (SOCl2)
and diethylamine using standard conditions. After purification
by silica-gel chromatography (1:1 EtOAc–hexane), 4a was ob-
tained as a colorless oil: yield 1.21 g (89%); 1H NMR (CDCl3),
δ: 7.29 (m, 1H), 7.18 (m, 1H), 6.93 (m, 1H), 3.79 (s, 3H),
3.59–3.54 (m, 2H), 3.14 (q, 2H, J = 7.02), 1.24 (t, 3H, J = 7.02),
1.02 (t, 3H, J = 7.02); 13C NMR (CDCl3), δ: 168.3, 154.7,
129.5, 126.9, 126.5, 120.3, 110.5, 55.0, 42.4, 38.4, 13.6, 12.5.
N,N-Diethyl-2-methoxy-6-methylbenzamide (5a)
Using the procedure described by Snieckus and co-workers
(17a), 5a was obtained as a colorless oil: yield 0.191 g (80%);
1H NMR (CDCl3), δ: 7.13–6.66 (m, 3H), 3.78–3.68 (m, 1H),
3.72 (s, 3H), 3.39–3.29 (m, 1H), 3.05 (q, 2H, J = 7.19), 2.17 (s,
3H), 1.18 (t, 3H, J = 7.33), 0.95 (t, 3H, J = 7.19); 13C NMR
(CDCl3), δ: 168.2, 155.1, 135.4, 128.9, 126.2, 122.3, 107.9,
55.3, 42.2, 38.3, 18.5, 13.6, 12.6.
3-Methyl-7-carboxy-8-hydroxy-2,3-dihydroisocoumarin
(7b)
To a stirred solution of 5b (1.20 g, 3.75 mmol) in dry THF
(75 mL) at –78°C under argon was added LDA (2.25 mL,
4.50 mmol, from 2.0 M in heptane–THF–ethylbenzene). After
© 1998 NRC Canada

916
Can. J. Chem. Vol. 76, 1998
stirring for 1 h at –78°C, the resultant burgundy red solution
was quenched with acetaldehyde. The mixture was allowed to
warm to room temperature, stirred for an additional 12 h,
washed with 1 N HCl (100 mL) and then extracted with
EtOAc. The organic layer was separated, dried over MgSO4,
and then concentrated under reduced pressure. The crude resi-
due was then refluxed in 6 N HCl for 4 days. The mixture was
filtered, washed with water, and dried under vacuum, produc-
ing 7b as a white solid: yield 0.528 g (63%); mp 233–235°C
from acetone (lit. (17c) 234–236°C); 1H NMR (DMSO-d6), δ:
7.96 (d, 1H, J = 7.9), 6.88 (d, 1H, J = 7.9), 4.70 (br, 1H), 3.00
(m, 2H), 1.39 (d, 3H, J = 6.27); 13C NMR (DMSO-d6), δ:
168.9, 165.4, 161.9, 146.9, 136.6, 117.9, 115.7, 110.9, 74.8,
34.5, 20.2.
Amide of ochratoxin B (9b)
To a stirred solution of 8 (0.264 g, 0.379 mmol) in dry acetoni-
trile–DMF (8:1, 45 mL) at 0°C under argon was added DCC
(0.094 g, 0.455 mmol) and N-hydroxysuccinimide (0.52 g,
0.455 mmol). The resulting mixture was allowed to warm to
room temperature and then stirred for an additional 1.5 h under
argon. L-Phenylalanineamide (0.062 g, 0.379 mmol) in pyri-
dine (2 mL) was then added in a dropwise fashion. The mixture
was stirred for an additional 12 h, then concentrated under
reduced pressure to leave a residue that was resuspended in
EtOAc (25 mL). The resulting solution was washed with three
20 mL portions of brine, dried over MgSO4, then concentrated
under diminished pressure to afford a crude product that was
purified by silica-gel chromatography. Elution with 80/20
EtOAc–hexanes produced 9b as a white residue: yield 0.027 g
1
3-Methyl-7-carboxy-8-acetate-2,3-dihydroisocoumarin (8)
To a suspension of 7b (0.200 g, 0.901 mmol) in acetic anhy-
dride (10 mL) was added phosphoric acid (0.10 mL). The re-
sulting solution was stirred for 24 h at 50°C. Concentration
under reduced pressure, followed by purification by silica-gel
chromatography (EtOAc) afforded 8 as a white solid: yield
0.233 g (98%); mp 136–137°C; 1H NMR (CDCl3), δ: 10.4 (br,
1H), 8.22 (d, 1H, J = 8.01), 7.24 (d, 2H, J = 8.01), 4.63 (m, 1H),
2.99 (m, 2H), 2.21 (s, 3H), 1.50 (d, 3H, J = 6.4); 13C NMR
(CDCl3), δ: 169.3, 168.6, 161.0, 146.8, 136.7, 125.1, 123.3,
119.4, 77.2, 74.3, 35.8, 21.0, 20.6. Anal. calcd. for C13H12O6:
C 59.09; H 4.58; found: C 59.03; H 4.62.
(20%); H NMR (CDCl3), δ: 12.4 (br, 1H), 8.53 (d, 1H, J =
7.1), 8.29 (d, 1H, J = 8.1), 7.24 (m, 5H), 6.82 (d, 1H, J = 8.1),
6.32 (s, 1H), 5.82 (s, 1H), 4.89 (m, 1H), 4.78 (m, 1H), 3.20 (m,
2H), 2.96 (m, 2H), 1.55 (d, 3H, J = 6.2); 13C NMR (CDCl3),
δ: 173.3, 170.2, 164.0, 160.3, 143.7, 138.7, 136.6, 129.3,
128.6, 126.9, 118.9, 118.5, 108.7, 54.9, 37.8, 34.5, 20.6; mass
spectrum (HREI), m/z: 368.13734 (C20H20N2O5 requires
368.1372).
5-Chloro-2-hydroxy-N-L-phenylalaninebenzamide
methyl ester (11)
To a stirred solution of the acetate of 5-chloro-2-hydroxyben-
zoic acid (1.00 g, 4.66 mmol) in dry acetonitrile–DMF (5:1,
55 mL) at 0°C under argon was added DCC (1.15 g,
5.59 mmol) and HOBT (0.755 g, 5.59 mmol). The mixture
was allowed to warm to room temperature, stirred for an addi-
tional 1.5 h, and then L-phenylalanine methyl ester (1.55 g,
4.66 mmol) in dry pyridine (3 mL) was added in a dropwise
fashion. The resulting mixture was stirred for an additional
12 h and then concentrated under diminished pressure. The
resulting residue was resuspended in EtOAc (25 mL), filtered,
and then washed with three 20 mL portions of brine. The or-
ganic layer was collected, dried over MgSO4, and concentrated
under reduced pressure to afford a crude product that was pu-
rified by silica-gel chromatography. Elution with 40/60
EtOAc–hexanes afforded 11 as a white solid: yield 0.9 g
Methyl ester of ochratoxin B (9a)
To a stirred solution of 8 (0.250 g, 0.87 mmol) in dry acetoni-
trile–DMF (10:3, 65 mL) at 0°C under argon was added DCC
(0.217 g, 1.05 mmol) and HOBT (0.141 g, 1.05 mmol). The
mixture was allowed to warm to room temperature and then
stirred for an additional 1.5 h. L-Phenylalanine methyl ester
(0.157 g, 0.87 mmol) in pyridine (2 mL) was then added in a
dropwise fashion. After stirring for 12 h, the solvent was re-
moved under reduced pressure. The resulting residue was re-
dissolved in EtOAc (25 mL), washed with three 20 mL
portions of brine, dried over MgSO4, and concentrated under
diminished pressure to yield a crude product. Purification by
silica-gel column chromatography (elution with EtOAc–hex-
ane, 80:20) afforded 9a as a colorless residue: yield 0.276 g
(83%); 1H NMR (CDCl3), δ: 12.7 (s, 1H), 8.53 (d, 1H, J = 6.4),
8.36 (d, 1H, J = 7.9), 7.27 (br, 5H), 6.84 (d, 1H, J = 7.9), 5.06
(m, 1H), 4.77 (m, 1H), 3.74(s, 3H), 3.24 (m, 2H), 2.99 (m, 2H),
1.56 (d, 3H, J = 6.4); 13C NMR (CDCl3), δ: 172.0, 170.3,
163.6, 160.5, 143.6, 139.0, 136.1, 129.3, 128.6, 127.1, 119.3,
118.4, 114.0, 108.7, 54.2, 52.3, 38.0, 34.6, 20.6; mass spec-
trum (HREI), m/z: 383.136 (C21H21NO6 requires 383.137).
1
(60%); mp 83–85°C. H NMR (CDCl3), δ: 11.9 (s, 1H),
7.40–6.60 (m, 9 H), 5.07 (m, 1H), 3.80 (s, 3H), 3.26 (m, 2H);
13C NMR (CDCl3), δ: 171.7, 168.4, 160.1, 135.3, 134.5, 129.2,
128.8, 127.5, 125.2, 123.5, 120.1, 114.7, 114.0, 53.3, 52.7,
37.8. Anal. calcd. for C17H16ClNO4: C 61.18; H 4.83; N 4.20;
found: C 61.07; H 4.87; N 4.16.
5-Chloro-2-hydroxy-N-L-phenylalaninebenzamide (12)
The hydrolysis of the methyl ester 11 (1.00 g, 3.00 mmol) into
the free acid 12 was achieved by reaction in a 3% LiOH solu-
tion in 3:1 H2O–MeOH (33). Recrystallization from
MeOH–H2O afforded 12 as a white solid: yield 0.67 g (70%);
mp 172–173°C. 1H NMR (acetone-d6), δ: 12.3 (br, 1H), 8.62
(d, 1H, J = 7.9), 7.92 (m, 1H), 7.54–7.10 (m, 6H), 6.94 (d, 1H,
J = 8.9), 5.01 (m, 1H), 3.4–3.1 (m, 2H); 13C NMR (acetone-d6),
δ: 173.0, 169.2, 160.6, 138.3, 134.6, 130.0, 129.2, 127.7,
127.5, 123.7, 120.3, 116.7, 55.0, 37.7. Anal. calcd. for
C16H14ClNO4: C 60.10; H 4.41; N 4.38; found: C 60.02; H
4.47; N 4.39.
Ochratoxin B (2)
The hydrolysis of the methyl ester 9a (0.383 g, 1.00 mmol)
into the free acid, OTB (2), was achieved by reaction in a 3%
LiOH solution in 3:1 H2O–MeOH. Recrystallization from ben-
zene afforded 2 as a white solid: yield 0.243 g (66%); mp
205–208°C (lit. (19) 220°C); 1H NMR (DMSO-d6), δ: 8.50 (d,
1H, J = 7.4), 8.06 (d, 1H, J = 7.9), 7.24 (m, 5H), 6.95 (d, 1H, J
= 7.9), 4.81 (m, 2H), 3.16 (m, 2H), 1.41 (d, 3H, J = 6.1); 13C
NMR (DMSO-d6), δ: 172.6, 169.1, 163.7, 159.6, 144.8, 137.0,
136.9, 129.3, 128.3, 126.7, 118.5, 118.3, 109.4, 76.1, 53.9,
36.7, 33.7, 20.2.
© 1998 NRC Canada

917
Ardus et al.
Metal binding
bromide (1 ug/mL). The gel was run at 110 V for 2 h and
visualized by UV illumination.
The ability of 1 and our synthesized derivatives to bind Fe(III)
and Cu(II) was studied by fluorescence spectroscopy at 25°C.
Stock solutions of the toxin samples were prepared in 1,4-di-
oxane. Concentrations were determined by UV-vis in ethanol
using extinction coefficients for the phenolic absorptions, ca.
330 nm. For Cu(II) binding, a solution (2 mL) containing
10 mM MES (pH 6.0), 100 mM NaClO4 buffer, or 10 mM
MOPS (pH 7.4), 100 mM NaClO4 buffer was utilized. Studies
on Fe(III) binding were performed in buffer solutions (2 mL)
containing 10 mM HCl KCl (pH 2.6), 100 mM NaClO4. In
each case, 2 µL of the toxin dioxane stock solution was added
for a final concentration of 10 µM. For determination of the
metal binding constants, quenching of the toxin’s emission
spectrum (emission ca. 440 nm, excitation ca. 330–380 nm)
was measured 5 min after each addition of Cu(II)(OAc)2 or
Fe(III)Cl3 to allow binding equilibration. Scatchard analysis
(20) of the quenching was performed using the following equa-
tion: (F/Fo) – 1)/[L] = (k11/ks)K11 – K11(F/Fo) where Fo is the
fluorescence intensity of free toxin, [L] is the ligand concen-
tration, and k11/ks are proportionality constants (20). From a
plot of (F/Fo) – 1/[L] versus F/Fo, an association constant was
derived from the negative slope. For Zn(OAc)2 binding at
pH 7.4 (10 mM MOPS, 100 mM NaClO4), emission spectra
were recorded at 400 nm, and the data was analyzed by the
Benesi–Hildebrand equation (20b, 38). Double-reciprocal
plots of 1/∆A (change in emission) versus 1/[Zn] afforded
straight lines from which the association constant (K11) was
determined from K11 = (y-intercept)/(slope) (20b).
Densitometric quantitation was performed using a Microtek
Scanmaker E6 equipped with PhotoImpact and UTHSCSA Im-
age Tool software. Supercoiled plasmid DNA values were cor-
rected by a factor of 1.3, based on average literature estimates
of lowering binding of ethidium to this structure (39).
Relaxation of supercoiled DNA by Cu(II) ochratoxin A
Reaction mixtures (20 µL total volume) contained 400 ng of
Form I DNA, 10 mM MOPS (pH 7.4), 50 mM NaCl, 20 µM
Cu(OAc)2, and 60 µM OTA. The OTA and Cu(OAc)2 were
premixed prior to addition of Form I DNA. Addition of sodium
ascorbate 10–250 µM initiated the reaction, and mixtures were
incubated at 37°C for 30 min, then quenched by the addition
of 4 µL loading buffer. Samples were loaded onto a 1%
agarose gel containing ethidium bromide (1 µg/mL). The gel
was run at 110 V for 2 h and visualized by UV illumination. In
a separate experiment, the quenching effect of catalase (1000
units/mL), sodium azide (100 mM), DMSO (1 M), and tert-bu-
tanol (1M) was examined.
Relaxation of supercoiled DNA by Cu(OP)2 in the
presence of ochratoxin A
Reaction mixtures (20 µL total volume) contained 400 ng of
Form I DNA, 10 mM MOPS (pH 7.4), 50 mM NaCl, 20 µM
Cu(OAc)2, 40 µM OP, and either 10 µM sodium ascorbate or
50–250 µM OTA (1). Reaction mixtures were incubated at
37°C for 30 min, then quenched by the addition of 4 µL load-
ing buffer. Samples were loaded onto a 1% agarose gel con-
taining ethidium bromide (1 µg/mL). The gel was run at 110 V
for 2 h and visualized by UV illumination. In a separate experi-
ment, the quenching effect of catalase (1000 units/mL) and
SOD (1000 units/mL) on the ability of 50 µM OTA to poten-
tiate DNA cleavage by Cu(II)(OP)2 was examined.
pKa determination
The pH measurements were obtained on a Fisher Scientific
Accumet pH meter 910 using standard glass electrodes. Cali-
bration was done using commercial buffers (BDH, pH 4.00,
7.00, and 10.00, all ± 0.01). The pKa of 12 (Scheme 2) was
determined using 0.1 M phosphate buffers, which were pre-
pared by dilution of a stock (1 M K2HPO4) solution with
deionized water. For (±)-mellein (7a), 0.1 M Na-
HCO3–Na2CO3 buffer solutions were utilized. All UV-vis
spectra were acquired at 25°C using a standard matched pair
of 1 cm quartz cells. In the case of 12, the sample and reference
were filled with 1.00 mL of buffer solution, and to the sample
cell was added 2 µL of a stock solution of 12 in 1,4-dioxane.
For 7a, 1 mL of carbonate buffer containing 20 vol.% 1,4-di-
oxane was used, and to the sample cell was added 2 µL of the
stock solution of 7a in 1,4-dioxane. The UV-visible spectra
were recorded by the overlay method in the wavelength range
of 250–500 nm.
Relaxation of supercoiled DNA by Cu(OP)2 in the
presence of OTA, OTB, and 12
Reaction mixtures (20 µL total volume) contained 400 ng of
Form I DNA, 10 mM MOPS (pH 7.4), 50 mM NaCl, 20 µM
Cu(OAc)2, 40 µM OP, and 50 µM OTA (1), OTB (2), or 12.
Reaction mixtures were incubated at 37°C for 30 min, then
quenched by the addition of 4 µL loading buffer. Samples were
loaded onto a 1% agarose gel containing ethidium bromide
(1 µg/mL). The gel was run at 110 V for 2 h and visualized by
UV illumination. In a separate experiment, the reactions de-
scribed above were repeated, but in 10 mM CHES buffer,
pH 8.6.
Relaxation of supercoiled DNA by FeIII EDTA in the
presence of OTA
Reaction mixtures (20 µL total volume) contained 400 ng
Form I DNA, 10 mM MOPS (pH 7.4), 50 mM NaCl, 20 µM
FeCl3, 20 µM EDTA, and either 50 µM sodium ascorbate or
50–250 µM OTA. Reaction mixtures were treated as described
in the previous experiments.
Relaxation of supercoiled DNA by Fe and Cu in the
presence of ochratoxin A
Reaction mixtures (20 µL total volume) contained 400 ng of
Form I DNA, 10 mM MOPS (pH 7.4), 50 mM NaCl, and 20 µM
each of OTA (1) and metal ion. The metal salts employed were
Cu(OAc)2, Fe(Cl)3, and Fe(NH4)2(SO4)2. For reactions with
iron, external reducing agents (200 µM sodium ascorbate) and
oxidants (200 µM HOOH) were also employed. Reaction mix-
tures were incubated at 37°C for varying lengths of time, then
quenched by the addition of 4 µL of loading buffer. Samples
were loaded onto a 1% agarose gel containing ethidium
Acknowledgment
We thank Dr. Fred W. Perrino (Department of Biochemistry,
Wake Forest University School of Medicine) for the sample of
© 1998 NRC Canada

918
Can. J. Chem. Vol. 76, 1998
16. J. Blair and G.T. Newbold. J. Chem. Soc. 2871 (1955).
17. (a) M. Watanabe, M. Sahara, M. Kubo, S. Furukawa, R.J.
Billedeau, and V. Snieckus. J. Org. Chem. 49, 742 (1984);
(b) M.P. Sibi, S. Chattopadyyay, J.W. Dankwardt, and V.
Snieckus. J. Am. Chem. Soc. 107, 6312 (1985).
18. S. Terao, M. Shiraishi, K. Kato, S. Ohkawa, Y. Ashida, and Y.
Maki. J. Chem. Soc. Perkin Trans. 1, 2909 (1982).
19. H. Xiao, R.R. Marquart, A.A. Frohlich, and Y.Z. Ling. J. Agric.
Food Chem. 43, 524 (1995).
20. (a) G. Scatchard. Ann. N.Y. Acad. Sci. 51, 660 (1949); (b) K.A.
Conners. Binding constants. The measurement of molecular
complex stability. John Wiley & Sons, New York. 1987.
pp. 339–343.
21. F.A. Cotton and G. Wilkinson. Advanced inorganic chemistry.
A comprehensive text. 4th ed. John Wiley & Sons, New York.
1980.
22. F.S. Chu. Arch. Biochem. Biophys. 147, 359 (1971).
23. F.S. Chu. CRC Crit. Rev. Toxicol. 2, 499 (1974).
24. A. Albert and E.P. Sergeant. The determination of ionization
constants. Chapman & Hall, London. 1962.
25. W. Knapp Pogozelski, T.J. McNeese, and T.D. Tullius. J. Am.
Chem. Soc. 117, 6428 (1995).
26. I.M.C.M. Rietjens, C. den Besten, R.P. Hanzlik, and P.J. van
plasmid DNA, Dr. Gary Byrd (RJR Tobacco Co.) for acquisi-
tion of the mass spec data, and Mr. Sumit Borah, Mr. Keith
Merritt, and Ms. Jennifer M. Yezek for preliminary experi-
ments. This work was supported by grant no. IRG from the
American Cancer Society and Wake Forest University.
References
1. K.J. van der Merwe, P.S. Steyn, L. Fourie, D.B. Scott, and J.J.
Theron. Nature (London), 205, 1112 (1965).
2. (a) T. Kuiper-Goodman and P.M. Scott. Biomed. Environ. Sci.
2, 179 (1989); (b) A.E. Pohland, S. Nesheim, and L. Friedman.
Pure Appl. Chem. 64, 1029 (1992).
3. (a) A.M. Bendele, W.W. Carlton, P. Krogh, and E.B. Lillehoj. J.
Natl. Cancer Inst. 75, 733 (1985); (b) M. Castegnaro, R. Plestina,
G. Dirheimer, I.N. Chernozemsky, and H. Bartsch. Mycotoxins,
Endem. Nephropathy Urinary Tract Tumours. Vol. 115. IARC
Sci. Publ., Oxford Univ. Press, Oxford, U.K. 1991; (c) R.R. Mar-
quardt and A.A. Frohlich. J. Anim. Sci. 70, 3968 (1992).
4. (a) Y. Grosse, I. Baudrimont, M. Castegnaro, A.-M. Betbeder,
E.E. Creppy, G. Dirheimer, and L.A. Pfohl. Chem. Biol. Interact.
95, 175 (1995); (b) P.S. Obrecht, Y. Grosse, L.A. Pfohl, and G.
Dirheimer. Arch. Toxicol. 70, 244 (1996).
5. (a) E.E. Creppy, A. Kane, G. Dirheimer, C. Lafarge-Frayssinet,
S. Mousset, and C. Frayssinet. Toxicol. Lett. 28, 29 (1985);
(b) J.C. Seegers, L.H. Bohmer, M.C. Kruger, M.L. Lottering,
and K.M. de A. Toxicol. Appl. Pharmacol. 129, 1 (1994).
6. C. Malaveille, G. Brun, and H. Bartsch. Mutat. Res. 307, 141
(1994).
7. H. Xiao, S. Madhyastha, R.R. Marquardt, S. Li, J.K. Vodela,
A.A. Frohlich, and B.W. Kemppainen. Toxicol. Appl. Pharma-
col. 137, 182 (1996).
8. (a) G.E. de Groene, I.G. Hassing, M.J. Blom, W. Seinen, G.J.
Fink, and G.J. Horbach. Cancer Res. 56, 299 (1996); (b) R.F.
Omar, H.V. Gelboin, and A.D. Rahimtula. Biochem. Pharmacol.
51, 207 (1996).
Bladeren. Chem. Res. Toxicol. 10, 629 (1997).
27. M.W. Bredenkamp, J.L.M. Dillen, P.H. van Rooyen, and P.S.
Steyn. J. Chem. Soc. Perkin Trans. 2, 1835 (1989).
28. H. Sigel and R.B. Martin. Chem. Rev. 82, 385 (1982).
29. (a) J. Balla, P.V. Bernhardt, P. Buglyo, P. Comba, T.W. Ham-
bley, R. Schmidlin, S. Stebler, and K. Varnagy. J. Chem. Soc.
Dalton Trans. 277 (1991); (b) F.V Pamatong, C.A. Detmer III,
and J.B. Bocarsly. J. Am. Chem. Soc. 118, 5339 (1996).
30. A. Stassinopoulos, G. Schulte, G.C. Papaefthymiou, and J.P.
Caradonna. J. Am. Chem. Soc. 113, 8686 (1991).
31. (a) D.P. Mack, B.L. Iverson, and P.B. Dervan. J. Am. Chem. Soc.
110, 7572 (1988); (b) C. Harford and B. Sarkar. Acc. Chem. Res.
30, 123 (1997).
32. H. Lund and M.M Baizer (Editors). Organic electrochemistry, an
introduction and a guide. 3rd ed. Revised and expanded. Marcel
Dekker, Inc., New York. 1991. pp. 615–635.
33. (a) K. Takehira, M. Shimiza, Y. Watanabe, H. Orita, and T.
Hayakawa. J. Chem. Soc. Chem. Commun. 1705 (1989); (b) Y.
Takizawa, T. Munakata, Y. Iwasa, and T. Mitsuhashi. J. Org.
Chem. 50, 4383 (1985).
34. (a) T. Ohe, T. Mashino, and M. Hirobe. Tetrahedron Lett. 36,
7681 (1995); (b) V.M. Samokyszyn, J.P. Freeman, K.R. Maddi-
pati, and R. Lloyd. Chem. Res. Toxicol. 9, 1368 (1996).
35. (a) Y.J. Abul-Hajj, K. Tabakovic, and I. Tabakovic. J. Am.
Chem. Soc. 117, 6144 (1995); (b) M.A. Lewis, D.G Yoerg, J.L.
Bolton, and J.A. Thompson. Chem. Res. Toxicol. 9, 1368 (1996).
36. D. Parsonage, H. Miller, R.P. Ross, and A. Claiborne. J. Biol.
Chem. 268, 3161 (1993).
9. I. Baudrimont, A.-M. Betbeder, A. Gharbi, L.A. Pfohl, G. Dirhe-
imer, and E.E. Creppy. Toxicology, 89, 101 (1994).
10. B.B. Hasinoff, A.D. Rahimtula, and R.F. Omar. Biochim. Bio-
phys. Acta, 1036, 78 (1990).
11. (a) D. Hoehler, R.R. Marquardt, A.R. McIntosh, and H. Xiao. J.
Biol. Chem. 271, 27 388 (1996); (b) D. Hoehler, R.R. Mar-
quardt, A.R. McIntosh, and G.M. Hatch. Biochim. Biophys.
Acta, 1357, 225 (1997).
12. (a) A.D. Rahimtula, J.C. Bereziat, V. Bussacchini-Griot, and H.
Bartsch. Biochem. Pharmacol. 37, 4469 (1988); (b) R.F. Omar,
A.D. Rahimtula, and H. Bartsh. J. Biochem. Toxicol. 6, 203
(1991).
13. (a) R.T. Scannell, J.R. Barr, V.S. Murty, K.S. Reddy, and S.M.
Hecht. J. Am. Chem. Soc. 110, 3650 (1988); (b) U.S. Singh, R.T.
Scannell, H. An, B.J. Carter, and S.M. Hecht. J. Am. Chem. Soc.
117, 12 691 (1995).
37. S.O. de Silva, M. Watanabe, and V. Snieckus. J. Org. Chem. 44,
4802 (1979).
38. H. Benesi and J.H. Hildebrand. J. Am. Chem. Soc. 71, 2703
(1949).
39. C.K. Mirabelli, C.H. Huang, and S.T. Crooke. Cancer Res. 40,
4173 (1980).
14. (a) Y. Li and M.A. Trush. Carcinogenesis, 14, 1303 (1993); (b)
Cancer Res. 54, 1895s (1994).
15. (a) D.S. Sigman. Acc. Chem. Res. 19, 180 (1986); (b) T.B. Thed-
erahn, M.D. Kuwabara, T.A. Larsen, and D.S. Sigman. J. Am.
Chem. Soc. 111, 4941 (1989).
© 1998 NRC Canada